KR101875489B1 - Method and system for automatic controlling of air conditioner by using an artificial intelligence - Google Patents

Abstract

Disclosed are an artificial intelligence (AI) system and an application thereof, wherein the AI system replicates functions of the human brain, such as recognition and judgment, by using machine learning algorithms such as deep learning. The objective of the present invention is to provide an apparatus and a method for automatically controlling an air conditioning system by predicting an air conditioning load. Moreover, the apparatus can predict outdoor air states such as temperature, humidity, and enthalpy in an air conditioning area, predict air conditioning loads, temperature, humidity, carbon dioxide concentration, air pollution levels for each hour of the next day by using the outdoor air states and building features, and automatically control the air conditioning system based on the predicted value.

Description

TECHNICAL FIELD [0001] The present invention relates to an apparatus and method for automatically controlling a cooling system,

The present disclosure relates to an artificial intelligence (AI) system and its application that simulate functions such as recognition and judgment of a human brain using a machine learning algorithm such as deep learning. The present invention relates to an optimal control method of an air conditioning system, and more particularly, to a method and system for optimizing an air conditioning system by predicting a heating / The present invention relates to an unmanned optimum control method for an air conditioning system that enables an air conditioning system to operate most efficiently and economically without an experienced driver, taking into consideration the performance of a device that varies greatly.

Recently, smart grid has become a hot topic. The original concept of the Smart Grid refers to a technology that optimizes energy efficiency by bi-directionally exchanging information between power suppliers and consumers by applying information technology to the existing unidirectional power grid, which was generated at the stage of power generation - transmission, distribution and sales. The goal is to efficiently operate the entire power system by connecting the power plant, transmission, distribution facilities and power consumers to information and communication networks and sharing information in both directions.

The ultimate goal of a smart grid implementation is to rationalize or optimize power consumption. As traditional fossil energy is depleted, it is about monitoring and controlling power usage. There are many ideas about how to implement a smart grid. As long as the ultimate goal of implementing smart grid is to optimize power consumption, the effect will be halved if power consumers, such as homes, offices, and factories, reduce power use and do not actively utilize natural energy. Therefore, in the present invention, a key element of the smart grid is seen as power saving.

Since the seasons with the greatest use of electricity are summer and winter, the use of electricity at this time should be reduced. The increasing use of electricity in summer and winter is due to the use of air conditioners and air conditioners that use electricity. Until now, it has been focused on increasing the efficiency of heating and cooling equipment, but it is more effective to reduce the need for cooling and heating by improving the surrounding environment in which these heating and cooling devices operate.

Explain winter as an example (summer is the opposite). The part that exchanges heat with the outside most in a home or an office is a window, and therefore, a blind (curtain) is installed so as to prevent heat from escaping. By blocking the windows with blinds, it is possible to prevent the heat from being lost through the windows, thereby keeping the room temperature as high as possible and reducing the heating burden.

However, when the weather is good even in winter, it is preferable to open the blinds. It is practically beneficial to open the blinds as the sunlight enters the room through the window and raises the room temperature. Conversely, it is advantageous to close the blinds when the weather is cloudy, snowy or sunny but the outside temperature is too low, and it is preferable to close the blinds in the evening after sunset.

However, since the weather fluctuates from time to time, it was necessary to judge whether or not to start the heating / cooling system or how to operate the system. Accordingly, until now, the person judged and operated the cooling / heating system manually. However, it is very cumbersome for a person to do, and it reduces the concentration of work in the office or factory, and there is no way to manage it when there is no person in the office or factory.

Therefore, users' needs to manage the air conditioning system optimally according to the temperature, humidity, air pollution degree, etc. have increased.

(Prior Art 1) Japanese Patent Registration No. 3783859 (registered on Mar. 24, 2006) (Prior Art 2) Korean Patent Laid-Open Publication No. 10-2015-0027500 (published on September 18, 2015)

According to an exemplary embodiment, an intelligent learning model is used to provide information (e.g., outside temperature, daylight information, dust information, and weather information) related to weather conditions from an external server over the Internet and analyze the information So as to keep the room temperature at a good level.

