CN113361772B - A method and device for predicting water volume in mixed flow pipe network - Google Patents
A method and device for predicting water volume in mixed flow pipe network Download PDFInfo
- Publication number
- CN113361772B CN113361772B CN202110630103.6A CN202110630103A CN113361772B CN 113361772 B CN113361772 B CN 113361772B CN 202110630103 A CN202110630103 A CN 202110630103A CN 113361772 B CN113361772 B CN 113361772B
- Authority
- CN
- China
- Prior art keywords
- monitoring data
- water volume
- minute
- rainfall
- data
- 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.)
- Active
Links
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 title claims abstract description 259
- 238000000034 method Methods 0.000 title claims abstract description 51
- 238000012544 monitoring process Methods 0.000 claims abstract description 269
- 238000012549 training Methods 0.000 claims abstract description 27
- 238000007418 data mining Methods 0.000 claims abstract description 12
- 238000007781 pre-processing Methods 0.000 claims abstract description 12
- 238000004140 cleaning Methods 0.000 claims abstract description 11
- 238000004364 calculation method Methods 0.000 claims description 18
- 238000012795 verification Methods 0.000 claims description 11
- 230000008569 process Effects 0.000 claims description 9
- 230000002159 abnormal effect Effects 0.000 claims description 8
- 238000013480 data collection Methods 0.000 claims description 8
- 238000007405 data analysis Methods 0.000 claims description 6
- 238000012806 monitoring device Methods 0.000 claims description 5
- 238000010276 construction Methods 0.000 claims description 4
- 238000010200 validation analysis Methods 0.000 claims description 4
- 238000007619 statistical method Methods 0.000 claims description 3
- 238000011084 recovery Methods 0.000 claims description 2
- 230000001186 cumulative effect Effects 0.000 claims 6
- 238000012545 processing Methods 0.000 abstract description 19
- 238000004458 analytical method Methods 0.000 abstract description 9
- 239000010865 sewage Substances 0.000 description 10
- 230000006870 function Effects 0.000 description 7
- 238000012216 screening Methods 0.000 description 7
- 238000003860 storage Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 4
- 230000004913 activation Effects 0.000 description 3
- 238000013528 artificial neural network Methods 0.000 description 3
- 238000004590 computer program Methods 0.000 description 3
- 241000251468 Actinopterygii Species 0.000 description 2
- 238000003491 array Methods 0.000 description 2
- 238000004422 calculation algorithm Methods 0.000 description 2
- 238000009826 distribution Methods 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 238000005065 mining Methods 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 238000009825 accumulation Methods 0.000 description 1
- 238000009412 basement excavation Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000005265 energy consumption Methods 0.000 description 1
- 238000009499 grossing Methods 0.000 description 1
- 230000008595 infiltration Effects 0.000 description 1
- 238000001764 infiltration Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01D—MEASURING NOT SPECIALLY ADAPTED FOR A SPECIFIC VARIABLE; ARRANGEMENTS FOR MEASURING TWO OR MORE VARIABLES NOT COVERED IN A SINGLE OTHER SUBCLASS; TARIFF METERING APPARATUS; MEASURING OR TESTING NOT OTHERWISE PROVIDED FOR
- G01D21/00—Measuring or testing not otherwise provided for
- G01D21/02—Measuring two or more variables by means not covered by a single other subclass
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/21—Design, administration or maintenance of databases
- G06F16/215—Improving data quality; Data cleansing, e.g. de-duplication, removing invalid entries or correcting typographical errors
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/24—Querying
- G06F16/245—Query processing
- G06F16/2458—Special types of queries, e.g. statistical queries, fuzzy queries or distributed queries
- G06F16/2465—Query processing support for facilitating data mining operations in structured databases
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/26—Government or public services
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Economics (AREA)
- Tourism & Hospitality (AREA)
- General Engineering & Computer Science (AREA)
- Strategic Management (AREA)
- Human Resources & Organizations (AREA)
- Quality & Reliability (AREA)
- General Business, Economics & Management (AREA)
- Marketing (AREA)
- Development Economics (AREA)
- Primary Health Care (AREA)
- Evolutionary Biology (AREA)
- Artificial Intelligence (AREA)
- General Health & Medical Sciences (AREA)
- Educational Administration (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Operations Research (AREA)
- Fuzzy Systems (AREA)
- Mathematical Physics (AREA)
- Probability & Statistics with Applications (AREA)
- Software Systems (AREA)
- Computational Linguistics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to a method and a device for predicting water quantity of a mixed flow pipe network. The method comprises the following steps: acquiring water quantity monitoring data of a drain pipe network sent by preset water quantity monitoring equipment, and preprocessing the water quantity monitoring data; the pretreatment comprises the following steps: cleaning the invalid data and optimizing the time resolution of the water volume monitoring data; carrying out data mining analysis processing on the pretreated water quantity monitoring data to construct a water quantity monitoring data item; training the water quantity monitoring data item to construct a multi-layer perceptron model; and predicting the water quantity of the pipe network by using a multi-layer perceptron model. The method improves the accuracy and the practicability of the water quantity prediction of the drainage pipe network.
