Pumped storage power station simulation test method compatible with multiple PLCs and computer device

Abstract

The application discloses a pumped storage power station simulation test method compatible with multiple PLCs and a computer device. The method comprises the following steps: a test platform matched with a target power station is constructed; the test platform is integrated with a data receiving model with a general data interface, a preset symmetric residual error network model, a preset long short-term memory model and multiple PLC communication protocols; the initial spatial characteristics of the target power station are processed based on the preset symmetric residual error network model to obtain first information; the target equipment parameter information, the actual environmental factor information and the first information of the target power station are analyzed based on a data fusion model to obtain a target operation strategy; the time sequence characteristics of the target power station are processed through the preset long short-term memory model to obtain future operation state prediction information; and a guiding operation strategy is obtained based on the future operation state prediction information and the target operation strategy. The method can provide a simulation test scheme with high compatibility, high test efficiency, high test precision and low maintenance cost.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to the technical field of pumped storage power station simulation, and particularly relates to a pumped storage power station simulation test method compatible with multiple PLCs and a computer device. BACKGROUND

[0002] Pumped storage power stations are important regulating tools in power grids, which can balance power grid loads and ensure stable operation of power systems by storing and releasing energy. PLC systems play a core role in the automation control of power stations due to their reliability and flexibility. However, compatibility problems between PLC systems of different manufacturers pose challenges to the simulation testing of power stations.

[0003] Specifically, different PLC manufacturers' systems may use different communication protocols and software interfaces, resulting in poor interoperability between systems and making it difficult to achieve seamless integration and testing. Due to compatibility problems, testers need to spend a lot of time to solve the communication and integration problems between systems, which reduces the efficiency and speed of testing. As technology develops and power station needs change, traditional simulation testing schemes are difficult to adapt to new PLC systems or software versions, resulting in increased maintenance and upgrade costs. In an environment where multiple systems coexist, data integration and analysis become more complex, which may lead to inaccuracy and inconsistency of test results. Traditional simulation testing schemes cannot completely simulate the complex environment and dynamic changes of pumped storage power stations in actual operation, affecting the reliability of test results. SUMMARY

[0004] Therefore, the embodiments of the present disclosure provide a pumped storage power station simulation test method compatible with multiple PLCs and a computer device, which can solve the problems of poor compatibility, low test efficiency, poor test accuracy, high maintenance cost, etc. of traditional simulation testing schemes.

[0005] In a first aspect, the embodiments of the present disclosure provide a pumped storage power station simulation test method compatible with multiple PLCs, which adopts the following technical scheme:

[0006] A test platform matched with a target power station is constructed, and the test platform is integrated with a data receiving model, a preset symmetric residual network model, a preset long short-term memory model, and multiple PLC communication protocols. The data receiving model is integrated with a general data interface having a data conversion function;

[0007] The initial spatial features in the target power station are processed based on the preset symmetric residual network model to obtain first information;

[0008] Based on the constructed data fusion model, the target device parameter information, the actual environmental factor information and the first information of the target power station are analyzed to obtain a target operation strategy of the target device under different environmental conditions and power station layouts;

[0009] The time sequence characteristics of the target power station are processed by the preset long short-term memory model to obtain future operation state prediction information of the target power station;

[0010] Based on the future operation state prediction information and the target operation strategy, a guiding operation strategy is obtained.

[0011] Optionally, the initial spatial characteristics in the target power station are processed based on the preset symmetric residual network model to obtain the first information, including:

[0012] The key spatial characteristics of the target power station are extracted, and the first spatial characteristics are obtained by preprocessing the key spatial characteristics;

[0013] Based on the first spatial characteristics, a spatial training set and a spatial test set are obtained;

[0014] A preset symmetric residual network model is constructed;

[0015] The preset symmetric residual network model is trained based on the spatial training set, and the trained preset symmetric residual network model is tested by the spatial test set;

[0016] Based on the trained preset symmetric residual network model, the first spatial characteristics are analyzed to obtain the first information;

[0017] The first information includes actual layout of several key spatial structures of the target power station, several potential abnormal problems and layout optimization scheme.

[0018] Optionally, the key spatial characteristics include at least one of power station geographical location characteristics, power station structure layout characteristics, power station equipment distribution characteristics, distance relationship between different equipment, equipment distribution density, important equipment layout parameters, hot area information, pipe layout, unit position and control room layout.

[0019] The important equipment layout parameters include at least one of generator set parameter information, transformer parameter information and switch device parameter information.

[0020] Optionally, the preset symmetric residual network model includes:

[0021] A network structure is determined; the network structure is a symmetrically designed structure; and a basic unit in the network structure is a residual block.

[0022] A plurality of convolutional layers are configured inside the residual block; the input channel number and the output channel number of the convolutional layers are the same, and the convolution kernel size and the step are consistent; the convolutional layers are 1x1 convolution or 3x3 convolution;

[0023] An activation function is added after the convolutional layer; the activation function includes ReLU, Leaky ReLU or ELU;

[0024] A batch normalization layer is added between the convolutional layer and the activation function;

[0025] A skip connection is added between the input and the output of the residual block, and the parameters in the skip connection share the parameters inside the residual block;

[0026] One of the convolutional layers is determined as the input layer of the network;

[0027] A plurality of residual blocks are stacked to form a main part of the network, wherein the output of each residual block is used as the input of the next residual block; a pooling layer or a transition layer is arranged between different residual blocks;

[0028] An output layer is configured; the output layer is a fully connected layer with a softmax activation function or a fully connected layer with a linear activation function;

[0029] A loss function, an optimizer and a preset learning rate are determined; the loss function includes one or more of a cross-entropy loss function and a mean square error loss function; the optimizer includes one or more of Adam and SGD.

[0030] Optionally, the processing of the time sequence characteristics of the target power station by the preset long short-term memory model to obtain the future operation state prediction information of the target power station includes:

[0031] Key time sequence characteristics of the target power station are extracted, and the key time sequence characteristics are preprocessed to obtain first time sequence characteristics;

[0032] A time sequence training set and a time sequence test set are obtained based on the first time sequence characteristics;

[0033] A preset long short-term memory model is constructed;

[0034] The preset long short-term memory model is trained based on the time sequence training set, and the trained preset long short-term memory model is tested through the time sequence test set;

[0035] Based on the trained preset long short-term memory model, the future operation state prediction information of the target power station is obtained.

[0036] Optionally, the construction of the preset long short-term memory model includes:

[0037] determine a model structure; the model structure comprises an input layer, a plurality of LSTM layers and an output layer, a shape of the input layer matches a dimension of input data; the LSTM layer comprises an input gate, a forgetting gate and an output gate, each of the LSTM layers has a hidden state and a cell state, and the hidden state and the cell state of each of the LSTM layers are transmitted to a next LSTM layer; the output layer is configured with an output unit number and an activation function;

[0038] configure the LSTM hidden unit number, the learning rate, the batch size, the loss function and the optimizer.

[0039] Optionally, the key time sequence features include at least one of power load changes, energy consumption information, energy output information, internal temperature changes of the power station, and internal humidity changes of the power station.

[0040] Optionally, the guidance operation strategy is obtained based on the future operation state prediction information and the target operation strategy.