According to one embodiment, it is possible to provide a method of automatically controlling a cooling system through a cooling load prediction, and the method can provide a method of automatically controlling a cooling system by estimating the temperature of the next day, the cloudiness, the humidity, the solar radiation amount, Measuring the condition of the air conditioning target area including the temperature, humidity, CO ₂ concentration and fine dust concentration of the air conditioning object area at predetermined time intervals, calculating the temperature of the air conditioning object area and the weather data Determining a first control value by comparing the temperature comparison value with a predetermined standard temperature, comparing the humidity of the air-conditioning object region with the humidity data of the vapor-phase data, and comparing the humidity comparison value the calculation, and the humidity value compared with a preset search step of determining the second control value is compared to a standard humidity, CO ₂ concentration of the air conditioning target area Compare with pre-set standard CO ₂ concentration range to calculate the CO ₂ concentration comparison value, the dust amount in the step of determining a third control value according to the CO comparison value _{concentration,} fine dust concentration of the air-conditioning target area and weather data, Determining the fourth control value according to the indoor pollution index, measuring the required calorific value and operation schedule of the cooling system connected to the coexistence target area, and connecting the air conditioning system to the cooling system The fifth control value is determined in consideration of the supply calorie of the supply system and the operation schedule, which depend on the energy charge, the operation condition, and the ambient condition depending on the time and season based on the cooling load based on the predicted supply calorie and cooling load of the supply system. A control for controlling the cooling system by combining the first control value, the second control value, the third control value, the fourth control value, and the fifth control value A step of controlling the operation of the cooling system in accordance with the control signal, a step of measuring a time amount of electric power according to the operation of the cooling system, and a mode of operation of the cooling system by time of day, And automatically controlling the operation of the cooling system according to the operation mode.

According to the embodiment, it is possible to control the cooling system efficiently in accordance with the spread of the various cooling systems.

According to an embodiment, automatic control that does not depend on the experience of the driver becomes possible, resulting in economical saving effect of the energy use cost.

According to one embodiment, the cooling load of the building can be predicted by the next day time zone by predicting the outside air conditions such as the temperature, humidity, and irradiation amount by the next day time zone. Also, the cooling system can be optimally controlled automatically in consideration of the predicted cooling load, energy intensity, and characteristics of the apparatus.

According to an embodiment, the power peak can be controlled by predicting and dispersing the power consumption amount, and a plan for power use can be efficiently established.

According to one embodiment, the cooling system can be controlled according to air purification of dust, dust, etc., and the cooling system can be monitored according to various environments.

According to one embodiment, an artificial intelligence learning model can be used to provide an optimal air conditioning system environment by utilizing vast amounts of data.

1 is a view for explaining an apparatus for automatically controlling a cooling system using a cooling load prediction according to an embodiment of the present invention.
2 is a functional block diagram of an apparatus and a connection system for automatically controlling a cooling system using a cooling load prediction according to an embodiment.
3 is a flow diagram of a method for automatically controlling a cooling system, in accordance with one embodiment.
FIG. 4 is a graph showing a change in heat transfer coefficient according to a change in total heat transfer coefficient of a window according to an embodiment.
5 is a graph showing a change in heat transfer coefficient according to a change in total heat transfer coefficient of a wall according to an embodiment.
FIG. 6 is a diagram illustrating data of an input feature map and filters 110-1 to 110-n provided as inputs in a deep learning deep run operation using an artificial neural network.

The following description and accompanying drawings are for understanding the operation according to the present invention, and parts that can be easily implemented by those skilled in the art can be omitted.

Further, the specification and drawings are not provided for the purpose of limiting the present invention, and the scope of the present invention should be determined by the claims. The terms used in this specification may be used to describe various elements in the technical idea of the present invention so as to most suitably represent the present invention, but the terms are defined by terms . Terms are used only for the purpose of distinguishing one component from another.

The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of the claims. The singular expressions include plural expressions unless the context clearly dictates otherwise. In the present application, the terms "comprises" or "having" and the like are used to specify that there is a feature, a number, a step, an operation, an element, a component or a combination thereof described in the specification, But do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, or combinations thereof.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings. Referring to the accompanying drawings, the same or corresponding elements are denoted by the same reference numerals, and a duplicate description thereof will be omitted.

1 is a view for explaining an apparatus for automatically controlling a cooling system using a cooling load prediction according to an embodiment of the present invention.