Description
Technical Field
The invention relates to the technical field of pipe network water quantity prediction, in particular to a mixed flow pipe network water quantity prediction method and device.
Background
In the urban drainage system, the drainage pipe network collects urban rain sewage and conveys the urban rain sewage to the processing unit, and plays an important role in going up and down. In reality, the water quantity of the drainage pipe network generally fluctuates, which can cause impact load on the sewage treatment unit, increase energy consumption or overflow, thereby increasing the operation difficulty of the sewage treatment unit. Therefore, the water quantity of the drainage pipe network is simulated and predicted, the running condition of the sewage treatment unit can be adjusted in advance, and the running parameters are optimized, so that the running of the sewage treatment unit is more efficient. In China, a large number of drainage pipe networks exist in a combined system, and meanwhile, the phenomenon of mixed joint exists in a split system pipe network. Therefore, the water quantity law of the drainage pipe network is influenced by factors such as the water law of production and living, seasons, date types (workdays, rest days) and the like, and the inflow and infiltration caused by rainfall has great influence on the water quantity of the drainage pipe network. The factors are complex and have certain randomness, the pipe network water quantity is difficult to predict by using a mechanism model, and the method is suitable for adopting a machine learning algorithm.
At present, the mode for predicting the water quantity of the drainage pipe network is as follows: the artificial fish swarm neural network is used for predicting the water inflow of the sewage plant with higher time precision by adopting an exponential smoothing model, but the consideration of rainfall factors is simpler, and the artificial fish swarm neural network is not suitable for a mixed flow system pipe network; the sewage quantity is predicted by adopting ELMAN neural network, and the mode considers factors such as domestic water, rainfall, continuous weather and the like, but the prediction time is a daily scale, and the precision is low. Therefore, high time accuracy predictions for mixed flow drainage systems remain to be broken through.
Disclosure of Invention
In view of the above, the invention aims to overcome the defects of the prior art and provide a method and a device for predicting the water quantity of a mixed flow pipe network. The problem of the present pipe network water yield prediction precision low is solved.
In order to achieve the above purpose, the invention adopts the following technical scheme:
a method for predicting the water quantity of a mixed flow pipe network comprises the following steps:
Acquiring water quantity monitoring data of a drain pipe network sent by preset water quantity monitoring equipment, and preprocessing the water quantity monitoring data; the pretreatment comprises the following steps: cleaning the invalid data and optimizing the time resolution of the water volume monitoring data;
Carrying out data mining analysis processing on the pretreated water quantity monitoring data to construct a water quantity monitoring data item;
Training the water quantity monitoring data item to construct a multi-layer perceptron model;
and predicting the water quantity of the pipe network by using the multi-layer perceptron model.
Optionally, the water volume monitoring data includes: water quantity minute-by-minute monitoring data, instantaneous rainfall minute-by-minute monitoring data, temperature minute-by-minute monitoring data, humidity minute-by-minute monitoring data and monitoring data acquisition time;
the obtaining of the water quantity monitoring data of the drain pipe network sent by the preset water quantity monitoring equipment comprises the following steps:
receiving minute-by-minute monitoring data of the water quantity monitored by a water meter arranged on a drain pipe network;
receiving the instantaneous rainfall minute-by-minute monitoring data monitored by a rain gauge arranged in a water collecting sheet area of a drainage pipe network, the temperature minute-by-minute monitoring data monitored by a thermometer, and the humidity minute-by-minute monitoring data monitored by a hygrometer;
And extracting specific acquisition time of the water quantity minute-by-minute monitoring data, the instantaneous rainfall minute-by-minute monitoring data, the temperature minute-by-minute monitoring data and the humidity minute-by-minute monitoring data as respective acquisition time of the monitoring data, and determining a date type corresponding to the acquisition time.
Optionally, the preprocessing the water quantity monitoring data includes:
carrying out data cleaning on the missing value, the abnormal value and the error value in the water quantity monitoring data;
Carrying out resolution reconstruction on the water quantity monitoring data after data cleaning, and adjusting the data time precision of the water quantity monitoring data from minute-level precision to hour-level precision; the water quantity minute-by-minute monitoring data are adjusted to water quantity hour-by-hour monitoring data, the instantaneous rainfall minute-by-minute monitoring data are adjusted to instantaneous rainfall minute-by-hour monitoring data, the temperature minute-by-minute monitoring data are adjusted to temperature minute-by-hour monitoring data, and the humidity minute-by-minute monitoring data are adjusted to humidity minute-by-hour monitoring data.
Optionally, the water volume monitoring data item includes: a numeric data item and a category data item;
The data mining analysis processing is carried out on the pretreated water quantity monitoring data, and a water quantity monitoring data item is constructed, and the method comprises the following steps:
calculating single-field accumulated rainfall according to the instantaneous rainfall time-by-time monitoring data and determining a rainfall influence stage;
Constructing the numerical data item according to the water quantity time-by-time monitoring data, the instantaneous rainfall time-by-time monitoring data, the temperature time-by-time monitoring data, the humidity time-by-time monitoring data, the single-field accumulated rainfall and the monitoring data acquisition time;
the category type data item is constructed from the rainfall impact stage and the date type.
