CN110135661B - A method and device for predicting water treatment capacity of demineralized water station of thermal power unit - Google Patents

Abstract

The present application provides a method and device for predicting the water treatment capacity of a desalted water station of a thermal power unit, wherein the method includes: obtaining the desalted water station type, boiler type, hydrogen production station type, and desalted water station of a target thermal power unit desalted water station. The duration of putting into operation, the power generation duration of the thermal power unit in multiple historical time periods, the power generation amount of the thermal power unit in each of the historical time periods, the estimated power generation duration of the thermal power unit in at least one future time period, the thermal power unit in the at least one The estimated power generation in the future time period; according to the obtained information and the pre-trained water treatment capacity prediction model of the demineralized water station of the thermal power unit, obtain the water treatment capacity of the target thermal power unit demineralized water station in each of the future time periods. The present application can realize the prediction of the water consumption of the demineralized water station of the thermal power unit in each period.

Description

Method and device for predicting water treatment capacity of demineralized water station of thermal power unit

technical field

The present application relates to the technical field of thermal power plant management, and in particular, to a method and device for predicting the water treatment capacity of a demineralized water station of a thermal power plant.

Background technique

Desalted water refers to the finished water obtained after removing suspended solids, colloids, inorganic cations, anions and other impurities in water by various water treatment processes. The demineralized water of thermal power units is mainly used for boiler make-up water.

At present, the demineralized water station for thermal power units is constructed according to the maximum water consumption during the entire service period of the thermal power unit. However, in the actual process, the water consumption of demineralized water will be different in different stages of the service period of different thermal power units, resulting in the current thermal power unit desalted water. There are problems such as waste of water treatment resources in the station; how to predict the water treatment capacity of the demineralized water station of the thermal power unit is an urgent problem to be solved at present.

SUMMARY OF THE INVENTION

In view of this, the purpose of the embodiments of the present application is to provide a method and device for predicting the water treatment capacity of the demineralized water station of a thermal power unit, which can predict the water consumption of the demineralized water station of a thermal power unit in each period.

In a first aspect, an embodiment of the present application provides a method for predicting the water treatment capacity of a desalination station of a thermal power unit, including:

Obtain the desalted water station type, boiler type, hydrogen production station type of the target thermal power unit desalted water station, the duration of the demineralized water station being put into operation, the power generation time of the thermal power unit in multiple historical time periods, and the thermal power unit in each historical time period. The power generation amount, the estimated power generation duration of the thermal power unit in at least one future time period, and the estimated power generation amount of the thermal power unit in the at least one future time period;

According to the type of demineralized water station, boiler type, and type of hydrogen production station, the duration of operation of the demineralized water station, the power generation time of the thermal power unit in multiple historical time periods, the power generation amount of the thermal power unit in each of the historical time periods, the thermal power generation The estimated power generation duration of the unit in at least one future time period, the estimated power generation amount of the thermal power unit in the at least one future time period, and the pre-trained water treatment capacity prediction model of the demineralized water station of the thermal power unit, obtain the target thermal power unit The amount of water treated by the brine station for each of said future time periods.

In an optional embodiment, the water treatment capacity prediction model of the demineralized water station of the thermal power unit includes a deep learning model;

The deep learning model includes: a recurrent neural network and a feature fusion network;

The deep learning model is trained as follows:

Obtain the sample desalted water station type, boiler type, and hydrogen production station type of the sample demineralized water station of multiple sample thermal power units, the time when the sample desalted water station has been put into operation, the power generation time of the sample thermal power unit in multiple sample historical time periods, the sample thermal power unit The power generation in each sample historical time period, the actual water treatment capacity of the sample thermal power unit demineralized water station in each sample historical time period;

According to the power generation duration of the sample thermal power unit in multiple sample historical time periods, the power generation amount of the sample thermal power unit in each sample historical time period, and according to the sequence of the sample historical time periods, a sample feature vector sequence is generated; , including a plurality of first sample feature vectors;

According to the sample demineralized water station type, boiler type, hydrogen production station type of the sample thermal power unit demineralized water station, and the duration of the sample demineralized water station into operation, the second sample feature vector of each sample thermal power unit demineralized water station is formed;

inputting the sequence of sample feature vectors into a cyclic neural network, and obtaining intermediate feature vectors corresponding to each of the first sample feature vectors;

For each first sample feature vector, the intermediate feature vector corresponding to the first sample feature vector and the second sample feature vector are spliced to form a sample splicing vector corresponding to the first sample feature vector;

Input the sample splicing vector corresponding to each first sample feature vector into the feature fusion network respectively, and obtain the water treatment capacity prediction result corresponding to the first sample feature vector;

The recurrent neural network and the feature fusion network are trained according to the water treatment capacity prediction results corresponding to each first sample feature vector and the actual water treatment capacity of the sample historical time period corresponding to each first sample feature vector respectively.

In an optional embodiment, the obtaining of the water treatment capacity of the demineralized water station of the target thermal power unit in a future time period includes:

According to the power generation duration of the thermal power unit in multiple historical time periods, the power generation amount of the thermal power unit in each of the historical time periods, the estimated power generation duration of the thermal power unit in at least one future time period, and the thermal power unit in the at least one future time period. For the estimated power generation, a sequence of feature vectors is generated according to the sequence of each of the historical time periods and the future time periods; the feature vector sequence includes a plurality of first feature vectors;

According to the demineralized water station type, boiler type, hydrogen production station type of the target thermal power unit demineralized water station, and the operating time of the demineralized water station, a second eigenvector is generated;

Inputting the feature vector sequence into a recurrent neural network to obtain an intermediate feature vector;

splicing each of the intermediate eigenvectors with the second eigenvector, respectively, to generate a splicing vector corresponding to each intermediate eigenvector, and input each of the splicing vectors into the feature fusion network to obtain the The water treatment capacity of the demineralized water station of the target thermal power unit in each of the future time periods.

In an optional embodiment, the method also includes:

According to the water treatment capacity of the demineralized water station of the target thermal power unit in the future time period, it is determined that the ultrafiltration device, activated carbon filtration system, reverse osmosis treatment system and ion exchange treatment system in the demineralized water station of the target thermal power unit will be in the future The operating time of the time period and the operating power.

In an optional embodiment, the method also includes:

According to the water treatment capacity of the target thermal power unit desalted water station in the future time period, the amount of discharged wastewater is determined.

