CN117592355A

CN117592355A - Start-stop control method and device for metal stack, computer equipment and storage medium

Info

Publication number: CN117592355A
Application number: CN202311420212.0A
Authority: CN
Inventors: 张涛; 石波; 吴治江; 黄勇; 王源; 李忠意; 李伟成; 何伯阳; 徐萌; 郭承旺; 郭伟; 熊国华; 王洪涛; 何文凯; 范建超; 吕跃跃; 方郁
Original assignee: China Nuclear Power Technology Research Institute Co Ltd; CGN Power Co Ltd
Current assignee: China Nuclear Power Technology Research Institute Co Ltd; CGN Power Co Ltd
Priority date: 2023-10-30
Filing date: 2023-10-30
Publication date: 2024-02-23

Abstract

The application relates to a start-stop control method and device for a metal stack, computer equipment and a storage medium. The method comprises the following steps: acquiring a first multi-dimensional time sequence corresponding to a target parameter according to operation data of a unit to be controlled in a first history period, and acquiring a modal component corresponding to the first multi-dimensional time sequence; constructing a first prediction model for loop temperature prediction and a second prediction model for reactor power prediction according to modal components corresponding to the first multidimensional time sequence; acquiring a real-time sequence corresponding to a target parameter, and respectively inputting the real-time sequence into a first prediction model and a second prediction model to acquire a loop temperature prediction sequence and a reactor power prediction sequence; and obtaining a predicted working condition of the starting separator according to the loop temperature predicted sequence and the reactor power predicted sequence, and performing start-stop control on the unit to be controlled according to the predicted working condition. The method can reduce the start-stop control cost of the metal stack.

Description

Start-stop control method and device for metal stack, computer equipment and storage medium

Technical Field

The present disclosure relates to the field of metal stack control technologies, and in particular, to a method, an apparatus, a computer device, a storage medium, and a computer program product for controlling start and stop of a metal stack.

Background

The starting separator is an important device in the starting and stopping stage of the metal stack, and has the main functions of absorbing the secondary loop working medium and maintaining the pressure value of the secondary loop working medium to be near a reference value. In the start-stop process of the metal stack, the working condition of starting the separator needs to be identified in real time, and decision information is provided for the start-stop control of the metal stack.

In the related art, expensive hardware instruments are usually relied on to identify the working condition of the starting separator, so that the starting and stopping control cost of the metal stack is high.

Disclosure of Invention

In view of the foregoing, it is desirable to provide a start-stop control method, apparatus, computer device, computer readable storage medium, and computer program product for a metal stack that can reduce the start-stop control cost of the metal stack.

In a first aspect, the present application provides a start-stop control method for a metal stack, including:

acquiring a first multi-dimensional time sequence corresponding to a target parameter according to operation data of a unit to be controlled in a first history period, and acquiring a modal component corresponding to the first multi-dimensional time sequence; the target parameters include a loop temperature and reactor power;

Constructing a first prediction model for loop temperature prediction according to the modal components corresponding to the first multi-dimensional time sequence, and constructing a second prediction model for reactor power prediction according to the modal components corresponding to the first multi-dimensional time sequence; the first prediction model and the second prediction model are obtained based on a plurality of single model combinations;

acquiring a real-time sequence corresponding to a target parameter, and respectively inputting the real-time sequence into a first prediction model and a second prediction model to acquire a loop temperature prediction sequence and a reactor power prediction sequence;

and obtaining a predicted working condition of the starting separator according to the loop temperature predicted sequence and the reactor power predicted sequence, and performing start-stop control on the unit to be controlled according to the predicted working condition.

In one embodiment, the target parameters further include at least one of a pressurizer pressure, a pressurizer water level, a circuit pressure, a reactor coolant outlet temperature, a reactor coolant inlet temperature, a coolant flow time, a circuit flow, a circuit boron concentration, a steam generator pressure, a steam generator water level, a high pressure cylinder exhaust pressure, a primary reheater extraction pressure, a low pressure cylinder inlet and outlet superheated steam temperature, a condenser operating pressure, a feedwater temperature, and a steam reheat.

In one embodiment, acquiring a first multidimensional time sequence corresponding to a target parameter includes:

acquiring an initial time sequence corresponding to each target parameter;

performing data preprocessing operation on the initial time sequence corresponding to each target parameter to obtain a sub-time sequence corresponding to each target parameter; the data preprocessing operation comprises missing value filling and outlier replacement;

and acquiring a first multidimensional time sequence according to the sub-time sequence corresponding to each target parameter.

In one embodiment, the single model includes a long-short-term memory LSTM model and an extreme learning machine ELM model; constructing a first prediction model for loop temperature prediction according to modal components corresponding to the first multidimensional time sequence, wherein the first prediction model comprises:

performing iterative training on the initial LSTM model and the initial ELM model according to the modal components corresponding to the first multidimensional time sequence to obtain a trained LSTM model and a trained ELM model;

acquiring an initial weighting coefficient;

and carrying out weighted fusion on the trained LSTM model and the trained ELM model according to the initial weighting coefficient to obtain an initial prediction model, and taking the initial prediction model as a first prediction model.

In one embodiment, constructing a first prediction model for loop temperature prediction according to the modal components corresponding to the first multidimensional time sequence, further includes:

Acquiring a second multidimensional time sequence corresponding to the target parameter according to the operation data of the unit to be controlled in the second history period;

and (3) adopting a sliding window mechanism, carrying out iterative updating on the initial prediction model according to the second multidimensional time sequence and the fusion error threshold value to obtain a target prediction model, and taking the target prediction model as a first prediction model.

In one embodiment, performing start-stop control on a unit to be controlled according to a predicted working condition includes:

under the condition that the predicted working condition is the same as the current working condition of the starting separator, determining a target control parameter according to the deviation value of the reactor outlet temperature, the deviation value of the steam pressure of the direct-current steam generator and the deviation value of the primary loop outlet flow;

under the condition that the predicted working condition is different from the current working condition of the starting separator, taking the preset control parameter corresponding to the predicted working condition as the target control parameter;

and according to the target control parameters, regulating the load regulating valve, the starting separator outlet isolation valve, the main water supply pump rotating speed, the main water supply low-load regulating valve, the main water supply pump water return regulating valve and the starting separator steam pipeline regulating valve.

