Simulation verification method and system for transformation design of intelligent pumped storage hydropower station

Abstract

The invention discloses an intelligent pumped storage hydropower station reconstruction design simulation verification method and system. The method comprises the following steps: establishing a digital twin model of the pumped storage hydropower station, wherein the digital twin model comprises a hydraulic structure model, an electromechanical equipment model and a control system model; importing a transformation design scheme into the digital twin model, and optimizing model parameters based on a deep learning algorithm and historical operation data; multi-working-condition simulation verification is carried out, dynamic response characteristics of all systems are analyzed, and the influence of the transformation design scheme on the overall performance of the power station is evaluated; and performing intelligent analysis on a simulation result based on machine learning, identifying potential risks and proposing optimization suggestions. The invention also discloses a system for realizing the method. According to the method, the deep learning algorithm is innovatively introduced to realize intelligent optimization of the model parameters, the workload of manual parameter adjustment is reduced, and the scheme optimization efficiency is improved.

Description

Technical Field

[0001] The present application belongs to the field of hydropower engineering technology, and specifically relates to a simulation verification method and system for the reconstruction design of an intelligent pumped-storage hydropower station. Background Art

[0002] Pumped-storage hydropower stations are important peak-shaving and frequency-regulating facilities in the power system, and their safe and stable operation is of great significance to the safety and stability of the power grid. As the service time of pumped-storage hydropower stations increases, problems such as equipment aging and performance degradation are becoming increasingly prominent, and they need to be transformed and upgraded to maintain their good operating conditions.

[0003] At present, the renovation design of pumped storage hydropower stations mainly relies on the experience of engineers and simple numerical calculations, which has the following problems: First, the feasibility verification of the renovation plan lacks comprehensiveness, and it is impossible to accurately evaluate the impact of the renovation plan on the overall performance of the hydropower station; second, the traditional simulation verification method has low calculation efficiency and is difficult to cope with complex multi-condition analysis needs; third, there is a lack of intelligent risk assessment methods, making it difficult to timely discover potential safety hazards. Therefore, an intelligent renovation design simulation verification method is urgently needed. Summary of the invention

[0004] In view of this, the present application provides an intelligent pumped-storage hydropower station renovation design simulation verification method and system, which solves the problems of insufficient comprehensiveness, low efficiency and backward risk assessment methods in the prior art of pumped-storage hydropower station renovation design simulation verification.

[0005] An embodiment of the present application provides a method for simulation and verification of the reconstruction design of an intelligent pumped-storage hydropower station, comprising: establishing a digital twin model of the pumped-storage hydropower station, the digital twin model comprising a hydraulic structure model, an electromechanical equipment model and a control system model; importing the reconstruction design scheme into the digital twin model, and optimizing the model parameters based on a deep learning algorithm and historical operating data; performing multi-operating condition simulation verification through the optimized digital twin model to obtain simulation results, the simulation results comprising: dynamic response characteristics of hydraulic structures, electromechanical equipment and control systems, and the impact of the reconstruction design scheme on the overall performance of the hydropower station; performing intelligent analysis of the simulation results based on machine learning, identifying potential risks and making optimization suggestions.

[0006] Optionally, establishing a digital twin model of the pumped-storage hydropower station includes: collecting actual operating data of the hydropower station, the operating data including geometric parameters of hydraulic structures, unit operating parameters and control system parameters; establishing a model of the hydraulic structures, including upper and lower reservoirs, pressure pipes and tailwater systems; establishing a model of the electromechanical equipment, including turbines, generators and main transformers; and establishing a model of the control system, including a speed governor, an excitation system and a monitoring system.

[0007] Optionally, the transformation design scheme is imported into the digital twin model, and the model parameters are optimized based on the deep learning algorithm and historical operation data, including: acquiring and preprocessing historical operation data, organizing the historical operation data according to the acquisition time, spatial position, parameter type and operation status to form a high-dimensional tensor containing time dimension, spatial dimension, parameter dimension and state dimension; decomposing the high-dimensional tensor by a convolution tensor decomposition method, and the convolution tensor decomposition includes CP decomposition and Tucker decomposition of the convolution kernel tensor; based on the decomposed high-dimensional tensor, constructing a multi-layer convolution network structure for parameter optimization, the network structure includes: an input layer for receiving input data, a feature extraction layer for extracting features, a feature fusion layer for fusing multi-dimensional features and an output layer for outputting optimized parameters; training the multi-layer convolution network structure by an adaptive learning rate strategy, and improving the training effect by batch normalization to obtain optimized model parameters.

[0008] Optionally, the high-dimensional tensor is decomposed using a convolutional tensor decomposition method, and also includes: using tensor ring decomposition to process cyclic features in the data; introducing tensor regularization constraints; applying residual connections in the network structure; adjusting tensor decomposition parameters; and determining the optimal tensor decomposition parameters based on model performance evaluation results.

[0009] Optionally, multi-operating condition simulation verification is performed through the optimized digital twin model, including: based on the digital twin model, designing a typical operating condition test plan including normal start and stop, accident conditions and phase adjustment operation; using the digital twin model, adopting a distributed computing method to perform multi-operating condition parallel simulation, and analyzing the dynamic response characteristics of hydraulic structures, electromechanical equipment and control systems under different operating conditions; storing the simulation results of the digital twin model in a simulation database; based on the simulation results of the digital twin model, generating a simulation verification report containing key performance indicators.

[0010] Optionally, the simulation results are intelligently analyzed based on machine learning to identify potential risks and make optimization suggestions, including: mapping key parameters of the transformation plan to a high-dimensional parameter space; constructing a Regge triangulation of the parameter space based on the correlation between the parameters to obtain a plurality of triangular units; defining a Regge metric tensor describing local geometric features on each triangular unit; calculating the discrete curvature of the simulation results in the parameter space based on the Regge metric tensor; determining the optimal path for parameter adjustment based on the distribution characteristics of the discrete curvature, and generating specific parameter optimization suggestions based on the optimal path.

[0011] Optionally, intelligent analysis of simulation results is performed based on machine learning to identify potential risks and make optimization suggestions, which also includes: analyzing areas with large discrete curvature in parameter space and defining them as parameter sensitive areas; calculating Regge metric characteristics of parameter sensitive areas, including principal curvature values ​​and principal curvature directions; identifying high curvature areas with abnormal discrete curvature based on the Regge metric characteristics as potential risk points of hydropower stations; constructing an optimization objective function containing constraints on parameter sensitive areas; using a geodesic search algorithm to perform parameter optimization based on the optimization objective function to obtain optimization parameters that avoid high-risk areas; and verifying the stability and reliability of the optimization results by evaluating changes in the Regge metric before and after optimization.