In the second aspect, the embodiment of the present application also provides a water treatment capacity prediction device of a desalination station of a thermal power unit, including:

The acquisition module is used to obtain the desalted water station type, boiler type, hydrogen production station type of the target thermal power unit desalted water station, the duration of the demineralized water station being put into operation, the power generation duration of the thermal power unit in multiple historical time periods, and the thermal power unit in each The power generation in the historical time period, the estimated power generation duration of the thermal power unit in at least one future time period, and the estimated power generation amount of the thermal power unit in the at least one future time period;

The prediction module is used for, according to the type of demineralized water station, the type of boiler, the type of hydrogen production station, the length of time when the demineralized water station is put into operation, the power generation time of the thermal power unit in multiple historical time periods, and the thermal power unit in each of the historical time periods. The estimated power generation time of the thermal power unit in at least one future time period, the estimated power generation amount of the thermal power unit in the at least one future time period, and the pre-trained water treatment capacity prediction model of the demineralized water station of the thermal power unit, obtain all The water treatment capacity of the demineralized water station of the target thermal power unit in each of the future time periods.

In an optional embodiment, the water treatment capacity prediction model of the demineralized water station of the thermal power unit includes a deep learning model;

The deep learning model includes: a recurrent neural network and a feature fusion network;

It also includes: a model training module for training the deep learning model in the following manner:

Obtain the sample desalted water station type, boiler type, and hydrogen production station type of the sample demineralized water station of multiple sample thermal power units, the time when the sample desalted water station has been put into operation, the power generation time of the sample thermal power unit in multiple sample historical time periods, the sample thermal power unit The power generation in each sample historical time period, the actual water treatment capacity of the sample thermal power unit demineralized water station in each sample historical time period;

According to the power generation duration of the sample thermal power unit in multiple sample historical time periods, the power generation amount of the sample thermal power unit in each sample historical time period, and according to the sequence of the sample historical time periods, a sample feature vector sequence is generated; , including a plurality of first sample feature vectors;

According to the sample demineralized water station type, boiler type, hydrogen production station type of the sample thermal power unit demineralized water station, and the duration of the sample demineralized water station into operation, the second sample feature vector of each sample thermal power unit demineralized water station is formed;

inputting the sequence of sample feature vectors into a cyclic neural network, and obtaining intermediate feature vectors corresponding to each of the first sample feature vectors;

For each first sample feature vector, the intermediate feature vector corresponding to the first sample feature vector and the second sample feature vector are spliced to form a sample splicing vector corresponding to the first sample feature vector;

Input the sample splicing vector corresponding to each first sample feature vector into the feature fusion network respectively, and obtain the water treatment capacity prediction result corresponding to the first sample feature vector;

The recurrent neural network and the feature fusion network are trained according to the water treatment capacity prediction results corresponding to each first sample feature vector and the actual water treatment capacity of the sample historical time period corresponding to each first sample feature vector respectively.

In an optional embodiment, the prediction module is used to obtain the water treatment capacity of the target thermal power unit demineralized water station in the future time period in the following manner:

According to the power generation duration of the thermal power unit in multiple historical time periods, the power generation amount of the thermal power unit in each of the historical time periods, the estimated power generation duration of the thermal power unit in at least one future time period, and the thermal power unit in the at least one future time period. For the estimated power generation, a sequence of feature vectors is generated according to the sequence of each of the historical time periods and the future time periods; the feature vector sequence includes a plurality of first feature vectors;

According to the demineralized water station type, boiler type, hydrogen production station type of the target thermal power unit demineralized water station, and the operating time of the demineralized water station, a second eigenvector is generated;

Inputting the feature vector sequence into a recurrent neural network to obtain an intermediate feature vector;

splicing each of the intermediate eigenvectors with the second eigenvector, respectively, to generate a splicing vector corresponding to each intermediate eigenvector, and input each of the splicing vectors into the feature fusion network to obtain the The water treatment capacity of the demineralized water station of the target thermal power unit in each of the future time periods.

In an optional embodiment, the device further includes:

The first determination module is used to determine the ultrafiltration device, activated carbon filtration system, reverse osmosis treatment system and ionic The operating time and operating power of the processing system for the future time period are exchanged.

In an optional embodiment, the device further includes:

The second determination module is configured to determine the amount of discharged waste water according to the water treatment amount of the target thermal power unit desalted water station in a future time period.

In a third aspect, embodiments of the present application further provide an electronic device, including: a processor, a memory, and a bus, where the memory stores machine-readable instructions executable by the processor, and when the electronic device runs, the processing A bus communicates between the processor and the memory, and when the machine-readable instructions are executed by the processor, the first aspect or the steps in any possible implementation manner of the first aspect are performed.

In a fourth aspect, an embodiment of the present application further provides a computer-readable storage medium, where a computer program is stored on the computer-readable storage medium, and the computer program is executed by a processor to execute the above-mentioned first aspect, or any one of the first aspect steps in a possible implementation.

In the embodiment of the present application, by acquiring the desalted water station type, boiler type, hydrogen production station type of the target thermal power unit desalted water station, the duration of the demineralized water station being put into operation, the power generation duration of the thermal power unit in multiple historical time periods, the thermal power unit in each The power generation of the historical time period, the estimated power generation duration of the thermal power unit in at least one future time period, and the estimated power generation of the thermal power unit in the at least one future time period; The water treatment capacity prediction model of the demineralized water station realizes the prediction of the water treatment capacity of the demineralized water station of the target thermal power unit in each of the future time periods. In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, the preferred embodiments are exemplified below, and are described in detail as follows in conjunction with the accompanying drawings.

Description of drawings

In order to illustrate the technical solutions of the embodiments of the present application more clearly, the following drawings will briefly introduce the drawings that need to be used in the embodiments. It should be understood that the following drawings only show some embodiments of the present application, and therefore do not It should be regarded as a limitation of the scope, and for those of ordinary skill in the art, other related drawings can also be obtained according to these drawings without any creative effort.

Fig. 1 shows a flow chart of a method for predicting the water treatment capacity of a demineralized water station of a thermal power unit provided by an embodiment of the present application;

2 shows a flowchart of a specific method for training a water treatment capacity prediction model in the water treatment capacity prediction method of the demineralized water station of the thermal power unit provided by the embodiment of the present application;

3 shows a flowchart of a specific method for obtaining the water treatment capacity of the target thermal power unit demineralized water station in the future time period in the method for predicting the water treatment capacity of the demineralized water station of the thermal power unit provided by the embodiment of the present application;

FIG. 4 shows a schematic diagram of a water treatment capacity prediction device of a demineralized water station of a thermal power unit provided by an embodiment of the present application;

FIG. 5 shows a schematic diagram of an electronic device provided by an embodiment of the present application.