Optionally, the calculating the single-field accumulated rainfall according to the time-by-time monitoring data of the instantaneous rainfall includes:
according to the instantaneous rainfall time-by-time monitoring data and the time-by-time rainfall historical data of the same pipe network water collecting sheet area obtained in advance, calculating the variation coefficients of the actual rainfall intervals at different preset rainfall minimum time intervals in a trial mode;
selecting a rainfall minimum time interval when the variation coefficient is 1 according to the trial calculation result;
dividing rainfall orders according to the minimum rainfall time interval;
and calculating the single-field accumulated rainfall according to the rainfall occasion and the instantaneous rainfall time-by-time monitoring data.
Optionally, the determining a rainfall impact stage includes:
screening the instantaneous rainfall time-by-time monitoring data according to a set standard, selecting the drought water volume data, and dividing the drought water volume data into date types; the date type includes: workdays and non-workdays;
Carrying out statistical analysis on the dry day water volume data to obtain the average water volume of the 24-hour dry day pipe network time by time;
obtaining the basic water quantity of the weekday arid-day pipe network and the basic water quantity of the non-weekday arid-day pipe network according to the average water quantity of the arid-day pipe network time by time and the date type;
And determining the rainfall influence stage by combining the rainfall field times, the workday drought day pipe network base water quantity and the non-workday drought day pipe network base water quantity.
Optionally, the training the water quantity monitoring data item to construct a multi-layer perceptron model includes:
constructing data characteristics according to the water quantity monitoring data items;
screening the data features by using a grid search method to obtain model trial calculation optimal parameters;
dividing a training set and a verification set for the water quantity monitoring data item;
Training an initial multi-layer perceptron model by using the training set according to the optimal parameters calculated by the model trial, judging a prediction result of the initial multi-layer perceptron model by using the verification set, and determining that the initial multi-layer perceptron model with the prediction result error meeting a preset condition is the multi-layer perceptron model.
Optionally, the step of screening the monitoring data of the instantaneous rainfall time by time according to a set standard, and selecting the data of the water volume in the dry days includes:
Determining whether rainfall exists on the date of the monitoring data acquisition moment corresponding to the instantaneous rainfall time-by-time monitoring data according to the instantaneous rainfall time-by-time monitoring data;
If rainfall exists, judging that the date is a rainy day;
and eliminating the time-by-time monitoring data of the instantaneous rainfall corresponding to the date and the time-by-time monitoring data of the instantaneous rainfall corresponding to the set days before and after the date, and taking the remaining time-by-time monitoring data of the instantaneous rainfall as the water volume data of the dry days.
Optionally, the rainfall influencing stage includes: a pre-rain stage, a mid-rain stage and a post-rain stage; the period from rainfall start to rainfall stop of a single rainfall is the period in the rain; the post-rain stage is a stage from the end of the period in rain to the recovery of the water quantity of the pipe network to the daily basic water quantity of the pipe network; the pre-rain stage is from a post-rain stage to the next rain stage.
A mixed flow pipe network water quantity prediction device comprises:
The monitoring data acquisition module is used for acquiring water quantity monitoring data of the drainage pipe network sent by preset water quantity monitoring equipment and preprocessing the water quantity monitoring data;
the data item construction module is used for carrying out data mining analysis processing on the pretreated water quantity monitoring data to construct a water quantity monitoring data item;
the model building module is used for training the water quantity monitoring data items and building a multi-layer perceptron model;
And the water quantity prediction module is used for predicting the water quantity of the pipe network by using the multi-layer perceptron model.
The technical scheme provided by the application can comprise the following beneficial effects:
The application discloses a method for predicting the water quantity of a mixed flow pipe network, which comprises the following steps: firstly, acquiring water quantity monitoring data of a drainage pipe network, and preprocessing the water quantity monitoring data; the water quantity monitoring data are detected by water quantity monitoring equipment arranged on a drainage pipe network. Then carrying out data mining analysis processing on the pretreated water quantity monitoring data to construct a water quantity monitoring data item; training the water quantity monitoring data item to construct a multi-layer perceptron model; and predicting the water quantity of the pipe network by using the multi-layer perceptron model. According to the method, firstly, invalid data in water quantity monitoring data of a drainage pipe network monitored in real time are cleaned, the time resolution of the data is improved, then the preprocessed water quantity monitoring data is subjected to excavation processing, a water quantity monitoring data item is constructed, then a multi-layer perceptron model is constructed according to the water quantity monitoring data item, the water quantity of the pipe network is predicted by using the model, the water quantity monitoring data item is cleaned and the resolution is reconstructed in the method, high-quality data in the water quantity monitoring data is selected, and then the water quantity is predicted by using the model constructed by the high-quality water quantity monitoring data item, so that the prediction accuracy of the multi-layer perceptron model is improved. Meanwhile, due to the high time resolution of the water quantity monitoring data, the time resolution of the prediction result is higher, and the practicability of water quantity prediction by using the multi-layer perceptron model is greatly improved.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described, it being obvious that the drawings in the following description are only some embodiments of the invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of a method for predicting the water quantity of a mixed-flow pipe network according to an embodiment of the invention;
FIG. 2 is a flow chart of a method of constructing water volume monitoring data items according to one embodiment of the present invention;
FIG. 3 is a flow chart of a method for constructing a model of a multi-layer perceptron in accordance with an embodiment of the present invention;
FIG. 4 is a block diagram of a device for predicting the water content of a mixed-flow pipe network according to an embodiment of the present invention;
FIG. 5 is a schematic diagram of the water content of the pipe network on dry days under working day and non-working day conditions according to an embodiment of the present invention;
FIG. 6 is a schematic illustration of a rain impact phase provided by an embodiment of the present invention;
FIG. 7 is a schematic diagram showing a comparison of model fitting values and measured values according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the technical solutions of the present invention will be described in detail below. It will be apparent that the described embodiments are only some, but not all, embodiments of the invention. All other embodiments, based on the examples herein, which are within the scope of the invention as defined by the claims, will be within the scope of the invention as defined by the claims.