[0012] Optionally, intelligent analysis of simulation results is performed based on machine learning to identify potential risks and make optimization suggestions, which also includes: using genetic algorithms to preliminarily optimize key parameters of the transformation plan; based on the preliminary optimization results, establishing a transformation process model that describes the dynamic change process of parameters; performing transition state analysis based on the transformation process model to evaluate the impact of parameter changes on the hydropower station; and evaluating the reliability of parameter optimization results by calculating reconstruction errors and analyzing the impact of different decomposition orders.

[0013] Optionally, the collection of actual operation data of the hydropower station includes: setting up data collection points, including key monitoring points of hydraulic structures, unit operation parameter monitoring points and control system monitoring points; establishing a data collection system to collect power station operation data in real time; and preprocessing the collected data, including data cleaning, outlier processing and data standardization.

[0014] The embodiment of the present application also provides an intelligent pumped-storage hydropower station reconstruction design simulation verification system, including: a model building module, used to establish a digital twin model of the pumped-storage hydropower station, the digital twin model including a hydraulic structure model, an electromechanical equipment model and a control system model; a parameter optimization module, used to import the reconstruction design scheme into the digital twin model, and optimize the model parameters based on a deep learning algorithm and historical operation data; a simulation calculation module, used to perform multi-operating condition simulation verification through the optimized digital twin model to obtain simulation results, the simulation results including: the dynamic response characteristics of hydraulic structures, electromechanical equipment and control systems, and the impact of the reconstruction design scheme on the overall performance of the hydropower station; an intelligent evaluation module, used to perform intelligent analysis of the simulation results based on machine learning, identify potential risks and make optimization suggestions.

[0015] An embodiment of the present application also provides a computer device, comprising: at least one processor; and a memory communicatively connected to 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 so that the at least one processor can execute the above-mentioned method for simulation verification of intelligent pumped-storage hydropower station reconstruction design.

[0016] An embodiment of the present application also provides a computer-readable storage medium, which stores computer instructions, and the computer instructions are used to enable a computer to execute the above-mentioned method for simulation verification of intelligent pumped-storage hydropower station reconstruction design.

[0017] An embodiment of the present application also provides a computer program product, including computer instructions, characterized in that when the computer instructions are executed by a processor, the steps of the method for simulation verification of the above-mentioned intelligent pumped-storage hydropower station reconstruction design are implemented.

[0018] This application has the following technical effects:

[0019] By building a digital twin model, a comprehensive simulation verification of the pumped-storage hydropower station renovation design was achieved, improving the accuracy and reliability of the renovation design;

[0020] The deep learning algorithm is used to optimize the model parameters, which improves the efficiency and accuracy of simulation verification;

[0021] Intelligent analysis based on machine learning can promptly identify potential risks and make optimization suggestions, thus improving the safety of renovation design. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention, and for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0023] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for describing the embodiments are briefly introduced below.

[0024] Figure 1 It is a flow chart of a simulation verification method for intelligent pumped storage hydropower station reconstruction design provided in an embodiment of the present application;

[0025] Figure 2 It is a structural schematic diagram of a digital twin model provided in an embodiment of the present application;

[0026] Figure 3It is a structural schematic diagram of a deep learning model based on convolution tensor decomposition provided in an embodiment of the present application;

[0027] Figure 4 It is a schematic diagram of the principle of Regge metric analysis provided in an embodiment of the present application. DETAILED DESCRIPTION

[0028] The embodiments of the present application are described below in conjunction with the drawings in the present application. It should be understood that the implementation methods described below in conjunction with the drawings are exemplary descriptions for explaining the technical solutions of the embodiments of the present application and do not constitute a limitation on the technical solutions of the embodiments of the present application.

[0029] It will be understood by those skilled in the art that, unless specifically stated, the singular forms "one", "said", and "the" used herein may also include plural forms. It should be further understood that the terms "including" and "comprising" used in the embodiments of the present application refer to that the corresponding features can be implemented as the presented features, information, data, steps, operations, elements and / or components, but do not exclude the implementation as other features, information, data, steps, operations, elements, components and / or combinations thereof supported by the technical field. It should be understood that when an embodiment of the present invention refers to an element being "connected" or "coupled" to another element, the one element may be directly connected or coupled to the other element, or may refer to the one element and the other element establishing a connection relationship through an intermediate element. In addition, the "connection" or "coupling" used herein may include wireless connection or wireless coupling. The term "and / or" used herein indicates at least one of the items defined by the term, for example, "A and / or B" may be implemented as "A", or as "B", or as "A and B".

[0030] In order to make the objectives, technical solutions and advantages of the present application clearer, the implementation methods of the present application will be further described in detail below with reference to the accompanying drawings.

[0031] The core of the present invention is to provide a method and system for generating intelligent software test data based on Internet public information. The present invention is further described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0032] It should be understood that the terms "including" and "having" and any variations thereof used herein are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device comprising a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products or devices.

[0033] In order to make the purpose, technical solution and advantages of the present application clearer, the technical solution of the present application will be clearly and completely described below in combination with the specific embodiments of the present application and the corresponding drawings. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present application.

[0034] The embodiment of the present application provides a simulation verification method for intelligent pumped storage hydropower station reconstruction design. Figure 1 The specific steps of the method include:

[0035] S1 builds a digital twin model of a pumped storage hydropower station:

[0036] The digital twin model is a virtual mapping of the physical hydropower station. It synchronizes the physical entity with the virtual model through real-time data interaction and dynamic updates. The model includes three sub-models: hydraulic structures, electromechanical equipment, and control systems. Collaborative simulation is achieved between them through data interfaces and information exchange mechanisms.

[0037] Specifically, a modeling method based on multi-physics field coupling is adopted to describe multiple physical fields such as hydraulics, mechanical dynamics and electrical control in a unified manner. For example, in the turbine model, it is necessary to consider the coupling relationship between hydraulic characteristics, mechanical characteristics and electrical characteristics at the same time, and to describe the interaction between various physical quantities by establishing a fluid-structure coupling equation group.