Detailed ways

In order to make the purposes, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only It is a part of the embodiments of the present application, but not all of the embodiments. The components of the embodiments of the present application generally described and illustrated in the drawings herein may be arranged and designed in a variety of different configurations. Thus, the following detailed description of the embodiments of the application provided in the accompanying drawings is not intended to limit the scope of the application as claimed, but is merely representative of selected embodiments of the application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without creative work fall within the protection scope of the present application.

In the process of construction of a thermal power plant, it is usually divided into multiple phases of construction. A certain number of thermal power units will be built in each phase of the project, and the thermal power units that have already been built have already been put into operation when the projects with a later construction time are under construction. Demineralized water is used as boiler make-up water for thermal power units, and is an indispensable raw material for thermal power units in the power generation process. As a production system for desalinated water, the demineralized water station is generally constructed according to the maximum water demand for all thermal power units in the multi-phase project of the thermal power plant to be put into operation. There are differences in different service stages of the current thermal power unit, resulting in the waste of water treatment resources in the current thermal power unit desalination station. Therefore, a method that can predict the water treatment volume of the desalination station with high accuracy has become an urgent problem to be solved at present.

Based on the above research, the present application provides a method and device for predicting the water treatment capacity of a demineralized water station of a thermal power unit. The value is input into the pre-trained water treatment capacity prediction model of the demineralized water station of the thermal power unit, and the water treatment capacity of the target demineralized water station of the thermal power unit in the future time period is obtained, so as to guide the water treatment work of the demineralized water station of the thermal power unit.

The defects existing in the above solutions are all the results obtained by the inventor after practice and careful research. Therefore, the discovery process of the above problems and the solutions proposed by the present application for the above problems hereinafter should be the inventors. Contributions made to this application during the course of this application.

The technical solutions in the present application will be clearly and completely described below with reference to the accompanying drawings in the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. The components of the present application generally described and illustrated in the drawings herein may be arranged and designed in a variety of different configurations. Thus, the following detailed description of the embodiments of the application provided in the accompanying drawings is not intended to limit the scope of the application as claimed, but is merely representative of selected embodiments of the application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without creative work fall within the protection scope of the present application.

It should be noted that like numerals and letters refer to like items in the following figures, so once an item is defined in one figure, it does not require further definition and explanation in subsequent figures.

In order to facilitate the understanding of this embodiment, a method for predicting the water treatment capacity of a demineralized water station of a thermal power unit disclosed in the embodiment of the present application is first introduced in detail. The execution body of the method is generally a computer device with computing capability.

Example 1

Referring to FIG. 1 , a flowchart of a method for predicting the water treatment capacity of a demineralized water station of a thermal power unit provided in Embodiment 1 of the present application, the method includes steps S101 to S102, wherein:

S101: Obtain the desalted water station type, boiler type, and hydrogen production station type of the target thermal power unit desalted water station, the duration of the demineralized water station being put into operation, the power generation duration of the thermal power unit in multiple historical time periods, and the thermal power unit in each historical period. The power generation amount of the time period, the estimated power generation duration of the thermal power unit in at least one future time period, and the estimated power generation amount of the thermal power unit in the at least one future time period.

In the specific implementation, there are many factors affecting the demineralized water station, such as the operating power of the thermal power unit, the type of the boiler in the thermal power unit, the type of the hydrogen production station in the thermal power unit, the length of time the demineralized water station is put into operation, the duration of the thermal power unit The duration of power generation, the power generation of thermal power units, etc., in addition, it also includes the ultrafiltration device, activated carbon filtration system, reverse osmosis treatment system and ion exchange treatment system in the desalination station. cause a certain impact. In the examples of this application, considering the normal operation of the ultrafiltration device, activated carbon filtration system, reverse osmosis treatment system and ion exchange treatment system in the demineralized water station, the application adopts the demineralized water station type, boiler type, hydrogen production station type, the duration of the demineralized water station being put into operation, the power generation duration of the thermal power unit in multiple historical time periods, the power generation amount of the thermal power unit in each of the historical time periods, the estimated power generation duration of the thermal power unit in at least one future time period, the thermal power generation duration The estimated power generation of the unit in the at least one future time period is used to predict the water treatment capacity of the demineralized water station of the thermal power unit.

Here, the number of historical time periods is determined according to the time when the demineralized water station is put into operation and the duration of each historical time period.

For example, if the method for predicting the water treatment capacity of a desalted water station of a thermal power unit provided in the embodiment of this application is used to predict the water treatment capacity of a demineralized station that has been put into use for a certain period of time in the future, if the predicted time is January 2018 1 day (prediction time, that is, the time for predicting the water treatment capacity of the target thermal power unit demineralized water station in at least one future time period using the method for predicting the water treatment capacity of the demineralized water station of the thermal power unit provided by the embodiment of the present application), historical time period The duration is 2 months, and the demineralized water station was put into operation on January 1, 2017, so there are 6 corresponding historical time periods, which are:

January 1, 2017 - February 29, 2017;

March 1, 2017 - April 30, 2017;

May 1, 2017 - June 30, 2017;

July 1, 2017 - August 31, 2017;

September 1, 2017 - October 31, 2017;

November 1, 2017 - December 31, 2017.

Correspondingly, the duration of each future time period is also 2 months. In the above example, when making predictions, the water treatment amount in the future time period after January 1, 2018 can be obtained.

For example, identified future time periods include:

January 1, 2018 - February 29, 2018;

March 1, 2018 - April 30, 2018.

If the method for predicting the water treatment capacity of desalination stations of thermal power units provided in the embodiment of the present application is used to predict the water treatment capacity of demineralized stations that have not been put into use for a certain period of time in the future, the corresponding historical period is empty.

Each future time period is at least one future time period after the demineralized water station is put into use.

For example, if the forecast time is January 1, 2018, and the demineralized water station is planned to be put into use on January 1, 2019, the future time period can include:

January 1, 20179 - February 29, 2019;

March 1, 2019 - April 30, 2019;

May 1, 2019 - June 30, 2019;

July 1, 2019 - August 31, 2019.

Specifically, the following methods can be used to determine the characteristic value of the target thermal power unit demineralized water station under the influence characteristics of each water treatment capacity:

(1) The type of demineralized water station, boiler type, and hydrogen production station type can be directly read from the management system of the thermal power unit.

The thermal power unit management system is a control system for managing thermal power units to generate electricity.

(2) The duration of the demineralized water station being put into operation: the time when the demineralized water station is put into operation and the predicted time can be used to determine the time when the demineralized water station is put into operation; time to determine the length of time the demineralized water station will be put into operation.

(3) The power generation of the thermal power unit in each of the historical time periods: it can also be directly read from the management system of the thermal power unit.

(4) The estimated power generation duration of the thermal power unit in at least one future time period and the estimated power generation amount of the thermal power unit in the at least one future time period are determined according to the actual power generation demand, or read from the relevant information of the power generation plan of the thermal power plant .