[0041] construct a manifold regularization model; the manifold regularization model is applicable to processing of different types of data, and a target manifold learning algorithm is configured in the manifold regularization model;

[0042] The first spatial feature and the first time sequence feature are used to train the manifold regularization model, and the trained manifold regularization model is used to analyze the future operation state prediction information and the target operation strategy to obtain an energy loss device area.

[0043] The guidance operation strategy is obtained based on the energy loss device area.

[0044] The guidance operation strategy includes one or more of a device optimization layout strategy, a device upgrade strategy and an energy scheduling strategy corresponding to the area.

[0045] In a second aspect, the embodiments of the present disclosure further provide a pumped storage power station simulation test system compatible with multiple PLCs, comprising:

[0046] a test platform construction module configured to construct a test platform matched with a target power station; the test platform is integrated with a data receiving model, a preset symmetric residual network model, a preset long short-term memory model and multiple PLC communication protocols, and the data receiving model is integrated with a general data interface having a data conversion function;

[0047] a spatial feature processing module configured to process initial spatial features in the target power station based on the preset symmetric residual network model to obtain first information;

[0048] The analysis module is configured to analyze the target device parameter information, the actual environmental factor information, and the first information of the target power station based on the constructed data fusion model, and obtain a target operation strategy of the target device under different environmental conditions and power station layouts.

[0049] The time sequence feature processing module is configured to process time sequence features of the target power station by using the preset long short-term memory model, and obtain future operation state prediction information of the target power station.

[0050] The acquisition module is configured to obtain a guidance operation strategy based on the future operation state prediction information and the target operation strategy.

[0051] In a third aspect, the embodiments of the present disclosure further provide a computer device, which adopts the following technical scheme:

[0052] The computer device comprises:

[0053] at least one processor; and

[0054] a memory connected with the at least one processor in communication; wherein

[0055] The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the compatible-PLC-compatible pumped storage power station simulation test method described in any one of the above.

[0056] In a fourth aspect, the embodiments of the present disclosure further provide a computer readable storage medium, which stores computer instructions for causing a computer to perform the compatible-PLC-compatible pumped storage power station simulation test method described in any one of the above.

[0057] In a fifth aspect, the embodiments of the present disclosure further provide a computer program product, which comprises computer programs / instructions, and the computer programs / instructions are executed by a processor to implement the steps of the method described in any one of the above.

[0058] The pumped storage power station simulation test method compatible with multiple PLCs disclosed in the application can adapt to different models and brands of PLCs by integrating multiple PLC communication protocols, so that the test platform can be widely applied to various pumped storage power stations, and the compatibility and universality are improved. Through the processing of the data receiving model, the symmetric residual network model and the long short-term memory model, the spatial features and time sequence features of the power station can be analyzed in detail, and the accuracy and reliability of data processing are improved. Through the analysis of the data fusion model, the target device parameters, environmental factors and other influencing factors can be more comprehensively considered to provide more reasonable operation strategies for the power station, and the operation efficiency and stability of the power station equipment are improved. Through the long short-term memory model, the future operation state of the power station is predicted, potential problems and risks can be found in advance, and corresponding prevention and adjustment measures can be formulated for the power station, and the safety and reliability of the power station are improved. Combined with the future operation state prediction information and the target operation strategy, a guiding operation strategy is provided for the power station, the operation of the power station is more intelligent and optimized, the overall operation efficiency and economy are improved, and more powerful support is provided for the stable operation of the power station.

[0059] The above description is only a summary of the technical solutions of the present disclosure. In order to more clearly understand the technical means of the present disclosure, the content of the specification can be implemented, and in order to make the above and other purposes, features and advantages of the present disclosure more obvious and easy to understand, the following preferred embodiments are described in detail below, and the accompanying drawings are described as follows. BRIEF DESCRIPTION OF DRAWINGS

[0060] In order to more clearly illustrate the technical solutions of the embodiments of the present disclosure, the following will briefly introduce the drawings needed to be used in the embodiments. Obviously, the drawings in the following description are only some embodiments of the present disclosure, and other drawings can be obtained by those skilled in the art without creating laborious work.

[0061] Figure 1 The flowchart of the pumped storage power station simulation test method compatible with multiple PLCs provided by the embodiments of the present disclosure.

[0062] Figure 2 The flowchart of the first information obtaining method provided by the embodiments of the present disclosure.

[0063] Figure 3 The flowchart of the construction method of the preset symmetric residual network model provided by the embodiments of the present disclosure.

[0064] Figure 4 The flowchart of the target power station future operation state prediction information obtaining method provided by the embodiments of the present disclosure.

[0065] Figure 5A flowchart of a construction method of a preset long short-term memory model provided by an embodiment of the present disclosure is shown.

[0066] Figure 6 A flowchart of an obtaining method of a guiding operation strategy provided by an embodiment of the present disclosure is shown.

[0067] Figure 7 A principle block diagram of a pumped storage power station simulation test system compatible with multiple PLCs provided by an embodiment of the present disclosure is shown.

[0068] Figure 8 A structure diagram of a computer device provided by an embodiment of the present disclosure is shown. DETAILED DESCRIPTION

[0069] The embodiments of the present disclosure will be described in detail below with reference to the drawings.

[0070] It should be apparent that the following describes embodiments of this disclosure by way of specific examples, and that one skilled in the art can readily derive other advantages and effects from the content disclosed in this specification. Obviously, the described embodiments are only a part of the embodiments of this disclosure, rather than all the embodiments. This disclosure can also be implemented or applied by other different specific embodiments, and various modifications or changes can be made to the details in this specification without departing from the spirit of this disclosure. It should be noted that the following embodiments and features in the embodiments can be combined with each other without conflict. Based on the embodiments in this disclosure, all other embodiments obtained by one of ordinary skill in the art without creative labor are within the scope of protection of this disclosure.

[0071] It should be noted that the various aspects of the embodiments described below are within the scope of the appended claims. It should be apparent that the aspects described herein can be embodied in a wide variety of forms and that any specific structure and / or function described herein is merely illustrative. Based on the teachings provided herein one skilled in the art will appreciate that one aspect described herein can be implemented independently of any other aspects and that the various aspects described herein can be combined in various ways. For example, an apparatus can be implemented or a method can be practiced using any number of the aspects set forth herein. In addition, such an apparatus can be implemented or such a method can be practiced using other structure and / or functionality in addition to or other than one or more of the aspects set forth herein.

[0072] It should also be noted that the drawings provided in the following embodiments are only to illustrate the basic concept of the present disclosure in a schematic manner, and only show the components related to the present disclosure in the drawings, not drawn according to the number, shape and size of the components in actual implementation, and the shape, number and ratio of each component in actual implementation can be arbitrarily changed, and the layout pattern of the components can be more complex.

[0073] Additionally, in the following description, specific details are provided to facilitate thorough understanding of examples. However, one skilled in the relevant art will appreciate that the described aspects can be practiced without these specific details.

[0074] Referring to Figure 1 The first aspect of the present application discloses a pumped storage power station simulation test method compatible with multiple PLCs, comprising:

[0075] S100, a test platform matching the target power station is constructed; the test platform integrates a data receiving model, a preset symmetric residual network model, a preset long short-term memory model, and multiple PLC communication protocols, and the data receiving model integrates a general data interface with data conversion function.