[0038] S1.1 Collect actual operation data of hydropower station:

[0039] This step adopts a layered and distributed data acquisition architecture to achieve all-round collection of hydropower station operation data. At the physical layer, various sensors and data acquisition equipment are arranged; at the network layer, a communication method combining industrial Ethernet and fieldbus is adopted; at the application layer, a unified data acquisition and management platform is adopted.

[0040] The data collection content includes: geometric parameters such as stress and strain, seepage, displacement, etc. of hydraulic structures; operating parameters such as speed, power, vibration, bearing temperature, etc. of the unit; and control parameters such as governor parameters, excitation voltage and current, etc. of the control system. For example, for the collection of turbine bearing temperature, a PT100 temperature sensor is used to transmit data in real time through the Modbus-TCP protocol, with a sampling period of 100ms.

[0041] The specific implementation is as follows:

[0042] S1.1.1 When setting data collection points, select monitoring locations according to the principle of "focused monitoring and reasonable layout". For example, strain gauges and displacement gauges are arranged at key locations such as the dam crest and dam foundation of the upper reservoir dam; temperature sensors and vibration sensors are arranged at key locations such as the unit bearings, stators, and rotors; and collection points for parameters such as voltage, current, and frequency are arranged at key links of the speed regulator and excitation system.

[0043] The arrangement of data collection points needs to take into account factors such as equipment layout, environmental conditions, and construction conditions to ensure the accuracy and reliability of the collected data. For example, the installation location of the vibration sensor should be selected at the rigid connection on the vibration transmission path, avoiding the node position.

[0044] The data acquisition system established in S1.1.2 adopts a three-layer architecture: the bottom layer is the field sensor network, which uses industrial Ethernet and field bus (such as Profibus, Modbus, etc.) to realize data transmission; the middle layer is the data acquisition server, which is responsible for data collection, caching and preprocessing; the upper layer is the data management platform, which realizes data storage, query and analysis functions.

[0045] The data acquisition system has real-time and reliability guarantee mechanisms, including data caching, breakpoint resume, data verification, etc. For example, when network communication is interrupted, the on-site equipment can temporarily store the data in the local cache and automatically upload it after communication is restored.

[0046] S1.1.3 Data preprocessing includes the following steps:

[0047] Data cleaning: Use median filtering, wavelet noise reduction and other methods to remove noise and interference in the collected data. For example, use db4 wavelet to perform 4-layer decomposition on the vibration signal to remove high-frequency noise;

[0048] Outlier processing: Identify outliers based on the 3σ criterion and box plot method, and process abnormal data by interpolation or deletion;

[0049] Data standardization: Use maximum and minimum value standardization or Z-score standardization method to unify data of different dimensions into the same scale range.

[0050] S1.2 Establish hydraulic structure model:

[0051] The hydraulic structure model is constructed based on the finite element method and computational fluid dynamics method, including the geometric model and mechanical model of the upper and lower reservoirs, pressure pipes and tailwater system. The model can simulate the stress distribution, deformation characteristics and water flow characteristics of hydraulic structures under different working conditions.

[0052] The Upper Reservoir Model focuses on the following aspects:

[0053] The dam structure model is established using three-dimensional solid elements, and the element type is a 20-node hexahedral isoparametric element, which can accurately describe complex geometric shapes;

[0054] A reservoir water pressure model was established, the dynamic water pressure effect caused by water level changes was considered, and the reservoir pressure distribution was calculated using the Westergaard added mass method.

[0055] Simulate the contact relationship between the dam body and the bedrock, use the surface-to-surface contact unit to describe the mechanical properties of the dam foundation contact surface, and set the normal stiffness and tangential stiffness parameters;

[0056] Considering the seepage effect, the saturated-unsaturated seepage theory is used to establish the seepage field model and calculate the pore water pressure distribution.

[0057] The pressure pipe model contains the following features:

[0058] Establish a three-dimensional solid model of the pipeline, taking into account the composite structural characteristics of the steel lining and concrete lining;

[0059] The fluid-solid coupling method is used to simulate the water hammer effect in the pressure pipe and calculate the transient pressure fluctuation;

[0060] Considering the elastic deformation of the pipeline, a stress-strain relationship model is established to analyze the structural stability of the pipeline;

[0061] Simulate the interaction between the pipeline and the surrounding rock mass, and set the contact surface parameters and boundary conditions.

[0062] The tailwater system model focuses on:

[0063] Establish a three-dimensional geometric model of the tailrace tunnel, taking into account the influence of the roughness of the tunnel wall;

[0064] Computational fluid dynamics methods are used to simulate tailwater flow characteristics and calculate velocity and pressure distributions;

[0065] Considering the influence of air mixing, a two-phase flow model is established to describe the water-gas two-phase flow state;

[0066] The free water surface flow at the tailrace outlet is simulated and the VOF method is used to track the free water surface.

[0067] Each sub-model is coupled through boundary conditions and data interfaces. For example, the upper reservoir model is connected to the pressure pipe model through the inlet water level and flow conditions, and the pressure pipe model is coupled to the tailwater system model through the outlet pressure and flow rate conditions.

[0068] S1.3 Establish electromechanical equipment model:

[0069] The electromechanical equipment model mainly includes the modeling of three core components: turbine, generator and main transformer. These models need to consider the coupling relationship between the electrical characteristics, mechanical characteristics and thermal characteristics of the equipment to accurately reflect the dynamic response characteristics of the equipment under various operating conditions.

[0070] First, the turbine model adopts a modeling method of multi-physics field coupling. In terms of hydraulic characteristics, a three-dimensional flow field model of the impeller flow channel is established, the Reynolds-averaged NS equation is used to describe the water flow movement, and the k-ε turbulence model is selected to simulate the turbulent characteristics of the flow. In terms of mechanical characteristics, the finite element method is used to establish the structural model of key components such as blades and shaft systems, and the stress distribution and deformation characteristics are calculated. Through fluid-solid coupling analysis, the characteristic curves of the turbine under different working conditions can be obtained, including flow-speed characteristics, efficiency-flow characteristics, etc.

[0071] Secondly, the generator model focuses on the coupling analysis of electromagnetic fields and thermal fields. In terms of electromagnetic characteristics, the finite element method is used to solve Maxwell's equations and calculate electromagnetic parameters such as air gap magnetic field distribution and stator current distribution. Equivalent circuit models of stator windings and rotor windings are established to analyze the voltage and current characteristics of the generator. In terms of thermal characteristics, various heat sources such as core loss and copper loss are considered, a three-dimensional temperature field model is established, and the temperature distribution of each component is analyzed. In addition, a mechanical vibration model needs to be established to calculate the vibration characteristics and critical speed of the shaft system.