According to an exemplary embodiment, an apparatus 100 for automatically controlling a cooling system using a cooling load prediction can perform an efficient cooling system control at a minimum operating cost without depending on the driver's experience. In addition, the apparatus 100 for automatically controlling the cooling system can control the cooling system by predicting the predicted cooling load, the energy intensity, the characteristics of the apparatus, the weather forecast, the current state of the air conditioning area, and the power state.

The apparatus 100 for automatically controlling the cooling system can automatically control the facilities of the building by monitoring facilities, power, lighting, access, environment, and disaster prevention of the building. Hereinafter, it will be described in detail using another drawing.

Artificial intelligence (AI) system is a computer system that implements human-level intelligence. Unlike existing Rule-based smart systems, AI is a system in which machines learn, judge and become smart. Artificial intelligence systems are increasingly recognized and improving their understanding of user preferences as they are used, and existing rule-based smart systems are gradually being replaced by deep-run-based artificial intelligence systems.

Artificial intelligence technology consists of element technologies that utilize deep learning and machine learning.

Machine learning is an algorithm technology that classifies / learns the characteristics of input data by itself. Element technology is a technology that simulates functions such as recognition and judgment of human brain using machine learning algorithms such as deep learning. Understanding, reasoning / prediction, knowledge representation, and motion control.

The various fields in which artificial intelligence technology is applied are as follows. Linguistic understanding is a technology for recognizing, applying, and processing human language / characters, including natural language processing, machine translation, dialogue system, query response, speech recognition / synthesis, and the like. Visual understanding is a technology for recognizing and processing objects as human vision, including object recognition, object tracking, image search, human recognition, scene understanding, spatial understanding, and image enhancement. Inference prediction is a technique for judging and logically inferring and predicting information, including knowledge / probability based reasoning, optimization prediction, preference base planning, and recommendation. Knowledge representation is a technology for automating human experience information into knowledge data, including knowledge building (data generation / classification) and knowledge management (data utilization). The motion control is a technique for controlling the autonomous travel of the vehicle and the motion of the robot, and includes motion control (navigation, collision, traveling), operation control (behavior control), and the like.

With the development of communication technology, various functions for exchanging with people have been provided to the user terminal, and contacts for communication such as messenger, e-mail, and social network service (SNS) have increased as well as contacts for voice calls and texts.

It is difficult to memorize a large number of contacts according to various kinds of communication channels. For convenience of users, each communication channel program provides a function of recommending contacts in various ways. However, most of the existing contact recommendation methods have been based on simple recommendation of contacts according to the number of calls and the duration of calls. Accordingly, the necessity of subdivided contact recommendation according to the user context has increased.

 2 is a functional block diagram of an apparatus and a connection system for automatically controlling a cooling system using a cooling load prediction according to an embodiment.

According to one embodiment, the present invention may provide an apparatus 100 for automatically controlling a cooling system through a cooling load prediction, the apparatus comprising: a network 250 for receiving data from a weather station server 200, cloudiness, air conditioning target containing humidity, solar radiation, temperature, humidity, CO ₂ concentration and particulate matter concentration of the dust amount, bunjinryang and atmospheric fine weather data receiver 110 for receiving weather data including the dust concentration, the air-conditioning target area A sensing unit 120 for measuring the condition of the area at predetermined time intervals, a temperature comparing unit 120 for comparing the temperature of the air-conditioning target area with the temperature data of the gas-phase data to calculate a temperature comparison value, A first controller 130 for determining a control value, a controller 130 for comparing the humidity of the air-conditioning object area with the humidity data of the weather data to calculate a humidity comparison value, In the second control unit (140),, CO2 concentration comparison value, and calculates the CO ₂ concentration comparison value by comparing the standard CO ₂ concentration range previously set, and the CO ₂ concentration of the air-conditioning target area, determining a second control value by comparing humidity The third control unit 150 determines a third control value. The third control unit 150 compares the fine dust concentration of the air-conditioning target area with the yellow amount, the dust amount, and the atmospheric fine dust concentration of the meteorological data to calculate the indoor pollution index. The fourth control unit 160 determines a fourth control value. The fourth control unit 160 estimates the amount of heat and the cooling load of the supply system connected to the cooling system by measuring the required calorific value and the operation schedule of the cooling system connected to the coexistence target area, A fifth control for determining the fifth control value in consideration of the supply calorie of the supply system and the operation schedule depending on the other energy charges, (170) and a first control value, a second control value, a third control value, a fourth control value, and a fifth control value to generate a control signal for controlling the cooling system, 300), measures the amount of electric power by time according to the operation of the cooling system, determines the operation mode of the cooling system by time zone on the basis of the electric energy amount per hour, automatically operates the cooling system according to the operation mode by time zone And a central control unit (180) for controlling the control unit.