[0076] Specifically, a test platform is built, integrating a data receiving model, a preset symmetric residual network model, a preset long short-term memory model, and multiple PLC communication protocols. The data receiving model should have a general data interface with data conversion function to adapt to different types of data input, i.e. it can receive data from different types of PLCs and perform effective processing and conversion, enabling data sharing and interaction with other models and systems, ensuring the efficiency and flexibility of data conversion and integration, and ensuring smooth data transmission and sharing.

[0077] Further, the data receiving model can include a data receiving unit, a data processing unit, and a data output unit. The data receiving unit is used to collect real-time data from different PLCs, the data processing unit is used to perform data conversion and cleaning, and the data output unit can pass the processed data to other models for further analysis.

[0078] The test platform can be compatible with multiple PLC devices, support different communication protocols, and ensure efficient data transmission and stable communication.

[0079] Common PLC communication protocols such as Modbus, Profibus, Ethernet / IP, etc. are integrated, and communication and data exchange with different PLC devices are realized through adapters or configuration files, etc.

[0080] In this embodiment, the general data interface can parse the original data into a general data format, convert the parsed data into a unified data exchange format, and package the converted data into a transmittable format such as JSON string or XML document, i.e. it can realize the mutual conversion and exchange of different PLC data formats.

[0081] By constructing a test platform matching the target power station, the simulation of the test environment and the integration of data can be realized, providing a perfect foundation for subsequent simulation testing, receiving, processing and analyzing different types of data, and providing reliable support for subsequent simulation testing and decision-making. Overall, such a test platform can improve the intelligent level and operation efficiency of the system, and provide important technical support for power station operation optimization and performance improvement.

[0082] S200, processing the initial spatial features in the target power station based on a preset symmetric residual network model to obtain first information.

[0083] The symmetric residual network model can effectively process spatial features and extract key information from data, providing useful references for subsequent analysis and decision-making.

[0084] S300, analyzing the target device parameter information, actual environmental factor information and first information of the target power station based on the constructed data fusion model to obtain the target operation strategy of the target device under different environmental conditions and power station layouts.

[0085] The data fusion model can effectively integrate data from different sources, help analyze the best operation strategy of the target device, and improve the performance and efficiency of the power station.

[0086] The target device parameter information can include the operating parameters of the pump unit, such as power, flow, efficiency curve, etc.

[0087] The actual environmental factor information can include environmental factors around the power station, such as water level, temperature, humidity, etc.

[0088] S400, processing the time sequence features of the target power station through a preset long short-term memory model to obtain future operation state prediction information of the target power station.

[0089] The long short-term memory model can effectively capture the long-term dependence relationship in time sequence data, improve the accurate prediction of future operation state, and help develop a reasonable operation strategy.

[0090] S500, obtaining a guidance operation strategy based on the future operation state prediction information and the target operation strategy.

[0091] By combining the future operation state prediction information and the target operation strategy, the guidance operation strategy is obtained to optimize the operation and performance of the power station. By developing a guidance operation strategy, intelligent management and optimization of the power station can be realized, energy utilization efficiency and equipment performance can be improved, and operation cost can be further reduced.

[0092] Specifically, by monitoring the changes in various parameters and the operating state of the power station in real time, adjustments can be made in a timely manner based on predictive information and target strategies. For example, when a load fluctuation in the power station is predicted, real-time adjustments to equipment parameters or load scheduling can be made to ensure stable operation of the power station.

[0093] Based on future state prediction information and target operating strategies, emergency response plans can be developed to prepare for possible emergencies. For example, in anticipation of future weather changes that may cause water level fluctuations, appropriate emergency operation procedures can be set in advance to ensure that the power station can quickly respond in emergency situations.

[0094] Based on actual monitoring data and operating results, the operating strategies of the power station can be continuously evaluated and optimized to improve the efficiency and performance of the power station and achieve a more sustainable operating mode.

[0095] Through the implementation of the above guiding operating strategies, power station managers can more scientifically and intelligently manage and operate the power station to cope with various complex operating conditions and challenges, maximize the efficiency, flexibility and reliability of the power station, and promote the effective use of clean energy and the development of intelligent energy systems.

[0096] The pumping storage power station simulation test method disclosed in the present application is compatible with multiple PLCs. By integrating multiple PLC communication protocols, it can adapt to different models and brands of PLCs, making the test platform widely applicable to various pumping storage power stations and improving compatibility and versatility. Through the processing of data receiving models, symmetric residual network models and long short-term memory models, the spatial features and time sequence features of the power station can be analyzed in detail, improving the accuracy and reliability of data processing. Through the analysis of the data fusion model, the target device parameters, environmental factors and other influencing factors can be more comprehensively considered to provide more reasonable operating strategies for the power station, improving the operating efficiency and stability of the power station equipment. Through the long short-term memory model to predict the future operating state of the power station, potential problems and risks can be discovered in advance to help the power station develop appropriate prevention and adjustment measures, improving the safety and reliability of the power station. Combined with future operating state prediction information and target operating strategies, the power station is provided with guiding operating strategies, making the operation of the power station more intelligent and optimized, improving the overall operating efficiency and economy.

[0097] In summary, this pumping storage power station simulation test method provides comprehensive and detailed analysis and guidance for power station operation through the integration of multiple models, effectively improving the level of power station operation management, improving the operating efficiency and reliability of the power station, and reducing operating risks and costs.

[0098] Reference Figure 2The method for obtaining the first information specifically comprises the following steps:

[0099] In S210, key spatial features of the target power station are extracted, and the key spatial features are preprocessed to obtain first spatial features.

[0100] Specifically, the key spatial features can be extracted from the power station data using data mining and pattern recognition techniques, and preprocessed, including data cleaning, feature extraction, standardized data format, noise removal, missing value filling, etc., to facilitate network processing.

[0101] The method disclosed in this step can ensure that the extracted spatial features are accurate and reliable, and provide a high-quality data basis for subsequent model training and analysis.

[0102] The key spatial features include at least one of power station geographic location features, power station structure layout features, power station equipment distribution features, distance relationships between different equipment, equipment distribution density, important equipment layout parameters, hotspot area information, pipeline layout, unit location, and control room layout.

[0103] The important equipment layout parameters preferably include at least one of generator set parameter information, transformer parameter information, and switch device parameter information.

[0104] In S220, a spatial training set and a spatial test set are obtained based on the first spatial features.

[0105] Specifically, the data in the random first spatial features can be divided into a training set and a test set according to a certain proportion, for example, a training set and a test set can be obtained by dividing according to a proportion of 7:3 or 8:2.

[0106] Alternatively, if there is a class imbalance in the data set (i.e., the first spatial features), stratified sampling can be considered to ensure that the class distribution in the training set and the test set is similar, which can better evaluate the performance of the model on different classes.

[0107] In this application, a suitable division scheme can be selected according to the specific situation, and the data distribution of the training set and the test set can be ensured to represent the entire data set, so as to effectively evaluate the generalization ability of the model.

[0108] In S230, a preset symmetric residual network model is constructed.

[0109] In S240, the preset symmetric residual network model is trained based on the spatial training set, and the trained preset symmetric residual network model is tested through the spatial test set.