In a second aspect, the present application further provides a start-stop control device for a metal stack, including:

The acquisition module is used for acquiring a first multi-dimensional time sequence corresponding to the target parameter according to the operation data of the unit to be controlled in the first history period and acquiring a modal component corresponding to the first multi-dimensional time sequence; the target parameters include a loop temperature and reactor power;

the construction module is used for constructing a first prediction model for loop temperature prediction according to the modal components corresponding to the first multi-dimensional time sequence, and constructing a second prediction model for reactor power prediction according to the modal components corresponding to the first multi-dimensional time sequence; the first prediction model and the second prediction model are obtained based on a plurality of single model combinations;

the prediction module is used for acquiring a real-time sequence corresponding to the target parameter, and inputting the real-time sequence into the first prediction model and the second prediction model respectively to acquire a loop temperature prediction sequence and a reactor power prediction sequence;

and the control module is used for acquiring a predicted working condition of the starting separator according to the loop temperature predicted sequence and the reactor power predicted sequence and performing start-stop control on the unit to be controlled according to the predicted working condition.

In a third aspect, the present application also provides a computer device comprising a memory and a processor, the memory storing a computer program, the processor implementing the following steps when executing the computer program:

In a fourth aspect, the present application also provides a computer readable storage medium having stored thereon a computer program which when executed by a processor performs the steps of:

In a fifth aspect, the present application also provides a computer program product comprising a computer program which, when executed by a processor, performs the steps of:

According to the start-stop control method, the device, the computer equipment, the storage medium and the computer program product of the metal stack, the first multi-dimensional time sequence corresponding to the target parameter is obtained according to the operation data of the unit to be controlled in the first historical period, the modal component corresponding to the first multi-dimensional time sequence is obtained, the first prediction model for loop temperature prediction and the second prediction model for reactor power prediction are constructed according to the modal component corresponding to the first multi-dimensional time sequence, further, the real-time sequence corresponding to the target parameter is obtained, the real-time sequence is respectively input into the first prediction model and the second prediction model, the loop temperature prediction sequence and the reactor power prediction sequence are obtained, the prediction working condition of the start-stop separator is obtained according to the loop temperature prediction sequence and the reactor power prediction sequence, and start-stop control is performed on the unit to be controlled according to the prediction working condition, so that the working condition of the start-stop control of the start-stop separator can be obtained according to the constructed prediction model, and the start-stop control of the unit to be controlled does not need to depend on hardware for working condition recognition, and therefore the start-stop control cost of the metal stack is reduced.

Drawings

In order to more clearly illustrate the embodiments of the present application or the technical solutions in the related art, the drawings that are required to be used in the embodiments or the related technical descriptions will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present application, and other drawings may be obtained according to the drawings without inventive effort for a person having ordinary skill in the art.

FIG. 1 is an application environment diagram of a start-stop control method for a metal stack in one embodiment;

FIG. 2 is a flow chart of a method for controlling start-stop of a metal stack according to one embodiment;

FIG. 3 is a schematic flow chart of tuning an initial predictive model in one embodiment;

FIG. 4 is a flow chart of a method for controlling start-stop of a metal stack according to another embodiment;

FIG. 5 is a block diagram of a start-stop control device for a metal stack in one embodiment;

fig. 6 is an internal structural diagram of a computer device in one embodiment.

Detailed Description

In order to make the objects, technical solutions and advantages of the present application more apparent, the present application will be further described in detail with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the present application.

The start-stop control method for the metal stack, provided by the embodiment of the application, can be applied to an application environment shown in fig. 1. Wherein the terminal 102 communicates with the server 104 via a network. The server 104 may construct a loop temperature prediction model and a reactor power prediction model according to historical operation data of the unit to be controlled, and obtain a loop temperature prediction sequence and a reactor power prediction sequence through the loop temperature prediction model and the reactor power prediction model in a real-time operation process of the unit to be controlled, and further obtain a prediction working condition of the starting separator according to the loop temperature prediction sequence and the reactor power prediction sequence, and perform start-stop control on the unit to be controlled according to the prediction working condition. The terminal 102 may be, but is not limited to, a device parameter acquiring and monitoring device in a nuclear power plant, and the server 104 may be implemented by a stand-alone server or a server cluster formed by a plurality of servers.

In an exemplary embodiment, as shown in fig. 2, a method for controlling start-stop of a metal stack is provided, and the method is applied to the server 104 in fig. 1 for illustration, and includes the following steps:

s202: acquiring a first multi-dimensional time sequence corresponding to a target parameter according to operation data of a unit to be controlled in a first history period, and acquiring a modal component corresponding to the first multi-dimensional time sequence; the target parameters include a loop temperature and reactor power.

The unit to be controlled refers to a metal stack generator set. The first historical period can be any historical operation period of the unit to be controlled, and the period length can be set according to the actual working condition of the unit to be controlled, for example, 24 hours is taken as one operation period.

Optionally, before the unit to be controlled is started and stopped, the server needs to construct a loop temperature prediction model and a reactor power prediction model.

In the process of model construction, model training data needs to be acquired first. The server obtains initial time sequence corresponding to each target parameter according to the operation data of the unit to be controlled in the first history period, and generates a multidimensional time sequence according to the initial time sequence corresponding to each target parameter.

The target parameters may include one or more relevant influencing parameters in addition to two parameters to be measured, namely a loop temperature and reactor power. The relevant influencing parameters include a pressure stabilizer pressure, a pressure stabilizer water level, a loop pressure, a reactor coolant outlet temperature, a reactor coolant inlet temperature, a coolant flow time, a loop flow, a loop boron concentration, a steam generator pressure, a steam generator water level, a high pressure cylinder exhaust pressure, a primary reheater extraction pressure, a low pressure cylinder inlet and outlet superheated steam temperature, a condenser operating pressure, a feedwater temperature, a steam reheat temperature and the like.

After the multi-dimensional time sequence is obtained, the server adopts an empirical mode decomposition (Empirical Mode Decomposition, EMD) algorithm to independently decompose each dimensional time sequence in the multi-dimensional time sequence to obtain the mode components of each dimensional time sequence, and recombines the mode components of each dimensional time sequence to obtain a plurality of mode components corresponding to the multi-dimensional time sequence.

