CN109917656B - Circulating cooling water minimum pressure difference energy-saving control system and method based on process medium multi-temperature target - Google Patents
Circulating cooling water minimum pressure difference energy-saving control system and method based on process medium multi-temperature target Download PDFInfo
- Publication number
- CN109917656B CN109917656B CN201910251616.9A CN201910251616A CN109917656B CN 109917656 B CN109917656 B CN 109917656B CN 201910251616 A CN201910251616 A CN 201910251616A CN 109917656 B CN109917656 B CN 109917656B
- Authority
- CN
- China
- Prior art keywords
- process medium
- water
- cooling
- medium temperature
- temperature
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
Images
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P80/00—Climate change mitigation technologies for sector-wide applications
- Y02P80/10—Efficient use of energy, e.g. using compressed air or pressurized fluid as energy carrier
Landscapes
- Feedback Control In General (AREA)
Abstract
本发明公开了一种基于工艺介质多温度目标的循环冷却水最小压差节能控制系统，包括循环冷却水系统和控制节能系统；控制节能系统包括工艺介质温度最不利点选择器、工艺介质温度控制器和给回水压差内环PID控制器；工艺介质温度最不利点选择器用于选择工艺介质温度最不利点；工艺介质温度控制器用于获取冷却给回水压差设定值；给回水压差内环PID控制器用于控制所有上塔阀的开度和给水泵机组的出口水流量。有益效果：使循环冷却水不需手工调节就能根据工业现场工艺介质降温需求自动改变循环冷却水冷却供给量，提高循环冷却水系统能源利用率，实现节能控制。
The invention discloses a circulating cooling water minimum pressure difference energy-saving control system based on multi-temperature targets of process medium, including a circulating cooling water system and a control energy-saving system; the control energy-saving system includes a process medium temperature most unfavorable point selector, a process medium temperature control system The inner loop PID controller of the pressure difference between the feed and return water; the most unfavorable point selector of the process medium temperature is used to select the most unfavorable point of the process medium temperature; the process medium temperature controller is used to obtain the set value of the cooling supply and return water pressure difference; the feed and return water The differential pressure inner loop PID controller is used to control the opening of all upper tower valves and the outlet water flow of the feed pump unit. Beneficial effects: The circulating cooling water can automatically change the cooling supply amount of the circulating cooling water according to the cooling demand of the industrial field process medium without manual adjustment, improve the energy utilization rate of the circulating cooling water system, and realize energy-saving control.
Description
Technical Field
The invention relates to the technical field of circulating cooling water in industrial production, in particular to a circulating cooling water minimum differential pressure energy-saving control system and method based on a process medium multi-temperature target.
Background
In the industrial production processes of chemical industry, electric power, metallurgy and the like, the system often generates a large amount of heat due to combustion, chemical reaction and the like according to needs, so that the temperature of equipment and the system is increased, the quality of a process medium in a production area is influenced, and serious economic loss is caused. The circulating cooling water system is a common temperature control engineering system for process media in an industrial production field, heat generated in the production process is conducted to the natural environment by using a heat transfer medium, the purpose of cooling is achieved, and the circulating cooling water system is wide in application. The pipe network topological structure of the circulating cooling water system is large in scale and complex in structure, the internal components of the pipe network are various, the system design mostly depends on experience, in order to meet production requirements, the water supply capacity and the cooling capacity are improved blindly according to the maximum load and a certain margin, and a large-horse-pull trolley phenomenon is often existed between the system design and actual industrial production requirements, so that a large amount of cooling resource waste is caused.
A traditional industrial circulating cooling water system adopts a plurality of constant-speed water pumps to supply water in parallel, and when the demand of the industrial production area for cooling water changes, the flow of the cooling water is controlled by adjusting outlet valves of the water pumps. However, in an actual industrial field, in order to prevent the current of the water pump motor from being increased due to the fact that the water pump motor is damaged due to the fact that the water pump outlet valve is opened, or the cooling effect of a production area is affected due to the fact that the water flow is reduced by only closing the valve through manual experience, the opening of the outlet valve of each water pump is usually set to be 50%, the great throttling loss is caused in the whole operation period, and the problem of cooling resource waste is not fundamentally solved. The water pump is used as a main energy consumption device of the circulating cooling water system, the electric quantity consumption is remarkable, the power consumption of the water pump of a large circulating cooling water system in one year is up to ten million yuan, and therefore the performance parameters and the operation working condition of the water pump can influence the economic and energy-saving effect of the whole system. The water pump operation management level in a water supply network system in China is relatively low, and the actual working condition of the water pump operation can deviate from the design working condition in most cases, so that the water pump operation efficiency is greatly reduced. The practice and exploration of a novel water pump impeller cutting technology [ J ] water supply and drainage, 2005,31(9):94-96 ] propose to regulate water flow by cutting the outer diameter of an impeller of a water pump, but the method is difficult to implement, has a limited flow regulation range, and cannot be regulated in a wide range following the change of flow demand. In the document 'Zhuanyang, Chua's bridge, closed water circulation system multi-pump parallel variable number regulation flow calculation and prediction [ J ] water pump technology, 2007(3):23-26 ], it is proposed to change the number of parallel pumps running to realize flow regulation. In the documents "Hickok H N.AdjustableSpeed- - -A Tool for Saving Energy resources in the PuM's, Fans, Blowers, and CoM' pressures [ J ]. IEEE Transactions on Industry Applications,1985, IA-21(1):124 and 136", it is proposed to change the water pump rotation Speed by using a variable frequency Speed regulation technology to regulate the water flow, the method has a wider regulation range and higher precision, does not increase the pipeline resistance additionally, can enable the water pump to keep working efficiently, and can save Energy fundamentally. The variable frequency speed regulation technology is combined with the control technology and is applied to the rotating speed regulation of the circulating cooling water pump motor to form a circulating cooling water pump variable frequency water supply control system.
However, the industrial heat exchanger has different arrangement heights and distances from the water supply outlet, and has different requirements on the water supply pressure of the water pump. In order to enable cooling water to reach each heat exchanger at a certain pressure, the pressure at the most unfavorable point of the pressure of a water supply network is tracked and controlled, the rotating speed of a variable-frequency adjusting water pump keeps the pressure at the point constant, and the hydraulic balance of the whole system is maintained, so that the variable-frequency water supply control scheme of the circulating cooling water pump which is most widely applied at present is adopted. The outlet pressure of the water pump in the scheme is changed along with the water consumption, so that the water pump is called as a circulating cooling water pump variable-frequency variable-pressure water supply control system.
The frequency conversion modification of the circulating cooling water pump has great improvement on the protection of a circulating water pump motor and the improvement of a power grid load power factor, but the energy saving rate of the running of the circulating water pump is still not high. This is because the following two factors are not considered when designing the circulating cooling water pump variable frequency and variable pressure water supply system:
firstly, the industrial production environment affects the temperature of cooling water, the environmental temperature in autumn and winter is obviously lower than that in spring and summer, and the environmental temperature at night and the environmental temperature in daytime have larger temperature difference. The heat exchange influence of the industrial production environment on the circulating cooling water system is embodied as follows: the lower the industrial production environment temperature is, the stronger the cooling capacity of a cooling tower in a circulating cooling water system is, the lower the water supply temperature cooled by the cooling tower is, the more obvious the cooling effect of cooling water with the same water flow on the same heat exchange load is, namely the smaller the required circulating cooling water flow is for the same heat exchange requirement.
Secondly, the cooling demand is influenced by the change of the industrial heat exchange load, the industrial heat exchange load in a slack production season is obviously smaller than that in a busy production season, and the heat exchange load is obviously larger than that in daily production after an industrial manufacturer receives an urgent order to increase the production. The heat exchange influence of the industrial heat exchange load change on the circulating cooling water system is embodied as follows: in order to ensure that the industrial production always meets the normal production requirements, the larger the industrial heat exchange load is, the larger the required cooling capacity is, and the larger the cooling water demand for the same temperature is.