[0110] The symmetric residual network model is trained using a spatial training set, and the trained model is verified and optimized using a spatial test set. Through training and testing, the model can accurately capture the correlation between the spatial features of the power station, improve the generalization ability and prediction accuracy of the model, and improve the accuracy of feature recognition and analysis.

[0111] In S250, the first spatial feature is analyzed based on the trained preset symmetric residual network model to obtain first information.

[0112] The first information includes the actual layout of several key spatial structures of the target power station, several potential abnormal problems, and layout optimization schemes.

[0113] The trained preset symmetric residual network model is used to analyze the first spatial feature to obtain information such as the actual layout of the key structure of the target power station, potential abnormal problems, and optimization schemes. By analyzing the first information, the structure and existing problems of the power station can be comprehensively understood, providing an important reference for formulating further optimization plans and improvement schemes, and improving the efficiency and reliability of the power station.

[0114] The scheme disclosed in this embodiment first extracts the key spatial features of the target power station through the preset symmetric residual network model, and pre-processes these key spatial features to obtain the first spatial feature. This step can help extract the features that have the greatest impact on the performance and structure of the power station, providing a basis for subsequent analysis and optimization. Then, based on the first spatial feature, the spatial training set and the spatial test set are divided according to appropriate methods to train and test the model. This step can ensure the generalization ability of the model on unknown data. The preset symmetric residual network model is constructed, which can better capture the complex relationships between spatial features and help improve the performance of the model. The preset symmetric residual network model is trained using the spatial training set, and the model is tested using the spatial test set to evaluate its performance. This step can help verify the effectiveness and accuracy of the model. Finally, based on the trained preset symmetric residual network model, the first spatial feature is analyzed to obtain first information, including the actual layout of the key spatial structure of the target power station, potential abnormal problems, and layout optimization schemes. These information can help optimize the structure and performance of the power station, and improve the efficiency and reliability of the power station.

[0115] In this scheme, the key spatial features are extracted and pre-processed, which helps to improve the understanding and abstraction ability of the model for power station data. Based on the preset symmetric residual network model, the complex relationships between features can be better captured, and the performance of the model can be improved. Through the division of the training and test sets, the generalization ability of the model can be effectively evaluated. By analyzing the first spatial feature, the problems existing in the power station can be found and optimization schemes can be proposed, thereby improving the overall efficiency and reliability of the power station.

[0116] Referring to Figure 3 , the method for constructing the preset symmetric residual network model specifically comprises the following steps:

[0117] S241, determining a network structure;

[0118] The network structure is a symmetrically designed structure.

[0119] The basic unit in the network structure is a residual block.

[0120] In this step, the network structure is a symmetrically designed structure, and the symmetry of this structure is conducive to the learning and understanding of the model; the residual block is used as the basic unit of the network, which helps to solve the problems of gradient vanishing and gradient explosion in the deep neural network and accelerate the model training.

[0121] S242, configuring a plurality of convolutional layers inside the residual block; the input channel number and the output channel number of the convolutional layers are the same, and the convolution kernel size and the step length are consistent.

[0122] The convolutional layer is 1x1 convolution or 3x3 convolution.

[0123] In this step, the input and output channel numbers of the convolutional layer are the same, and the convolution kernel size and the step length are consistent, so that the size of the feature map can be kept stable; the 1x1 or 3x3 convolution kernel is used to help capture features of different scales.

[0124] S243, adding an activation function after the convolutional layer; the activation function includes ReLU, Leaky ReLU or ELU.

[0125] The use of activation functions such as ReLU, Leaky ReLU or ELU can introduce nonlinearity and increase the expression capacity of the model.

[0126] S244, adding a batch normalization layer between the convolutional layer and the activation function.

[0127] Batch normalization can accelerate model convergence, reduce the problem of gradient vanishing, and enhance the generalization ability of the model.

[0128] S245, adding a skip connection between the input and the output of the residual block, and the parameters in the skip connection are shared with the parameters inside the residual block.

[0129] The skip connection helps to alleviate the problem of gradient vanishing and enables information to be transmitted more quickly; parameter sharing can improve the parameter efficiency of the model.

[0130] S246, determining a convolutional layer as the input layer of the network;

[0131] The plurality of residual blocks are stacked to form a network main part, wherein the output of each residual block is taken as the input of the next residual block; a pooling layer or a transition layer is arranged between different residual blocks.

[0132] The pooling layer or the transition layer can reduce the dimension of the feature map, reduce the calculation complexity, and improve the calculation efficiency of the model.

[0133] S247, configuring an output layer; the output layer is a fully connected layer with a softmax activation function or a fully connected layer with a linear activation function.

[0134] The output layer uses a fully connected layer with a softmax activation function or a fully connected layer with a linear activation function, which can adapt to different types of tasks.

[0135] S248, determining a loss function, an optimizer, and a preset learning rate;

[0136] The loss function includes one or more of a cross-entropy loss function and a mean square error loss function; the optimizer includes one or more of Adam and SGD.

[0137] The loss function suitable for the task is selected, such as a cross-entropy loss function for a classification task and a mean square error loss function for a regression task; a suitable optimizer is selected, such as Adam and SGD, which can accelerate the convergence speed and improve the model performance; the preset learning rate can effectively control the update speed of the model parameters and avoid oscillation in the training process.

[0138] In this embodiment, the symmetric design is beneficial to simplify the network structure, improve the interpretability and training efficiency of the network; the residual block can alleviate the gradient vanishing problem and accelerate the model convergence speed; the stacking of multiple convolutional layers can increase the non-linear expression ability of the network; the convolution kernel with the same input and output channel number can keep the size of the feature map stable and reduce information loss; the activation function introduces nonlinearity and enhances the expression ability of the network; the batch normalization layer helps to accelerate the model training, improve the stability and generalization ability of the model; the skip connection helps to alleviate the gradient vanishing problem and accelerate information transmission; parameter sharing can improve the utilization rate of parameters and reduce overfitting; by stacking multiple residual blocks, the network depth can be increased and the feature abstraction ability can be improved; the pooling layer or the transition layer can reduce the feature dimension and improve the calculation efficiency; the design of the output layer selects appropriate activation functions according to task requirements, which can better adapt to different types of tasks; selecting a suitable loss function can evaluate the model performance and guide the model training; a suitable optimizer can accelerate the optimization process of the model parameters and improve the model convergence speed; setting a suitable learning rate can avoid oscillation in the training process and ensure the convergence performance of the model.

[0139] In summary, the overall scheme comprehensively utilizes a series of techniques and optimization methods of deep learning, which can improve the performance of the model, speed up the training of the model, enhance the stability and generalization ability of the model, so as to better apply to practical problems and provide more effective solutions.

[0140] Further, the constructed preset symmetric residual network model also contains a pooling layer, which is used to reduce the spatial dimension of the feature map, reduce the calculation amount, and increase the receptive field at the same time. The receptive field refers to the response range of neurons to input data. Through the pooling layer, the receptive field can be expanded, which helps the network to better capture the features of the input data. In this way, the complexity of model training can be reduced and the computing efficiency can be improved while maintaining the effective information.

[0141] Further, a Dropout layer can also be added at an appropriate position in the preset symmetric residual network model to prevent overfitting. Specifically, L2 regularization can be applied to penalize large weights and improve the generalization ability of the model.