S204: constructing a first prediction model for loop temperature prediction according to the modal components corresponding to the first multi-dimensional time sequence, and constructing a second prediction model for reactor power prediction according to the modal components corresponding to the first multi-dimensional time sequence; the first prediction model and the second prediction model are based on a combination of a plurality of single models.

The single models may include, but are not limited to, long and short term memory network (Long Short Term Memory, LSTM) models, extreme learning machine (Extreme Learning Machine, ELM) models, back propagation neural network (Back Propagation Neural Network, BPNN) models, and support vector regression (Support Vector Regression, SVR) models, among others.

Wherein the multiple single models are combined in ways including but not limited to voting, weighted averaging, stacking, weighted fusion.

Optionally, after obtaining the modal component corresponding to the first multi-dimensional time sequence, the server first takes the modal component corresponding to the first multi-dimensional time sequence as input, takes a loop temperature prediction sequence as output, and performs iterative training on the plurality of single models until the prediction error of each single model is smaller than a set threshold value, so as to obtain a plurality of trained loop temperature prediction single models. Then, a plurality of trained single loop temperature prediction models are combined to obtain a first prediction model.

In addition, the server also needs to iteratively train the plurality of single models by taking the modal component corresponding to the first multidimensional time sequence as input and the reactor power prediction sequence as output until the prediction error of each single model is smaller than a set threshold value, so as to obtain a plurality of trained reactor power prediction single models. Then, a plurality of trained reactor power singles models are combined to obtain a second predictive model.

By combining the single models to obtain a combined prediction model and carrying out parameter prediction according to the combined model, the prediction accuracy of the primary loop temperature and the reactor power can be improved.

S206: and acquiring a real-time sequence corresponding to the target parameter, and respectively inputting the real-time sequence into the first prediction model and the second prediction model to acquire a loop temperature prediction sequence and a reactor power prediction sequence.

Optionally, after model training is completed, the server acquires a real-time multidimensional time sequence corresponding to the target parameter according to operation data in a real-time operation process of the unit to be controlled, inputs the real-time multidimensional time sequence into the first prediction model to acquire a loop temperature prediction sequence, and inputs the real-time multidimensional time sequence into the second prediction model to acquire a reactor power prediction sequence.

S208: and obtaining a predicted working condition of the starting separator according to the loop temperature predicted sequence and the reactor power predicted sequence, and performing start-stop control on the unit to be controlled according to the predicted working condition.

Optionally, a plurality of start-stop working conditions and judging conditions corresponding to the start-stop working conditions are preset in the server. After obtaining a loop temperature prediction sequence and a reactor power prediction sequence, the server determines a prediction working condition of the starting separator in a future time period according to the loop temperature prediction sequence, the reactor power prediction sequence and the judging conditions corresponding to each starting and stopping condition, and further performs starting and stopping control on the unit to be controlled according to the prediction working condition.

In an alternative embodiment, the preset start-stop conditions include a start-up condition 1, a start-up condition 2, a start-up condition 3, a start-up condition 4 and a shutdown condition 1. The distinguishing mode of each start-stop working condition is as follows:

(1) Pile starting working condition 1: the temperature of the primary loop is raised to 0% FP power of the reactor

(2) Pile starting working condition 2: reactor 0% fp power ramp to 8% fp power

(3) Stack starting working condition 3: the 8% FP power of the reactor is increased to 16% FP power

(4) Stack starting working condition 4: reactor 16% fp power ramp to 20% fp power

(5) Shutdown condition 1: reactor 20% fp power reduction to 0% fp power

According to the start-stop control method of the metal stack, the first multi-dimensional time sequence corresponding to the target parameter is obtained according to the operation data of the unit to be controlled in the first historical period, the modal component corresponding to the first multi-dimensional time sequence is obtained, the first prediction model for loop temperature prediction and the second prediction model for reactor power prediction are constructed according to the modal component corresponding to the first multi-dimensional time sequence, further, the real-time sequence corresponding to the target parameter is obtained, the real-time sequence is respectively input into the first prediction model and the second prediction model, a loop temperature prediction sequence and a reactor power prediction sequence are obtained, the prediction working condition of the start-stop separator is obtained according to the loop temperature prediction sequence and the reactor power prediction sequence, start-stop control is carried out on the unit to be controlled according to the prediction working condition, so that the start-stop control is carried out on the unit to be controlled according to the constructed prediction model, and a hardware instrument for working condition recognition is not needed, and start-stop control cost of the metal stack is reduced.

In one embodiment, obtaining a first multidimensional time sequence corresponding to a target parameter includes: acquiring an initial time sequence corresponding to each target parameter; performing data preprocessing operation on the initial time sequence corresponding to each target parameter to obtain a sub-time sequence corresponding to each target parameter; the data preprocessing operation comprises missing value filling and outlier replacement; and acquiring a first multidimensional time sequence according to the sub-time sequence corresponding to each target parameter.

Optionally, due to the fact that the historical operation data are in the process of collecting and recording, missing values and abnormal values may exist in the initial time sequence corresponding to the target parameters due to equipment failure, communication equipment sudden faults or manual brake-off and other reasons. Therefore, the server needs to perform data preprocessing operation on the initial time sequence to adapt to the input data requirement of the model.

For the missing values, under the similar working conditions of different historical operation periods, the data generally have certain similarity, so that corresponding data in a plurality of different historical periods before and after the first historical period can be obtained, an average value is obtained, and the missing values in the initial time sequence are filled by the obtained average value.

For abnormal values, a transverse contrast correction and a longitudinal contrast correction can be performed, which are specifically as follows:

(1) And (3) transverse contrast correction: setting a first change threshold by taking adjacent time data in the initial time sequence as a reference, and if the data at the time t in the initial time sequence exceeds the first change threshold compared with the adjacent time data, namely:

the data is smoothed by the average value of the data at the adjacent time, that is:

wherein,for the first change threshold, ++>Is the value at time t in the initial timing sequence, < >>Is in +.>Value of time of day->Is in +.>A value of the time of day.

(2) Longitudinal contrast correction: setting a second variation threshold by taking data in different historical operation periods as reference, if the initial time sequence is inThe time data and different history running period are +.>The data at the moment exceeds a second change threshold, namely:

the data in the initial time sequence is corrected by adopting the same time data in different historical operation periods, namely:

wherein,for the second change threshold, +.>Is in +.>Value of time of day->For data in different historical operating periods +. >A value of the time of day.