Fig. 1 is a flowchart of a method for predicting water quantity of a mixed-flow pipe network according to an embodiment of the present invention. Referring to fig. 1, a method for predicting water quantity of a mixed flow pipe network includes:
Step 101: acquiring water quantity monitoring data of a drain pipe network sent by preset water quantity monitoring equipment, and preprocessing the water quantity monitoring data; the pretreatment comprises the following steps: cleaning the invalid data and optimizing the time resolution of the water volume monitoring data.
In this embodiment, when predicting the water volume of the drainage pipe network, the relevant water volume monitoring data of the drainage pipe network needs to be obtained first, for example: water quantity minute-by-minute monitoring data, instantaneous rainfall minute-by-minute monitoring data, temperature minute-by-minute monitoring data, humidity minute-by-minute monitoring data and monitoring data acquisition time. The water quantity minute-by-minute monitoring data are obtained by monitoring a water meter arranged on a drain pipe network, the instantaneous rainfall minute-by-minute monitoring data are obtained by monitoring a rain gauge arranged in a water collecting sheet area of the drain pipe network, the temperature minute-by-minute monitoring data are obtained by monitoring a thermometer arranged in the water collecting sheet area of the drain pipe network, and the humidity minute-by-minute monitoring data are obtained by monitoring the thermometer. Meanwhile, the specific time of the acquisition of the monitoring data is extracted to be the respective time of the acquisition of the monitoring data, namely the year, month, day and time; and then dividing the data into working days and non-working days according to the date type of the monitored data. It should be noted that, in this embodiment, the drainage pipe network may be a branch pipe, a main pipe or a main pipe, and the selection of a specific drainage pipe network is not fixed, and may be determined according to the actual situation, so long as the above data required for water volume prediction can be monitored.
After the water quantity monitoring data of the drainage pipe network are obtained, the water quantity monitoring data are preprocessed, and specifically, missing data, abnormal data and error data in the water quantity monitoring data are cleaned. The missing data processing specifically comprises the following steps: if the rainfall data, the temperature data or the humidity data are missing, directly discarding the data nodes with missing data; if the drainage pipe network water quantity monitoring data is missing, filling the missing data through a naive Bayesian algorithm. The error data processing is specifically as follows: and eliminating data exceeding the measuring range of the monitoring equipment without practical significance, and then discarding or filling the data according to the missing data processing. The processing of the abnormal data is specifically as follows: describing data distribution by adopting a mathematical statistics method to obtain overall distribution characteristics of the data, mining specific data to find abnormal data, and finally performing professional judgment by combining domain knowledge to determine whether abnormal values are removed or reserved.
Then, the time accuracy of the water volume monitoring data after data cleaning is adjusted from minute-level accuracy to hour-level accuracy, and the water volume monitoring data is stored in an hour-by-hour time sequence.
Step 102: and carrying out data mining analysis processing on the pretreated water quantity monitoring data to construct a water quantity monitoring data item.
Step 103: training the water quantity monitoring data item, and constructing a multi-layer perceptron model.
Step 104: and predicting the water quantity of the pipe network by using a multi-layer perceptron model.
Step 102, namely performing data mining analysis processing on the preprocessed water quantity monitoring data, and constructing a water quantity monitoring data item according to the following specific process:
fig. 2 is a flow chart of a method for constructing water volume monitoring data items according to an embodiment of the present invention. Referring to fig. 2, constructing the water volume monitoring data item includes:
step 201: and calculating the single-field accumulated rainfall according to the instantaneous rainfall time-by-time monitoring data and determining the rainfall influence stage.