[0072] The main transformer model includes an electromagnetic field model and a thermal field model. In the electromagnetic field model, the magnetic circuit method is used to establish the equivalent circuit of the transformer and calculate parameters such as leakage reactance and excitation current. Through finite element analysis, the magnetic field distribution of the winding and the core can be obtained. In the thermal field model, the oil circulation cooling system is considered, a convection heat transfer model is established, and the temperature distribution of each component of the transformer is analyzed.

[0073] It should be noted that there is a close coupling relationship between the various parts of the electromechanical equipment model. For example, the output torque of the turbine directly affects the speed and output power of the generator, and the electromagnetic force of the generator affects the vibration characteristics of the shaft system. Therefore, a reasonable data exchange mechanism needs to be established when building the model to ensure the coordination between the sub-models.

[0074] S1.4 Establish control system model:

[0075] The control system model mainly includes three parts: speed regulator, excitation system and monitoring system. A hierarchical and distributed modeling architecture is adopted to achieve the coordination of various control loops. In the embodiment of the present application, the control system model should reflect the dynamic characteristics of each controller, and also consider the transmission characteristics of the control signal and the implementation process of the control algorithm.

[0076] The speed governor model adopts a proportional-integral-differential (PID) control structure, including an electro-hydraulic conversion system, a hydraulic system, and a mechanical actuator. First, a guide vane opening control loop model is established, including a speed feedback link, a PID controller, an electro-hydraulic servo valve, and a hydraulic actuator. Among them, the electro-hydraulic servo valve uses a second-order oscillation link to describe its dynamic characteristics, and the hydraulic actuator considers the influence of factors such as oil pressure changes and piston movement.

[0077] The excitation system model includes an automatic voltage regulator (AVR) and a limiter. The AVR adopts a multi-loop feedback control structure, including a voltage outer loop and a current inner loop. The role of the power angle stabilizer (PSS) is considered in the model, and the dynamic stability of the system is improved by introducing a compensation signal for power changes. The limiter model includes protection functions such as over-excitation limit, under-excitation limit, and V / Hz limit.

[0078] The monitoring system model adopts the architecture of distributed control system (DCS), including field control layer, process control layer and monitoring layer. When modeling, the focus is on the realization of functions such as data acquisition, signal processing, control algorithm execution and human-computer interaction. By establishing a communication network model, the data transmission process between different levels is simulated, including the influence of network characteristics such as communication delay and data packet loss.

[0079] In summary, by establishing a complete digital twin model (digital twin with a structure such as Figure 2 ), which can realize the comprehensive mapping of the physical entity of the pumped-storage hydropower station and provide basic support for subsequent simulation verification.

[0080] S2 imports the transformation design into the digital twin model and optimizes it:

[0081] The core of this step is to integrate the transformation design plan with the digital twin model, and optimize the model parameters through deep learning algorithms to improve the accuracy and reliability of the model.

[0082] S2.1 Acquire and preprocess historical operation data:

[0083] In the embodiment of the present application, the operation data is first extracted from the power station operation history database, including multi-dimensional information such as equipment operation parameters, control system status, environmental parameters, etc. These data are systematically organized according to the acquisition time, spatial location, parameter type and operation status to form a four-dimensional tensor structure. The time dimension reflects the change law of the parameter over time, the spatial dimension describes the distribution characteristics of the parameter at different monitoring points, the parameter dimension contains different types of monitoring data, and the state dimension corresponds to different operating conditions.

[0084] The data is processed in segments using a sliding time window method, and the window length is determined based on the changing characteristics of the parameters. For example, for fast-changing parameters (such as pressure pulsation), a smaller time window (such as 1 second) is used, while for slowly changing parameters (such as bearing temperature), a larger time window (such as 1 minute) is used. In addition, the spectral characteristics of the data are analyzed by Fourier transform to identify and filter out interference signals.

[0085] S2.2 uses the convolution tensor decomposition method:

[0086] In the embodiments of the present application, Figure 3 As shown in the figure, a multi-level convolutional tensor decomposition strategy is adopted to extract the essential features of the data by combining CP decomposition and Tucker decomposition. CP decomposition decomposes the original tensor into the sum of several rank-1 tensors, each component representing a basic pattern. Tucker decomposition introduces a core tensor that can capture the interaction between different dimensions.

[0087] In order to process the cyclic features in the data, the tensor loop decomposition is introduced. Specifically, the data in the time dimension is regarded as a periodic signal, and a circular convolution kernel is constructed through a circular shift operation to extract the periodic change pattern. At the same time, tensor regularization constraints are introduced, including L1 norm constraints and nuclear norm constraints, to suppress overfitting.

[0088] Residual connections are set in the network structure to alleviate the gradient vanishing problem of deep networks. The optimal tensor decomposition parameters, including the rank of CP decomposition and the core tensor dimension of Tucker decomposition, are determined by cross-validation. For example, when processing turbine vibration data, the change curve of the reconstruction error can be analyzed to select the decomposition parameters that can balance the computational complexity and accuracy requirements.

[0089] S2.3 build a multi-layer convolutional network structure:

[0090] The multi-layer convolutional network in the embodiment of the present application adopts an encoder-decoder structure. The input layer receives multi-dimensional features after tensor decomposition and extracts local features through multi-layer convolution operations. The feature extraction layer uses convolution kernels of different scales to capture multi-scale spatiotemporal correlations. The feature fusion layer uses an attention mechanism to perform adaptive weighting according to the importance of different features.

[0091] The convolution layer uses a depth-separable convolution structure to reduce the number of parameters while maintaining the model's expressiveness. Jump connections are set between feature maps to achieve the fusion of low-level features and high-level features. Finally, the output layer reconstructs the optimized model parameters through deconvolution operations.

[0092] S2.4 adopts an adaptive learning rate strategy:

[0093] This application adopts an adaptive learning rate adjustment strategy based on the Adam optimizer. The initial value of the learning rate is determined by grid search and dynamically adjusted according to the change of the loss function during training. When the loss function decreases slowly, the learning rate decay mechanism is triggered to achieve refined parameter tuning.