And after finishing preprocessing operation on the initial time sequence corresponding to each target parameter to obtain the sub-time sequence corresponding to each target parameter, the server constructs a multi-dimensional time sequence according to the sub-time sequence corresponding to each target parameter.

In this embodiment, the data integrity and effectiveness of the multi-dimensional time sequence can be improved by acquiring the initial time sequence corresponding to each target parameter, performing data preprocessing operation on the initial time sequence corresponding to each target parameter, and further acquiring the first multi-dimensional time sequence according to the sub-time sequence corresponding to each target parameter.

In one embodiment, the single model includes a long-short-term memory LSTM model and an extreme learning machine ELM model; constructing a first prediction model for loop temperature prediction according to modal components corresponding to the first multidimensional time sequence, wherein the first prediction model comprises: performing iterative training on the initial LSTM model and the initial ELM model according to the modal components corresponding to the first multidimensional time sequence to obtain a trained LSTM model and a trained ELM model; acquiring an initial weighting coefficient; and carrying out weighted fusion on the trained LSTM model and the trained ELM model according to the initial weighting coefficient to obtain an initial prediction model, and taking the initial prediction model as a first prediction model.

Optionally, considering the advantage that the LSTM model can better capture long-term dependencies in sequence data through cell states and gating mechanisms, and the advantage that the ELM model has a fast learning speed and a strong generalization ability, the LSTM model and the ELM model can be selected as a single model for weighted fusion.

After obtaining the modal component corresponding to the first multi-dimensional time sequence, the server takes the modal component corresponding to the first multi-dimensional time sequence as input, takes a loop temperature prediction sequence as output, and carries out iterative training on the initial LSTM model and the initial ELM until the prediction error is smaller than a set threshold value, so as to obtain a trained LSTM model and a trained ELM model. And obtaining an initial weighting coefficient, and carrying out weighted fusion on the trained LSTM model and the trained ELM model to obtain a first initial prediction model corresponding to the loop temperature. In an alternative embodiment, the first initial prediction model corresponding to the loop temperature may be directly used as the first prediction model.

The training pattern of the initial predictive model of reactor power may be referred to the training pattern of the initial predictive model of a loop temperature. Specifically, in an alternative embodiment, the single model includes an LSTM model and an ELM model, and constructing a second prediction model for reactor power prediction according to a modal component corresponding to the first multidimensional time series includes: according to the modal components corresponding to the first multidimensional time sequence, performing iterative training on the initial LSTM model and the initial ELM model, obtaining a trained LSTM model and a trained ELM model, obtaining an initial weighting coefficient, performing weighted fusion on the trained LSTM model and the trained ELM model according to the initial weighting coefficient to obtain a second initial prediction model, and taking the second initial prediction model as the second prediction model, wherein the specific implementation is the same as that in the embodiment, and is not repeated here.

Wherein the initial weighting coefficients may be obtained by:

(1) Arithmetic averaging method

Wherein,the weighting factor for the i-th single model is represented, and m represents the number of single models.

(2) Prediction error square sum reciprocal method

Wherein,weight coefficient representing the i-th single model, < +.>Represents the absolute error of the ith single model at time t,/->Observation value representing time t ++>The predicted value of the ith single model at the time t is represented, m represents the number of single models, and N represents the time length of a predicted sequence.

In this embodiment, by performing iterative training on the initial LSTM model and the initial ELM model according to the modal components corresponding to the first multidimensional time sequence, a trained LSTM model and a trained ELM model are obtained, and further, an initial weighting coefficient is obtained, and according to the initial weighting coefficient, the trained LSTM model and the trained ELM model are subjected to weighted fusion, so as to obtain an initial prediction model, so that data prediction can be performed according to the weighted fused combined prediction model, and thus, the prediction accuracy of the data can be improved.

In one embodiment, constructing a first prediction model for loop temperature prediction according to the modal components corresponding to the first multi-dimensional time sequence, further includes: acquiring a second multidimensional time sequence corresponding to the target parameter according to the operation data of the unit to be controlled in the second history period; and (3) adopting a sliding window mechanism, carrying out iterative updating on the initial prediction model according to the second multidimensional time sequence and the fusion error threshold value to obtain a target prediction model, and taking the target prediction model as a first prediction model.

Optionally, fig. 3 is a schematic diagram of a tuning flow of the initial prediction model, as shown in fig. 3, where the server first obtains a second multidimensional time sequence corresponding to the target parameter according to the operation data of the unit to be controlled in the second history period.

When the first initial prediction model predicts the ith data in the second multidimensional time sequence, the server judges whether the ith data is last-bit data or not. When the ith data is last data, completing model tuning to obtain a first target prediction model, and taking the first target prediction model as a first prediction model; and under the condition that the ith data is not last data, updating the ith data into a sliding window to serve as new sample data, and then judging whether a combined prediction error corresponding to the ith data is larger than a preset fusion error threshold value. When the combined prediction error is not greater than the fusion error threshold, the model continues to predict the (i+1) th data in the second multidimensional time sequence, and the model is not updated; and after the combined prediction error is larger than the fusion error threshold, adjusting the model parameters and the model weights according to the new samples accumulated in the sliding window, and continuously predicting the (i+1) th data.

The training mode of the target prediction model of the reactor power can refer to the training mode of the target prediction model of a loop temperature. Specifically, in an alternative embodiment, the constructing a second prediction model for predicting the reactor power according to the modal component corresponding to the first multidimensional time sequence further includes: according to the operation data of the unit to be controlled in the second history period, a second multidimensional time sequence corresponding to the target parameter is obtained, a sliding window mechanism is adopted, and according to the second multidimensional time sequence and the fusion error threshold value, the second initial prediction model is subjected to iterative updating to obtain a second target prediction model, the second target prediction model is used as the second prediction model, and the specific implementation is the same as that in the embodiment and is not repeated here.

In this embodiment, a sliding window mechanism is adopted to obtain a second multidimensional time sequence corresponding to the target parameter according to the operation data of the unit to be controlled in the second historical period, and the initial prediction model is iteratively updated according to the second multidimensional time sequence and the fusion error threshold value to obtain the target prediction model, so that the model can adapt to the dynamic characteristics of the variable working condition process, and the prediction accuracy of the data is improved.