Wherein, calculating the single-field accumulated rainfall comprises:
Firstly, according to instantaneous rainfall time-by-time monitoring data and time-by-time rainfall historical data, calculating variation coefficients of rainfall actual intervals at different preset rainfall minimum time intervals in a trial mode; the variation coefficient is the ratio of standard deviation to average value, and then the rainfall minimum time interval with the trial-calculation time variation coefficient of 1 is selected. The preset rainfall minimum time interval during trial calculation can be determined by personnel according to actual conditions, the step needs to carry out trial calculation for a plurality of times, and a plurality of different preset rainfall minimum time intervals are set so as to ensure that the variation coefficient is 1. Meanwhile, in order to improve the accuracy of the variation coefficient, the data used in trial calculation not only includes the instantaneous rainfall time-by-time monitoring data, but also uses the time-by-time rainfall historical data in the same water catchment area as the instantaneous rainfall time-by-time monitoring data.
And then dividing rainfall occasions according to the minimum rainfall time interval. And in the same rainfall, adding all the instantaneous rainfall monitoring data time by time before the current moment to be used as a single-field accumulated rainfall.
Meanwhile, the determining of the rainfall influence stage specifically comprises the following steps:
Firstly, screening the instantaneous rainfall monitoring data time by time, and selecting the water volume data of the dry days. During screening, the rainy days are divided into rainy days as long as the rainy days exist on the same day, and the rest rainy days are dry days; removing the water volume data of the dry days on the rainy day and the set days before and after the rainy day, and dividing the date types of the water volume data of the rest dry days into two types of working days and non-working days; carrying out statistical analysis on the pipe network water quantity of the residual dry day water quantity data to obtain the average water quantity of the 24-hour dry day pipe network time by time, thereby obtaining the working day dry day pipe network base water quantity and the non-working day dry day pipe network base water quantity; and finally, determining a rainfall influence stage by combining rainfall scene times, the basic water quantity of the workday drought day pipe network and the basic water quantity of the non-workday drought day pipe network. The specific number of days of the set days in this embodiment is not fixed, and may be specifically determined according to the region where the pipe network is located and different cities.
Specifically, the rainfall influencing stage includes: a pre-rain stage, an in-rain stage and a post-rain stage. In the period of rain, namely, the period from the beginning of single rainfall to the stopping of rain, the water quantity and the water level of the pipe network are affected by rain drop; recovering the pipe network water quantity from the end of the period in rain to the daily basic water quantity of the pipe network, defined as a post-rain phase; the period from the post-rain period to the next mid-rain period is defined as the pre-rain period.
Step 202: the water quantity time-by-time monitoring data, the instantaneous rainfall time-by-time monitoring data, the temperature time-by-time monitoring data, the humidity time-by-time monitoring data, the single-field accumulated rainfall and the monitoring data acquisition time are used for constructing a numerical data item.
Step 203: the category data item is constructed from the rainfall impact phase and the date type.
In more detail, step 103, namely training the water quantity monitoring data item, the specific implementation process of constructing the multi-layer perceptron model is as follows:
FIG. 3 is a flow chart of a method for constructing a model of a multi-layer perceptron in accordance with an embodiment of the present invention. Referring to fig. 3, training the water volume monitoring data item to construct a multi-layer perceptron model includes:
Step 301: and constructing data characteristics according to the water quantity monitoring data items. The data features are used as input data of the multi-layer perceptron model. The numerical data item is directly used as the data characteristic, and the category data item is used as the data characteristic after vector coding. And then, carrying out standardized processing on the data to enable the data format to meet the model requirement.
Step 302: and screening the data characteristics by using a grid search method to obtain optimal parameters of model trial calculation. Wherein the parameters include: the number of input data nodes, the prediction step length, the number of layers of the multi-layer perceptron, the number of nodes of each layer, the activation function, the iteration number and the like.
In the process of trial calculation of the optimal parameters, the user sets various parameter combinations, for example: setting the number of data nodes to be 8, 12, 16, 20 and 24, the prediction step length to be 2,3 and 4, the number of layers of the multi-layer perceptron to be 1,2 and 3, the number of nodes of each layer to be 8, 16, 24 and 32, the activation function to be sigmod, tanh, reLU, the iteration number to be 100, 200, 300 and 400, and the total number of choices to be 2160, wherein each choice is substituted into the model for trial calculation. Finally obtaining the optimal parameters of model trial calculation.
Step 303: and carrying out training set and verification set division on the water quantity monitoring data items.
Dividing the optimal parameters obtained in the steps into a non-rainfall stage data node and a rainfall stage data node, then randomly extracting the data nodes respectively, and dividing a training set and a verification set to balance the data proportion of the non-rainfall stage (a pre-rain stage) and the rainfall stage (a mid-rain stage and a post-rain stage).
Step 304: training an initial multi-layer perceptron model by utilizing the training set according to the optimal parameters calculated by the model trial calculation, judging the prediction result of the initial multi-layer perceptron model by utilizing the verification set, and determining that the initial multi-layer perceptron model with the prediction result error meeting the preset condition is the multi-layer perceptron model.
After an initial model is obtained through a training set, an initial multi-layer perceptron model is utilized to conduct prediction by utilizing a verification set, then the relative average Absolute Error (RMAE, relative Mean Ablute Error) is selected to judge the advantages and disadvantages of the prediction result according to the prediction result of the data of the verification set, the model is continuously subjected to iterative optimization, the accuracy of the model is improved, and finally the initial model with the optimal prediction result is used as the multi-layer perceptron model.