[0094] During the training process, batch normalization technology is used to standardize the input of each layer to speed up the training convergence. At the same time, the dropout mechanism is introduced to prevent overfitting, and the dropout rate is dynamically adjusted according to the depth of the network layer. In addition, the early stopping strategy is adopted. When the performance indicators on the validation set have not improved for multiple consecutive epochs, the training process is automatically terminated.

[0095] Through the above steps, the intelligent optimization of model parameters is achieved, and the accuracy of the digital twin model is improved. The optimized model can more accurately reflect the actual operation status of the hydropower station and provide a reliable basis for subsequent simulation verification.

[0096] S3 uses the optimized digital twin model to perform multi-condition simulation verification:

[0097] This application adopts a systematic simulation verification method to comprehensively evaluate the feasibility and effectiveness of the transformation design scheme through simulation analysis of multiple working conditions and multiple scenarios.

[0098] When designing a typical operating condition test plan, the key operating states of the hydropower station are considered. The first is the normal start-stop condition, including pumping start, power generation start, normal shutdown and other processes, focusing on analyzing the dynamic response characteristics of the equipment. The second is the accident condition, such as load shedding, emergency shutdown, power grid failure, etc., to verify the protection function and safety performance of the system. The third is the phase adjustment operation condition, analyzing the stability and regulation characteristics of the unit at different speeds. In addition, special conditions, such as sudden changes in reservoir water levels and extreme meteorological conditions, need to be considered.

[0099] During the simulation calculation process, a distributed parallel computing architecture is used to improve computing efficiency. Complex simulation tasks are decomposed into multiple subtasks and executed in parallel on the computing cluster. For example, for the flow field calculation of a turbine, the regional decomposition method is used to divide the calculation domain into multiple sub-regions, and each computing node is responsible for the calculation of a sub-region. Data exchange and synchronization between nodes are achieved through the message passing interface (MPI).

[0100] The simulation result storage adopts a distributed database architecture to support efficient access to massive data. The database design includes time series data tables, spatial data tables, and relational data tables to achieve unified management of multi-dimensional data. A data indexing mechanism is established to improve query efficiency. At the same time, data compression storage is achieved to reduce storage space occupancy.

[0101] The generation of simulation verification reports adopts a combination of template and intelligent methods. A standardized report template is established, including performance indicator analysis, dynamic characteristics evaluation, safety performance verification, etc. Through data mining and feature extraction, abnormal changes in key performance indicators are automatically identified, and analysis conclusions and optimization suggestions are generated.

[0102] This application adopts a systematic simulation verification method to comprehensively evaluate the feasibility and effectiveness of the transformation design scheme through simulation analysis of multiple working conditions and multiple scenarios. Take a pumped storage power station unit transformation project as an example:

[0103] In the verification of normal start-stop conditions, the pumping start process is analyzed first. During the simulation process, the following parameters are focused on: the dynamic response curve of the unit speed increasing from 0rpm to 500rpm when the guide vane opening gradually increases from 0% to 85%; the change in the inlet and outlet pressure of the turbine, especially the pressure fluctuation that may occur at the beginning of the startup; the bearing vibration and temperature change, etc. For example, through simulation, it is found that when the speed reaches 300rpm, the bearing vibration value exceeds 0.12mm due to the rapid increase in the guide vane opening, and the adjustment curve of the guide vane opening needs to be optimized.

[0104] For the verification of accident conditions, the emergency shutdown condition is taken as an example. The unit is simulated to have an emergency shutdown due to a sudden failure when operating at rated load (300MW). The water hammer pressure during the rapid closing of the guide vane (closing time 3 seconds) was simulated and analyzed, and it was found that the maximum pressure rise of the pressure pipeline reached 1.8MPa, exceeding the design limit of 1.5MPa. Based on this, a pressure release channel was added to the modification plan during the guide vane closing process to control the maximum pressure rise within 1.3MPa.

[0105] For the phase-shifting operation condition, the focus is on verifying the operating characteristics of the unit at low speed (about 350rpm) and high speed (about 550rpm). Through simulation analysis, it is found that during high-speed phase-shifting operation, the bearing temperature rises rapidly, from 45°C in normal operation to 65°C. Through analysis, it is found that the flow of the bearing cooling system is insufficient, and accordingly, the automatic adjustment function of the bearing cooling water flow is added to the transformation plan.

[0106] In the special working condition verification, the sudden drop of the water level in the upper reservoir was simulated. For example, when the water level dropped from the normal high water level of 732m to 722m within 10 minutes, the dynamic response characteristics of the unit were analyzed. The simulation results showed that due to the change in water level, a vortex appeared at the water inlet, causing flow fluctuations and unit vibration. To address this problem, the design of the water inlet diversion device was added to the transformation plan.

[0107] When the distributed computing method is used to perform parallel simulation of the above working conditions, the computing tasks are distributed to 64 computing nodes. For example, for the calculation of the turbine flow field, the flow channel area is divided into 2 million grid cells, and each computing node is responsible for the calculation of about 30,000 grid cells. Through MPI, data exchange between nodes is realized, and the computing efficiency is improved by about 15 times.

[0108] After the simulation results are stored in the distributed database, an index mechanism based on time series is established. For example, for the speed data, a data block is created every 5 minutes, and each data block contains 300 sampling points. This storage structure improves the query efficiency by about 8 times and the data compression rate reaches 70%.

[0109] When generating a simulation verification report, the system automatically extracts key performance indicators. For example, for bearing vibration data, the vibration root mean square value, spectrum peak and envelope characteristics are calculated to identify the 2nd frequency vibration anomaly during the startup process. Based on these analysis results, the report automatically generates optimization suggestions for bearing modification, including specific measures such as adjusting bearing clearance and optimizing lubrication oil circuits.

[0110] The above simulation verification process not only verified the feasibility of the transformation plan, but also discovered potential problems through multi-condition analysis, providing an important basis for the optimization of the transformation plan. Through distributed computing and intelligent analysis, the verification efficiency was greatly improved and the reliability of the verification results was guaranteed.

[0111] S4 performs intelligent analysis of simulation results based on machine learning:

[0112] This application adopts geometric learning theory to construct a parameter space analysis framework, and uses the Regge metric tensor to describe the geometric characteristics of parameter changes to achieve intelligent identification and evaluation of potential risks.

[0113] like Figure 4 In the parameter mapping process, a high-dimensional parameter space is first constructed. The key parameters of the transformation scheme, such as turbine guide vane opening, generator excitation current, governor PID parameters, etc., are used as the basis vectors of the space. The principal component analysis (PCA) is used to reduce the dimension and retain the main feature directions. In the parameter space after dimensionality reduction, each point represents a specific set of parameter configurations.