In one embodiment, the cooling system 300 may include a device for controlling the cold temperature, the air purifying control, and the oxygen concentration control of the air conditioning area, such as the air conditioner, the freezer, and the air conditioner. But is not limited thereto.

The network 250 according to an exemplary embodiment may be a wireless LAN, a Wi-Fi, a 3G, a 4G, a WiMAX, a WiBro, a wireless mesh network, a UWB Speed Downlink Packet Access). ≪ / RTI >

The meteorological station server 200 can measure meteorological data including temperature, cloudiness, humidity, solar radiation amount, dust amount, dust amount and atmospheric fine dust concentration measured according to a predetermined time every day. The weather station server 200 may provide the meteorological data measured according to a predetermined time period to the apparatus 100 for automatically controlling the cooling system. The weather station server 200 can provide weather data of a specific area or a specific time zone to the apparatus 100 for automatically controlling the cooling system.

3 is a flow diagram of a method for automatically controlling a cooling system, in accordance with one embodiment.

(AI) system comprising a processor for performing a convolution operation of data of an input feature map and a weight value of a filter using a systolic array by means of an automatic start Can be provided.

In step S301, the apparatus 100 for automatically controlling the cooling system according to an embodiment obtains weather data including the temperature, the cloudiness, the humidity, the solar radiation amount, the yellow dust amount, the dust amount and the atmospheric fine dust concentration of the next day from the weather station server .

In step S302, the apparatus 100 for automatically controlling the cooling system according to one embodiment measures the condition of the air-conditioning object area including the temperature, humidity, CO ₂ concentration, and fine dust concentration of the air- . The system that automatically controls the cooling system can measure the condition of the area to be air-conditioned by using duct temperature detector, duct temperature / humidity detector, real differential pressure indicator, indoor temperature / humidity detector, pipe temperature detector and CO2 detection sensor .

In step S303, the apparatus 100 for automatically controlling the cooling system according to an embodiment calculates the temperature comparison value by comparing the temperature of the air-conditioning object area with the temperature data of the vapor-phase data, The first control value can be obtained as an appropriate temperature of the air conditioning object region from the learned first learning model by adding information on cloudiness, humidity, solar radiation amount, dust amount, dust amount and atmospheric fine dust concentration.

In step S304, the apparatus 100 for automatically controlling the cooling system according to an embodiment compares the humidity of the air-conditioning object area with the humidity data of the gas-phase data, calculates the humidity comparison value, and compares the humidity comparison value and the preset standard humidity The second control value can be determined as the appropriate humidity of the air-conditioning object region from the learned second learning model by adding information about the temperature, the cloudiness, the solar radiation amount, the yellow dust amount, the dust amount and the atmospheric fine dust concentration.

When the temperature comparison value and the preset standard temperature ratio are larger than the humidity comparison value and the preset standard humidity ratio, the second control value is changed to 0, and the ratio of the temperature comparison value and the predetermined standard temperature Is smaller than the ratio of the humidity comparison value and the standard humidity set in advance, the first control value can be changed to zero.

* In step S305, one embodiment ExamplesExamples device 100 to automatically control the air-conditioning system according compares a preset standard CO ₂ concentration range of the CO ₂ concentration of the air-conditioning target area, and calculates the CO ₂ concentration comparison value, CO ₂ The third control value can be determined from the learned third learning model by adding information on temperature, cloudiness, humidity, solar radiation amount, dust amount, dust amount and atmospheric fine dust concentration as an appropriate CO ₂ concentration of the target area have.

In step S306, the apparatus 100 for automatically controlling the cooling system according to an embodiment compares the concentration of fine dust in the air-conditioning target area with the amount of yellow dust in the air-conditioning data, the amount of dust and the concentration of air fine dust, It is possible to determine a fourth control value as an appropriate indoor pollution index of the air conditioning object zone from the learned fourth learning model by adding information on temperature, cloudiness, humidity, solar radiation amount, dust amount, dust amount and atmospheric fine dust concentration.