In the embodiment, the accuracy of prediction is improved by taking high-quality data as a basis; semi-automatic preprocessing of the monitoring data is realized, and the quality of the data is improved, so that the accuracy of prediction is improved. Meanwhile, the influence of rainfall on the water quantity of the pipeline is deeply considered in the water quantity monitoring data in the embodiment, and the prediction is more reasonable; based on the actual monitoring data of the drainage pipe network and the rainfall, the influence of rainfall on the pipe network water quantity is deeply analyzed, the actual situation of reality is fully reflected, and the reliability of the calculation result is high.
Meanwhile, the time resolution of the monitoring data is optimized from a minute level to an hour level, so that the randomness and the fluctuation of the data are reduced while the prediction timeliness is basically not sacrificed, the real situation is more closed, and the practical application value of the data is improved; moreover, the prediction result in the application reaches the hour precision and can guide the operation of the drainage system: the running state of the processing unit is adjusted in advance, and the running parameters of the processing unit are optimized, so that the running of the processing unit is more efficient.
The multi-layer perceptron model trained in the application has strong robustness; the parameter selection is carried out by combining the monitoring data analysis and the grid search method, so that the whole process is simple and convenient, and the popularization is easy. The pre-storing method is suitable for different levels of pipe networks and has strong universality. The difference of incoming water rules of the branch pipes, the main pipes and the main pipes is considered, the model is respectively constructed, the parameters are selected for prediction, the universality is strong, and the application range is wide.
In order to more clearly describe the water quantity prediction method in the present application, an example will now be described:
the sewage treatment system is characterized in that a water inlet source at the tail end of a sewage plant is taken as a diversion system drainage system of a certain river basin in the north, and a small amount of mixed joint and partial combined system areas exist in the system to introduce the system as a scene. The specific implementation process is as follows:
1. monitoring data collection, preprocessing, storage
Aiming at the drainage system, arranging a water meter on a main pipe to obtain main pipe water quantity minute-by-minute monitoring data; and arranging a rain gauge with a temperature and humidity function in a water collecting sheet area of the drainage system to obtain rainfall, temperature and humidity monitoring data every minute.
Preprocessing the collected monitoring data to remove error values and abnormal values;
Reconstructing the time resolution of the preprocessed monitoring data into an hour level, wherein each data node comprises a numerical data item (instantaneous rainfall, pipe network water quantity, temperature, humidity and specific time) and a category data item (date type and rainfall stage), the date type is divided into working days and non-working days, and the specific time is divided according to a 24-hour system from 0 hour to 23 hours.
2. Monitoring data analysis and information mining
Analyzing the daily water quantity fluctuation rule analysis of the pipe network to obtain the basic water quantity of the pipe network on dry days, wherein the basic water quantity of the pipe network on dry days is respectively the basic water quantity of the pipe network on working days and non-working days as shown in figure 5;
Combining the rainfall monitoring data and the rainfall duration data to obtain a rainfall interval time variation coefficient of 1, wherein the rainfall interval is 13h, and dividing rainfall occasions according to the rainfall interval time variation coefficient;
According to the division of rainfall orders, obtaining single-field accumulated rainfall, and adding the single-field accumulated rainfall as a numerical data item into the data items of the data nodes;
And judging a rainfall influence stage (as shown in figure 6) by combining the division of rainfall fields and the basic water quantity of the drought days of the pipe network: a rainy stage, namely a stage from rainfall start to rainfall stop of a single rainfall; recovering the pipe network water quantity from the end of the period in rain to the daily basic water quantity of the pipe network, defined as a post-rain phase; the period from the post-rain period to the next mid-rain period is defined as the pre-rain period.
3. Constructing a multi-layer perceptron model to predict the water quantity of a pipeline network
For the type data in the data node, constructing data characteristics by adopting a single-hot coding mode; for numerical data in the data nodes, the numerical data is directly used as data characteristics;
Data is normalized X ′ ＝(X-Xstd)/Xmean, wherein: X ′ is the input value after normalization, X is the original input value, Xstd and Xmean are standard deviation and average value;
Determining multi-layer perceptron model parameters using a grid search method, comprising: the input time sequence data step length is 8, the prediction step length of the main pipe is 2, the number of layers of the multi-layer perceptron is 3, the number of nodes of each layer is 16, the activation function is ReLU, and the iteration number is 300;
Randomly extracting data, wherein 80% of samples are used as training sets, 20% of samples are used as verification sets, and the data proportion of a non-rainfall stage (a pre-rain stage) and a rainfall stage (a mid-rain stage and a post-rain stage) is ensured to be 1:1 during sampling;
Comparison of model fit and measured values as in fig. 7, the relative mean error for the current validation set was 10%. With the accumulation of data, iterative optimization is continuously carried out on the model, and the accuracy of the model is continuously improved.
The embodiment of the invention also provides a device for predicting the water quantity of the mixed-flow pipe network. Please see the examples below.