Disclosure of Invention
Aiming at the problems, the invention provides a circulating cooling water minimum pressure difference energy-saving control system and method based on a process medium multi-temperature target. The cooling supply quantity of the circulating cooling water is automatically changed according to the cooling requirement of the industrial field process medium, and the energy-saving control is realized. In order to achieve the purpose, the invention adopts the following specific technical scheme:
a circulating cooling water minimum pressure difference energy-saving control system based on a process medium multi-temperature target comprises a circulating cooling water system and a control energy-saving system; the circulating cooling water system comprises N cooling towers, a water suction pool, a water supply pump unit, a water outlet pipe group, M 'heat exchangers and a water return pipe group, wherein the N cooling towers, the water suction pool, the water supply pump unit, the water outlet pipe group, the M' heat exchangers and the water return pipe group are arranged along a circulating water path; a cooling pool is arranged in the cooling tower; the water outlet pipe group comprises N water inlet pipes of the water suction tanks, L water outlet pipes of the water suction tanks, a water supply main pipe and M' water supply branch pipes; the water return pipe group comprises M' water return branch pipes, a water return main pipe and N upper tower water return pipes; any cooling tower is correspondingly connected with the water suction pool through one water suction pool inlet pipe, the water suction pools are connected with the water supply main pipe through L water suction pool outlet pipes in parallel, and the water supply main pipe supplies water to M 'heat exchangers in a one-to-one correspondence manner through M' water supply branch pipes; a water feeding pump is arranged on the water outlet pipe of the water suction pool in parallel; any heat exchanger is correspondingly connected with the water return main pipe through a water return branch pipe, and the water return main pipe is correspondingly connected with N cooling towers one by one through N upper tower water return pipes which are arranged in parallel; process medium temperature sensors are arranged in the M' heat exchangers and used for acquiring real-time temperature detection values of various process media;
a cooling water supply temperature sensor and a cooling water supply pressure sensor are arranged on the water supply header pipe, the cooling water supply temperature sensor is used for acquiring a cooling water supply temperature detection value, and the cooling water supply pressure sensor is used for acquiring a cooling water supply pressure detection value; a cooling return water temperature sensor and a cooling return water pressure sensor are arranged on the return water main pipe, the cooling return water temperature sensor is used for acquiring a cooling return water temperature detection value, and the cooling return water pressure sensor is used for acquiring a cooling return water pressure detection value; an upper tower valve is respectively arranged on the N upper tower return water pipes;
the key technology is as follows: the control energy-saving system comprises a process medium temperature least-unfavorable point selector, a process medium temperature controller and a water supply and return pressure difference inner ring PID controller; the process medium temperature minimum point selector obtains a process medium temperature minimum point according to a process medium temperature deviation value sequence collected by the circulating cooling water system; the process medium temperature controller is used for obtaining a cooling water supply and return water pressure difference set value by combining field operation data according to the least unfavorable point of the process medium temperature;
obtaining a cooling water supply pressure detection value and a cooling water return pressure detection value which are obtained by the detection of the circulating cooling water system, and then obtaining a cooling water supply water pressure difference detection value; the difference between the cooling water supply and return pressure difference detection value and the cooling water supply and return pressure difference set value is obtained to obtain a cooling water supply and return pressure difference deviation value; and the water supply and return pressure difference inner ring PID controller adjusts the opening of all the upper tower valves according to the cooling water supply and return pressure difference deviation value, so that the outlet water flow of the water supply pump unit is changed.
Through the design, on the basis of a circulating cooling water system, a circulating cooling water minimum differential pressure energy-saving control system based on a process medium multi-temperature target is established by combining water pump variable-frequency variable-pressure water supply control. The difference value of the process medium online temperature detection value and the temperature set value is called process medium temperature deviation, and the minimum point of the temperature deviation of each process medium in the whole heat exchanger group is called the process medium temperature least-beneficial point. The circulating cooling water system controls the temperature of all process media in an industrial field to be maintained in a normal, safe and energy-saving production temperature range according to the temperature deviation and deviation change rate of the most unfavorable point of the process media temperature and the real-time detection value of the cooling water supply temperature, determines the required amount of cooling water flow in real time, adjusts an upper tower valve and changes the outlet amount of cooling water of a water pump. Finally, on the basis of meeting the core requirement on the production process medium temperature control, the energy conservation of a water pump motor in the circulating cooling water system is further realized fundamentally.
The further technical scheme is as follows: the process medium temperature deviation value sequence comprises M' process medium temperature deviation values, and any process medium temperature deviation value is equal to the difference value between the corresponding process medium real-time temperature detection value and the corresponding process medium temperature set value; the process medium temperature deviation change rate sequence comprises M' process medium temperature deviation change rates, wherein any one of the process medium temperature deviation change rates is the ratio of the temperature change value of two adjacent detection time periods corresponding to the process medium to the last detection time period.
By adopting the scheme, the temperature acquisition of the process media is realized by combining with a circulating cooling water system, the difference values in all the process media are obtained by combining with the detection value and the set value, and the difference values are used for searching the heat exchanger and the process media which are closest to the temperature set threshold value to control the upper tower valve of the whole system, so that each process medium is ensured to be kept in a safe and energy-saving temperature range.
A control method of a circulating cooling water minimum differential pressure energy-saving control system based on multiple temperature targets of process media is characterized by comprising the following specific steps:
s1: setting a sampling period, and carrying out field operation data acquisition on the circulating cooling water system;
the circulating cooling water system collects real-time temperature detection values of M 'process media in M' heat exchangers on site through the process media temperature sensor; the circulating cooling water system collects a cooling water supply temperature detection value on site through the cooling water supply temperature sensor; the circulating cooling water system collects a cooling water supply pressure detection value on site through the cooling water supply pressure sensor; the circulating cooling water system collects a cooling return water temperature detection value on site through the cooling return water temperature sensor; the circulating cooling water system collects a cooling return water pressure detection value on site through the cooling return water pressure sensor;
s2: obtaining M 'process medium temperature deviation values and M' corresponding process medium temperature deviation change rates according to M 'process medium real-time temperature detection values acquired on site and M' corresponding process medium temperature set values; m' process medium temperature deviation values form the process medium temperature deviation value sequence; m' process medium temperature deviation change rate values form the process medium temperature deviation change rate sequence;
s3: the process medium temperature least-benefit point selector selects the temperature deviation minimum value in the process medium temperature deviation value sequence as the process medium temperature least-benefit point, and obtains the process medium temperature deviation change rate corresponding to the process medium temperature least-benefit point and the corresponding heat exchanger;
s4: the process medium temperature controller acquires a process medium temperature deviation value and a process medium temperature deviation change rate of a corresponding heat exchanger according to the least favorable point of the process medium temperature, and inputs field operation data of the corresponding heat exchanger into a corresponding process medium temperature prediction model to obtain a cooling water supply and return water differential pressure set value and a process medium temperature predicted value;
s5: the cooling water supply pressure detection value and the cooling water return pressure detection value are subjected to difference to obtain a cooling water supply and return pressure difference detection value; and the difference between the set value of the cooling water supply and return pressure difference and the detected value of the cooling water supply and return pressure difference is used for obtaining a cooling water supply and return pressure difference deviation value, and the cooling water supply and return pressure difference deviation value is sent to the water supply and return pressure difference inner ring PID controller to adjust the opening of the upper tower valve, so that the outlet water flow of the water supply and supply pump unit is changed.
According to the method, the closed loop for controlling the water temperature of the heat exchanger in the production area monitors the cooling condition of each production area to be cooled in real time, and the opening degree of the upper tower valve is adjusted to accurately control the required cooling water amount in the heat exchanger of each production area to be cooled, so that the waste of cooling resources is reduced while the temperature of the process medium is controlled within a safe production range, energy is saved, and the cooling effect is better. When the opening of the upper tower valve reaches the maximum and the temperature of the process medium cannot be effectively controlled, a measure for increasing the refrigeration effect of the cooling tower can be taken.