In step S307, the apparatus 100 for automatically controlling the cooling system according to an embodiment measures the required calorific value and the operation schedule of the cooling system connected to the area to be air-conditioned, and predicts the supply heat quantity and the cooling load of the supply system connected to the cooling system The fifth learning model, which is learned in consideration of the supply calorie of the supply system and the operation schedule, which varies depending on the energy charge, the operation condition, and the outside air condition by time and season, based on the cooling load, Value can be determined.

In step S308, the apparatus 100 for automatically controlling the cooling system according to an embodiment synthesizes a first control value, a second control value, a third control value, a fourth control value, and a fifth control value, the data of the input feature map is systolic arrayed by changing the order of the data of the input feature map so that data having the same value is repeated a predetermined number of times according to a sequential clock using a mapper And the data of the mapped input feature map and the weight value of the filter are input to the systolic array operator to perform a convolution operation to cumulatively accumulate the output values of the convolution operation to output an output feature map feature map, so that a control signal for controlling the cooling system can be generated.

The step of generating a control signal by changing the input order of the input data so that the same two input data values of the input feature map are repeated and input to the systolic array operator, The control signal can be generated by repeating the process of changing the input order of the data of the input feature map so as to be repeatedly input to the systolic array operator by the same number of times as the size N. [

For example, the mapper may change the input order of the input data so that the same two input data values of the input feature map are repeatedly input to the systolic array operator.

The apparatus for controlling the cooling system may include a mapper, a systolic computing unit, and an adder.

It is possible to generate a control signal for reducing the amount of power consumed in performing the arithmetic operation and increasing the arithmetic operation speed through mapping which changes the order of data inputted to the systolic array operator.

An apparatus for controlling a cooling system according to the present invention includes an unmapper for acquiring position information on the systolic array of data identified as zero among data of the input feature map mapped by a mapper in a systolic array, Mapper control circuit for controlling the accumulator to cumulatively add output values for the respective regions of the input feature map based on position information of the obtained zero-value data, the adder-tree control circuit ).

The mapper inputs the data of the input feature map or the weight value of the filter so that the data having the same value is repeated to the systolic operator according to the sequential clock, A mapping can be performed. In one embodiment, the mapper may change the input sequence of data of the input feature map so that the data of the input feature map is repeatedly input to the multiplication addition operator in the systolic operator with the number of times equal to the size (N) of the filter . For example, if the size of the filter is 3 x 3, the mapper maps the input feature map to the systolic array so that the input data of x ₁₁ , x ₁₂ , and x ₁₃ are repeatedly input three times to the multiply adder . However, the present invention is not limited to this, and the mapper may map the input feature map such that the data of the input feature map is repeated twice and input to the multiply-add operator.

The systolic operator may include a plurality of multiplication and addition processors (Processing Elements). Each of the plurality of multiplication addition operators may be configured as a flip-flop. Each of the plurality of multiplication addition operation units may perform an operation of multiplying input data input according to a sequential clock having a constant latency and a weight value of a filter, respectively, and adding the multiplied values.

The accumulator can accumulate the output values obtained by multiplying the data of the input feature map by the weight value of the filter and generate an output feature map. The accumulator selects an output value calculated by each of the plurality of multiplication addition operators of the systolic array operator based on the combination of the data of the input feature map and the weight value of the filter that is changed as the filter is strided on the input feature map . The accumulator can generate an output feature map by cumulatively summing the selected output values.

Here, the control direction of the entire cooling system can be determined by the control signal.

In step S309, the apparatus 100 for automatically controlling the cooling system according to an embodiment can control the operation of the cooling system according to the control signal, and can measure the amount of power according to the operation of the cooling system.

In step S310, the apparatus 100 for automatically controlling the cooling system according to an embodiment determines the operation mode of the cooling system on a time-by-time basis based on the amount of electric power per hour, automatically operates the cooling system according to the operation mode for each time zone Can be controlled.

The apparatus 100 for automatically controlling the cooling system can generate control signals differently according to the operation mode for each time zone, and the power usage is dispersed, so that the power peak of the cooling system can be controlled.