In one embodiment, the method for controlling the start and stop of the unit to be controlled according to the predicted working condition comprises the following steps: under the condition that the predicted working condition is the same as the current working condition of the starting separator, determining a target control parameter according to the deviation value of the reactor outlet temperature, the deviation value of the steam pressure of the direct-current steam generator and the deviation value of the primary loop outlet flow; under the condition that the predicted working condition is different from the current working condition of the starting separator, taking the preset control parameter corresponding to the predicted working condition as the target control parameter; and according to the target control parameters, regulating the load regulating valve, the starting separator outlet isolation valve, the main water supply pump rotating speed, the main water supply low-load regulating valve, the main water supply pump water return regulating valve and the starting separator steam pipeline regulating valve.

The server is pre-configured with corresponding control parameters for each start-stop working condition of the start-up separator.

Optionally, after determining the predicted working condition of the starting separator, the server determines a corresponding control strategy according to the predicted working condition and the current operating working condition of the starting separator.

And under the condition that the predicted working condition is the same as the current operating condition, the server acquires a reactor outlet temperature deviation value, a direct current steam generator vapor pressure deviation value and a loop outlet flow deviation value according to the reactor outlet temperature feedback value and the set value, the direct current steam generator vapor pressure feedback value and the set value and the loop outlet flow feedback value and the set value respectively. And further, according to the reactor outlet temperature deviation value, the direct current steam generator steam pressure deviation value and the loop outlet flow deviation value, obtaining feedback control parameters, and according to the feedback control parameters, carrying out feedback adjustment on the load adjusting valve, the starting separator outlet isolation valve, the main feed water pump rotating speed, the main feed water low-load adjusting valve, the main feed water pump return water adjusting valve and the starting separator steam pipeline adjusting valve.

Under the condition that the predicted working condition is different from the current operating condition, the server takes the preset control parameter corresponding to the predicted working condition as a target control parameter, and adjusts the load regulating valve, the starting separator outlet isolation valve, the main water supply pump rotating speed, the main water supply low-load regulating valve, the main water supply pump water return regulating valve and the starting separator steam pipeline regulating valve according to the target control parameter.

In this embodiment, the target control parameter is adjusted through the feedback mechanism under the condition that the starting separator does not switch the working condition, and the target control parameter is adjusted through the preset parameter under the condition that the starting separator switches the working condition, so that the load regulating valve, the starting separator outlet isolation valve, the main feed water pump rotating speed, the main feed water low-load regulating valve, the main feed water pump return water regulating valve and the starting separator steam pipeline regulating valve are regulated according to the target control parameter, so that the unit to be controlled can be automatically controlled in the starting and stopping stage, and the safe and stable operation of the unit to be controlled is ensured.

In one embodiment, as shown in fig. 4, there is provided a start-stop control method of a metal stack, the method comprising the steps of:

and acquiring an initial time sequence corresponding to each target parameter according to the operation data of the unit to be controlled in the first history period. The target parameters include a loop temperature and reactor power, and may further include one or more of a stabilizer pressure, a stabilizer water level, a loop pressure, a reactor coolant outlet temperature, a reactor coolant inlet temperature, a coolant flow time, a loop flow, a loop boron concentration, a steam generator pressure, a steam generator water level, a high pressure cylinder exhaust pressure, a primary reheater extraction pressure, a low pressure cylinder inlet and outlet superheated steam temperature, a condenser operating pressure, a feedwater temperature, and a steam reheat.

And carrying out data preprocessing operation on the initial time sequence corresponding to each target parameter to obtain a sub-time sequence corresponding to each target parameter, wherein the data preprocessing operation comprises missing value filling and abnormal value replacement.

And acquiring a first multidimensional time sequence according to the sub-time sequence corresponding to each target parameter, and acquiring a modal component corresponding to the first multidimensional time sequence.

And carrying out iterative training on the initial LSTM model and the initial ELM model according to the modal components corresponding to the first multidimensional time sequence, obtaining a trained LSTM model and a trained ELM model, obtaining an initial weighting coefficient, and carrying out weighted fusion on the trained LSTM model and the trained ELM model according to the initial weighting coefficient to obtain an initial prediction model.

According to the running data of the unit to be controlled in the second historical period, a second multidimensional time sequence corresponding to the target parameter is obtained, a sliding window mechanism is adopted, and according to the second multidimensional time sequence and the fusion error threshold value, the initial prediction model is subjected to iterative updating to obtain a target prediction model, and the target prediction model is used as the first prediction model or the second prediction model.

And acquiring a real-time sequence corresponding to the target parameter, and respectively inputting the real-time sequence into the first prediction model and the second prediction model to acquire a loop temperature prediction sequence and a reactor power prediction sequence.

And obtaining the predicted working condition of the starting separator according to the loop temperature predicted sequence and the reactor power predicted sequence.

Under the condition that the predicted working condition is the same as the current working condition of the starting separator, determining a target control parameter according to the deviation value of the reactor outlet temperature, the deviation value of the steam pressure of the direct-current steam generator and the deviation value of the primary loop outlet flow; and under the condition that the predicted working condition is different from the current working condition of the starting separator, taking the preset control parameter corresponding to the predicted working condition as the target control parameter.

It should be understood that, although the steps in the flowcharts related to the embodiments described above are sequentially shown as indicated by arrows, these steps are not necessarily sequentially performed in the order indicated by the arrows. The steps are not strictly limited to the order of execution unless explicitly recited herein, and the steps may be executed in other orders. Moreover, at least some of the steps in the flowcharts described in the above embodiments may include a plurality of steps or a plurality of stages, which are not necessarily performed at the same time, but may be performed at different times, and the order of the steps or stages is not necessarily performed sequentially, but may be performed alternately or alternately with at least some of the other steps or stages.

Based on the same inventive concept, the embodiment of the application also provides a start-stop control device for the metal stack, which is used for realizing the start-stop control method for the metal stack. The implementation scheme of the device for solving the problem is similar to that described in the above method, so the specific limitation in the embodiments of the start-stop control device for one or more metal stacks provided below can be referred to the limitation of the start-stop control method for the metal stack in the above description, and will not be repeated here.