S102: According to the type of the demineralized water station, the type of the boiler, and the type of the hydrogen production station, the duration of the demineralized water station being put into operation, the power generation duration of the thermal power unit in multiple historical time periods, and the power generation amount of the thermal power unit in each of the historical time periods , the estimated power generation duration of the thermal power unit in at least one future time period, the estimated power generation amount of the thermal power unit in the at least one future time period, and the pre-trained thermal power unit demineralized water treatment capacity prediction model of the station, to obtain the target thermal power The water treatment capacity of the demineralized water station of the unit in each of the future time periods.

In the specific implementation, the water treatment capacity prediction model of the demineralized water station of the thermal power unit may include: logistic regression model, autoregressive model, moving average model, autoregressive moving average model, integrated moving average autoregressive model, generalized autoregressive conditional difference model Either a variance model or a deep learning model.

For different types of water treatment volume prediction models, there can be different training methods.

A: For the case where the water treatment capacity prediction model of the demineralized water station of the thermal power unit includes a deep learning model, the deep learning model includes: a recurrent neural network and a feature fusion network.

Referring to Figure 2, the embodiment of the present application provides a specific method for training a water treatment capacity prediction model, including:

S201: Obtain the sample demineralized water station type, boiler type, and hydrogen production station type of the sample demineralized water stations of a plurality of sample thermal power units, the time when the sample demineralized water stations have been put into operation, the power generation duration of the sample thermal power units in multiple sample historical time periods, and the samples The power generation of the thermal power unit in each sample historical time period, and the actual water treatment capacity of the sample thermal power unit demineralized water station in each sample historical time period.

S202: According to the power generation duration of the sample thermal power unit in a plurality of sample historical time periods, the power generation amount of the sample thermal power unit in each sample historical time period, and according to the sequence of the sample historical time periods, generate a sample feature vector sequence; the sample feature vector The sequence includes a plurality of first sample feature vectors.

Wherein, each first sample feature vector corresponds to a sample historical time period, and the number of sample historical time periods is consistent with the total number of historical time periods and future time periods in the following S102.

Here, if the sample thermal power unit is of the same type as the target thermal power unit, the sample features can be formed by the power generation duration of the sample thermal power unit in multiple sample historical time periods and the power generation amount of the sample thermal power unit in each of the sample historical time periods. vector sequence.

In another embodiment, if the type of the sample thermal power unit is different from the target thermal power unit, the sample feature vector sequence may also include other related information of the generator unit, such as the power generated in each sample historical time period.

S203: According to the sample demineralized water station type, boiler type, hydrogen production station type of the sample thermal power unit demineralized water station, and the running time of the sample demineralized water station, form the second sample feature vector of each sample thermal power unit demineralized water station.

S204: Input the sequence of sample feature vectors into a recurrent neural network, and obtain intermediate feature vectors corresponding to each of the first sample feature vectors respectively.

S205: For each first sample feature vector, splicing the intermediate feature vector corresponding to the first sample feature vector and the second sample feature vector to form a sample splicing vector corresponding to the first sample feature vector.

S206: Input the sample splicing vector corresponding to each first sample feature vector into the feature fusion network respectively, and obtain a water treatment amount prediction result corresponding to the first sample feature vector.

S207: Train a recurrent neural network and a feature fusion network according to the water treatment capacity prediction results corresponding to each first sample feature vector and the actual water treatment capacity of the sample historical time period corresponding to each first sample feature vector respectively.

Here, the recurrent neural network and the feature fusion network are trained according to the water treatment capacity prediction results corresponding to each first sample feature vector respectively, and the actual water treatment capacity of the sample historical time period corresponding to each first sample feature vector respectively, For each first sample feature vector, according to the output force prediction result corresponding to the first sample feature vector and the actual water treatment result of the corresponding sample historical time period, the cross entropy loss of the deep learning model is obtained, and according to the deep learning The cross-entropy loss of the model, adjust the parameters of the cyclic neural network and the feature fusion network, and finally make the cyclic neural network and the feature fusion network predict the water treatment volume prediction result for each sample historical time period, and the corresponding historical time period of each sample. The difference between the actual water treatment volumes is less than the preset difference threshold.

After the water treatment capacity prediction model is obtained by training, the water treatment capacity of the target thermal power unit in at least one future time period can be obtained based on the water treatment capacity prediction model.

Specifically, as shown in FIG. 3 , an embodiment of the present application also provides a method for obtaining the target thermal power unit demineralized water station in a future time period under the condition that the water treatment capacity prediction model of the demineralized water station of the thermal power unit includes a deep learning model specific ways of treating the water volume, including:

S301: According to the power generation duration of the thermal power unit in multiple historical time periods, the power generation amount of the thermal power unit in each of the historical time periods, the estimated power generation duration of the thermal power unit in at least one future time period, the thermal power unit in the at least one future time period According to the estimated power generation amount of each segment, a sequence of feature vectors is generated according to the sequence of each historical time segment; the sequence of feature vectors includes a plurality of first feature vectors.

Here, since the sum of the number of historical time periods and future time periods is the same as the number of sample historical time periods, in the formed feature vector sequence, the number of first feature vectors is also the same as the number of first sample feature vectors . Among them, the first feature vector corresponds to the historical time period and the future time period one-to-one.

That is, if the demineralized water station of the target thermal power unit is put into operation for 6 months, the duration of the historical time period and the future time period are both 2 months, and the sum of the number of historical time periods and future time periods is 10, The number of corresponding historical time periods is 3, and the number of future time periods is 7. What is obtained is the amount of water treated by the demineralized station for each of the 7 future time periods in the future.

If the demineralized water station of the target thermal power unit has not been put into operation, the duration of the historical time period and the future time period are both 2 months, and the sum of the number of historical time periods and future time periods is 10, the corresponding historical time period The number is 0, and the number of future time periods is 10. What is obtained is the amount of water treated in each of the 20 future time periods after the demineralized station is put into operation in the future.

S302: Generate a second feature vector according to the type of demineralized water station, the type of boiler, the type of hydrogen production station, and the operating time of the demineralized water station;

S303: Input the feature vector sequence into a recurrent neural network to obtain an intermediate feature vector;

S304: splicing each of the intermediate feature vectors with the second feature vector respectively, generating a splicing vector corresponding to each intermediate eigenvector, and inputting each of the splicing vectors into the feature fusion network to obtain The water treatment capacity of the demineralized water station of the target thermal power unit in each of the future time periods.