Still further, the specific step of finding the minimum value in the sequence of the process medium temperature deviation values as the process medium temperature least favorable point by using the process medium temperature least favorable point selector in the step S3 includes:
s31: initializing, setting M 'technological medium temperature deviation values to form a difference group, totaling M' technological medium temperature deviation values, and enabling W to bek＝M’；k＝1
S32: let Wk+1W is equal to Wk + Xk+1Can be divided by M' evenly, X is a gap with large filling difference; and X is equal to 0-M' -1;
s33: calculating Wk+2＝Wk+1/M’
S34: from Wk+2In the group, the minimum value is found out from the M' process medium temperature deviation values of each group by adopting a cross comparison method to obtain Wk+2A process medium temperature deviation value;
s35: judgment of Wk+2Whether it is equal to 1; if so, taking the process medium temperature deviation value as the least unfavorable point of the process medium temperature; otherwise, let k be k + 2; return is made to step S32.
And a most unfavorable point selector is additionally arranged, and a large number of minimum values of temperature deviations of the process medium are searched to determine the most unfavorable point of the process medium temperature and pay attention to the real-time performance and accuracy of the selection of the most unfavorable point of the process medium.
Still further, the building steps of the M' process medium temperature prediction model deep learning neural networks in the step S4 are as follows:
s411: numbering M' heat exchangers, acquiring historical data generated by running of a circulating cooling water system in X sampling periods, and screening the acquired historical data according to screening conditions to obtain training data of a process medium temperature prediction model in the heat exchangers;
s412: determining characteristic variables from historical data, taking process medium temperature deviation, process medium temperature deviation change rate, cooling water supply temperature detection value and cooling water supply and return pressure difference detection value in the historical data of the heat exchanger I as input data, performing data normalization processing to obtain a normalized data set, and dividing the normalized data set into a training sample set and a test sample set;
s413: performing greedy unsupervised pre-training on the training sample set layer by layer based on a stacked automatic encoder to obtain a weight matrix W of an input layer and a hidden layer initialized by a deep learning neural network and a threshold matrix B of the input layer and the hidden layer;
s414: and (3) fine adjustment of parameters: finely adjusting a weight matrix W of an input layer and a hidden layer initialized by a deep learning neural network and a threshold matrix B of the input layer and the hidden layer until the iteration times reach the maximum value of the iteration times to obtain an initial process medium temperature prediction model based on a stacking automatic encoder;
s415: and (5) evaluating the initial process medium temperature prediction model obtained in the step (S414) by using the test sample data set to obtain a process medium temperature prediction model based on the stacking automatic encoder.
Compared with the traditional shallow neural network, the process medium temperature prediction model based on the stacked automatic encoder effectively solves a series of problems caused by the random initialization of the parameters of the traditional neural network, can effectively mine the implicit relation of each data, and greatly improves the accuracy of process medium temperature prediction control in the industrial production field; compared with the traditional stack automatic encoder method, the method can improve the prediction control precision of the temperature of the process medium in the industrial production field. The method is applied to the energy-saving control of the circulating cooling water system in the industrial production field, can quickly and accurately give the set value of the cooling water supply and return water pressure difference, is beneficial to industrial production field management personnel to master the temperature change trend of the process medium in each heat exchanger in real time, and most importantly, reduces the energy consumption of the circulating cooling water system.
Still further, the characteristic variables comprise a process medium temperature deviation value, a process medium temperature deviation change rate, a process medium real-time temperature detection value, a cooling feed water temperature detection value and a cooling feed water return pressure difference detection value in the circulating cooling water system in any heat exchanger.
Still further, the screening condition is historical data that the process medium temperature is within a safe and energy-saving temperature value interval, and the safe and energy-saving temperature value interval is within a process medium temperature threshold value interval.
The difference between the safe and energy-saving temperature value interval and the process medium temperature threshold value interval is set after induction according to the historical experience technology of technicians.
Further, in step S413, when greedy unsupervised pre-training is performed on the training sample set layer by layer, the training sample set is divided into P groups of small batch training samples, training is performed in sequence, a Dropout technique is adopted, a part of neurons are randomly selected to pause working, iteration is performed in sequence, training is performed layer by layer, and weight matrix W, input layer threshold matrix B and hidden layer threshold matrix B of the input layer and hidden layer initialized by the neural network are deeply learned.
In step S413, the stacked automatic encoder sae (stacked automatic encoder) is a typical deep learning neural network, and its basic constituent unit is an Automatic Encoder (AE), and its network structure is composed of an encoder and a decoder: the input vector is mapped to a feature vector in the hidden layer by the encoder, and then the feature phasor is reconstructed to the original input vector by the decoder.
When given an input sample set X ═ XiI is more than or equal to 1 and less than or equal to N, wherein N is the total number of samples, and x isiThe ith training sample in the sample set has dimension n. Let H ═ HiI is more than or equal to 1 and less than or equal to N is a hidden layer characteristic vector set, hiThe feature vector corresponding to the ith sample has a dimension M', and the coding relationship between X and H is:
H＝sf(WX+B)
in the formula: w is a weight matrix of the input layer and the hidden layer; b is an input layer and hidden layer threshold matrix; sfFor the neuron activation function of the encoder, a Sigmoid function is usually used, which has good feature recognition:
sf(z)＝1/(1+exp(-z))
in the formula: z is the input vector.
The decoder is the inverse operation of the encoder, and takes the characteristic vector of the hidden layer as the input vector
In order to output the set of vectors,
and an output vector corresponding to the ith sample is obtained, and the dimension is n, so that the expression of the decoder is as follows:
in the formula: w' is a weight matrix of the hidden layer and the output layer; b' is a hidden layer and a transport layerA threshold matrix of the layer; sg is the neuron activation function of the decoder. The automatic encoder achieves the purpose of feature extraction by minimizing the reconstruction error between the output vector and the input vector, and the formula of the reconstruction error is as follows:
and continuously adjusting the network weight and the threshold by using a gradient descent algorithm to reduce the reconstruction error, wherein the formula is as follows:
in the formula: l is the learning rate;
to represent
Calculating the deviation of the weight W;
to represent
The bias is calculated for the threshold B.
The SAE is a deep learning neural network model composed of stacked AE, and the output of the lower AE is used as the input of the upper AE. A gradual abstraction of features is achieved by stacking of AEs, ultimately resulting in more compact, useful features. WnIs a weight matrix of the n-1 st hidden layer and the n-th hidden layer, BnThe threshold matrix is the n-1 st hidden layer and the nth hidden layer. The training process comprises two steps of greedy layer-by-layer unsupervised pre-training and supervised fine tuning. Greedy non-supervision pre-training layer by layer obtains the initial weight and the threshold value of the network through layer by layer training, the input of the bottom layer AE is original data, and the output data of the hidden layer is used as the input data of the upper layer AE.
After the layered pre-training is completed, the hidden layers are stacked, and the relationship between the input data and the output data is expressed as follows:
in the formula: f is an activation function, xiW, B are respectively the network initialization weight and threshold value obtained by pre-training layer by layer,
is the predicted value of the ith sample. Constructing an error loss function of the actual value and the predicted value, wherein the formula is as follows:
in the formula: n is the total number of samples, yiIs the actual value of the ith sample. The whole network weight and the threshold value are finely adjusted through the back propagation from top to bottom, and the error between the predicted value and the actual value is reduced.
The SAE can effectively extract high-order features of data through greedy unsupervised pre-training layer by layer, better approaches complex functions, reduces parameter optimization space, can quickly obtain network parameters, and improves deep feature learning capability of a neural network.