In an exemplary embodiment, as shown in fig. 5, there is provided a start-stop control device of a metal stack, including: an acquisition module 510, a construction module 520, a prediction module 530, and a control module 540, wherein:

the obtaining module 510 is configured to obtain a first multidimensional time sequence corresponding to the target parameter according to operation data of the unit to be controlled in the first history period, and obtain a modal component corresponding to the first multidimensional time sequence; the target parameters include a loop temperature and reactor power;

the construction module 520 is configured to construct a first prediction model for loop temperature prediction according to the modal component corresponding to the first multi-dimensional time sequence, and construct a second prediction model for reactor power prediction according to the modal component corresponding to the first multi-dimensional time sequence; the first prediction model and the second prediction model are obtained based on a plurality of single model combinations;

The prediction module 530 is configured to obtain a real-time sequence corresponding to the target parameter, and input the real-time sequence into the first prediction model and the second prediction model respectively to obtain a loop temperature prediction sequence and a reactor power prediction sequence;

the control module 540 is configured to obtain a predicted working condition of the starting separator according to a loop temperature prediction sequence and a reactor power prediction sequence, and perform start-stop control on the unit to be controlled according to the predicted working condition.

In one embodiment, the obtaining module 510 is further configured to obtain a first multidimensional time sequence corresponding to the target parameter, including: acquiring an initial time sequence corresponding to each target parameter; performing data preprocessing operation on the initial time sequence corresponding to each target parameter to obtain a sub-time sequence corresponding to each target parameter; the data preprocessing operation comprises missing value filling and outlier replacement; and acquiring a first multidimensional time sequence according to the sub-time sequence corresponding to each target parameter.

In one embodiment, the building module 520 is further configured to iteratively train the initial LSTM model and the initial ELM model according to the modal component corresponding to the first multidimensional time sequence, to obtain a trained LSTM model and a trained ELM model; acquiring an initial weighting coefficient; and carrying out weighted fusion on the trained LSTM model and the trained ELM model according to the initial weighting coefficient to obtain an initial prediction model, and taking the initial prediction model as a first prediction model.

In one embodiment, the construction module 520 is further configured to obtain a second multidimensional time sequence corresponding to the target parameter according to the operation data of the unit to be controlled in the second history period; and (3) adopting a sliding window mechanism, carrying out iterative updating on the initial prediction model according to the second multidimensional time sequence and the fusion error threshold value to obtain a target prediction model, and taking the target prediction model as a first prediction model.

In one embodiment, the control module 540 is further configured to determine the target control parameter based on the reactor outlet temperature deviation value, the once-through steam generator vapor pressure deviation value, and the loop outlet flow deviation value, if the predicted operating condition is the same as the current operating condition at which the separator is activated; under the condition that the predicted working condition is different from the current working condition of the starting separator, taking the preset control parameter corresponding to the predicted working condition as the target control parameter; and according to the target control parameters, regulating the load regulating valve, the starting separator outlet isolation valve, the main water supply pump rotating speed, the main water supply low-load regulating valve, the main water supply pump water return regulating valve and the starting separator steam pipeline regulating valve.

The modules in the start-stop control device of the metal stack can be realized in whole or in part by software, hardware and a combination thereof. The above modules may be embedded in hardware or may be independent of a processor in the computer device, or may be stored in software in a memory in the computer device, so that the processor may call and execute operations corresponding to the above modules.

In one exemplary embodiment, a computer device is provided, which may be a server, the internal structure of which may be as shown in fig. 6. The computer device includes a processor, a memory, an Input/Output interface (I/O) and a communication interface. The processor, the memory and the input/output interface are connected through a system bus, and the communication interface is connected to the system bus through the input/output interface. Wherein the processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The database of the computer equipment is used for storing preset control parameters and other data corresponding to the start-up and stop conditions. The input/output interface of the computer device is used to exchange information between the processor and the external device. The communication interface of the computer device is used for communicating with an external terminal through a network connection. The computer program, when executed by a processor, implements a method for controlling start-stop of a metal stack.

It will be appreciated by those skilled in the art that the structure shown in fig. 6 is merely a block diagram of some of the structures associated with the present application and is not limiting of the computer device to which the present application may be applied, and that a particular computer device may include more or fewer components than shown, or may combine certain components, or have a different arrangement of components.

In one exemplary embodiment, a computer device is provided comprising a memory and a processor, the memory having stored therein a computer program, the processor when executing the computer program performing the steps of: acquiring a first multi-dimensional time sequence corresponding to a target parameter according to operation data of a unit to be controlled in a first history period, and acquiring a modal component corresponding to the first multi-dimensional time sequence; the target parameters include a loop temperature and reactor power; constructing a first prediction model for loop temperature prediction according to the modal components corresponding to the first multi-dimensional time sequence, and constructing a second prediction model for reactor power prediction according to the modal components corresponding to the first multi-dimensional time sequence; the first prediction model and the second prediction model are obtained based on a plurality of single model combinations; acquiring a real-time sequence corresponding to a target parameter, and respectively inputting the real-time sequence into a first prediction model and a second prediction model to acquire a loop temperature prediction sequence and a reactor power prediction sequence; and obtaining a predicted working condition of the starting separator according to the loop temperature predicted sequence and the reactor power predicted sequence, and performing start-stop control on the unit to be controlled according to the predicted working condition.

In one embodiment, the target parameters involved in the processor executing the computer program further include at least one of a pressurizer pressure, a pressurizer water level, a circuit pressure, a reactor coolant outlet temperature, a reactor coolant inlet temperature, a coolant flow time, a circuit flow, a circuit boron concentration, a steam generator pressure, a steam generator water level, a high pressure cylinder exhaust pressure, a primary reheater extraction pressure, a low pressure cylinder inlet and outlet superheated steam temperature, a condenser operating pressure, a feedwater temperature, and a steam reheat.

In one embodiment, the processor when executing the computer program further performs the steps of: acquiring an initial time sequence corresponding to each target parameter; performing data preprocessing operation on the initial time sequence corresponding to each target parameter to obtain a sub-time sequence corresponding to each target parameter; the data preprocessing operation comprises missing value filling and outlier replacement; and acquiring a first multidimensional time sequence according to the sub-time sequence corresponding to each target parameter.