Further, the step of determining the fourth control value may include detecting a number of users in the air-conditioning object area, establishing a first heat amount corresponding to a temperature change rate in the air-conditioning object area and a second heat amount corresponding to a heat amount change amount entering and leaving the air- And a fourth control value is determined based on Equation (1)

[Equation 1]

는 단위 면적당 일사량(W/㎡),

는 창의 차폐계수,

는 질량흐름률(kg/s),

는 내부 발열량(W)을 나타낼 수 있다.Where ρ is density (kg / m 3), c _p is specific heat (W / (kgK)), V is volume (m 3), T is temperature (K), h is heat transfer coefficient, U is integrated heat transfer coefficient / (M2K)), A is the area (m2),

(W / m < 2 >) per unit area,

The window shielding coefficient,

Is the mass flow rate (kg / s),

Can represent the internal heating value (W).

The method may further include a sterilizing and deodorizing step of automatically controlling the power applied to the blower to automatically sterilize and deodorize the air conditioning subject area when the fourth control value is measured to be higher than a predetermined threshold value.

The disinfection and deodorization step of the present invention includes the steps of disinfection and deodorization using potassium carbonate (K2CO3), sodium carbonate (Na2CO3), potassium chloride (KCl), D-maltitol (C12H24O11), sodium sulfate (Na2SO4), zinc gluconate (C12H22O14ZnH2O ) Is diluted in a ratio of 1: 9 to a harmful substance decomposition product having a weight ratio of 13 to 20: 4 to 5: 4 to 6: 2 to 7: 1 to 3: 7 to 8, And the air is sprayed to the air-conditioning object area at a speed that is higher than the air-

More preferably, the step of sterilizing and deodorizing is carried out in the same manner as that of the first step except that potassium carbonate (K2CO3), sodium carbonate (Na2CO3), potassium chloride (KCl), D-maltitol (C12H24O11), sodium sulfate (Na2SO4), zinc gluconate (C12H22O14ZnH2O 3) is diluted in a ratio of 1: 9 to a harmful substance decomposition product having a weight ratio of 15: 5: 6: 5: 2: 8, and then sprayed to the air conditioning object region at a rate of 60 ml / min have.

FIG. 4 is a graph showing a change in heat transfer coefficient according to a change in total heat transfer coefficient of a window according to an embodiment.

FIG. 4 shows the relationship between the total heat transfer coefficient and the heat transfer coefficient for various kinds of windows (23 in the present experiment) provided by the EnergyPlus database in order to obtain the coefficient of heat transfer coefficient for the window, and the total heat transfer coefficient and the heat transfer coefficient .

5 is a graph showing changes in heat transfer coefficient according to changes in the total heat transfer coefficient of the wall.

In the present invention, in order to obtain the load factor coefficient considering the structure of various walls constituting the building, the structure of the wall constituting the building is analyzed as a variable and each variable is divided into three conditions. In this case, In order to solve this problem, the number of cases is excessively increased. In order to solve this problem, the number of walls (27 in this experiment) is set as an analytical model by using an experimental design method, Similarly, the relationship between the total heat transfer coefficient and the coefficient of heat transfer coefficient was examined. As a result, it was confirmed that the total heat transfer coefficient and the heat transfer coefficient can be expressed linearly in the wall as in Fig. .

FIG. 6 is a graph illustrating a change in the solar radiation characteristic coefficient according to the variation of the solar radiation acquisition coefficient of a window according to an exemplary embodiment.

In order to obtain the solar radiation characteristic coefficient for the window, the present invention examined the relationship between the solar radiation characteristic coefficient and the solar radiation characteristic coefficient for various types of windows provided in the EnergyPlus database (207 in this experiment) 6.

When curve fitting is performed on the graph shown in FIG. 6, the solar radiation characteristic coefficient for the window can be obtained.

Here, constants C ₅ , C ₆ and C ₇ can be obtained by using a building energy simulation program for various types of windows, respectively, and SC _win is a solar shielding coefficient of an external shading device installed in a window, And i is the number of the six sidewalls surrounding the building's heating and cooling space, and j is the number of windows that form one bearing surface of the building.

The apparatus 100 for automatically controlling the cooling system can transmit the state of the current air conditioning subject area to the external terminal. In addition, the apparatus 100 for automatically controlling the cooling system can transmit control information to an external terminal.