Compared with a shallow machine learning algorithm, when the deep network processes high-dimensional data, the overfitting problem is more easily generated due to the complex network structure of the deep network, so that the generalization capability of the model is limited.
Dropout is a mainstream anti-overfitting technique, and its basic idea is: during model training, a part of nodes are randomly selected to be out of operation, the nodes store the weight of the last iteration, and the output is set to be 0. And the selected nodes restore the weight values reserved before in the process of the next iteration, and part of the nodes are randomly selected again to repeat the process. The network structure is changed to a certain extent in each iteration process, and by adopting a Dropout technology, part of neurons selected randomly do not work temporarily, so that the combined action among specific nodes is reduced, the dependence of network output on the states of the specific nodes is reduced, and overfitting is prevented.
Further, when the initial process medium temperature prediction model is evaluated, the improved stack automatic encoder after test training is tested by using a test sample data set, and the average percent error (MAPE) is used as a standard for measuring the evaluation performance of the improved stack automatic encoder, wherein the expression is as follows:
in the formula: y isi、
And respectively obtaining an actual value of the pressure difference of the circulating cooling water supply and return water of the ith sample and a predicted value obtained through SAE evaluation. In step S414, when the parameters are fine-tuned, a top-down small-batch RM' SProp optimization method is used to fine-tune the weight matrix W threshold matrix B initialized by the deep learning neural network.
The method comprises the following specific steps: the global learning rate l and the decay rate ρ are set, and the initial cumulative variable r1 is 0 and r2 is 0.
Selecting a small batch of data sets containing M' samples from the training set, and calculating the gradient according to an error loss function:
calculating the cumulative squared gradient, as shown in equation (9):
in the formula: as an element-by-element product symbol.
Updating the weight and threshold parameters, respectively:
and when the iteration times reach the requirements, stopping operation, otherwise, returning to the step 2 to continue to execute the calculation.
Further describing, in step S5, the specific control content of the supply and return water pressure difference inner ring PID controller on the opening of the upper tower valve and the outlet water flow of the water supply pump unit is as follows:
if the deviation value P of the cooling supply and return water differential pressure is greater than 0, the supply and return water differential pressure inner ring PID controller sends an upper tower valve amplification control signal for controlling the upper tower valve to increase the valve opening; if the cooling water supply and return pressure difference deviation value P is equal to 0, the water supply and return pressure difference inner ring PID controller keeps the original control state;
if the deviation value P of the cooling supply and return water pressure difference is less than 0, the supply and return water pressure difference inner ring PID controller sends a control signal for adjusting the upper tower valve to be small, and the control signal is used for controlling the upper tower valve to reduce the opening of the valve; through the design, the opening degree of the upper tower valve is adjusted by combining the differential pressure deviation value of cooling water supply and return water, so that the temperature of all process media is ensured to be within the threshold value. And ensures the environmental protection performance.
The invention has the beneficial effects that: the cooling water supply quantity of the circulating cooling water can be automatically changed according to the cooling requirement of the industrial field process medium without manual adjustment, the energy utilization rate of the circulating cooling water system is improved, and energy-saving control is realized.
Drawings
FIG. 1 is a process diagram of a circulating cooling water system of a chemical plant A;
FIG. 2 is a block diagram of a variable-frequency and variable-pressure water supply control of a water supply pump unit in a chemical plant A;
FIG. 3 is a schematic diagram of a heat exchanger set in a water synthesis area in a circulating cooling water system of a chemical plant A;
FIG. 4 is a flow chart for finding the worst point of process media temperature;
FIG. 5 is a control block diagram of a recirculated cooling water minimum differential pressure energy saving control system based on a process media multiple temperature target;
FIG. 6 is a flow chart of a recirculated cooling water minimum differential pressure energy saving control based on process media multiple temperature targets;
FIG. 7 is a flow chart of a process media multiple temperature target set point switching multiple parameter predictive control algorithm;
FIG. 8 is a block diagram of an automatic encoder;
FIG. 9 is a stacked self-encoder architecture;
fig. 10 is a Dropout self-encoder architecture.
Detailed Description
The following provides a more detailed description of the embodiments and the operation of the present invention with reference to the accompanying drawings.
A minimum differential pressure energy-saving control system of circulating cooling water based on multiple temperature targets of process media comprises a circulating cooling water system and an energy-saving control system.
In this embodiment, the chemical plant a is taken as an example, and the circulating cooling water energy saving control is performed. As can be seen from fig. 1, a process diagram of a circulating cooling water system of a chemical plant a, which can be seen in combination with fig. 1, includes N cooling towers, a water suction pool, a water supply pump unit, a water outlet pipe group, M' heat exchangers and a water return pipe group, which are arranged along a circulating water path; a cooling pool is arranged in the cooling tower; in this embodiment, M' is a positive integer, there are three heat exchanger groups: a synthesis zone heat exchanger group, a urea zone heat exchanger group and a power zone heat exchanger group.
The water outlet pipe group comprises N water inlet pipes of the water suction tanks, L water outlet pipes of the water suction tanks, a water supply main pipe and M' water supply branch pipes; in the present embodiment, the number of heat exchangers M' is 50. Wherein, 50 heat exchangers are divided into three heat exchanger groups, namely a synthesis area heat exchanger group, a urea heat exchanger group and a power heat exchanger group. The three heat exchanger groups have 50 heat exchangers in total. In this embodiment, the water return pipe group includes M' water return branch pipes, a water return main pipe, and N upper tower water return pipes;
in this embodiment, any one of the cooling towers is connected to the water suction pool through one water inlet pipe of the water suction pool, the water suction pool is connected to the water supply main pipe through L water outlet pipes of the water suction pool in parallel, and the water supply main pipe supplies water to M 'heat exchangers through M' water supply branch pipes in a one-to-one correspondence manner; a water feeding pump is arranged on the water outlet pipe of the water suction pool in parallel;
in this embodiment, any one of the heat exchangers is correspondingly connected with the water return main pipe through one water return branch pipe, and the water return main pipe is correspondingly connected with the N cooling towers one by one through N upper tower water return pipes which are parallel;
in this embodiment, process medium temperature sensors are arranged in each of the M' heat exchangers, and the process medium temperature sensors are used for acquiring real-time temperature detection values of various process media;
in the embodiment, a cooling feed water temperature sensor and a cooling feed water pressure sensor are arranged on the feed water header pipe, the cooling feed water temperature sensor is used for acquiring a cooling feed water temperature detection value, and the cooling feed water pressure sensor is used for acquiring a cooling feed water pressure detection value;
in this embodiment, a cooling return water temperature sensor and a cooling return water pressure sensor are arranged on the return water main pipe, the cooling return water temperature sensor is used for acquiring a cooling return water temperature detection value, and the cooling return water pressure sensor is used for acquiring a cooling return water pressure detection value;
in this embodiment, N upper tower return pipes are respectively provided with an upper tower valve;
in this embodiment, as can be seen from fig. 2, 3 and 5, the control energy-saving system includes a process medium temperature least-benefit selector, a process medium temperature controller and a supply-return water pressure difference inner loop PID controller; the process medium temperature minimum point selector obtains a process medium temperature minimum point according to a process medium temperature deviation value sequence collected by the circulating cooling water system;
the process medium temperature controller is used for obtaining a cooling water supply and return water pressure difference set value by combining field operation data according to the least unfavorable point of the process medium temperature;
obtaining a cooling water supply pressure detection value and a cooling water return pressure detection value which are obtained by the detection of the circulating cooling water system, and then obtaining a cooling water supply water pressure difference detection value; the difference between the cooling water supply and return pressure difference detection value and the cooling water supply and return pressure difference set value is obtained to obtain a cooling water supply and return pressure difference deviation value; and the water supply and return pressure difference inner ring PID controller cools the water supply and return pressure difference deviation value to adjust and control the opening of all the upper tower valves, so that the outlet water flow of the water supply pump unit is changed.