[0114] Based on the correlation between parameters, the Delaunay triangulation algorithm is used to construct the Regge triangulation. On each triangular unit, the Regge metric tensor reflecting the local geometric characteristics is defined. The metric tensor contains the direction and amplitude information of the parameter change and can describe the local curvature characteristics of the parameter space. By solving the Einstein field equations, the discrete curvature distribution in the parameter space is calculated.

[0115] In the parameter optimization process, the geodesic search algorithm is used to find the optimal path. Based on the geometric structure of the parameter space, the algorithm avoids high-risk areas and finds a stable parameter adjustment path. For example, when adjusting the guide vane opening of the turbine, the curvature distribution of the parameter space is analyzed to avoid dangerous parameter areas that may cause water hammer effect.

[0116] Specifically, in the embodiment of the present application, in order to better understand the geometric characteristics of the parameter space and its application in the optimization of hydropower station parameters, the specific implementation process of Regge triangulation and geodesic search is now described in detail:

[0117] First, take the optimization of turbine guide vane opening as an example. The key parameters involved include guide vane opening θ (0-100%), unit speed n (0-500rpm) and pipeline pressure p (0-2MPa). These three parameters constitute a three-dimensional parameter space. In practical applications, the parameter space is discretized into a finite number of sampling points, each of which represents a specific set of parameter configurations. For example, a point is taken for every 5% guide vane opening, a point is taken for every 20rpm speed, and a point is taken for every 0.1MPa pressure.

[0118] The Delaunay triangulation algorithm is used to divide the parameter space. The core of this algorithm is to ensure that the circumscribed circle of any triangle does not contain other sampling points. The specific steps are as follows:

[0119] 1. First, construct the initial triangle to cover all sampling points;

[0120] 2. Insert the sampling points one by one and find the triangle containing the point;

[0121] 3. Split the triangle into three new triangles;

[0122] 4. Check whether the newly generated triangle meets the Delaunay criterion. If not, perform edge exchange operation;

[0123] 5. Repeat steps 2-4 until all sampling points have been processed.

[0124] The Regge metric tensor G is defined on each triangular unit and its expression is:

[0125] G=[g11 g12 g13]

[0126] [g21 g22 g23]

[0127] [g31 g32 g33]

[0128] Among them, gij represents the metric relationship between parameters. For example, g11 represents the influence of the change of the guide vane opening on the system response, and g12 represents the coupling relationship between the guide vane opening and the speed. Specifically, by analyzing the historical operation data and simulation results, the influence of parameter changes on the system performance is statistically analyzed, and the components of the metric tensor are constructed.

[0129] Based on the Regge metric tensor, the discrete curvature in the parameter space can be calculated. On each triangular unit, the calculation formula of the discrete curvature R is:

[0130] R=det(G) / A

[0131] Where det(G) is the determinant of the metric tensor, and A is the area of ​​the triangle. A larger curvature value indicates that the parameter change in this area has a significant impact on the system performance. For example, in the process of the guide vane opening rapidly decreasing from 80% to 20%, due to the influence of the water hammer effect, the curvature value of this area is significantly higher than that of other areas.

[0132] In the parameter optimization process, the geodesic search algorithm is used to find the optimal path. A geodesic is the shortest path between two points in the parameter space.

[0133] The specific implementation steps of the algorithm are:

[0134] 1. Determine the starting point (current parameter configuration) and the end point (target parameter configuration);

[0135] 2. Construct a path search grid in parameter space;

[0136] 3. For each possible path in the grid, calculate its geodesic length;

[0137] 4. Also consider the maximum curvature value on the path and avoid high-risk areas;

[0138] 5. Select the path with a smaller geodesic length and a maximum curvature value within a safe range as the optimal path.

[0139] Taking the guide vane opening adjustment as an example, assume that the current opening is adjusted from 85% to the target opening of 30%. The traditional method may directly reduce the opening linearly, but this may cause severe water hammer effect. Through the geodesic search algorithm, the optimal path obtained is:

[0140] 1. First, slowly reduce it to 70%, at which point the curvature value is small and the system responds smoothly;

[0141] 2. Use a smaller adjustment step in the 60%-40% area because the curvature value in this area is larger;

[0142] 3.Finally adjust it to 30% smoothly.

[0143] In this way, the stability of parameter adjustment is guaranteed, and risks such as water hammer effect are effectively avoided. Practice has proved that after adopting this method, the maximum pressure fluctuation during the adjustment process is reduced by about 40%, which greatly improves the safety of adjustment.

[0144] For the analysis of parameter-sensitive areas, focus on areas with large discrete curvature. Calculate the principal curvature values ​​and principal curvature directions of these areas to identify the direction in which parameter changes have the greatest impact on system performance. For example, when analyzing the speed regulator parameters, it can be found that the integral time constant in the PID parameters has a greater impact on system stability.

[0145] In addition, this application also uses genetic algorithms for parameter optimization. When designing the fitness function, multiple performance indicators are considered, such as regulation quality, energy conversion efficiency, equipment life, etc. Through crossover and mutation operations, the parameter combination is continuously optimized. During the optimization process, a state transition model that describes the dynamic changes of parameters is established to evaluate the transition characteristics during parameter adjustment.

[0146] Specifically, in the embodiment of the present application, parameter sensitive area analysis and genetic algorithm optimization are two important links that complement each other. The following is a detailed description of the speed regulator PID parameter optimization as a specific case:

[0147] The parameter sensitive area analysis process is as follows:

[0148] First, the PID parameter space is constructed, including the proportional coefficient Kp (0.5-5), the integral time constant Ti (0.1-2s) and the differential time constant Td (0-0.5s). In this three-dimensional parameter space, the local geometric feature description has been established by Regge triangulation. For each parameter sensitive area, the principal curvature values ​​κ1, κ2, κ3 and the corresponding principal curvature directions v1, v2, v3 are calculated.

[0149] Taking a typical working condition as an example, in the area near Ti=0.2s, the principal curvature values ​​are:

[0150] κ1=2.35(Ti direction)

[0151] κ2=0.82(Kp direction)

[0152] κ3=0.31(Td direction)

[0153] This shows that in this region, the change of the integral time constant Ti has the greatest impact on the system stability.