The apparatus 100 for automatically controlling the cooling system can receive the control information from the external terminal and automatically control the cooling system according to the control information.

The external terminal includes a personal computing device capable of wireless networking such as a smart phone, a smart pad, a PDA (Personal Digital Assistants), a laptop, a slate PC, and a portable PC can do.

FIG. 6 is a diagram illustrating data of an input feature map and filters 110-1 to 110-n provided as inputs in a deep learning deep run operation using an artificial neural network.

The artificial intelligence (AI) algorithm including Deep Learning and the like inputs input data 10 to an artificial neural network ANN and outputs output data 30 through arithmetic operations such as convolution . An artificial neural network can refer to a computational architecture that models a biological brain. Within an artificial neural network, the neurons of the brain are connected to each other and collectively act to process input data. Examples of various types of neural networks include a Convolutional Neural Network (CNN), a Recurrent Neural Network (RNN), a Deep Belief Network (DBN), a Restricted Boltzman (RBM) method, but the present invention is not limited thereto. In a feed-forward neural network, neurons in a neural network have links to other neurons. Such connections can be extended in one direction, for example in the forward direction, through a neural network.

Referring to FIG. 6, input data 10 is input to an artificial neural network, and output data 30 is transmitted through a Convolution Neural Network (CNN) 20 including one or more layers Output structure is shown. The artificial neural network may be a Deep Neural Network having two or more layers.

Convolutional neural network 20 may be used to extract "features" such as borders, line colors, etc. from input data 10. [ Convolutional neural network 20 may include a plurality of layers. Each layer can receive data, process data input to the layer, and generate data output from the layer. The data output from the layer may be a feature map generated by convoluting an input image or an input feature map with a weight value of one or more filters, have. The initial layers of the convolutional neural network 20 may be operated to extract low level features such as edges or gradients from the input. The following layers of the convolutional neural network 20 can extract progressively more complex features such as eyes, nose, and the like in the image.

An input feature map 100 and a plurality of filters 110-1 through 110-n provided as inputs to the artificial neural network are shown. The input feature map 100 may be, but is not limited to, a set of pixel values or numerical data of an image input to the neural network. In FIG. 1 (b), the input feature map 100 can be defined as the pixel value of the image to be learned through the artificial neural network. For example, the input feature map 100 may have 256 x 256 pixels and K depth. However, the values are exemplary and the pixel size of the input feature map 100 is not limited to the above example.

The filters 110-1 to 110-n may be formed of N pieces. Each of the plurality of filters 110-1 to 110-n may include a weight value of n by n (nxn). For example, each of the plurality of filters 110-1 to 110-n may have a depth value of K and a pixel of 3x3. However, the size of the filter is exemplary and the size of each of the plurality of filters 110-1 to 110-n is not limited to the above example.

Meanwhile, the above-described embodiments can be implemented in a general-purpose digital computer that can be created as a program that can be executed in a computer and operates the program using a medium readable by a computer. In addition, the structure of the data used in the above-described embodiment can be recorded on a computer-readable medium through various means. Furthermore, the above-described embodiments may be embodied in the form of a recording medium including instructions executable by a computer, such as program modules, being executed by a computer. For example, methods implemented with software modules or algorithms may be stored in a computer readable recording medium, such as code or program instructions, which the computer can read and execute.

The particular implementations described in this disclosure are by way of example only and are not intended to limit the scope of the present disclosure in any way. For brevity of description, descriptions of conventional electronic configurations, control systems, software, and other functional aspects of the systems may be omitted.

Claims (2)