[0142] Reference Figure 4 The method for obtaining the future operation state prediction information of the target power station specifically includes the following steps:

[0143] S410, extracting key time sequence features of the target power station, preprocessing the key time sequence features to obtain first time sequence features.

[0144] Specifically, appropriate feature extraction techniques (such as sliding window method, Fourier transform, etc.) can be used to extract key time sequence features of the target power station, and then these features can be preprocessed (such as standardization, normalization, etc.) to obtain first time sequence features.

[0145] In this step, extracting key features can reduce data dimension and alleviate model complexity; preprocessing can make data easier to be processed by the model and improve model performance.

[0146] Preferably, the key time sequence features include at least one of power load change, energy consumption information, energy output information, internal temperature change of the power station, internal humidity change of the power station, reservoir water level change, and environmental conditions.

[0147] Specifically, these time sequence data can be collected from various monitoring systems of the power station to ensure the integrity and accuracy of the data.

[0148] S420, obtaining a time sequence training set and a time sequence test set based on the first time sequence features.

[0149] Specifically, the extracted first time sequence features can be divided into a training set and a test set according to a certain division rule, for training and testing of the model; through this step, the generalization ability of the model can be evaluated, and the performance of the model on unknown data can be verified.

[0150] S430, a preset long short-term memory model is constructed.

[0151] The preset long short-term memory model is suitable for processing time series data and can capture long-term dependencies in the data, making it suitable for time series prediction tasks.

[0152] S440, the preset long short-term memory model is trained based on the time series training set, and the trained preset long short-term memory model is tested based on the time series test set.

[0153] The preset long short-term memory model is trained using the time series training set, and then the model is tested on the time series test set to evaluate the performance and accuracy of the model. Through training and testing, the prediction ability of the model can be evaluated, the parameters of the model can be optimized, and the generalization ability of the model can be improved.

[0154] S450, based on the trained preset long short-term memory model, the future operating state prediction information of the target power station is obtained.

[0155] Using the trained LSTM model, the future time series data is predicted to obtain the future operating state prediction information of the target power station, which can accurately predict the future state of the target power station, help to discover potential problems in advance, and perform preventive maintenance to improve the operating efficiency and reliability of the power station.

[0156] In this embodiment, the trained preset long short-term memory model can accurately predict the future operating state of the target power station, helping operators make decisions and adjustments in advance; the automated prediction model can reduce labor costs and quickly and effectively provide prediction information for the future state of the power station; through accurate future state prediction, potential problems or abnormal situations can be discovered in advance, and appropriate warning and preventive measures can be taken to reduce operational risks; through accurate prediction of future operating states, power station operation and maintenance plans can be optimized to improve the operating efficiency and reliability of the power station; the model prediction results can provide decision support for management, enabling them to better understand the current state and future trend of the power station and develop decision-making strategies; through training and testing of the time series training set, model parameters can be continuously optimized to improve prediction accuracy, making the model more in line with actual conditions.

[0157] In summary, the overall scheme processes the time sequence features of the target power station by combining the LSTM model, which can provide effective decision support for power station management and operation, improve the operating efficiency and reliability of the power station, reduce potential risks, and bring better economic benefits and management effects.

[0158] Referring to Figure 5 The method for constructing the preset long short-term memory model specifically comprises the following steps:

[0159] S431, determining a model structure;

[0160] The model structure comprises an input layer, a plurality of LSTM layers and an output layer, the shape of the input layer matches the dimension of input data; the LSTM layer comprises an input gate, a forgetting gate and an output gate, each LSTM layer has a hidden state and a cell state, and the hidden state and the cell state of each LSTM layer are transmitted to the next LSTM layer; the output layer is configured with an output unit number and an activation function.

[0161] Specifically, the feature dimension of the input layer matches the dimension of the input data, which can ensure that the data can be smoothly transmitted into the model; the model comprises a plurality of LSTM layers to capture time sequence information, and finally an output layer is used to output a prediction result; the specific structure of the LSTM can effectively capture the long-term dependence of time sequence data; the hidden state and the cell state can transmit information and model the data; the output unit number of the output layer should be determined according to the task requirement, and the activation function can be selected according to the task type, such as sigmoid for binary classification and softmax for multi-classification.

[0162] Further, the unit number of each LSTM layer is usually 12, 256, 512, etc.

[0163] S432, configuring the LSTM hidden unit number, the learning rate, the batch size, the loss function and the optimizer.

[0164] The configured LSTM hidden unit number determines the capacity size of the model, and affects the learning ability and accuracy of the model.

[0165] The configuration of the learning rate can control the update speed of the model parameters in the training process, which will not cause oscillation due to being too high, nor slow convergence due to being too low. Further, learning rate decay or learning rate scheduler can be considered.

[0166] The batch size defines the number of samples used in each iteration during the training process, which affects the training speed and stability of the model. Specifically, the batch size can be determined according to the available computing resources, and common choices are 32, 64, 128, etc.

[0167] The loss function is used to evaluate the gap between the model output and the true label, and a suitable loss function for the task can be selected to guide the model optimization. Specifically, the loss function can be selected according to the task type, such as mean square error (MSE) for regression and cross-entropy for classification.

[0168] Selecting a suitable optimizer to update the model parameters, commonly including Adam, SGD, RMSprop, etc., can accelerate convergence and improve model performance.

[0169] In this embodiment, a suitable LSTM model structure can be designed according to task requirements and data characteristics, which can effectively capture time series information and improve prediction accuracy. By reasonably setting the number of hidden units, learning rate and other parameters, the model performance can be optimized, the training process can be accelerated, and the prediction accuracy can be improved.

[0170] Further, in the constructed preset long short-term memory model, dropout or L2 regularization can also be used to prevent overfitting.

[0171] Referring to Figure 6 The method for obtaining the guiding operation strategy specifically includes the following steps:

[0172] S510, a manifold regularization model is constructed; the manifold regularization model is suitable for processing different types of data, and the manifold regularization model is configured with a target manifold learning algorithm.

[0173] Specifically, a manifold regularization model suitable for processing different types of data is constructed, which is configured with a target manifold learning algorithm. At the same time, a suitable manifold learning algorithm is selected to learn the structure of the data and highlight important features in the data. These algorithms can map high-dimensional data to low-dimensional manifold space to better understand the characteristics and rules of the data.

[0174] In the manifold regularization model, key parameters of the model such as regularization strength and neighborhood size can be set to adjust the sensitivity of the model to the internal structure of the data.

[0175] In this step, the manifold regularization model is suitable for processing different types of data, which helps to discover the potential structure inside the data; through manifold learning, important features in the data can be effectively extracted to help subsequent analysis and decision making.

[0176] S520, the first spatial feature and the first time series feature are used to train the manifold regularization model, and the trained manifold regularization model is used to analyze the future operation state prediction information and the target operation strategy to obtain the energy loss equipment area.

[0177] Specifically, the first spatial feature and the first time series feature are used to train the manifold regularization model, so that the model can learn the structure of the data, and the trained model is used to analyze the future operation state prediction information and the target operation strategy.

[0178] For example, assuming that in a certain area of the power plant, due to unreasonable device layout, there is a large loss of energy during transmission, by analyzing the fused data (i.e., analyzing the future running state prediction information and the target running strategy) through the trained manifold regularization model, the problem area can be pointed out.