[0154] The specific analysis results show:

[0155] 1. When Ti decreases from 0.2s to 0.15s, the system overshoot increases rapidly from 15% to 35%;

[0156] 2. When Ti increases from 0.2s to 0.25s, the system adjustment time is extended from 2.5s to 4.8s;

[0157] 3. Changes of Kp and Td of the same magnitude have relatively little impact on system performance.

[0158] Genetic algorithm optimization design is more comprehensive and systematic. First, design the chromosome encoding scheme:

[0159] Chromosome structure = [Kp_bit, Ti_bit, Td_bit]

[0160] in:

[0161] Kp_bit: 10-bit binary, mapping range 0.5-5

[0162] Ti_bit: 8-bit binary, mapping range 0.1-2s

[0163] Td_bit: 6-bit binary, mapping range 0-0.5s

[0164] The design of the fitness function considers multiple performance indicators:

[0165] F=w1*(1-Os / Os_max)+w2*(1-Ts / Ts_max)+w3*Ef+w4*L

[0166] in:

[0167] Os: Overshoot

[0168] Ts: Adjustment time

[0169] Ef: Energy conversion efficiency

[0170] L: Factor affecting equipment life expectancy

[0171] w1-w4: weight coefficient

[0172] Specific evaluation index calculation method:

[0173] 1. The overshoot Os is obtained by simulating the step response of the system;

[0174] 2. The adjustment time Ts is defined as the time when the system response enters the ±2% band;

[0175] 3. Energy conversion efficiency Ef takes into account the combined effects of turbine efficiency and generator efficiency;

[0176] 4. Equipment life impact factor L is evaluated based on equipment stress and vibration level.

[0177] Specific parameter settings of genetic algorithm:

[0178] Population size: 100

[0179] Crossover probability: 0.85

[0180] Mutation probability: 0.01

[0181] Maximum number of generations: 200

[0182] During the optimization process, a state transition model is established to describe the parameter change process:

[0183] x(k+1)=Ax(k)+Bu(k)

[0184] y(k)=Cx(k)+Du(k)

[0185] Among them, the state variable x includes speed deviation, power deviation, etc.; the input u is the change of PID parameters; and the output y is the system performance index.

[0186] Here is an example of a typical iteration in the optimization process:

[0187] 1. Initial PID parameters: Kp = 2.0, Ti = 0.5s, Td = 0.1s

[0188] 2. Calculate fitness: F = 0.72

[0189] 3. After the crossover operation, a new parameter combination is obtained: Kp = 2.2, Ti = 0.3s, Td = 0.1s

[0190] 4. State transition model prediction:

[0191] Speed ​​fluctuation reduced by 12%

[0192] Power fluctuation reduced by 8%

[0193] 15% reduction in adjustment time

[0194] 5. Recalculate fitness: F = 0.85

[0195] 6. Keep new parameter combinations due to improved fitness

[0196] After 200 generations of iterative optimization, the optimal PID parameters are finally obtained:

[0197] Kp=2.5, Ti=0.28s, Td=0.08s

[0198] Performance indicators of this group of parameters:

[0199] Overshoot: 12.5% ​​(meets the requirement of <15%)

[0200] Adjustment time: 2.2s (meets the requirement of <3s)

[0201] Energy conversion efficiency: 92.8% (increased by 1.2 percentage points)

[0202] Equipment vibration level: reduced by 25%

[0203] By combining parameter sensitivity analysis with genetic algorithm optimization, the accuracy of the optimization direction is guaranteed and the search for the global optimal solution is realized. The PID parameters finally obtained not only meet the dynamic performance requirements, but also take into account the improvement of equipment life and operating efficiency. In practical applications, this method significantly improves the regulation quality of the speed regulator and significantly improves the overall performance of the system.

[0204] Finally, by calculating the model reconstruction error and analyzing the influence of different decomposition orders on the optimization results, the reliability of parameter optimization is verified. An evaluation index system is established, including stability index, economic index and safety index, to comprehensively evaluate the effectiveness of the optimization results.

[0205] Through the above intelligent analysis process, potential risks in the transformation plan can be discovered in a timely manner, and targeted optimization suggestions can be given, providing important guarantees for the safe and stable operation of the hydropower station.

[0206] The embodiment of the present application also provides an intelligent pumped-storage hydropower station reconstruction design simulation verification system, including: a model building module, used to establish a digital twin model of the pumped-storage hydropower station, the digital twin model including a hydraulic structure model, an electromechanical equipment model and a control system model; a parameter optimization module, used to import the reconstruction design scheme into the digital twin model, and optimize the model parameters based on a deep learning algorithm and historical operation data; a simulation calculation module, used to perform multi-operating condition simulation verification through the optimized digital twin model to obtain simulation results, the simulation results including: the dynamic response characteristics of hydraulic structures, electromechanical equipment and control systems, and the impact of the reconstruction design scheme on the overall performance of the hydropower station; an intelligent evaluation module, used to perform intelligent analysis of the simulation results based on machine learning, identify potential risks and make optimization suggestions.

[0207] An embodiment of the present application also provides a computer device, comprising: at least one processor; and a memory communicatively connected to 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 so that the at least one processor can execute the above-mentioned method for simulation verification of intelligent pumped-storage hydropower station reconstruction design.

[0208] An embodiment of the present application also provides a computer-readable storage medium, which stores computer instructions, and the computer instructions are used to enable a computer to execute the above-mentioned method for simulation verification of intelligent pumped-storage hydropower station reconstruction design.

[0209] An embodiment of the present application also provides a computer program product, including computer instructions, characterized in that when the computer instructions are executed by a processor, the steps of the method for simulation verification of the above-mentioned intelligent pumped-storage hydropower station reconstruction design are implemented.

[0210] The above is only a preferred embodiment of the present application, and does not limit the scope of the patent of the present application. Any modification, equivalent replacement and improvement made within the spirit and principle of the present application should be included in the scope of the patent protection of the present application. It should be understood that although the flowchart of the embodiment of the present application indicates each operation step by arrows, the implementation order of these steps is not limited to the order indicated by the arrows. Unless there is a clear description in this article, in some implementation scenarios of the embodiment of the present application, the implementation steps in each flowchart can be executed in other orders according to demand. In addition, some or all of the steps in each flowchart can include multiple sub-steps or multiple stages based on the actual implementation scenario. Some or all of these sub-steps or stages can be executed at the same time, and each sub-step or stage in these sub-steps or stages can also be executed at different times. In different scenarios at the execution time, the execution order of these sub-steps or stages can be flexibly configured according to demand, and the embodiment of the present application does not limit this.