Referring to fig. 3, a schematic diagram of a heat exchanger group in a synthesis area is shown, cooling feed water enters each heat exchanger in an industrial production area through a water outlet pipe group, a temperature difference exists between a process medium in the heat exchanger and the cooling feed water, the process medium continuously transfers heat to the cooling feed water to realize self temperature reduction, and the cooling feed water flows out of the heat exchanger group carrying heat exchange quantity and is sent to a water return pipe group to become cooling return water. The whole heat exchange process must ensure that the temperatures of different process media in each heat exchanger always meet the respective production requirements, so that all process media in an industrial production field need to be subjected to real-time temperature detection and control.
Preferably, the process medium temperature deviation value sequence comprises M' process medium temperature deviation values, the process medium temperature deviation values are sequentially calculated according to the serial numbers of the heat exchangers, and the process medium temperature deviation value sequence is obtained according to the serial numbers. The temperature deviation value of any process medium is equal to the difference value between the real-time temperature detection value corresponding to the process medium and the temperature set value corresponding to the process medium;
the process medium temperature deviation change rate sequence comprises M' process medium temperature deviation change rates, wherein any process medium temperature deviation change rate is the ratio of the temperature change value of two adjacent detection time periods corresponding to the process medium to the last detection time period.
In this embodiment, the process media temperature deviation value corresponds to a process media temperature deviation change rate.
Referring to fig. 6, a control method of a circulating cooling water minimum differential pressure energy-saving control system based on multiple temperature targets of a process medium is characterized by comprising the following specific steps:
s1: setting a sampling period, and carrying out field operation data acquisition on the circulating cooling water system;
the circulating cooling water system collects real-time temperature detection values of M 'process media in M' heat exchangers on site through the process media temperature sensor;
the circulating cooling water system collects a cooling water supply temperature detection value on site through a cooling water supply temperature sensor;
the circulating cooling water system collects a cooling water supply pressure detection value on site through a cooling water supply pressure sensor;
the circulating cooling water system collects a cooling return water temperature detection value on site through a cooling return water temperature sensor;
the circulating cooling water system collects a cooling backwater pressure detection value on site through a cooling backwater pressure sensor;
s2: obtaining M 'process medium temperature deviation values and corresponding M' process medium temperature deviation change rates according to M 'process medium real-time temperature detection values acquired in the step S1 on site and M' process medium temperature set values; forming a process medium temperature deviation value sequence by the M' process medium temperature deviation values; m' process medium temperature deviation change rate values form a process medium temperature deviation change rate sequence;
s3: the process medium temperature least-benefit point selector selects the temperature deviation minimum value in the process medium temperature deviation value sequence as the process medium temperature least-benefit point, and obtains the process medium temperature deviation change rate corresponding to the process medium temperature least-benefit point and the corresponding heat exchanger;
s4: the process medium temperature controller acquires a process medium temperature deviation value and a process medium temperature deviation change rate of a corresponding heat exchanger according to the least favorable point of the process medium temperature, and inputs field operation data of the corresponding heat exchanger into a corresponding process medium temperature prediction model to obtain a cooling water supply and return water differential pressure set value and a process medium temperature predicted value;
s5: the cooling water supply pressure detection value and the cooling water return pressure detection value are subjected to difference to obtain a cooling water supply and return pressure difference detection value; and the difference between the set value of the cooling water supply and return pressure difference and the detected value of the cooling water supply and return pressure difference is used for obtaining a cooling water supply and return pressure difference deviation value, and the cooling water supply and return pressure difference deviation value is sent to the water supply and return pressure difference inner ring PID controller to adjust the opening of the upper tower valve, so that the outlet water flow of the water supply pump unit is changed. In the present embodiment, the feed pump is a centrifugal pump, and the centrifugal pump is connected to the frequency converter. In step S5, the supply and return water pressure difference inner ring PID controller adjusts the opening of the upper tower valve, so that the specific control content of changing the outlet water flow of the water supply pump set is as follows:
if the deviation value P of the cooling supply and return water differential pressure is greater than 0, the supply and return water differential pressure inner ring PID controller sends an upper tower valve amplification control signal for controlling the upper tower valve to increase the valve opening; the water supply and return pressure difference inner ring PID controller also sends a water supply pump unit outlet water flow reduction control signal for reducing the water pump unit outlet water flow; if the cooling water supply and return pressure difference deviation value P is equal to 0, the water supply and return pressure difference inner ring PID controller keeps the original control state; if the deviation value P of the cooling supply and return water pressure difference is less than 0, the supply and return water pressure difference inner ring PID controller sends a control signal for adjusting the upper tower valve to be small, and the control signal is used for controlling the upper tower valve to reduce the opening of the valve; and the supply and return water pressure difference inner ring PID controller also sends a water supply pump unit outlet water flow increasing control signal for increasing the water pump unit outlet water flow. As can be seen from the first table, the heat exchangers of the circulating cooling water system of the chemical plant A are listed, wherein DSC computer screen temperature display points are arranged in the first table after heat exchange of the process medium.
According to the requirements of the heat exchangers listed in the table, the design indexes and the control ranges of the process media in each heat exchanger are different, so that the outlet temperature of each process medium is detected one by one to realize the overall control of the temperature of the process media in order to ensure that all the process media are maintained in the respective normal production temperature range. The heat exchangers on the industrial production site are huge in quantity, the process media are various in types, and the difference value between the temperature detection value of each process media and the corresponding temperature set value is calculated on line to form a real-time updated process media temperature deviation sequence. If the method of calculating the cooling demand of each process medium and then determining the adjustment amount of the upper tower valve is adopted, the control process is complex and has large redundancy, the adjustment delay of the cooling water flow due to large calculation amount can be caused, and the control energy-saving effect is influenced.
Therefore, the invention provides the most unfavorable point for selecting and controlling the temperature of the process medium in real time, wherein the most unfavorable point of the temperature of the process medium refers to the process medium corresponding to the minimum value in the difference sequence of the detected temperature and the set temperature of each process medium, namely the process medium with the highest possibility of exceeding the set temperature range in each heat exchanger is controlled in real time, and the temperature of the whole process medium is equivalently controlled. With the continuous change of the temperature of the process medium along with the proceeding of industrial production, the temperature deviation and the deviation change rate of the process medium at the most unfavorable point are updated in real time, so that the temperature deviation and the deviation change rate of the process medium at the most unfavorable point are updated in real time as the input of the process medium temperature controller.
As can be seen from fig. 4, the specific steps of finding the minimum value in the sequence of the process medium temperature deviation values as the process medium temperature least favorable point by using the process medium temperature least favorable point selector in step S3 are as follows:
s31: initializing, setting M 'technological medium temperature deviation values to form a difference group, totaling M' technological medium temperature deviation values, and enabling W to bek＝M’；k＝1；
S32: let Wk+1W is equal to Wk + Xk+1Can be divided by M' evenly, X is a gap with large filling difference; and X is equal to 0-M' -1; s33: calculating Wk+2＝Wk+1/M’
S34: from Wk+2In the group, the minimum value is found out from the M' process medium temperature deviation values of each group by adopting a cross comparison method to obtain Wk+2A process medium temperature deviation value;
s35: judgment of Wk+2Whether it is equal to 1; if so, taking the process medium temperature deviation value as the least unfavorable point of the process medium temperature; otherwise, let k be k + 2; return is made to step S32.
Further, as can be seen from fig. 7, in step S4, the step of establishing the deep learning neural network of any heat exchanger is: s411: historical data generated by the operation of the circulating cooling water system in X sampling periods are used as training data of a process medium temperature prediction model in the heat exchanger after the obtained historical data are screened according to screening conditions; wherein X is a positive integer, and in this embodiment, X is equal to 1000.