Fig. 4 is a block diagram of a device for predicting the water quantity of a mixed-flow pipe network according to an embodiment of the present invention. Referring to fig. 4, a device for predicting water quantity of a mixed flow pipe network includes:
The monitoring data acquisition module 401 is configured to acquire water volume monitoring data of a drain pipe network sent by a preset water volume monitoring device, and perform pretreatment on the water volume monitoring data.
The data item construction module 402 is configured to perform data mining analysis processing on the preprocessed water volume monitoring data, and construct a water volume monitoring data item.
The model building module 403 is configured to train the water volume monitoring data item to build a multi-layer perceptron model.
The water quantity prediction module 404 is configured to predict a pipe network water quantity by using the multi-layer perceptron model.
Wherein, the data item construction module 402 is specifically configured to: calculating single-field accumulated rainfall according to the instantaneous rainfall time-by-time monitoring data and determining a rainfall influence stage; constructing the numerical data item according to the water quantity time-by-time monitoring data, the instantaneous rainfall time-by-time monitoring data, the temperature time-by-time monitoring data, the humidity time-by-time monitoring data, the single-field accumulated rainfall and the monitoring data acquisition time; the category type data item is constructed from the rainfall impact stage and the date type.
The model building module 403 is specifically configured to: constructing data characteristics according to the water quantity monitoring data items; screening the data features by using a grid search method to obtain model trial calculation optimal parameters; dividing a training set and a verification set for the water quantity monitoring data item; training an initial multi-layer perceptron model by using the training set according to the optimal parameters calculated by the model trial, judging a prediction result of the initial multi-layer perceptron model by using the verification set, and determining that the initial multi-layer perceptron model with the prediction result error meeting a preset condition is the multi-layer perceptron model.
The device builds a multi-layer perceptron model by monitoring the water quantity monitoring data of the drainage pipe network, and the device is based on high-quality data, so that the prediction accuracy is improved, the influence of rainfall on the water quantity of the pipe network is deeply considered, the prediction is more reasonable, meanwhile, the prediction time resolution is higher, and the practical application value is high.
In order to more clearly describe a hardware system for implementing the embodiment of the invention, the embodiment of the invention also provides a device for predicting the water quantity of the mixed-flow pipe network, which corresponds to the method for predicting the water quantity of the mixed-flow pipe network provided by the embodiment of the invention. Please see the examples below.
A mixed flow pipe network water quantity prediction device, comprising:
a processor, and a memory coupled to the processor;
the storage is used for storing a computer program, and the computer program is at least used for executing the method for predicting the water quantity of the mixed flow pipe network; the processor is configured to invoke and execute the computer program in the memory.
It is to be understood that the same or similar parts in the above embodiments may be referred to each other, and that in some embodiments, the same or similar parts in other embodiments may be referred to.
It should be noted that in the description of the present invention, the terms "first," "second," and the like are used for descriptive purposes only and are not to be construed as indicating or implying relative importance. Furthermore, in the description of the present invention, unless otherwise indicated, the meaning of "plurality" means at least two.
Any process or method descriptions in flow charts or otherwise described herein may be understood as representing modules, segments, or portions of code which include one or more executable instructions for implementing specific logical functions or steps of the process, and further implementations are included within the scope of the preferred embodiment of the present invention in which functions may be executed out of order from that shown or discussed, including substantially concurrently or in reverse order, depending on the functionality involved, as would be understood by those reasonably skilled in the art of the present invention.
It is to be understood that portions of the present invention may be implemented in hardware, software, firmware, or a combination thereof. In the above-described embodiments, the various steps or methods may be implemented in software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, may be implemented using any one or combination of the following techniques, as is well known in the art: discrete logic circuits having logic gates for implementing logic functions on data signals, application specific integrated circuits having suitable combinational logic gates, programmable Gate Arrays (PGAs), field Programmable Gate Arrays (FPGAs), and the like.
Those of ordinary skill in the art will appreciate that all or a portion of the steps carried out in the method of the above-described embodiments may be implemented by a program to instruct related hardware, where the program may be stored in a computer readable storage medium, and where the program, when executed, includes one or a combination of the steps of the method embodiments.
In addition, each functional unit in the embodiments of the present invention may be integrated in one processing module, or each unit may exist alone physically, or two or more units may be integrated in one module. The integrated modules may be implemented in hardware or in software functional modules. The integrated modules may also be stored in a computer readable storage medium if implemented in the form of software functional modules and sold or used as a stand-alone product.
The above-mentioned storage medium may be a read-only memory, a magnetic disk or an optical disk, or the like.
In the description of the present specification, a description referring to terms "one embodiment," "some embodiments," "examples," "specific examples," or "some examples," etc., means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the present invention. In this specification, schematic representations of the above terms do not necessarily refer to the same embodiments or examples. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.
While embodiments of the present invention have been shown and described above, it will be understood that the above embodiments are illustrative and not to be construed as limiting the invention, and that variations, modifications, alternatives and variations may be made to the above embodiments by one of ordinary skill in the art within the scope of the invention.