In one embodiment, the processor when executing the computer program further performs the steps of: performing iterative training on the initial LSTM model and the initial ELM model according to the modal components corresponding to the first multidimensional time sequence to obtain a trained LSTM model and a trained ELM model; acquiring an initial weighting coefficient; and carrying out weighted fusion on the trained LSTM model and the trained ELM model according to the initial weighting coefficient to obtain an initial prediction model, and taking the initial prediction model as a first prediction model.

In one embodiment, the processor when executing the computer program further performs the steps of: acquiring a second multidimensional time sequence corresponding to the target parameter according to the operation data of the unit to be controlled in the second history period; and (3) adopting a sliding window mechanism, carrying out iterative updating on the initial prediction model according to the second multidimensional time sequence and the fusion error threshold value to obtain a target prediction model, and taking the target prediction model as a first prediction model.

In one embodiment, the processor when executing the computer program further performs the steps of: under the condition that the predicted working condition is the same as the current working condition of the starting separator, determining a target control parameter according to the deviation value of the reactor outlet temperature, the deviation value of the steam pressure of the direct-current steam generator and the deviation value of the primary loop outlet flow; under the condition that the predicted working condition is different from the current working condition of the starting separator, taking the preset control parameter corresponding to the predicted working condition as the target control parameter; and according to the target control parameters, regulating the load regulating valve, the starting separator outlet isolation valve, the main water supply pump rotating speed, the main water supply low-load regulating valve, the main water supply pump water return regulating valve and the starting separator steam pipeline regulating valve.

In one embodiment, a computer readable storage medium is provided having a computer program stored thereon, which when executed by a processor, performs the steps of: acquiring a first multi-dimensional time sequence corresponding to a target parameter according to operation data of a unit to be controlled in a first history period, and acquiring a modal component corresponding to the first multi-dimensional time sequence; the target parameters include a loop temperature and reactor power; constructing a first prediction model for loop temperature prediction according to the modal components corresponding to the first multi-dimensional time sequence, and constructing a second prediction model for reactor power prediction according to the modal components corresponding to the first multi-dimensional time sequence; the first prediction model and the second prediction model are obtained based on a plurality of single model combinations; acquiring a real-time sequence corresponding to a target parameter, and respectively inputting the real-time sequence into a first prediction model and a second prediction model to acquire a loop temperature prediction sequence and a reactor power prediction sequence; and obtaining a predicted working condition of the starting separator according to the loop temperature predicted sequence and the reactor power predicted sequence, and performing start-stop control on the unit to be controlled according to the predicted working condition.

In one embodiment, the target parameters involved in the execution of the computer program by the processor further include at least one of a pressurizer pressure, a pressurizer water level, a loop pressure, a reactor coolant outlet temperature, a reactor coolant inlet temperature, a coolant flow time, a loop flow, a loop boron concentration, a steam generator pressure, a steam generator water level, a high pressure cylinder exhaust pressure, a primary reheater extraction pressure, a low pressure cylinder inlet and outlet superheated steam temperature, a condenser operating pressure, a feedwater temperature, and a steam reheat.

In one embodiment, the computer program when executed by the processor further performs the steps of: acquiring an initial time sequence corresponding to each target parameter; performing data preprocessing operation on the initial time sequence corresponding to each target parameter to obtain a sub-time sequence corresponding to each target parameter; the data preprocessing operation comprises missing value filling and outlier replacement; and acquiring a first multidimensional time sequence according to the sub-time sequence corresponding to each target parameter.

In one embodiment, the computer program when executed by the processor further performs the steps of: performing iterative training on the initial LSTM model and the initial ELM model according to the modal components corresponding to the first multidimensional time sequence to obtain a trained LSTM model and a trained ELM model; acquiring an initial weighting coefficient; and carrying out weighted fusion on the trained LSTM model and the trained ELM model according to the initial weighting coefficient to obtain an initial prediction model, and taking the initial prediction model as a first prediction model.

In one embodiment, the computer program when executed by the processor further performs the steps of: acquiring a second multidimensional time sequence corresponding to the target parameter according to the operation data of the unit to be controlled in the second history period; and (3) adopting a sliding window mechanism, carrying out iterative updating on the initial prediction model according to the second multidimensional time sequence and the fusion error threshold value to obtain a target prediction model, and taking the target prediction model as a first prediction model.

In one embodiment, the computer program when executed by the processor further performs the steps of: under the condition that the predicted working condition is the same as the current working condition of the starting separator, determining a target control parameter according to the deviation value of the reactor outlet temperature, the deviation value of the steam pressure of the direct-current steam generator and the deviation value of the primary loop outlet flow; under the condition that the predicted working condition is different from the current working condition of the starting separator, taking the preset control parameter corresponding to the predicted working condition as the target control parameter; and according to the target control parameters, regulating the load regulating valve, the starting separator outlet isolation valve, the main water supply pump rotating speed, the main water supply low-load regulating valve, the main water supply pump water return regulating valve and the starting separator steam pipeline regulating valve.

In one embodiment, a computer program product is provided comprising a computer program which, when executed by a processor, performs the steps of: acquiring a first multi-dimensional time sequence corresponding to a target parameter according to operation data of a unit to be controlled in a first history period, and acquiring a modal component corresponding to the first multi-dimensional time sequence; the target parameters include a loop temperature and reactor power; constructing a first prediction model for loop temperature prediction according to the modal components corresponding to the first multi-dimensional time sequence, and constructing a second prediction model for reactor power prediction according to the modal components corresponding to the first multi-dimensional time sequence; the first prediction model and the second prediction model are obtained based on a plurality of single model combinations; acquiring a real-time sequence corresponding to a target parameter, and respectively inputting the real-time sequence into a first prediction model and a second prediction model to acquire a loop temperature prediction sequence and a reactor power prediction sequence; and obtaining a predicted working condition of the starting separator according to the loop temperature predicted sequence and the reactor power predicted sequence, and performing start-stop control on the unit to be controlled according to the predicted working condition.