B: The water treatment capacity prediction models for the demineralized water station of the thermal power unit include: logistic regression model, autoregressive model, moving average model, autoregressive moving average model, integrated moving average autoregressive model, and generalized autoregressive conditional heteroscedasticity model In any one of the situations, referring to Figure 4, the water treatment capacity prediction model can be trained in the following manner:

Obtain the sample desalted water station type, boiler type, and hydrogen production station type of the sample demineralized water station of multiple sample thermal power units, the time when the sample desalted water station has been put into operation, the power generation time of the sample thermal power unit in multiple sample historical time periods, the sample thermal power unit The power generation in each sample historical time period and the actual water treatment capacity of the sample thermal power unit demineralized water station in each sample historical time period.

According to the sample demineralized water station type, boiler type, and hydrogen production station type of multiple sample thermal power unit demineralized water stations, each sample thermal power unit demineralized water station is classified.

For each classification, each sample thermal power unit desalted water station in the classification is regarded as the target thermal power unit desalted water station. The power generation in the historical time period and the actual water treatment capacity of the sample thermal power unit demineralized water station in each sample historical time period are used to train the basic prediction model corresponding to the classification, and the water treatment capacity prediction model corresponding to the classification is obtained.

Specifically, it is any one of the basic prediction model logistic regression model, autoregressive model, moving average model, autoregressive moving average model, integrated moving average autoregressive model, and generalized autoregressive conditional heteroscedasticity model.

The base predictive model can be trained in the following ways:

Taking the power generation duration of each target thermal power unit's demineralized water station in multiple sample historical time periods and the power generation of the sample thermal power unit in each sample historical time period as the independent variable matrix, the sample thermal power unit's demineralized water station in each sample historical time period The actual water treatment volume is used as the dependent variable matrix, and the parameter matrix of each independent variable and dependent variable in the basic prediction model is solved to obtain the parameters of each independent variable and the parameter of the dependent variable, and finally the water treatment volume prediction model is obtained.

When using this water treatment capacity prediction model to predict the water treatment capacity of the target thermal power unit desalination station, the following methods can be used:

According to the desalted water station type, boiler type and hydrogen production station type of the target thermal power unit desalted water station, determine the water treatment capacity prediction model corresponding to the target thermal power unit desalted water station.

The power generation duration of the target thermal power unit demineralized water station thermal power unit in a plurality of historical time periods, the power generation amount of the thermal power unit in each of the described historical time periods, the estimated power generation duration of the thermal power unit in at least one future time period, the thermal power unit in the described The estimated power generation in at least one future time period is used as the value of the independent variable, which is input into the water treatment amount prediction model, and the water treatment amount corresponding to each future time period is obtained.

In addition, in another embodiment of the present application, the provided method for predicting the water treatment capacity of a demineralized water station of a power unit further includes:

S103: According to the water treatment capacity of the demineralized water station of the target thermal power unit in the future time period, determine the ultrafiltration device and the activated carbon filtration system in the demineralized water station of the target thermal power unit, the reverse osmosis treatment system and the ion exchange treatment system in the The operating time and operating power of the future time period.

Here, there is a mapping relationship between the water treatment capacity of the demineralized water station of the target thermal power unit and the running time and operating power of the ultrafiltration device, activated carbon filtration system, reverse osmosis treatment system and ion exchange treatment system in the demineralized water station of the target thermal power unit .

After the water treatment capacity of the target thermal power unit desalted water station in the future time period is determined, the ultrafiltration device, activated carbon filtration system, reverse osmosis treatment system and The operating time and operating power of the ion exchange treatment system for the future time period.

In addition, in another embodiment of the present application, the provided method for predicting the water treatment capacity of a demineralized water station of a power unit further includes:

S104: Determine the amount of discharged wastewater according to the water treatment amount of the target thermal power unit desalted water station in a future time period.

Here, there is a mapping relationship between the water treatment capacity of the demineralized water station of the target thermal power unit and the discharge amount of wastewater. The amount of wastewater discharged to guide the related work of the wastewater treatment system.

In addition, in another embodiment of the present application, the provided method for predicting the water treatment capacity of the demineralized water station of the thermal power unit further includes: based on the water treatment capacity of the target thermal power unit demineralized water station in each of the future time periods, allocating the target Water treatment resources for demineralized water stations of thermal power plants.

In the embodiment of the present application, by acquiring the desalted water station type, boiler type, hydrogen production station type of the target thermal power unit desalted water station, the duration of the demineralized water station being put into operation, the power generation duration of the thermal power unit in multiple historical time periods, the thermal power unit in each The power generation of the historical time period, the estimated power generation duration of the thermal power unit in at least one future time period, and the estimated power generation of the thermal power unit in the at least one future time period; The water treatment capacity prediction model of the demineralized water station realizes the prediction of the water treatment capacity of the demineralized water station of the target thermal power unit in each of the future time periods.

Based on the same inventive concept, the embodiment of the present application also provides a water treatment capacity prediction device for the demineralized water station of the thermal power unit corresponding to the method for predicting the water treatment capacity of the demineralized water station of the thermal power unit. The principle is similar to that of the above-mentioned method for predicting the water treatment capacity of the demineralized water station of the thermal power unit in the embodiment of the present application. Therefore, the implementation of the device can be referred to the implementation of the method, and the repetition will not be repeated.

Embodiment 2

Referring to FIG. 4, it is a schematic diagram of a water treatment capacity prediction device of a demineralized water station of a thermal power unit provided in the second embodiment of the present application. The device includes: an acquisition module 41 and a prediction module 42; wherein,

The acquisition module 41 is used to acquire the desalted water station type, boiler type, and hydrogen production station type of the target thermal power unit desalted water station, the duration of the demineralized water station being put into operation, the power generation duration of the thermal power unit in multiple historical time periods, and the thermal power unit in The power generation amount of each of the historical time periods, the estimated power generation duration of the thermal power unit in at least one future time period, and the estimated power generation amount of the thermal power unit in the at least one future time period;

The prediction module 42 is used for, according to the type of the demineralized water station, the type of the boiler, the type of the hydrogen production station, the duration of the demineralized water station being put into operation, the power generation duration of the thermal power unit in a plurality of historical time periods, and the thermal power unit in each of the historical time periods. The power generation amount of the period, the estimated power generation duration of the thermal power unit in at least one future time period, the estimated power generation amount of the thermal power unit in the at least one future time period, and the pre-trained thermal power unit demineralized water treatment capacity prediction model of the station, obtain The water treatment capacity of the demineralized water station of the target thermal power unit in each of the future time periods.