S412: taking the process medium temperature deviation, the process medium temperature deviation change rate, the cooling water supply temperature detection value and the cooling water supply and return water pressure difference detection value in the historical data of the heat exchanger I as input data according to the determined characteristic variables from the historical data, carrying out data normalization processing to obtain a normalized data set, and dividing the normalized data set into a training sample set and a test sample set;
s413: performing greedy unsupervised pre-training on the training sample set layer by layer based on a stacked automatic encoder to obtain a weight matrix W of an input layer and a hidden layer initialized by a deep learning neural network and a threshold matrix B of the input layer and the hidden layer;
s414: and (3) fine adjustment of parameters: finely adjusting a weight matrix W of an input layer and a hidden layer initialized by a deep learning neural network and a threshold matrix B of the input layer and the hidden layer until the iteration times reach the maximum value of the iteration times to obtain an initial process medium temperature prediction model based on a stacking automatic encoder;
s415: and (5) evaluating the initial process medium temperature prediction model obtained in the step (S414) by using the test sample data set to obtain a process medium temperature prediction model based on the stacking automatic encoder.
As can be seen from fig. 7, the characteristic variables include a process medium temperature deviation value ei and a process medium temperature deviation change rate Δ e in any one of the heat exchangersiA real-time temperature detection value Tci of the process medium, a cooling water supply temperature detection value Tgs and a cooling water supply and return pressure difference detection value P in the circulating cooling water systemΔj。
In this embodiment, the screening condition is historical data that the process medium temperature is within a safe and energy-saving temperature value interval, and the safe and energy-saving temperature value interval is within a process medium temperature threshold value interval.
In this embodiment, the difference between the safe and energy-saving temperature value interval and the process medium temperature threshold value interval is 4 ℃.
In step S413, when greedy unsupervised pre-training is performed on the training sample set layer by layer, the training sample set is divided into P groups of small batch training samples, training is performed in sequence, a Dropout technique is adopted, a part of neurons are randomly selected to suspend working, iteration is performed in sequence, training is performed layer by layer, and weight matrix W, input layer threshold matrix B and hidden layer threshold matrix B of the input layer and hidden layer initialized by the neural network are deeply learned. In step S414, when the parameters are fine-tuned, a top-down small-batch RMSProp optimization method is used to perform fine-tuning on the weight matrix W of the input layer and the hidden layer initialized by the deep learning neural network, and the threshold matrix B of the input layer and the hidden layer.
The pre-control algorithm is directly oriented to field measurement data of an industrial circulating cooling water system, and nonlinear mapping relations among the measurement data of each heat exchanger, including temperature deviation and temperature deviation change rate of process media in each heat exchanger, a cooling water supply temperature detection value, a cooling water supply and return water differential pressure set value and control quantity prediction quantity, namely the cooling water supply and return water differential pressure set value are respectively established through a deep-layer framework. A two-stage off-line training learning method of pre-training-parameter fine tuning is adopted, and a Dropout technology and a RMSPROP technology are introduced to optimize the process medium temperature model parameters in each heat exchanger. The trained model can extract high-order characteristics beneficial to the process medium temperature prediction control effect by means of a hidden mode of deep structure mining data. In addition, the method can improve the generalization capability of the model through unsupervised training of a large number of unlabeled samples.
The number of hidden layers of the deep learning neural network trained by the process medium in the constructed SAE heat exchanger and the number of neurons of each hidden layer have certain influence on the evaluation precision and the off-line training time. Taking the field detection data of the circulating cooling water system as sample input data, and setting the number of neurons in the hidden layer by layer: the optimal number of the neurons of the hidden layer at the layer 1 is determined and fixed, then a layer is added to determine the optimal number of the neurons of the hidden layer at the layer 2, and the like is repeated until the average percentage error (MAPE) is not increased any more.
In order to make the technical solution of the present invention clearer, the principle of the stack automatic encoder used in the present invention is explained. SAE is a typical deep learning neural network, and its basic constituent unit is an Automatic Encoder (AE), and its network structure is shown in fig. 8. An Automatic Encoder (AE) network structure is shown in fig. 8, and is composed of an encoder and a decoder: the input vector is mapped to a feature vector in the hidden layer by the encoder, and then the feature phasor is reconstructed to the original input vector by the decoder.
When given an input sample set X ═ XiI is more than or equal to 1 and less than or equal to N, wherein N is the total number of samples, and x isiThe ith training sample in the sample set has dimension n. Let H ═ HiI is more than or equal to 1 and less than or equal to N is a hidden layer characteristic vector set, hiThe feature vector corresponding to the ith sample has a dimension M', and the coding relationship between X and H is:
H＝sf(WX + B) wherein: w is a weight matrix of the input layer and the hidden layer; b is an input layer and hidden layer threshold matrix; sf is the neuron activation function of the encoder, and a sigM' oid function is usually adopted, which has good feature identification:
sf(z)＝1/(1+exp(-z))
in the formula: z is the input vector.
The decoder is the inverse operation of the encoder, and takes the characteristic vector of the hidden layer as the input vector
In order to output the set of vectors,
and an output vector corresponding to the ith sample is obtained, and the dimension is n, so that the expression of the decoder is as follows:
in the formula: w' is a weight matrix of the hidden layer and the output layer; b' is a threshold matrix of a hidden layer and an output layer; sg is the neuron activation function of the decoder.
The automatic encoder achieves the purpose of feature extraction by minimizing the reconstruction error between the output vector and the input vector, and the formula of the reconstruction error is as follows:
and continuously adjusting the network weight and the threshold by using a gradient descent algorithm to reduce the reconstruction error, wherein the formula is as follows:
in the formula: l is the learning rate;
to represent
Calculating the deviation of the weight W;
to represent
The bias is calculated for the threshold B.
The SAE is a deep learning neural network model composed of stacked AE, and the output of the lower AE is used as the input of the upper AE. Gradual abstraction of features is achieved by stacking of AEs, eventually leading to more compact, useful features, as shown in fig. 9 for a stacked self-encoder architecture. WnIs a weight matrix of the n-1 st hidden layer and the n-th hidden layer, BnThe threshold matrix is the n-1 st hidden layer and the nth hidden layer. The training process comprises two steps of greedy layer-by-layer unsupervised pre-training and supervised fine tuning. Greedy non-supervision pre-training layer by layer obtains the initial weight and the threshold value of the network through layer by layer training, the input of the bottom layer AE is original data, and the output data of the hidden layer is used as the input data of the upper layer AE.
After the layered pre-training is completed, the hidden layers are stacked, and the relationship between the input data and the output data is expressed as follows:
in the formula: f is an activation function, xiW, B are respectively the network initialization weight and threshold value obtained by pre-training layer by layer,
is the predicted value of the ith sample. Constructing an error loss function of the actual value and the predicted value, wherein the formula is as follows:
in the formula: n is the total number of samples, yiIs the actual value of the ith sample. The whole network weight and the threshold value are finely adjusted through the back propagation from top to bottom, and the error between the predicted value and the actual value is reduced.
The SAE can effectively extract high-order features of data through greedy unsupervised pre-training layer by layer, better approaches complex functions, reduces parameter optimization space, can quickly obtain network parameters, and improves deep feature learning capability of a neural network.
Compared with a shallow machine learning algorithm, when the deep network processes high-dimensional data, the overfitting problem is more easily generated due to the complex network structure of the deep network, so that the generalization capability of the model is limited. Dropout is a mainstream anti-overfitting technique, and its basic idea is: during model training, a part of nodes are randomly selected to be out of operation, the nodes store the weight of the last iteration, and the output is set to be 0. And the selected nodes restore the weight values reserved before in the process of the next iteration, and part of the nodes are randomly selected again to repeat the process. The network structure will change to some extent in each iteration process, and by adopting the Dropout technology, part of neurons selected randomly do not work temporarily, as shown in fig. 10, the coaction among specific nodes is reduced, the dependence of the network output on the states of the specific nodes is reduced, and overfitting is prevented.