[0211] The above is only an optional implementation method for some implementation scenarios of the present application. It should be pointed out that for ordinary technicians in this technical field, without departing from the technical concept of the solution of the present application, other similar implementation methods based on the technical ideas of the present application are also within the protection scope of the embodiments of the present application.

Claims

1. An intelligent pumped storage hydropower station reconstruction design simulation verification method, characterized in that: include: Establishing a digital twin model of the pumped-storage hydropower station, wherein the digital twin model includes a hydraulic structure model, an electromechanical equipment model, and a control system model; Importing the transformation design plan into the digital twin model, and optimizing the model parameters based on the deep learning algorithm and historical operation data; The optimized digital twin model is used to perform multi-operating condition simulation verification to obtain simulation results, which include: the dynamic response characteristics of hydraulic structures, electromechanical equipment and control systems, and the impact of the transformation design scheme on the overall performance of the hydropower station; Intelligently analyze simulation results based on machine learning to identify potential risks and make optimization suggestions.

2. The method according to claim 1, characterized in that: Establishing a digital twin model of the pumped storage hydropower station includes: Collecting actual operation data of the hydropower station, the operation data including geometric parameters of hydraulic structures, unit operation parameters and control system parameters; Establishing the hydraulic structure model, including upper and lower reservoirs, pressure pipes and tailwater system; Establishing a model of the electromechanical equipment, including a turbine, a generator and a main transformer; The control system model is established, including a speed regulator, an excitation system and a monitoring system.

3. The method according to claim 1, characterized in that Import the transformation design plan into the digital twin model and optimize the model parameters based on the deep learning algorithm and historical operation data, including: Acquire and preprocess historical operation data, organize the historical operation data according to acquisition time, spatial location, parameter type and operation status, and form a high-dimensional tensor including time dimension, spatial dimension, parameter dimension and state dimension; Decomposing the high-dimensional tensor by using a convolution tensor decomposition method, wherein the convolution tensor decomposition includes performing CP decomposition and Tucker decomposition on the convolution kernel tensor; Based on the decomposed high-dimensional tensor, a multi-layer convolutional network structure for parameter optimization is constructed, wherein the network structure includes: an input layer for receiving input data, a feature extraction layer for extracting features, a feature fusion layer for fusing multi-dimensional features, and an output layer for outputting optimization parameters; The multi-layer convolutional network structure is trained using an adaptive learning rate strategy, and the training effect is improved through batch normalization to obtain optimized model parameters.

4. The method according to claim 3, characterized in that The convolution tensor decomposition method is used to decompose the high-dimensional tensor, which also includes: Use tensor loop decomposition to handle cyclic features in the data; Introducing tensor regularization constraints; Apply residual connections in the network structure; Adjust tensor decomposition parameters; Based on the model performance evaluation results, the optimal tensor decomposition parameters are determined.

5. The method according to claim 1, characterized in that The optimized digital twin model is used to perform multi-condition simulation verification, including: Based on the digital twin model, design a typical operating condition test plan including normal start-stop, accident conditions and phase adjustment operation; The digital twin model is used to perform multi-operating condition parallel simulation using a distributed computing method to analyze the dynamic response characteristics of hydraulic structures, electromechanical equipment, and control systems under different operating conditions; Storing the simulation results of the digital twin model in a simulation database; Based on the simulation results of the digital twin model, a simulation verification report including key performance indicators is generated.

6. The method according to claim 1, characterized in that Intelligently analyze simulation results based on machine learning to identify potential risks and make optimization suggestions, including: Map the key parameters of the transformation plan into a high-dimensional parameter space; Based on the correlation between parameters, the Regge triangulation of the parameter space is constructed to obtain multiple triangular units; A Regge metric tensor describing local geometric features is defined on each triangular unit; Calculate the discrete curvature of the simulation result in parameter space based on the Regge metric tensor; According to the distribution characteristics of the discrete curvature, an optimal path for parameter adjustment is determined, and specific parameter optimization suggestions are generated based on the optimal path.

7. The method according to claim 6, characterized in that Intelligent analysis of simulation results based on machine learning to identify potential risks and make optimization suggestions, including: Analyze the area with large discrete curvature in parameter space and define it as parameter sensitive area; Calculate the Regge metric features of parameter-sensitive areas, including the principal curvature value and principal curvature direction; Based on the Regge metric feature, high curvature areas with discrete curvature anomalies are identified as potential risk points of the hydropower station; Construct an optimization objective function that includes parameter-sensitive region constraints; Using a geodesic search algorithm, based on the optimization objective function, parameters are optimized to obtain optimization parameters that avoid high-risk areas; By evaluating the changes in the Regge metric before and after optimization, the stability and reliability of the optimization results are verified.

8. The method according to claim 1, characterized in that Intelligent analysis of simulation results based on machine learning to identify potential risks and make optimization suggestions, including: Genetic algorithms are used to preliminarily optimize the key parameters of the transformation plan; According to the preliminary optimization results, a transformation process model describing the dynamic change process of parameters is established; Performing transition state analysis based on the transformation process model to evaluate the impact of parameter changes on the hydropower station; The reliability of the parameter optimization results is evaluated by calculating the reconstruction error and analyzing the influence of different decomposition orders.

9. The method according to claim 2, characterized in that: The actual operation data of the hydropower station is collected, including: Set up data collection points, including key monitoring points of hydraulic structures, unit operation parameter monitoring points and control system monitoring points; Establish a data acquisition system to collect power plant operation data in real time; The collected data are preprocessed, including data cleaning, outlier processing and data standardization.

10. An intelligent pumped storage hydropower station reconstruction design simulation verification system, characterized in that: include: A model building module, used to establish a digital twin model of the pumped-storage hydropower station, wherein the digital twin model includes a hydraulic structure model, an electromechanical equipment model, and a control system model; A parameter optimization module, used to import the transformation design scheme into the digital twin model and optimize the model parameters based on the deep learning algorithm and historical operation data; A simulation calculation module is used to perform multi-operating condition simulation verification through the optimized digital twin model to obtain simulation results, which include: dynamic response characteristics of hydraulic structures, electromechanical equipment and control systems, and the impact of the transformation design scheme on the overall performance of the hydropower station; Intelligent evaluation module, which is used to intelligently analyze simulation results based on machine learning, identify potential risks and make optimization suggestions.

Images