In the embodiment of the present application, by acquiring the desalted water station type, boiler type, hydrogen production station type of the target thermal power unit desalted water station, the duration of the demineralized water station being put into operation, the power generation duration of the thermal power unit in multiple historical time periods, the thermal power unit in each The power generation of the historical time period, the estimated power generation duration of the thermal power unit in at least one future time period, and the estimated power generation of the thermal power unit in the at least one future time period; The water treatment capacity prediction model of the demineralized water station realizes the prediction of the water treatment capacity of the demineralized water station of the target thermal power unit in each of the future time periods.

In a possible implementation manner, the water treatment capacity prediction model of the demineralized water station of the thermal power unit includes a deep learning model;

The deep learning model includes: a recurrent neural network and a feature fusion network;

It also includes: a model training module 43 for training the deep learning model in the following manner:

Obtain the sample desalted water station type, boiler type, and hydrogen production station type of the sample demineralized water station of multiple sample thermal power units, the time when the sample desalted water station has been put into operation, the power generation time of the sample thermal power unit in multiple sample historical time periods, the sample thermal power unit The power generation in each sample historical time period, the actual water treatment capacity of the sample thermal power unit demineralized water station in each sample historical time period;

According to the power generation duration of the sample thermal power unit in multiple sample historical time periods, the power generation amount of the sample thermal power unit in each sample historical time period, and according to the sequence of the sample historical time periods, a sample feature vector sequence is generated; , including a plurality of first sample feature vectors;

According to the sample demineralized water station type, boiler type, hydrogen production station type of the sample thermal power unit demineralized water station, and the duration of the sample demineralized water station into operation, the second sample feature vector of each sample thermal power unit demineralized water station is formed;

inputting the sequence of sample feature vectors into a cyclic neural network, and obtaining intermediate feature vectors corresponding to each of the first sample feature vectors;

For each first sample feature vector, the intermediate feature vector corresponding to the first sample feature vector and the second sample feature vector are spliced to form a sample splicing vector corresponding to the first sample feature vector;

Input the sample splicing vector corresponding to each first sample feature vector into the feature fusion network respectively, and obtain the water treatment capacity prediction result corresponding to the first sample feature vector;

The recurrent neural network and the feature fusion network are trained according to the water treatment capacity prediction results corresponding to each first sample feature vector and the actual water treatment capacity of the sample historical time period corresponding to each first sample feature vector respectively.

In a possible implementation manner, the prediction module 42 is used to obtain the water treatment capacity of the target thermal power unit demineralized water station in the future time period in the following manner:

According to the power generation duration of the thermal power unit in multiple historical time periods, the power generation amount of the thermal power unit in each of the historical time periods, the estimated power generation duration of the thermal power unit in at least one future time period, and the thermal power unit in the at least one future time period. For the estimated power generation, a sequence of feature vectors is generated according to the sequence of each of the historical time periods and the future time periods; the feature vector sequence includes a plurality of first feature vectors;

According to the demineralized water station type, boiler type, hydrogen production station type of the target thermal power unit demineralized water station, and the operating time of the demineralized water station, a second eigenvector is generated;

Inputting the feature vector sequence into a recurrent neural network to obtain an intermediate feature vector;

splicing each of the intermediate eigenvectors with the second eigenvector, respectively, to generate a splicing vector corresponding to each intermediate eigenvector, and input each of the splicing vectors into the feature fusion network to obtain the The water treatment capacity of the demineralized water station of the target thermal power unit in each of the future time periods.

In a possible implementation manner, the device further includes: a first determination module 44 for determining the amount of water in the demineralized water station of the target thermal power unit according to the water treatment capacity of the demineralized water station of the target thermal power unit in a future time period. Operation time and operation power of ultrafiltration device, activated carbon filtration system, reverse osmosis treatment system and ion exchange treatment system in said future time period.

In a possible implementation, the device further includes:

The second determination module 45 is configured to determine the amount of discharged wastewater according to the water treatment amount of the target thermal power unit desalted water station in a future time period.

For the description of the processing flow of each module in the apparatus and the interaction flow between the modules, reference may be made to the relevant descriptions in the foregoing method embodiments, which will not be described in detail here.

Embodiment 3

The provided schematic diagram of the structure of the computer equipment 500 includes:

The processor 51, the memory 52, and the bus 53; the memory 52 is used to store execution instructions, including the memory 521 and the external memory 522; the memory 521 here is also called internal memory, which is used to temporarily store the operation data in the processor 51, and The data exchanged by the external memory 522 such as the hard disk, the processor 51 exchanges data with the external memory 522 through the memory 521, and when the computer device 500 is running, the processor 51 and the memory 52 communicate through the bus 53, so that The processor 61 executes the following instructions in the user mode:

Obtain the desalted water station type, boiler type, hydrogen production station type of the target thermal power unit desalted water station, the duration of the demineralized water station being put into operation, the power generation time of the thermal power unit in multiple historical time periods, and the thermal power unit in each historical time period. The power generation amount, the estimated power generation duration of the thermal power unit in at least one future time period, and the estimated power generation amount of the thermal power unit in the at least one future time period;

According to the type of demineralized water station, boiler type, and type of hydrogen production station, the duration of operation of the demineralized water station, the power generation time of the thermal power unit in multiple historical time periods, the power generation amount of the thermal power unit in each of the historical time periods, the thermal power generation The estimated power generation duration of the unit in at least one future time period, the estimated power generation amount of the thermal power unit in the at least one future time period, and the pre-trained water treatment capacity prediction model of the demineralized water station of the thermal power unit, obtain the target thermal power unit The amount of water treated by the brine station for each of said future time periods.

In a possible implementation, in the instructions executed by the processor 51, the water treatment capacity prediction model of the demineralized water station of the thermal power unit includes a deep learning model;

The deep learning model includes: a recurrent neural network and a feature fusion network;

The deep learning model is trained as follows:

Obtain the sample desalted water station type, boiler type, and hydrogen production station type of the sample demineralized water station of multiple sample thermal power units, the time when the sample desalted water station has been put into operation, the power generation time of the sample thermal power unit in multiple sample historical time periods, the sample thermal power unit The power generation in each sample historical time period, the actual water treatment capacity of the sample thermal power unit demineralized water station in each sample historical time period;

According to the power generation duration of the sample thermal power unit in multiple sample historical time periods, the power generation amount of the sample thermal power unit in each sample historical time period, and according to the sequence of the sample historical time periods, a sample feature vector sequence is generated; , including a plurality of first sample feature vectors;

According to the sample demineralized water station type, boiler type, hydrogen production station type of the sample thermal power unit demineralized water station, and the duration of the sample demineralized water station into operation, the second sample feature vector of each sample thermal power unit demineralized water station is formed;

inputting the sequence of sample feature vectors into a cyclic neural network, and obtaining intermediate feature vectors corresponding to each of the first sample feature vectors;

For each first sample feature vector, the intermediate feature vector corresponding to the first sample feature vector and the second sample feature vector are spliced to form a sample splicing vector corresponding to the first sample feature vector;

Input the sample splicing vector corresponding to each first sample feature vector into the feature fusion network respectively, and obtain the water treatment capacity prediction result corresponding to the first sample feature vector;

The recurrent neural network and the feature fusion network are trained according to the water treatment capacity prediction results corresponding to each first sample feature vector and the actual water treatment capacity of the sample historical time period corresponding to each first sample feature vector respectively.