Claims (6)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110630103.6A CN113361772B (en) | 2021-06-07 | 2021-06-07 | A method and device for predicting water volume in mixed flow pipe network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110630103.6A CN113361772B (en) | 2021-06-07 | 2021-06-07 | A method and device for predicting water volume in mixed flow pipe network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113361772A CN113361772A (en) | 2021-09-07 |
| CN113361772B true CN113361772B (en) | 2024-11-26 |
Family
ID=77532635
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110630103.6A Active CN113361772B (en) | 2021-06-07 | 2021-06-07 | A method and device for predicting water volume in mixed flow pipe network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113361772B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116698112B (en) * | 2023-01-31 | 2023-12-01 | 三峡智慧水务科技有限公司 | Quantitative analysis method for mixing of sewage pipeline external water based on automatic temperature and water level monitoring data |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108376318A (en) * | 2018-02-28 | 2018-08-07 | 清华大学 | A kind of drainage pipeline networks, which becomes a mandarin, infiltrates appraisal procedure and system |
| CN110110890A (en) * | 2019-03-28 | 2019-08-09 | 杭州电子科技大学 | Day wastewater quantity prediction method based on ELMAN neural network |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR100420717B1 (en) * | 2001-05-03 | 2004-03-02 | (주)동명기술공단종합건축사사무소 | Method for detecting leakage of service water tube and its system |
| JP4481920B2 (en) * | 2004-12-14 | 2010-06-16 | 株式会社東芝 | Rainwater drainage support system, rainwater drainage support method, and rainwater drainage control system |
| KR102225241B1 (en) * | 2019-04-30 | 2021-03-11 | 경북대학교 산학협력단 | Apparatus and method for real-time flooding prediction |
| CN110646867A (en) * | 2019-08-28 | 2020-01-03 | 北京无线电计量测试研究所 | Urban drainage monitoring and early warning method and system |
| CN111210093B (en) * | 2020-03-05 | 2023-05-09 | 重庆森鑫炬科技有限公司 | Daily water consumption prediction method based on big data |
| CN112408524B (en) * | 2020-12-04 | 2022-09-16 | 北控水务(中国)投资有限公司 | High-load processing system, method, device and equipment for pipe network regulation and storage coupling water plant |
-
2021
- 2021-06-07 CN CN202110630103.6A patent/CN113361772B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108376318A (en) * | 2018-02-28 | 2018-08-07 | 清华大学 | A kind of drainage pipeline networks, which becomes a mandarin, infiltrates appraisal procedure and system |
| CN110110890A (en) * | 2019-03-28 | 2019-08-09 | 杭州电子科技大学 | Day wastewater quantity prediction method based on ELMAN neural network |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113361772A (en) | 2021-09-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP4807565B2 (en) | Flow prediction device | |
| CN110008301B (en) | Regional geological disaster susceptibility prediction method and device based on machine learning | |
| CN109543147B (en) | A method for rapid nonlinear diagnosis and attribution analysis of rainfall-runoff relationship in a watershed | |
| CN115495991A (en) | Rainfall interval prediction method based on time convolution network | |
| CN113762618B (en) | Lake water level forecasting method based on multi-factor similarity analysis | |
| CN101807045A (en) | Data-based urban sewage pumping station system modeling method | |
| Cui et al. | Configurational entropy theory for streamflow forecasting | |
| Gogien et al. | Assessing the impact of climate change on Combined Sewer Overflows based on small time step future rainfall timeseries and long-term continuous sewer network modelling | |
| CN113361772B (en) | A method and device for predicting water volume in mixed flow pipe network | |
| CN115204688A (en) | Comprehensive evaluation method of drainage system health | |
| Roodsari et al. | A comparison of SAC‐SMA and Adaptive Neuro‐fuzzy Inference System for real‐time flood forecasting in small urban catchments | |
| JP5574769B2 (en) | Unknown water monitoring device and unknown water monitoring method | |
| CN108615092A (en) | A method of the sewage treatment plant inflow amount prediction based on exponential smoothing model | |
| CN113095694B (en) | Rainfall sand transportation model construction method suitable for multiple landform type areas | |
| CN113988210A (en) | Distorted data restoration method, device and storage medium for structural monitoring sensor network | |
| CN115267950B (en) | Initial rainwater quantification method and device, storage medium and electronic equipment | |
| CN113792367B (en) | PySWMM-based drainage system multi-source inflow infiltration and outflow dynamic estimation method | |
| CN113506000B (en) | Method for calculating storm overflow risk of urban water collecting area | |
| KR100538445B1 (en) | Total management/monitoring system for sewer and method thereof | |
| CN118674281A (en) | Reservoir group flood control risk space-time evolution identification method and system under runoff change condition | |
| CN114370611B (en) | Pipe network inflow and infiltration monitoring method and terminal | |
| CN109978263A (en) | A kind of parks' water system water level method for early warning | |
| Dent et al. | Automated calibration using optimization techniques with SWMM RUNOFF | |
| KR20040054906A (en) | The System & Method For Managing Sewer-Pipe | |
| JP4326408B2 (en) | Sewage inflow prediction device |
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 |