[0179] In this step, the future running state prediction information and the target running strategy are combined to more comprehensively analyze the characteristics and potential laws of the data; by analyzing the relationship between various data, problems and optimization spaces in device operation can be found.

[0180] S530, obtaining a guiding running strategy based on the energy loss device area, the guiding running strategy including one or more of the device optimization layout strategy, the device upgrade strategy, and the energy scheduling strategy corresponding to the area.

[0181] Specifically, according to the energy loss device area obtained by analysis, a corresponding running strategy is formulated, including a device optimization layout strategy, a device upgrade strategy, an energy scheduling strategy, etc., that is, in this step, a targeted running strategy can be formulated according to the actual situation and analysis results, which is more targeted and effective; by optimizing device layout, upgrading devices, adjusting energy scheduling, etc., device utilization can be improved, energy loss can be reduced, and overall operation efficiency can be improved.

[0182] Through the obtained guiding running strategy, the power plant manager can re-plan the device layout of the area or upgrade the related devices to reduce energy loss. After implementing the optimization measures, the energy consumption and output data are continuously monitored to evaluate the effect of the optimization measures.

[0183] The method disclosed in this embodiment formulates a guiding running strategy according to the situation of the energy loss device area by analyzing the future running state and the target running strategy, which can be tailored according to the actual situation to improve the targeting and effectiveness; through the device optimization layout strategy in the guiding running strategy, the device layout can be reasonably planned to improve the collaborative efficiency between devices and reduce energy waste; the device upgrade strategy in the guiding running strategy can upgrade or replace the devices according to the actual situation to improve the performance and operation efficiency of the devices; the formulation of the energy scheduling strategy can help reasonably allocate energy resources, improve energy utilization efficiency, and reduce energy costs; through the implementation of the overall scheme, the energy loss device area can be found and improved, the device utilization rate can be improved, the energy loss can be reduced, and the operating costs can be reduced; the measures of optimizing device layout, upgrading devices, and reasonably scheduling energy can improve the overall operation efficiency, reduce the failure rate, and enhance the stability and reliability of device operation; the implementation of the guiding running strategy can reduce human intervention, improve the automation level, and reduce management costs.

[0184] In summary, the overall solution provides more targeted and guiding operation strategies for operation management through scientific data analysis and model construction, which helps to optimize equipment operation state, improve energy utilization efficiency, reduce operation cost, and enhance overall operation efficiency and management level.

[0185] In this embodiment, the manifold regularization technique effectively integrates the results of spatial analysis and time series analysis, providing a comprehensive and detailed operation simulation view for pumped storage power stations. This comprehensive view helps to deeply understand the operation mechanism of the power station and guide the optimization and improvement of the power station. Through the above steps, we can effectively combine the spatial and temporal characteristics of pumped storage power stations to provide more accurate and comprehensive simulation tests for the operation of the power station, which helps to improve the efficiency and reliability of the power station operation, and provides important decision support for the management and maintenance of the power station.

[0186] Further, the future operation state prediction information can include: predicting the trend of changes in power station output power, water level, pressure and other parameters in the future time period, predicting the load condition of the power station in the future time period and the corresponding power grid response, and the predictable operation state of different environmental conditions, power station layout, etc.

[0187] The target operation strategy can include: determining the optimal equipment parameter configuration according to the actual environmental factors and power station layout, setting the target equipment operation mode, optimizing the control strategy to achieve the best power station operation benefit, designing the corresponding operation scheme for different environmental conditions, including load scheduling, optimization of energy storage, etc.

[0188] For example, according to the long-term operation goal and short-term demand of the power station, the operation goal can be set, such as maximizing energy conversion efficiency, minimizing operation cost, etc.

[0189] The equipment operation mode can be set, such as giving priority to power generation during the peak load period of the power grid and pumping during the trough period.

[0190] The guiding operation strategy can include: adjusting the power station operation parameters according to the future state prediction information, making preparations in advance, setting the operation goal and equipment operation mode according to the target operation strategy, monitoring the power station operation state in real time, and making adjustments according to the prediction information and target strategy to maximize the efficiency and performance of the power station.

[0191] For example, it can cooperate with the power grid operator to conduct load scheduling according to the real-time demand and predicted load of the power grid. During the peak demand period of the power grid, priority is given to power generation; during the trough period, pumping energy storage is performed.

[0192] The risk of the operation strategy can be evaluated, including equipment failure, environmental changes, etc., and corresponding risk mitigation measures can be developed; emergency plans can be established to deal with unpredictable events such as extreme weather or equipment failure; optimization algorithms such as linear programming, dynamic programming, etc. can be used to optimize the operation strategy of the power station; multi-objective optimization strategies can be developed considering cost, efficiency and environmental impact.

[0193] In this application, based on the future state prediction information and the target operation strategy, an effective guidance operation strategy can be developed to achieve efficient, stable and sustainable development of the target power station.

[0194] Example 1: Adaptation of Siemens and Schneider PLC simulation test, the specific steps are as follows:

[0195] 1) Construct a test platform for adapting Siemens and Schneider PLC, integrate data receiving model, symmetric residual network model and long short-term memory model.

[0196] 2) Based on the symmetric residual network model, process the initial spatial features to obtain the first information of the power station.

[0197] 3) Data fusion analyzes the parameter information, environmental factors and first information of Siemens and Schneider PLC system, and develops a target operation strategy.

[0198] 4) Use the long short-term memory model to process the time sequence features and predict the future operation state of the power station.

[0199] 5) According to the prediction information and the target operation strategy, develop a guidance operation strategy, test the compatibility and operation effect of different PLC systems.

[0200] The test goal of this embodiment is to verify the effectiveness of this scheme in adapting to different PLC systems, optimize the operation strategy of the power station, and improve the performance and efficiency of the power station.

[0201] Through this example, the compatibility test of Siemens and Schneider PLC system can be realized.

[0202] Example 2: Simulation test under different load conditions, the specific steps are as follows:

[0203] 1) Set the simulation parameters under high load and low load conditions, including load size, environmental factors, etc.

[0204] 2) Run the simulation test to obtain the operation state data of the power station under different load conditions.

[0205] 3) Compare and analyze the simulation results with the actual data to evaluate the accuracy and reliability of the simulation model.

[0206] The test objective of this example is to verify the simulation performance under different load conditions, providing a reference for the power station to adapt to different load situations in actual operation.

[0207] This example demonstrates the simulation results of power station operation under high and low load conditions and compares them with actual data for analysis.

[0208] Example 3: System robustness testing using simulated faults, with the following specific steps:

[0209] 1) Simulate different system fault conditions such as equipment failure, communication interruption, etc., to trigger abnormal situations in the power station PLC system.

[0210] 2) Monitor the system's response to faults, record the fault handling process, and analyze the system's flexibility and robustness.

[0211] 3) Based on the fault test results, adjust and optimize the system to improve the efficiency and recovery ability of the system in handling faults.

[0212] The test objective of this example is to test the system's robustness through simulated faults, evaluate the reliability and stability of the power station PLC system, provide improvement suggestions for system design and operation, and ensure that the power station can still operate efficiently in the face of various abnormal situations.

[0213] In this example, it is proved that the method disclosed in this application can test the response ability and fault handling efficiency of the power station PLC system by simulating different system fault conditions.