In a possible implementation manner, in the instructions executed by the processor 51, the obtaining of the water treatment capacity of the target thermal power unit demineralized water station in the future time period includes:

According to the power generation duration of the thermal power unit in multiple historical time periods, the power generation amount of the thermal power unit in each of the historical time periods, the estimated power generation duration of the thermal power unit in at least one future time period, and the thermal power unit in the at least one future time period. For the estimated power generation, a sequence of feature vectors is generated according to the sequence of each of the historical time periods and the future time periods; the feature vector sequence includes a plurality of first feature vectors;

According to the demineralized water station type, boiler type, hydrogen production station type of the target thermal power unit demineralized water station, and the operating time of the demineralized water station, a second eigenvector is generated;

Inputting the feature vector sequence into a recurrent neural network to obtain an intermediate feature vector;

splicing each of the intermediate eigenvectors with the second eigenvector, respectively, to generate a splicing vector corresponding to each intermediate eigenvector, and input each of the splicing vectors into the feature fusion network to obtain the The water treatment capacity of the demineralized water station of the target thermal power unit in each of the future time periods.

In a possible implementation manner, in the instructions executed by the processor 51, the method further includes:

According to the water treatment capacity of the demineralized water station of the target thermal power unit in the future time period, it is determined that the ultrafiltration device, activated carbon filtration system, reverse osmosis treatment system and ion exchange treatment system in the demineralized water station of the target thermal power unit will be in the future The operating time of the time period and the operating power.

In a possible implementation manner, in the instructions executed by the processor 51, the method further includes:

According to the water treatment capacity of the target thermal power unit desalted water station in the future time period, the amount of discharged wastewater is determined.

In addition, the embodiments of the present application also provide a computer-readable storage medium, on which a computer program is stored, and when the computer program is run by the processor, the thermal power unit demineralized water station described in the above method embodiment is executed. The steps of the water treatment capacity prediction method.

The computer program product of the method for predicting the water treatment capacity of the demineralized water station of the thermal power unit provided by the embodiment of the present application includes a computer-readable storage medium storing program codes, and the instructions included in the program codes can be used to execute the above method embodiments. For the steps of the method for predicting the water treatment capacity of the demineralized water station of the thermal power unit, reference may be made to the above method embodiments, which will not be repeated here.

Those skilled in the art can clearly understand that, for the convenience and brevity of description, for the specific working process of the system and device described above, reference may be made to the corresponding process in the foregoing method embodiments, which will not be repeated here. In the several embodiments provided in this application, it should be understood that the disclosed system, apparatus and method may be implemented in other manners. The apparatus embodiments described above are only illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components may be combined or Can be integrated into another system, or some features can be ignored, or not implemented. On the other hand, the shown or discussed mutual coupling or direct coupling or communication connection may be through some communication interfaces, indirect coupling or communication connection of devices or units, which may be in electrical, mechanical or other forms.

The units described as separate components may or may not be physically separated, and components displayed as units may or may not be physical units, that is, may be located in one place, or may be distributed to multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution in this embodiment.

In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically alone, or two or more units may be integrated into one unit.

The functions, if implemented in the form of software functional units and sold or used as stand-alone products, may be stored in a processor-executable non-volatile computer-readable storage medium. Based on this understanding, the technical solution of the present application can be embodied in the form of a software product in essence, or the part that contributes to the prior art or the part of the technical solution. The computer software product is stored in a storage medium, including Several instructions are used to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (Read-Only Memory, ROM), random access memory (Random Access Memory, RAM), magnetic disk or optical disk and other media that can store program codes.

Finally, it should be noted that the above-mentioned embodiments are only specific implementations of the present application, and are used to illustrate the technical solutions of the present application, rather than limit them. The embodiments describe the application in detail, and those of ordinary skill in the art should understand that: any person skilled in the art can still modify the technical solutions described in the foregoing embodiments within the technical scope disclosed in the application. Or can easily think of changes, or equivalently replace some of the technical features; and these modifications, changes or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions in the embodiments of the application, and should be covered in this application. within the scope of protection. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims (8)