A method for automatically controlling a cooling system using an artificial intelligence (AI) system, comprising a processor for performing a convolution operation of data of an input feature map and a weight value of a filter using a systolic array ,
Obtaining meteorological data including the temperature, cloudiness, humidity, solar radiation amount, dust amount, dust amount and atmospheric fine dust concentration of the next day from the weather station server;
Measuring the condition of the air conditioning subject area including the temperature, humidity, CO ₂ concentration and fine dust concentration of the air conditioning subject area at intervals of 30 minutes;
Wherein the control unit compares the temperature of the air-conditioning target area with the temperature data of the gas-phase data to calculate a temperature comparison value, and compares the temperature comparison value with a preset standard temperature, Acquiring a first control value as an appropriate temperature of the region to be air-conditioned from the learned first learning model by adding information on the concentration;
Comparing the humidity of the air-conditioning object area with the humidity data of the gas-phase data to calculate a humidity comparison value, and comparing the humidity comparison value with a preset standard humidity, wherein the humidity, cloudiness, amount of dust, Determining a second control value as an appropriate humidity of the air conditioning object region from the learned second learning model by adding information on the concentration;
If the ratio of the temperature comparison value and the predetermined standard temperature is greater than the ratio of the humidity comparison value and the predetermined standard humidity, the second control value is changed to 0, and the ratio of the temperature comparison value and the predetermined standard temperature Changing the first control value to 0 if the humidity comparison value is smaller than a predetermined standard humidity ratio;
Calculating the air-conditioning target area of the comparative CO ₂ concentration and the advance to set comparison a standard CO ₂ concentration range of CO ₂ concentration value, but compared to the CO ₂ concentration, temperature, cloud cover, humidity, solar radiation, dust amount, bunjinryang and atmospheric fine Determining a third control value as an appropriate CO ₂ concentration of the air conditioning object region from the learned third learning model by adding information on the dust concentration;
The indoor pollution index is calculated by comparing the fine dust concentration of the air-conditioning subject area with the yellowish amount, the dust amount and the atmospheric fine dust concentration of the gas-phase data, and the indoor pollution index is calculated. Determining a fourth control value as an appropriate indoor pollution index of the area to be air-conditioned from the learned fourth learning model;
And estimating a supply heat amount and a cooling load of the supply system connected to the cooling system by measuring a required heat quantity and an operation schedule of the cooling system connected to the air conditioning object area, Determining a fifth control value, which is a control value of the cooling system, from a learned fifth learning model in consideration of a calorie supply amount and an operation schedule of the supply system that varies depending on the outdoor air condition;
The first control value, the second control value, the third control value, the fourth control value, and the fifth control value are synthesized by using a mapper, Wherein the input feature map data is mapped to a systolic array by changing the order of data of the input feature map so that the data having the input feature map is repeated a predetermined number of times and input to the multiplication addition operator, The weight of the filter is input to the systolic array operator to perform a convolution operation to accumulate the convoluted output values to generate an output feature map, Generating a control signal for controlling the control signal;
Controlling the operation of the cooling system in accordance with the control signal and measuring the amount of power in time according to the operation of the cooling system;
Determining the operating mode of the cooling system on a time-based basis based on the time-based power amount, and automatically controlling the operation of the cooling system according to the operating mode by time zone,
Wherein the step of generating the control signal comprises:
And changing the input order of the input data so that the same two input data values of the input feature map are repeatedly input to the systolic array operator, and when the data of the input feature map is equal to the size N of the filter Generating the control signal by repeating a process of repeatedly inputting the data of the input feature map to be input to the systolic array operator,
Wherein the determining the fourth control value comprises:
Based on Equation (1), which is established based on a first heat amount corresponding to a temperature change rate in the air-conditioning object region and a second heat amount according to a heat amount change amount entering and leaving the air- And a control value is determined,
[Equation 1]

(Where, ρ is the density (kg / ㎥), c _p is the specific heat (W / (㎏K)), V is the volume (㎥), T is temperature (K), h is the heat transfer coefficient, U is an integrated heat transfer coefficient ( W / (m2K)), A is the area (m2),

(W / m < 2 >) per unit area,

The window shielding coefficient,

Is the mass flow rate (kg / s),

Represents the internal calorific value (W).)
And a sterilizing and deodorizing step of automatically sterilizing and deodorizing the area to be air-conditioned by automatically controlling power applied to the blower when the fourth control value is measured to be higher than a predetermined threshold value. To automatically control the cooling system. The method according to claim 1,
In the sterilization and deodorization step,
The weight ratio of potassium carbonate (K2CO3), sodium carbonate (Na2CO3), potassium chloride (KCl), D-maltitol (C12H24O11), sodium sulfate (Na2SO4) and zinc gluconate (C12H22O14ZnH2O To 5: 4 to 6: 2 to 7: 1 to 3: 7 to 8 at a ratio of 1: 9, and then spraying the mixture at the rate of 30 to 70 ml / min to the air- A method for automatically controlling a cooling system using an artificial intelligence system.

Images