[0214] Example 4: Real-time optimization scheduling of intelligent energy management system, with the following specific steps:

[0215] 1) Establish an intelligent energy management system, integrating power station data, energy market information, and user demand.

[0216] 2) Real-time monitoring of power station operation status and environmental conditions, collecting real-time data and transmitting to the system.

[0217] 3) Based on real-time data and future prediction information, optimize the scheduling of power station operation parameters, including load scheduling, energy storage optimization, etc.

[0218] 4) Implement a guidance operation strategy, adjust the power station operation mode according to the latest situation to ensure efficient and stable operation of the power station.

[0219] The test objective of this example is to improve the efficiency and performance of the power station through real-time optimization scheduling, reduce energy costs, meet user demand, and interact well with the power grid to promote the development of intelligent energy systems.

[0220] Example 5: Emergency response and recovery capability test, the specific steps are as follows:

[0221] 1) Set up simulation test scenarios under different emergency situations, such as sudden load increase, equipment failure, etc.

[0222] 2) Trigger test scenarios to evaluate the emergency response speed, response capability and recovery efficiency of the power station.

[0223] 3) Analyze test results, identify existing problems and bottlenecks, and propose improvement suggestions and optimization measures.

[0224] The test goal of this example is to evaluate the emergency response and recovery capability of the power station under emergency situations, improve the emergency response plan and mechanism, and ensure that the power station can quickly respond and recover under emergency situations.

[0225] Through the above examples, the application effect and practical value of the simulation test method disclosed in the application in different aspects can be comprehensively verified, providing guidance and reference for the construction and operation of intelligent energy systems, and promoting the efficient use and sustainable development of smart grids and clean energy.

[0226] Referring to Figure 7 , the second aspect of the application discloses a pumped storage power station simulation test system 10 compatible with multiple PLCs, comprising:

[0227] A test platform construction module 11 is used to construct a test platform matched with the target power station; the test platform integrates a data receiving model, a preset symmetric residual network model, a preset long short-term memory model, and multiple PLC communication protocols, and the data receiving model integrates a general data interface with data conversion function;

[0228] A spatial feature processing module 12 is used to process the initial spatial features in the target power station based on the preset symmetric residual network model to obtain first information;

[0229] An analysis module 13 is used to analyze the target device parameter information, actual environmental factor information and first information of the target power station based on the constructed data fusion model to obtain the target operation strategy of the target device under different environmental conditions and power station layout;

[0230] A time sequence feature processing module 14 is used to process the time sequence features of the target power station through the preset long short-term memory model to obtain the future operation state prediction information of the target power station;

[0231] An acquisition module 15 is used to obtain the guiding operation strategy based on the future operation state prediction information and the target operation strategy.

[0232] A computer device according to an embodiment of the present disclosure includes a memory and a processor. The memory is configured to store non-transitory computer readable instructions. Specifically, the memory can include one or more computer program products, which can include various forms of computer readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory may, for example, include random access memory (RAM), cache memory, and / or the like. The non-volatile memory may, for example, include read-only memory (ROM), hard disk, flash memory, and / or the like.

[0233] The processor can be a central processing unit (CPU) or other form of processing unit that has data processing and / or instruction execution capabilities, and can control other components in the computer device to perform desired functions. In one embodiment of the present disclosure, the processor is configured to execute the computer readable instructions stored in the memory, so that the computer device performs all or part of the steps of the simulation test method for pumped storage power station compatible with multiple PLCs according to the aforementioned embodiments of the present disclosure.

[0234] Those skilled in the art will understand that, in order to solve the technical problem of how to obtain a good user experience effect, the embodiment can also include well-known structures such as a communication bus, an interface, and the like, which should also be included in the protection scope of the present disclosure.

[0235] As Figure 8 A structural schematic diagram of a computer device according to an embodiment of the present disclosure is shown. It shows a structural schematic diagram suitable for implementing the computer device in the embodiment of the present disclosure. Figure 8 The computer device shown is merely an example and should not impose any limitation on the functions and use range of the embodiment of the present disclosure.

[0236] As Figure 8 As shown, the computer device can include a processor (such as a central processing unit, a graphics processing unit, and the like), which can perform various appropriate actions and processes according to programs stored in a read-only memory (ROM) or loaded into a random access memory (RAM) from a storage device. In the RAM, various programs and data required for the operation of the computer device are also stored. The processor, the ROM, and the RAM are connected to each other through a bus. An input / output (I / O) interface is also connected to the bus.

[0237] Generally, the following devices can be connected to the I / O interface: input devices including, for example, sensors or visual information acquisition devices; output devices including, for example, display screens; storage devices including, for example, magnetic tapes, hard disks, and the like; and communication devices. The communication devices can allow the computer device to communicate with other devices (such as edge computing devices) wirelessly or by wire to exchange data. Although Figure 8Computer devices with various devices are shown, but it is understood that not all of the devices shown are required to implement or possess the devices. More or less devices can alternatively be implemented or possessed.

[0238] In particular, according to embodiments of the present disclosure, the processes described above with reference to the flowcharts can be implemented as a computer software program. For example, embodiments of the present disclosure include a computer program product comprising a computer program carried on a non-transitory computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network through a communication device, or installed from a storage device, or installed from a ROM. When the computer program is executed by a processor, all or part of the steps of the compatible multi-PLC pumped storage power station simulation test method of the embodiments of the present disclosure are performed.

[0239] Detailed descriptions of the embodiments can refer to the corresponding descriptions in the foregoing embodiments, which will not be repeated here.

[0240] The computer-readable storage medium according to the embodiments of the present disclosure has non-transitory computer-readable instructions stored thereon. When the non-transitory computer-readable instructions are run by a processor, all or part of the steps of the compatible multi-PLC pumped storage power station simulation test method of the embodiments of the present disclosure are performed.

[0241] The computer-readable storage medium described above includes, but is not limited to, optical storage media (such as CD-ROM and DVD), magneto-optical storage media (such as MO), magnetic storage media (such as magnetic tape or mobile hard disk), media with built-in rewritable non-volatile memory (such as memory card), and media with built-in ROM (such as ROM cartridge).

[0242] Detailed descriptions of the embodiments can refer to the corresponding descriptions in the foregoing embodiments, which will not be repeated here.

[0243] The basic principles of the present disclosure are described above in conjunction with specific embodiments, but it should be noted that the advantages, advantages, effects, etc. mentioned in the present disclosure are only examples and are not limiting, and these advantages, advantages, effects, etc. cannot be considered as the must-have of each embodiment of the present disclosure. In addition, the above specific details of the disclosure are only for the purpose of example and for the purpose of understanding, and are not limiting, and the above details do not limit the present disclosure to the must-use specific details.

[0244] In this disclosure, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. The block diagrams of the devices, apparatuses, equipment, and systems involved in this disclosure are only illustrative examples and do not intend to require or imply that they must be connected, arranged, and configured in the manner shown in the block diagrams. As those skilled in the art will recognize, these devices, apparatuses, equipment, and systems can be connected, arranged, and configured in any way. Words such as "comprising", "including", "having", etc. are open-ended words, meaning "including but not limited to", and can be used interchangeably with it. The word "or" and "and" used here refer to the word "and / or", and can be used interchangeably with it, unless the context clearly indicates otherwise. The phrase "such as" used here refers to the phrase "such as but not limited to", and can be used interchangeably with it.