1. a water treatment capacity prediction method of thermal power unit desalinated water station, is characterized in that, comprises: Obtain the desalted water station type, boiler type, hydrogen production station type of the target thermal power unit desalted water station, the duration of the demineralized water station being put into operation, the power generation time of the thermal power unit in multiple historical time periods, and the thermal power unit in each historical time period. The power generation amount, the estimated power generation duration of the thermal power unit in at least one future time period, and the estimated power generation amount of the thermal power unit in the at least one future time period; According to the type of demineralized water station, boiler type, and type of hydrogen production station, the duration of operation of the demineralized water station, the power generation time of the thermal power unit in multiple historical time periods, the power generation amount of the thermal power unit in each of the historical time periods, the thermal power generation The estimated power generation duration of the unit in at least one future time period, the estimated power generation amount of the thermal power unit in the at least one future time period, and the pre-trained water treatment capacity prediction model of the demineralized water station of the thermal power unit, obtain the target thermal power unit the water treatment capacity of the brine station for each of said future time periods; The water treatment capacity prediction model of the demineralized water station of the thermal power unit includes a deep learning model; The deep learning model includes: a recurrent neural network and a feature fusion network; The deep learning model is trained as follows: Obtain the sample desalted water station type, boiler type, and hydrogen production station type of the sample demineralized water station of multiple sample thermal power units, the time when the sample desalted water station has been put into operation, the power generation time of the sample thermal power unit in multiple sample historical time periods, the sample thermal power unit The power generation in each sample historical time period, the actual water treatment capacity of the sample thermal power unit demineralized water station in each sample historical time period; According to the power generation duration of the sample thermal power unit in multiple sample historical time periods, the power generation amount of the sample thermal power unit in each sample historical time period, and according to the sequence of the sample historical time periods, a sample feature vector sequence is generated; , including a plurality of first sample feature vectors; According to the sample demineralized water station type, boiler type, hydrogen production station type of the sample thermal power unit demineralized water station, and the duration of the sample demineralized water station into operation, the second sample feature vector of each sample thermal power unit demineralized water station is formed; inputting the sequence of sample feature vectors into a cyclic neural network, and obtaining intermediate feature vectors corresponding to each of the first sample feature vectors; For each first sample feature vector, the intermediate feature vector corresponding to the first sample feature vector and the second sample feature vector are spliced to form a sample splicing vector corresponding to the first sample feature vector; Input the sample splicing vector corresponding to each first sample feature vector into the feature fusion network respectively, and obtain the water treatment capacity prediction result corresponding to the first sample feature vector; The recurrent neural network and the feature fusion network are trained according to the water treatment capacity prediction results corresponding to each first sample feature vector and the actual water treatment capacity of the sample historical time period corresponding to each first sample feature vector respectively. 2. The method according to claim 1, wherein the obtaining the water treatment capacity of the demineralized water station of the target thermal power unit in a future time period comprises: According to the power generation duration of the thermal power unit in multiple historical time periods, the power generation amount of the thermal power unit in each of the historical time periods, the estimated power generation duration of the thermal power unit in at least one future time period, and the thermal power unit in the at least one future time period. For the estimated power generation, a sequence of feature vectors is generated according to the sequence of each of the historical time periods and the future time periods; the feature vector sequence includes a plurality of first feature vectors; According to the demineralized water station type, boiler type, hydrogen production station type of the target thermal power unit demineralized water station, and the operating time of the demineralized water station, a second eigenvector is generated; Inputting the feature vector sequence into a recurrent neural network to obtain an intermediate feature vector; splicing each of the intermediate eigenvectors with the second eigenvector, respectively, to generate a splicing vector corresponding to each intermediate eigenvector, and input each of the splicing vectors into the feature fusion network to obtain the The water treatment capacity of the demineralized water station of the target thermal power unit in each of the future time periods. 3. The method according to claim 1, wherein the method further comprises: According to the water treatment capacity of the demineralized water station of the target thermal power unit in the future time period, it is determined that the ultrafiltration device, activated carbon filtration system, reverse osmosis treatment system and ion exchange treatment system in the demineralized water station of the target thermal power unit will be in the future The operating time of the time period and the operating power. 4. The method according to claim 1, wherein the method further comprises: According to the water treatment capacity of the target thermal power unit desalted water station in the future time period, the amount of discharged wastewater is determined. 5. A water treatment capacity prediction device of a demineralized water station of thermal power unit, is characterized in that, comprises: The acquisition module is used to obtain the desalted water station type, boiler type, hydrogen production station type of the target thermal power unit desalted water station, the duration of the demineralized water station being put into operation, the power generation duration of the thermal power unit in multiple historical time periods, and the thermal power unit in each The power generation in the historical time period, the estimated power generation duration of the thermal power unit in at least one future time period, and the estimated power generation amount of the thermal power unit in the at least one future time period; The prediction module is used for, according to the type of demineralized water station, the type of boiler, the type of hydrogen production station, the length of time the demineralized water station is put into operation, the power generation time of the thermal power unit in multiple historical time periods, and the thermal power unit in each of the historical time periods. The power generation capacity of the thermal power unit, the estimated power generation duration of the thermal power unit in at least one future time period, the estimated power generation amount of the thermal power unit in the at least one future time period, and the pre-trained water treatment capacity prediction model of the thermal power unit demineralized water station, obtain all the water treatment capacity of the demineralized water station of the target thermal power unit in each of the future time periods; The water treatment capacity prediction model of the demineralized water station of the thermal power unit includes a deep learning model; The deep learning model includes: a recurrent neural network and a feature fusion network; It also includes: a model training module for training the deep learning model in the following manner: Obtain the sample desalted water station type, boiler type, and hydrogen production station type of the sample demineralized water station of multiple sample thermal power units, the time when the sample desalted water station has been put into operation, the power generation time of the sample thermal power unit in multiple sample historical time periods, the sample thermal power unit The power generation in each sample historical time period, the actual water treatment capacity of the sample thermal power unit demineralized water station in each sample historical time period; According to the power generation duration of the sample thermal power unit in multiple sample historical time periods, the power generation amount of the sample thermal power unit in each sample historical time period, and according to the sequence of the sample historical time periods, a sample feature vector sequence is generated; , including a plurality of first sample feature vectors; According to the sample demineralized water station type, boiler type, hydrogen production station type of the sample thermal power unit demineralized water station, and the duration of the sample demineralized water station into operation, the second sample feature vector of each sample thermal power unit demineralized water station is formed; inputting the sequence of sample feature vectors into a cyclic neural network, and obtaining intermediate feature vectors corresponding to each of the first sample feature vectors; For each first sample feature vector, the intermediate feature vector corresponding to the first sample feature vector and the second sample feature vector are spliced to form a sample splicing vector corresponding to the first sample feature vector; Input the sample splicing vector corresponding to each first sample feature vector into the feature fusion network respectively, and obtain the water treatment capacity prediction result corresponding to the first sample feature vector; The recurrent neural network and the feature fusion network are trained according to the water treatment capacity prediction results corresponding to each first sample feature vector and the actual water treatment capacity of the sample historical time period corresponding to each first sample feature vector respectively. 6. device according to claim 5, is characterized in that, described prediction module, is used to obtain the water treatment capacity of described target thermal power unit demineralized water station in the future time period in the following manner: According to the power generation duration of the thermal power unit in multiple historical time periods, the power generation amount of the thermal power unit in each of the historical time periods, the estimated power generation duration of the thermal power unit in at least one future time period, and the thermal power unit in the at least one future time period. For the estimated power generation, a sequence of feature vectors is generated according to the sequence of each of the historical time periods and the future time periods; the feature vector sequence includes a plurality of first feature vectors; According to the demineralized water station type, boiler type, hydrogen production station type of the target thermal power unit demineralized water station, and the operating time of the demineralized water station, a second eigenvector is generated; Inputting the feature vector sequence into a recurrent neural network to obtain an intermediate feature vector; splicing each of the intermediate feature vectors with the second feature vector, respectively, to generate a splicing vector corresponding to each intermediate eigenvector, and input each of the splicing vectors into the feature fusion network to obtain the The water treatment capacity of the demineralized water station of the target thermal power unit in each of the future time periods. 7. The device according to claim 5, wherein the device further comprises: The first determination module is used to determine the ultrafiltration device, activated carbon filtration system, reverse osmosis treatment system and ionic The operating time and operating power of the processing system for the future time period are exchanged. 8. The device according to claim 5, characterized in that, the device further comprises: The second determination module is configured to determine the amount of discharged waste water according to the water treatment amount of the target thermal power unit desalted water station in a future time period.

Images