Those skilled in the art will appreciate that implementing all or part of the above described methods may be accomplished by way of a computer program stored on a non-transitory computer readable storage medium, which when executed, may comprise the steps of the embodiments of the methods described above. Any reference to memory, database, or other medium used in the various embodiments provided herein may include at least one of non-volatile and volatile memory. The nonvolatile Memory may include Read-Only Memory (ROM), magnetic tape, floppy disk, flash Memory, optical Memory, high density embedded nonvolatile Memory, resistive random access Memory (ReRAM), magnetic random access Memory (Magnetoresistive Random Access Memory, MRAM), ferroelectric Memory (Ferroelectric Random Access Memory, FRAM), phase change Memory (Phase Change Memory, PCM), graphene Memory, and the like. Volatile memory can include random access memory (Random Access Memory, RAM) or external cache memory, and the like. By way of illustration, and not limitation, RAM can be in the form of a variety of forms, such as static random access memory (Static Random Access Memory, SRAM) or dynamic random access memory (Dynamic Random Access Memory, DRAM), and the like. The databases referred to in the various embodiments provided herein may include at least one of relational databases and non-relational databases. The non-relational database may include, but is not limited to, a blockchain-based distributed database, and the like. The processors referred to in the embodiments provided herein may be general purpose processors, central processing units, graphics processors, digital signal processors, programmable logic units, quantum computing-based data processing logic units, etc., without being limited thereto.

The technical features of the above embodiments may be arbitrarily combined, and all possible combinations of the technical features in the above embodiments are not described for brevity of description, however, as long as there is no contradiction between the combinations of the technical features, they should be considered as the scope of the description.

The above examples only represent a few embodiments of the present application, which are described in more detail and are not to be construed as limiting the scope of the present application. It should be noted that it would be apparent to those skilled in the art that various modifications and improvements could be made without departing from the spirit of the present application, which would be within the scope of the present application. Accordingly, the scope of protection of the present application shall be subject to the appended claims.

Claims

1. A method for controlling start-stop of a metal stack, the method comprising:

constructing a first prediction model for loop temperature prediction according to the modal components corresponding to the first multi-dimensional time sequence, and constructing a second prediction model for reactor power prediction according to the modal components corresponding to the first multi-dimensional time sequence; the first prediction model and the second prediction model are obtained based on a plurality of single models;

Acquiring a real-time sequence corresponding to the target parameter, and respectively inputting the real-time sequence into the first prediction model and the second prediction model to acquire a loop temperature prediction sequence and a reactor power prediction sequence;

and obtaining a predicted working condition of a starting separator according to the primary loop temperature predicted sequence and the reactor power predicted sequence, and performing start-stop control on the unit to be controlled according to the predicted working condition.

2. The method of claim 1, wherein the target parameters further comprise at least one of a pressurizer pressure, a pressurizer water level, a circuit pressure, a reactor coolant outlet temperature, a reactor coolant inlet temperature, a coolant flow time, a circuit flow, a circuit boron concentration, a steam generator pressure, a steam generator water level, a high pressure cylinder exhaust pressure, a primary reheater extraction pressure, a low pressure cylinder inlet and outlet superheated steam temperature, a condenser operating pressure, a feedwater temperature, and a steam reheat.

3. The method according to claim 1, wherein the obtaining the first multi-dimensional time sequence corresponding to the target parameter includes:

Acquiring an initial time sequence corresponding to each target parameter;

and acquiring the first multidimensional time sequence according to the sub-time sequence corresponding to each target parameter.

4. The method of claim 1, wherein the single model comprises a long-short-term memory LSTM model and an extreme learning machine ELM model; the constructing a first prediction model for predicting a loop temperature according to the modal components corresponding to the first multidimensional time sequence includes:

acquiring an initial weighting coefficient;

and carrying out weighted fusion on the trained LSTM model and the trained ELM model according to the initial weighting coefficient to obtain an initial prediction model, and taking the initial prediction model as the first prediction model.

5. The method of claim 4, wherein constructing a first prediction model for loop temperature prediction from the modal components corresponding to the first multi-dimensional time series sequence further comprises:

Acquiring a second multidimensional time sequence corresponding to the target parameter according to the operation data of the unit to be controlled in a second history period;

and adopting a sliding window mechanism, carrying out iterative updating on the initial prediction model according to the second multidimensional time sequence and the fusion error threshold value to obtain a target prediction model, and taking the target prediction model as the first prediction model.

6. The method of claim 1, wherein the performing start-stop control on the unit to be controlled according to the predicted operating condition comprises:

under the condition that the predicted working condition is the same as the current working condition of the starting separator, determining a target control parameter according to a reactor outlet temperature deviation value, a direct current steam generator vapor pressure deviation value and a loop outlet flow deviation value;

under the condition that the predicted working condition is different from the current working condition of the starting separator, taking a preset control parameter corresponding to the predicted working condition as a target control parameter;

and according to the target control parameters, regulating a load regulating valve, a starting separator outlet isolation valve, a main water supply pump rotating speed, a main water supply low-load regulating valve, a main water supply pump water return regulating valve and a starting separator steam pipeline regulating valve.

7. A start-stop control device for a metal stack, the device comprising:

the acquisition module is used for acquiring a first multidimensional time sequence corresponding to the target parameter according to the operation data of the unit to be controlled in the first history period and acquiring a modal component corresponding to the first multidimensional time sequence; the target parameters include a loop temperature and reactor power;

the construction module is used for constructing a first prediction model for loop temperature prediction according to the modal components corresponding to the first multi-dimensional time sequence, and constructing a second prediction model for reactor power prediction according to the modal components corresponding to the first multi-dimensional time sequence; the first prediction model and the second prediction model are obtained based on a plurality of single models;

the prediction module is used for acquiring a real-time sequence corresponding to the target parameter, inputting the real-time sequence into the first prediction model and the second prediction model respectively, and acquiring a loop temperature prediction sequence and a reactor power prediction sequence;

and the control module is used for acquiring a predicted working condition of the starting separator according to the loop temperature predicted sequence and the reactor power predicted sequence, and performing start-stop control on the unit to be controlled according to the predicted working condition.

8. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that the processor implements the steps of the method of any of claims 1 to 6 when the computer program is executed.

9. A computer readable storage medium, on which a computer program is stored, characterized in that the computer program, when being executed by a processor, implements the steps of the method of any of claims 1 to 6.

10. A computer program product comprising a computer program, characterized in that the computer program, when being executed by a processor, implements the steps of the method of any of claims 1 to 6.