[0245] In addition, as used herein, the "or" used in the listing of items starting with "at least one" indicates a separate listing. For example, the listing of "at least one of A, B, or C" means A or B or C, or AB or AC or BC, or ABC (that is, A and B and C). In addition, the term "exemplary" does not mean that the described examples are preferred or better than other examples.

[0246] It should also be noted that in the systems and methods of this disclosure, each component or each step can be decomposed and / or recombined. These decompositions and / or recombinations should be regarded as equivalent solutions of this disclosure.

[0247] Various changes, substitutions, and alterations to the technologies described herein can be made without departing from the teachings defined by the appended claims. In addition, the scope of the claims of this disclosure is not limited to the specific aspects of the processes, machines, manufactures, compositions of events, means, methods, and acts described above. Current or later-developed processes, machines, manufactures, compositions of events, means, methods, or acts that perform substantially the same function or achieve substantially the same result as the corresponding aspects described herein can be utilized. Thus, the appended claims include such processes, machines, manufactures, compositions of events, means, methods, or acts within their scope.

[0248] The above description of the disclosed aspects is provided to enable any person skilled in the art to make or use this disclosure. Various modifications to these aspects will be readily apparent to those skilled in the art, and the general principles defined herein can be applied to other aspects without departing from the scope of this disclosure. Therefore, this disclosure is not intended to be limited to the aspects shown herein, but rather to the broadest scope consistent with the principles and novel features disclosed herein.

[0249] The above description of the disclosed aspects is provided to enable any person skilled in the art to make or use this disclosure. Various modifications to these aspects will be readily apparent to those skilled in the art, and the general principles defined herein can be applied to other aspects without departing from the scope of this disclosure. Therefore, this disclosure is not intended to be limited to the aspects shown herein, but rather to the broadest scope consistent with the principles and novel features disclosed herein.

[0250] The foregoing description has been presented for the purposes of illustration and description. Furthermore, the description is not intended to limit the embodiments of the disclosure to the forms disclosed herein. Although the various example aspects and embodiments have been discussed and illustrated, various modifications, alterations, additions, and omissions can be made by those skilled in the art.

Claims

1. A method for simulating and testing a pumped storage power station compatible with multiple PLCs, characterized in that, The application relates to a power station operation strategy optimization method and device. The method comprises the following steps: A test platform matched with a target power station is constructed; the test platform is integrated with a data receiving model, a preset symmetric residual network model, a preset long short-term memory model and multiple PLC communication protocols; the data receiving model is integrated with a general data interface with data conversion function; Initial spatial features in the target power station are processed based on the preset symmetric residual network model to obtain first information; Target equipment parameter information, actual environmental factor information and the first information of the target power station are analyzed based on a constructed data fusion model to obtain a target operation strategy of the target equipment under different environmental conditions and power station layouts; Time sequence features of the target power station are processed through the preset long short-term memory model to obtain future operation state prediction information of the target power station; Based on the future operation state prediction information and the target operation strategy, a guiding operation strategy is obtained. The first information is obtained by preprocessing extracted key spatial features of the target power station to obtain first spatial features, obtaining a spatial training set and a spatial test set based on the first spatial features, constructing a preset symmetric residual network model, training the preset symmetric residual network model based on the spatial training set, testing the trained preset symmetric residual network model through the spatial test set, and analyzing the first spatial features based on the trained preset symmetric residual network model to obtain the first information; the first information includes actual layouts of several key spatial structures of the target power station, several potential abnormal problems and layout optimization schemes. The future operation state prediction information of the target power station is obtained by preprocessing extracted key time sequence features of the target power station to obtain first time sequence features, obtaining a time sequence training set and a time sequence test set based on the first time sequence features, constructing a preset long short-term memory model, training the preset long short-term memory model based on the time sequence training set, testing the trained preset long short-term memory model through the time sequence test set, and obtaining the future operation state prediction information of the target power station based on the trained preset long short-term memory model. The guiding operation strategy is obtained based on the future operation state prediction information and the target operation strategy by constructing a manifold regularization model, training the manifold regularization model by using the first spatial features and the first time sequence features, analyzing the future operation state prediction information and the target operation strategy by using the trained manifold regularization model to obtain an energy loss equipment area, and obtaining the guiding operation strategy based on the energy loss equipment area; the guiding operation strategy includes one or more of device optimization layout strategies, device upgrade strategies and energy scheduling strategies corresponding to the area. 2.The compatible-PLC-pumped storage power station simulation test method according to claim 1, wherein, The key space features include at least one of power station geographical location features, power station structure layout features, power station equipment distribution features, distance relationships between different equipment, equipment distribution density, important equipment layout parameters, hotspot area information, pipeline layout, unit location, and control room layout. The important equipment layout parameters include at least one of generator set parameter information, transformer parameter information, and switch device parameter information. 3.The compatible-PLC-pumped storage power station simulation test method according to claim 1, wherein, The preset symmetric residual network model is constructed, including: determining a network structure; the network structure is a symmetrically designed structure, and a basic unit in the network structure is a residual block; a plurality of convolution layers are configured inside the residual block; the input channel number and the output channel number of the convolution layer are the same, and the convolution kernel size and the step are consistent; the convolution layer is 1x1 convolution or 3x3 convolution; an activation function is added after the convolution layer; the activation function includes ReLU, Leaky ReLU, or ELU; a batch normalization layer is added between the convolution layer and the activation function; a skip connection is added between the input and the output of the residual block, and the parameters in the skip connection are shared with the parameters inside the residual block; one of the convolution layers is determined as an input layer of the network; a plurality of residual blocks are stacked to form a main part of the network, wherein the output of each residual block is used as the input of the next residual block; a pooling layer or a transition layer is arranged between different residual blocks; an output layer is configured; the output layer is a fully connected layer with a softmax activation function or a fully connected layer with a linear activation function; a loss function, an optimizer, and a preset learning rate are determined; the loss function includes one or more of a cross-entropy loss function and a mean square error loss function; the optimizer includes one or more of Adam and SGD. 4.The compatible-PLC-pumped storage power station simulation test method according to claim 3, characterized in that, The preset long short-term memory model is constructed, including: determining a model structure; the model structure includes an input layer, a plurality of LSTM layers, and an output layer; the shape of the input layer matches the dimension of the input data; the LSTM layer includes an input gate, a forgetting gate, and an output gate; each LSTM layer has a hidden state and a cell state, and the hidden state and the cell state of each LSTM layer are transmitted to the next LSTM layer; the output layer is configured with an output unit number and an activation function; the LSTM hidden unit number, the learning rate, the batch size, the loss function, and the optimizer are configured.

5. A computer apparatus, characterized by The computer device includes: at least one processor; and a memory connected in communication with the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the compatible multi-PLC pumped storage power station simulation test method of any one of claims 1-4.

6. A computer-readable storage medium, characterized in that, The computer readable storage medium stores computer instructions for causing a computer to execute the compatible multi-PLC pumped storage power station simulation test method of any one of claims 1-4.

7. A computer program product comprising computer instructions, characterized in that, The computer instructions, when executed by the processor, implement the steps of the method of any one of claims 1-4.

Images