In the parameter fine-tuning stage, a top-down small-batch RMSProp optimization method is adopted to fine-tune the weight and the threshold of the network until the iteration number reaches a set value, and the method specifically comprises the following steps:
in step 1, a global learning rate l and a decay rate ρ are set, and an initialization cumulative variable r1 is 0 and r2 is 0.
and 3, calculating the accumulated square gradient as shown in the formula (9):
in the formula: as an element-by-element product symbol.
And 4, respectively updating the weight and the threshold parameters:
and 5, stopping operation when the iteration times meet the requirements, and otherwise, returning to the step 2 to continue to execute the calculation.
The process medium temperature prediction control algorithm can effectively represent the process medium temperature deviation and the deviation change rate, and the complex function between the industrial circulating cooling water supply temperature and the cooling water supply pressure set value, quickly and accurately carry out prediction control on the process medium temperature on the industrial production site, and meanwhile, the algorithm has good generalization capability and strong prediction adaptability to process media with different temperature change characteristics. Compared with manual experience calculation and adjustment, the accuracy of process medium temperature control is greatly improved, the calculation time overhead is saved, and industrial production field management personnel can master the process medium temperature change trend in each heat exchanger in real time; the process medium temperature prediction control algorithm comprises the steps of off-line training and on-line application of a process medium temperature prediction model in a heat exchanger based on a stack automatic encoder to give a cooling water supply and return water differential pressure set value of a circulating cooling water system: the stack automatic encoder belongs to a deep learning neural network, and compared with the traditional shallow neural network, the stack automatic encoder effectively solves a series of problems caused by the random initialization of the parameters of the traditional neural network, can effectively mine the implicit relation of each data, and greatly improves the accuracy of the prediction control of the temperature of the process medium in the industrial production field; compared with the traditional stack automatic encoder method, the improved stack automatic encoder algorithm for parameter fine adjustment by utilizing the RMSProp optimization method has better generalization capability and can improve the prediction control precision of the industrial production field process medium temperature. The method is applied to the energy-saving control of the circulating cooling water system in the industrial production field, can quickly and accurately give the set value of the cooling water supply and return water pressure difference, is beneficial to industrial production field management personnel to master the temperature change trend of the process medium in each heat exchanger in real time, and most importantly, reduces the energy consumption of the circulating cooling water system.
It should be noted that the above description is not intended to limit the present invention, and the present invention is not limited to the above examples, and those skilled in the art may make variations, modifications, additions or substitutions within the spirit and scope of the present invention.
Claims (7)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910251616.9A CN109917656B (en) | 2019-03-29 | 2019-03-29 | Circulating cooling water minimum pressure difference energy-saving control system and method based on process medium multi-temperature target |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910251616.9A CN109917656B (en) | 2019-03-29 | 2019-03-29 | Circulating cooling water minimum pressure difference energy-saving control system and method based on process medium multi-temperature target |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109917656A CN109917656A (en) | 2019-06-21 |
| CN109917656B true CN109917656B (en) | 2022-03-01 |
Family
ID=66967827
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910251616.9A Expired - Fee Related CN109917656B (en) | 2019-03-29 | 2019-03-29 | Circulating cooling water minimum pressure difference energy-saving control system and method based on process medium multi-temperature target |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109917656B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112434298B (en) * | 2021-01-26 | 2021-07-06 | 浙江大学 | Network threat detection system based on self-encoder integration |
| CN114139405B (en) * | 2022-01-18 | 2025-03-28 | 光大环保技术装备（常州）有限公司 | Method, system and device for controlling the temperature of cooling medium in circulating cooling system |
| CN116301117B (en) * | 2023-05-24 | 2023-10-20 | 荣耀终端有限公司 | Temperature control method and electronic equipment |
| CN117234071B (en) * | 2023-11-16 | 2024-01-26 | 中控技术股份有限公司 | Circulating water system energy-saving optimization method and device based on temperature and pressure control |
Citations (23)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5691896A (en) * | 1995-08-15 | 1997-11-25 | Rosemount, Inc. | Field based process control system with auto-tuning |
| US5701554A (en) * | 1994-06-10 | 1997-12-23 | Seiko Epson Corporation | Fixing apparatus having controller for setting a target temperature and for estimating the amount of heat transferred to a pressure roller |
| JP2002147914A (en) * | 2000-11-13 | 2002-05-22 | Sanyo Engineering Kk | Low temperature cool-storage medium |
| US6619066B1 (en) * | 1999-02-24 | 2003-09-16 | Hachiyo Engineering Co., Ltd. | Heat pump system of combination of ammonia cycle carbon dioxide cycle |
| JP2008232467A (en) * | 2007-03-16 | 2008-10-02 | Toshiba Corp | Air-conditioning control system |
| CN101382773A (en) * | 2008-10-17 | 2009-03-11 | 东华大学 | Anti-time-delay pseudo-proportional-integral control system and method |
| CN101893296A (en) * | 2009-05-22 | 2010-11-24 | 富士电机系统株式会社 | Precise temperature adjustment system and its control device |
| CN102147146A (en) * | 2011-04-22 | 2011-08-10 | 黄真银 | Digital integrated intelligent control system of central air conditioner and adjusting method thereof |
| CN202133298U (en) * | 2011-05-03 | 2012-02-01 | 戴军 | Energy-saving control unit for condenser system of electric power plant |
| WO2012049702A1 (en) * | 2010-10-12 | 2012-04-19 | 三菱電機株式会社 | Air conditioner |
| CN102980272A (en) * | 2012-12-08 | 2013-03-20 | 珠海派诺科技股份有限公司 | Air conditioner system energy saving optimization method based on load prediction |
| CN104566765A (en) * | 2013-10-16 | 2015-04-29 | 嘉日国际集团控股有限公司 | Overall energy saving control method of central air conditioner |
| WO2015136172A1 (en) * | 2014-03-13 | 2015-09-17 | Peugeot Citroen Automobiles Sa | Method for predicting the thermal status of a starter |
| CN205242674U (en) * | 2015-11-26 | 2016-05-18 | 天津大学 | Envelope adjustable laboratory that loads |
| CN106152250A (en) * | 2016-08-16 | 2016-11-23 | 河南柴油机重工有限责任公司 | Blood circulation that a kind of heat exchange station automatically controls and circulation controlling means |
| CN106440549A (en) * | 2016-05-30 | 2017-02-22 | 靳凯 | Control system for increasing comprehensive energy efficiency ratio of central air conditioning system |
| CN107192097A (en) * | 2017-06-06 | 2017-09-22 | 珠海格力电器股份有限公司 | Control method and device for air-conditioning chilled water system |
| CN107976088A (en) * | 2016-10-24 | 2018-05-01 | 李宁 | A kind of closed cycle cooling process and device |
| JP6324560B1 (en) * | 2017-02-22 | 2018-05-16 | クラフトワーク株式会社 | Ice making apparatus and ice production method |
| CN108507245A (en) * | 2018-04-04 | 2018-09-07 | 重庆大学 | A kind of the full-automatic circulating cooling control system and control method of nesting closed loop |
| CN108643278A (en) * | 2018-06-22 | 2018-10-12 | 安徽机电职业技术学院 | A kind of energy-saving Township water supply system and its water-supply control |
| CN108990383A (en) * | 2018-08-15 | 2018-12-11 | 北京建筑大学 | A kind of data center's air-conditioning system forecast Control Algorithm |
| CN109460861A (en) * | 2018-10-18 | 2019-03-12 | 扬州大学 | A kind of chemical circulation cooling water system water pump fan combination frequency conversion optimization operating scheme determines method |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP5040213B2 (en) * | 2006-08-15 | 2012-10-03 | 東京エレクトロン株式会社 | Heat treatment apparatus, heat treatment method and storage medium |
| US20090037142A1 (en) * | 2007-07-30 | 2009-02-05 | Lawrence Kates | Portable method and apparatus for monitoring refrigerant-cycle systems |
-
2019
- 2019-03-29 CN CN201910251616.9A patent/CN109917656B/en not_active Expired - Fee Related
Patent Citations (23)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5701554A (en) * | 1994-06-10 | 1997-12-23 | Seiko Epson Corporation | Fixing apparatus having controller for setting a target temperature and for estimating the amount of heat transferred to a pressure roller |
| US5691896A (en) * | 1995-08-15 | 1997-11-25 | Rosemount, Inc. | Field based process control system with auto-tuning |
| US6619066B1 (en) * | 1999-02-24 | 2003-09-16 | Hachiyo Engineering Co., Ltd. | Heat pump system of combination of ammonia cycle carbon dioxide cycle |
| JP2002147914A (en) * | 2000-11-13 | 2002-05-22 | Sanyo Engineering Kk | Low temperature cool-storage medium |
| JP2008232467A (en) * | 2007-03-16 | 2008-10-02 | Toshiba Corp | Air-conditioning control system |
| CN101382773A (en) * | 2008-10-17 | 2009-03-11 | 东华大学 | Anti-time-delay pseudo-proportional-integral control system and method |
| CN101893296A (en) * | 2009-05-22 | 2010-11-24 | 富士电机系统株式会社 | Precise temperature adjustment system and its control device |
| WO2012049702A1 (en) * | 2010-10-12 | 2012-04-19 | 三菱電機株式会社 | Air conditioner |
| CN102147146A (en) * | 2011-04-22 | 2011-08-10 | 黄真银 | Digital integrated intelligent control system of central air conditioner and adjusting method thereof |
| CN202133298U (en) * | 2011-05-03 | 2012-02-01 | 戴军 | Energy-saving control unit for condenser system of electric power plant |
| CN102980272A (en) * | 2012-12-08 | 2013-03-20 | 珠海派诺科技股份有限公司 | Air conditioner system energy saving optimization method based on load prediction |
| CN104566765A (en) * | 2013-10-16 | 2015-04-29 | 嘉日国际集团控股有限公司 | Overall energy saving control method of central air conditioner |
| WO2015136172A1 (en) * | 2014-03-13 | 2015-09-17 | Peugeot Citroen Automobiles Sa | Method for predicting the thermal status of a starter |
| CN205242674U (en) * | 2015-11-26 | 2016-05-18 | 天津大学 | Envelope adjustable laboratory that loads |
| CN106440549A (en) * | 2016-05-30 | 2017-02-22 | 靳凯 | Control system for increasing comprehensive energy efficiency ratio of central air conditioning system |
| CN106152250A (en) * | 2016-08-16 | 2016-11-23 | 河南柴油机重工有限责任公司 | Blood circulation that a kind of heat exchange station automatically controls and circulation controlling means |
| CN107976088A (en) * | 2016-10-24 | 2018-05-01 | 李宁 | A kind of closed cycle cooling process and device |
| JP6324560B1 (en) * | 2017-02-22 | 2018-05-16 | クラフトワーク株式会社 | Ice making apparatus and ice production method |
| CN107192097A (en) * | 2017-06-06 | 2017-09-22 | 珠海格力电器股份有限公司 | Control method and device for air-conditioning chilled water system |
| CN108507245A (en) * | 2018-04-04 | 2018-09-07 | 重庆大学 | A kind of the full-automatic circulating cooling control system and control method of nesting closed loop |
| CN108643278A (en) * | 2018-06-22 | 2018-10-12 | 安徽机电职业技术学院 | A kind of energy-saving Township water supply system and its water-supply control |
| CN108990383A (en) * | 2018-08-15 | 2018-12-11 | 北京建筑大学 | A kind of data center's air-conditioning system forecast Control Algorithm |
| CN109460861A (en) * | 2018-10-18 | 2019-03-12 | 扬州大学 | A kind of chemical circulation cooling water system water pump fan combination frequency conversion optimization operating scheme determines method |
Non-Patent Citations (6)
| Title |
|---|
| Design of LT-MED Seawater Desalination Temperature Control System Based on Dynamic Matrix Predictive Fuzzy PID Control Algorithm;Hongli Liu，等;《Proceedings of 2018 IEEE International Conference on Mechatronics and Automation》;20180808;第293-297页 * |
| Evaluation_of_Differential_Pressure_Setpoint_of_Chilled_Water_Pumps_in_Clean_Room_HVAC_Systems_for_Energy_Savings_in_High-Tech_Industries;Chun-Lien Su，等;《IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS》;20130630;第49卷(第3期);第1015-1022页 * |
| The_effects_of_temperature_gradient_on_transient_behavior_of_low-range_differential_air_pressure_calibration_setup;Muhammad Nazar Ul Islam，等;《2015 IEEE International Conference on Industrial Technology (ICIT)》;20151231;第1488-1494页 * |
| 中央空调冷却水、冷冻水系统的变频节能分析;曹华，等;《变频调速与应用》;20141231;第41卷(第4期);第62-66页 * |
| 中央空调房间温度智能PID控制的仿真研究;唐鑫，等;《计算机仿真》;20100531;第27卷(第5期);第140-144页 * |
| 基于AM3517的换热站节能控制系统研究;孔令坤;《中国优秀硕士学位论文全文数据库工程科技Ⅱ辑》;20141215(第（2014年）12期);第C038-326页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109917656A (en) | 2019-06-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109917656B (en) | Circulating cooling water minimum pressure difference energy-saving control system and method based on process medium multi-temperature target | |
| CN111256294B (en) | Model prediction-based optimization control method for combined operation of water chilling unit | |
| Xu et al. | Mid-term prediction of electrical energy consumption for crude oil pipelines using a hybrid algorithm of support vector machine and genetic algorithm | |
| CN110288164B (en) | Predictive control method for building air-conditioning refrigeration station system | |
| CN110008575B (en) | Multi-parameter predictive control algorithm for switching process medium multi-temperature target set values of circulating cooling water system | |
| CN110392515A (en) | A kind of Cooling and Heat Source equipment room energy-conserving control method and system based on historical data | |
| Yabanova et al. | Development of ANN model for geothermal district heating system and a novel PID-based control strategy | |
| CN109945402A (en) | A kind of central air-conditioning water system energy saving method | |
| CN112901449B (en) | Air compressor system energy consumption optimization method based on machine learning | |
| CN113011662A (en) | System and method for integrated combined optimized scheduling of network and river of regional enterprise plant | |
| CN102183972B (en) | Method for controlling water level of reservoir of urban drainage system | |
| CN1869532A (en) | Automatic control method for central cold supply system | |
| CN110567101A (en) | Water chiller high-energy-efficiency control method based on support vector machine model | |
| CN108921352A (en) | A kind of wet metallurgical leaching process optimization method with bounded-but-unknown uncertainty | |
| CN113959071A (en) | Machine learning assistance-based operation control optimization method for centralized water chilling unit air conditioning system | |
| CN106545968A (en) | A kind of two grades of chilled water energy-saving control systems of central air-conditioning open type | |
| CN112949186B (en) | Method for predicting wax precipitation point temperature of wax-containing crude oil based on SSA-LSSVM model | |
| CN105068582B (en) | DCS-based high-level water collection cooling tower water level maintaining control method | |
| Deng et al. | Optimal control of chilled water system with ensemble learning and cloud edge terminal implementation | |
| CN112431772A (en) | Optimal dispatching operation method for flood prevention and drainage pump station group in polder region | |
| CN111814861B (en) | Online cooling control method based on double self-learning models | |
| WO2022233101A1 (en) | Intelligent optimization control device for low-temperature thermal system | |
| CN115577828A (en) | A data-driven modeling and optimization method for group control of air-conditioning and refrigeration station systems | |
| CN113642256B (en) | Practical gate opening and closing optimization method based on theme database | |
| CN113123959B (en) | Intelligent water quantity scheduling system for multi-stage pumping station |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20220301 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |