A method and system for intelligent management of a hydropower station

Abstract

The application belongs to the technical field of intelligent management, and discloses a kind of hydropower station intelligent management method and system.The method comprises the following steps: cloud data center, data analysis model and hydropower station intelligent management model are constructed;Intelligent management base station, according to the physical entity data of target physical entity, constructs OPC UA instance;Data acquisition device, real-time monitoring data are collected and written into OPC UA instance;Intelligent management base station, using data analysis model, carries out data analysis;Intelligent management base station, using hydropower station intelligent management model, carries out hydropower station intelligent management;Intelligent management base station, real-time OPC UA information data, real-time data analysis result and real-time hydropower station intelligent management strategy are encrypted and uploaded to cloud data center for distributed storage.The application solves the problems of insufficient data processing capacity, lack of adaptive ability and excessive manual intervention in the prior art.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of intelligent management, and particularly relates to an intelligent management method and system for a hydropower station. BACKGROUND

[0002] With the transformation of global energy structure and the promotion of the "double carbon" target, hydropower, as a clean and renewable energy form, becomes increasingly important. A hydropower station is a comprehensive engineering facility that can convert water energy into electric energy. It generally includes a reservoir formed by water retaining and water releasing structures, a diversion system, a power plant building, and mechanical and electrical equipment. The high water level of the reservoir flows into the plant through the diversion system to drive the hydroelectric generator set to generate electric energy, which is then input into the power grid through the booster transformer, switch station and transmission line.

[0003] The safe operation of a hydropower station is of great importance. An intelligent management system can monitor equipment status in real time, predict potential faults, and timely maintenance and repair to reduce the risk of accidents and ensure the stability and safety of power supply. Through automation and intelligent management, the need for manual operation and inspection can be reduced, and the operation and maintenance cost can be reduced. For example, using drones and automated inspection robots for equipment inspection not only improves inspection efficiency, but also reduces labor costs.

[0004] The existing intelligent management technology for hydropower stations has the following defects:

[0005] 1) The existing intelligent management technology for hydropower stations cannot handle large-scale real-time data, resulting in inaccurate and timely data analysis, and insufficient processing capacity for complex data, making it difficult to extract valuable information from massive data;

[0006] 2) The existing intelligent management technology for hydropower stations lacks sufficient adaptive ability and cannot adapt to the dynamic changes of the hydropower station operating environment, and has poor adaptability to specific conditions of different hydropower stations such as operating conditions and equipment types;

[0007] 3) Although the existing intelligent management technology for hydropower stations has achieved partial automation, the existing system still requires manual intervention in many cases, increasing the operating cost and error probability, and the manual decision-making process may be affected by subjective factors, resulting in unscientific decision-making. SUMMARY

[0008] In order to solve the problems of insufficient data processing capacity, lack of adaptive ability and excessive manual intervention in the existing technology, the application aims to provide an intelligent management method and system for a hydropower station.

[0009] The technical solution adopted by the application is:

[0010] An intelligent management method for a hydropower station, comprising the following steps:

[0011] The cloud data center uses an artificial intelligence algorithm to build a data analysis model and a hydropower station intelligent management model, and deploys the model into an intelligent management base station of each hydropower station connected to the cloud data center.

[0012] The intelligent management base station builds a corresponding OPC UA instance in a pre-set OPC UA server according to the physical entity data of the target physical entity.

[0013] The data acquisition device acquires real-time monitoring data of the target physical entity of the hydropower station, and uses an OPC UA protocol to write the real-time monitoring data into the corresponding OPC UA instance in the OPC UA server.

[0014] The intelligent management base station extracts real-time OPC UA information data of the OPC UA instance, and uses the data analysis model to perform data analysis on the real-time OPC UA information data to obtain real-time data analysis results.

[0015] The intelligent management base station uses the hydropower station intelligent management model to perform hydropower station intelligent management according to the real-time data analysis results, and obtains and executes real-time hydropower station intelligent management strategies.

[0016] The intelligent management base station encrypts and uploads the real-time OPC UA information data, the real-time data analysis results, and the real-time hydropower station intelligent management strategies to the cloud data center for distributed storage.

[0017] Further, the target physical entity includes a water turbine, a generator, a transformer, a switch station and power distribution equipment, a gate and flood discharge system, and a dam and hydraulic structure.

[0018] Further, the physical entity data includes basic information, monitoring indicators, indicator attributes, and indicator relationships of the target physical entity.

[0019] The real-time monitoring data includes real-time electrical signal monitoring data, real-time environmental monitoring data, real-time mechanical monitoring data, and real-time communication monitoring data.

[0020] Further, the cloud data center uses an artificial intelligence algorithm to build a data analysis model and a hydropower station intelligent management model, and deploys the model into an intelligent management base station of each hydropower station connected to the cloud data center, including the following steps:

[0021] The cloud data center acquires historical monitoring data of a plurality of historical physical entities, and pre-processes the data to obtain a plurality of pre-processed historical monitoring data.

[0022] The cloud data center uses an OPC UA protocol to perform graph structure conversion on the plurality of pre-processed historical monitoring data to obtain a plurality of historical OPC UA information graphs.

[0023] According to a plurality of historical OPC UA information graphs, using a deep learning algorithm, a data analysis model is constructed, a plurality of historical data analysis results are generated, and first model metadata of the data analysis model is extracted;

[0024] According to a plurality of historical data analysis results, using an enhanced reinforcement learning algorithm, a smart management model of the hydropower station is constructed, and second model metadata of the smart management model of the hydropower station is extracted;

[0025] The first model metadata and the second model metadata are transmitted to a smart management base station of each hydropower station connected to a cloud data center;

[0026] The smart management base station reconstructs the model according to the first model metadata and the second model metadata to obtain a reconstructed data analysis model and a reconstructed smart management model of the hydropower station.

[0027] Further, the data analysis model is constructed based on a GCN-LSTM algorithm, and the data analysis model comprises a graph feature extraction module constructed based on a GCN algorithm and a data analysis module constructed based on an LSTM algorithm connected in sequence.

[0028] Further, the smart management model of the hydropower station is constructed based on an MPO-DQN algorithm, and the smart management model of the hydropower station comprises a meta-policy optimization module constructed based on an MPO algorithm and a reinforcement learning module constructed based on a DQN algorithm, the reinforcement learning module is provided with an agent, a deep Q network and an experience replay pool, the agent is connected with the deep Q network, the experience replay pool and the meta-policy optimization module respectively, and the meta-policy optimization module is connected with the experience replay pool.

[0029] Further, the smart management base station constructs a corresponding OPC UA instance in a pre-set OPC UA server according to the physical entity data of the target physical entity, comprising the following steps:

[0030] The smart management base station defines a plurality of object nodes according to the basic information and the monitoring indexes in the physical entity data of any target physical entity;

[0031] According to the index attributes in the physical entity data of the same target physical entity, the attribute nodes of each node are defined, and the node relationship between the object nodes is defined according to the index relationship in the physical entity data;

[0032] All target physical entities managed by the smart management base station are traversed to obtain a plurality of object nodes, a plurality of attribute nodes and a plurality of node relationships of each target physical entity;

[0033] Integrate the several object nodes, the several attribute nodes and the several node relationships of all target physical entities, build an OPC UA information model using the OPC UA protocol, and extract corresponding third model metadata;

[0034] Send the third model metadata to an OPC UA server of an intelligent management base station, and build a corresponding OPC UA instance in an address space of the OPC UA server according to the third model metadata.

[0035] Further, the intelligent management base station extracts real-time OPC UA information data of the OPC UA instance, and uses a data analysis model to perform data analysis on the real-time OPC UA information data to obtain real-time data analysis results, including the following steps:

[0036] The intelligent management base station extracts real-time OPC UA information data of the OPC UA instance, and performs graph structure conversion on the real-time OPC UA information data to obtain a real-time OPC UA information graph.

[0037] The intelligent management base station extracts real-time graph features of the real-time OPC UA information graph using a graph feature extraction module of the data analysis model, and inputs the real-time graph features into the data analysis module.

[0038] The intelligent management base station uses the data analysis module to perform data analysis according to the real-time graph features to obtain real-time data analysis results.

[0039] Further, the intelligent management base station uses a hydropower station intelligent management model to perform hydropower station intelligent management according to the real-time data analysis results, obtains and executes real-time hydropower station intelligent management strategies, including the following steps:

[0040] The intelligent management base station extracts real-time meta-strategy optimization experience of a meta-strategy optimization module from an experience replay pool of the hydropower station intelligent management model according to the real-time data analysis results.

[0041] The intelligent management base station generates real-time network parameters of a deep Q network of a reinforcement learning module using the meta-strategy optimization module according to the real-time meta-strategy optimization experience.

[0042] The intelligent management base station adjusts the deep Q network of the reinforcement learning module according to the real-time network parameters to obtain an adjusted reinforcement learning module.

[0043] The intelligent management base station extracts real-time reinforcement learning experience of the reinforcement learning module from the experience replay pool of the hydropower station intelligent management model according to the real-time data analysis results.

[0044] The intelligent management base station uses the adjusted reinforcement learning module to perform hydropower station intelligent management according to the real-time data analysis results and the real-time reinforcement learning experience to obtain real-time hydropower station intelligent management strategies.

[0045] Based on the real-time intelligent management strategy of the hydropower station, real-time control commands are generated for the target physical entity and the target auxiliary control equipment, and then sent to the target physical entity and the target auxiliary control equipment.

[0046] A hydropower station intelligent management system is provided to realize an intelligent management method for hydropower stations. The system includes a cloud data center and several intelligent management base stations. The cloud data center is communicatively connected to several intelligent management base stations. Each intelligent management base station is set up in a corresponding hydropower station. The intelligent management base station is communicatively connected to several data acquisition devices, several target physical entities, and target auxiliary control equipment of the corresponding hydropower station. The intelligent management base station is equipped with an OPC UA server.

[0047] The beneficial effects of this invention are as follows:

[0048] This invention discloses an intelligent management method and system for hydropower stations. The data analysis model and intelligent management model for hydropower stations, constructed using artificial intelligence algorithms, can process and analyze massive amounts of data in real time and efficiently, thereby improving the accuracy and timeliness of data analysis. The intelligent management model can automatically adjust the management strategy of the hydropower station according to its operating status and specific conditions, improving adaptability and management efficiency for different hydropower stations. The data analysis model can uncover deep features of monitoring data, significantly improving the accuracy and reliability of fault prediction. By constructing a unified OPC UA information model, standardized conversion between different data sources is achieved, simplifying the data processing flow and improving processing efficiency. By enhancing automation and intelligence, manual intervention is significantly reduced, lowering operating costs and error rates, making the decision-making process more scientific and objective.

[0049] Other beneficial effects of the present invention will be further explained in the specific embodiments. Attached Figure Description

[0050] Figure 1 This is a flowchart of the intelligent management method for hydropower stations in this invention.

[0051] Figure 2 This is a structural block diagram of the intelligent management system for hydropower stations in this invention. Detailed Implementation

[0052] The present invention will be further explained below with reference to the accompanying drawings and specific embodiments.

[0053] Example 1:

[0054] like Figure 1 As shown in the figure, this embodiment provides a smart management method for hydropower stations, including the following steps:

[0055] S1: a cloud data center, using an artificial intelligence algorithm, constructing a data analysis model and a hydropower station intelligent management model, and deploying the model to an intelligent management base station of each hydropower station connected to the cloud data center, including the following steps:

[0056] S1-1: the cloud data center collects historical monitoring data of a plurality of historical physical entities, and pre-processes the data to obtain a plurality of pre-processed historical monitoring data;

[0057] S1-2: using an Object Linking and Embedding for Process Control Unified Architecture (OPC UA) protocol of a process control unified architecture, performing graph structure conversion on the plurality of pre-processed historical monitoring data to obtain a plurality of historical OPC UA information graphs;

[0058] S1-3: using a deep learning algorithm, constructing a data analysis model according to the plurality of historical OPC UA information graphs, generating a plurality of historical data analysis results, and extracting first model metadata of the data analysis model;

[0059] The data analysis model is constructed based on a Graph Convolutional Network (GCN)-Long Short-Term Memory (LSTM) algorithm, and the data analysis model includes a graph feature extraction module constructed based on a GCN algorithm and a data analysis module constructed based on an LSTM algorithm connected in sequence;

[0060] The GCN network performs feature propagation on the OPC UA information graph through an operation similar to convolution, extracts node features of each object node and attribute node in the OPC UA information graph, and edge features between the object nodes and between the object nodes and the attribute nodes, to form graph features, and the LSTM network performs prediction analysis according to the graph features;

[0061] S1-4: using an enhanced reinforcement learning algorithm, constructing a hydropower station intelligent management model according to the plurality of historical data analysis results, and extracting second model metadata of the hydropower station intelligent management model;

[0062] The intelligent management model of the hydropower station is constructed based on a Meta-Policy Optimization (MPO)-Deep Q Network (DQN) algorithm, and the intelligent management model of the hydropower station comprises a Meta-Policy Optimization module constructed based on the MPO algorithm and a reinforcement learning module constructed based on the DQN algorithm, the reinforcement learning module is provided with an agent, a deep Q network and an experience replay pool, the agent is connected with the deep Q network, the experience replay pool and the Meta-Policy Optimization module respectively, and the Meta-Policy Optimization module is connected with the experience replay pool;

[0063] The Meta-Policy Optimization module is used for optimizing the deep Q network parameters in the reinforcement learning module, so that the parameters can quickly adapt to new, never-seen tasks, improve the generalization ability of the model, and update the deep Q network based on previous learning experience even under never-seen data; through Meta-Policy Optimization, the DQN algorithm can converge faster, reduce the amount of data and time required for training, and significantly improve the learning efficiency; the reinforcement learning module outputs the intelligent management strategy of the hydropower station according to the data analysis result;

[0064] According to the analysis result of a plurality of historical data, an enhanced reinforcement learning algorithm is used to construct an intelligent management model of a hydropower station, and second model metadata of the intelligent management model of the hydropower station is extracted, comprising the following steps:

[0065] S1-4-1: The intelligent management problem of the hydropower station is taken as a simulation environment of the DQN algorithm, and the network parameter generation problem of the DQN algorithm is divided into a plurality of subtasks of the MPO algorithm according to different intelligent management scenarios of the hydropower station;

[0066] S1-4-2: An initial Meta-Policy Optimization module is constructed using the MPO algorithm, and a plurality of historical user portraits are inputted, and the training is performed under a plurality of subtasks to obtain a final Meta-Policy Optimization module, and a plurality of historical Meta-Policy Optimization experiences are generated;

[0067] S1-4-3: The final Meta-Policy Optimization module is used to output the historical network parameters of the deep Q network, and the initial reinforcement learning module is constructed using the DQN algorithm according to the historical network parameters;

[0068] S1-4-4: According to the analysis result of the historical data, a state space is defined, an action space is defined according to a preset intelligent management action of the hydropower station, and a reward function is set;

[0069] S1-4-5: Based on the state space, the action space and the reward function, a plurality of historical data analysis results are inputted, the agent and the deep Q network of the initial reinforcement learning module are trained to obtain a final reinforcement learning module, and a plurality of historical reinforcement learning experiences are generated;

[0070] S1-4-6: Store the historical meta-strategy optimization experience and the historical reinforcement learning experience into the experience replay pool of the final reinforcement learning module, and integrate the final meta-strategy optimization module and the final reinforcement learning module to obtain the intelligent management model of the hydropower station;

[0071] S1-4-7: Extract the second model metadata of the intelligent management model of the hydropower station;

[0072] S1-5: Store the first model metadata and the second model metadata into the intelligent management base station of each hydropower station connected to the cloud data center;

[0073] S1-6: The intelligent management base station reconstructs the model according to the first model metadata and the second model metadata to obtain the reconstructed data analysis model and the reconstructed intelligent management model of the hydropower station;

[0074] S2: The intelligent management base station constructs the corresponding OPC UA instance in the pre-set OPC UA server according to the physical entity data of the target physical entity, including the following steps:

[0075] S2-1: The intelligent management base station defines a plurality of object nodes according to the basic information and monitoring indicators in the physical entity data of any target physical entity;

[0076] The target physical entity includes a hydraulic turbine, a generator, a transformer, a switch station and a power distribution device, a gate and a flood discharge system, and a dam and a hydraulic structure;

[0077] The physical entity data includes the basic information, monitoring indicators, indicator attributes and indicator relationships of the target physical entity;

[0078] S2-2: According to the indicator attributes in the physical entity data of the same target physical entity, define the attribute nodes of each node, and according to the indicator relationships in the physical entity data, define the node relationships between the object nodes;

[0079] S2-3: Traverse all target physical entities managed by the intelligent management base station to obtain a plurality of object nodes, a plurality of attribute nodes and a plurality of node relationships of each target physical entity;

[0080] S2-4: Integrate the plurality of object nodes, the plurality of attribute nodes and the plurality of node relationships of all target physical entities, use the OPC UA protocol to construct the OPC UA information model, and extract the corresponding third model metadata;

[0081] S2-5: Send the third model metadata to the OPC UA server of the intelligent management base station, and construct the corresponding OPC UA instance in the address space of the OPC UA server according to the third model metadata;

[0082] S3: a data acquisition device acquires real-time monitoring data of a target physical entity of the hydropower station, and writes the real-time monitoring data into a corresponding OPC UA instance in an OPC UA server using an OPC UA protocol;

[0083] The real-time monitoring data includes real-time electrical signal monitoring data, real-time environmental monitoring data, real-time mechanical monitoring data, and real-time communication monitoring data;

[0084] S4: an intelligent management base station extracts real-time OPC UA information data of the OPC UA instance, and uses a data analysis model to perform data analysis on the real-time OPC UA information data to obtain real-time data analysis results, including the following steps:

[0085] S4-1: the intelligent management base station extracts real-time OPC UA information data of the OPC UA instance, and performs graph structure conversion on the real-time OPC UA information data to obtain a real-time OPC UA information graph;

[0086] S4-2: a graph feature extraction module of the data analysis model extracts real-time graph features of the real-time OPC UA information graph, and inputs the real-time graph features into a data analysis module;

[0087] S4-3: the data analysis module performs data analysis according to the real-time graph features to obtain real-time data analysis results;

[0088] The real-time data analysis results include operation state analysis results of the target physical entity, potential fault prediction conditions, abnormal index analysis results, and missing data acquisition conditions;

[0089] S5: the intelligent management base station uses a hydropower station intelligent management model to perform hydropower station intelligent management according to the real-time data analysis results, obtains and executes real-time hydropower station intelligent management strategies, including the following steps:

[0090] S5-1: the intelligent management base station extracts real-time meta-strategy optimization experience of a meta-strategy optimization module in an experience replay pool of the hydropower station intelligent management model according to the real-time data analysis results;

[0091] S5-2: the meta-strategy optimization module generates real-time network parameters of a deep Q network of a reinforcement learning module according to the real-time meta-strategy optimization experience;

[0092] S5-3: the deep Q network of the reinforcement learning module is adjusted according to the real-time network parameters to obtain an adjusted reinforcement learning module;

[0093] S5-4: real-time reinforcement learning experience of the reinforcement learning module is extracted in the experience replay pool of the hydropower station intelligent management model according to the real-time data analysis results;

[0094] S5-5: Based on real-time data analysis results and real-time reinforcement learning experience, an adjusted reinforcement learning module is used to conduct intelligent management of hydropower stations, resulting in a real-time intelligent management strategy for hydropower stations, including the following steps:

[0095] S5-5-1: The state space of the reinforcement learning module is updated and adjusted based on the real-time data analysis results, resulting in the updated state space S'=[s'1,...,s' i" ,...,s' I' ], where s' i" For the updated i"th state, i" is the state indicator and I' is the total number of dimensions in the state space;

[0096] S5-5-2: Based on real-time reinforcement learning experience, update and adjust the action space of the reinforcement learning module to obtain the updated action space A'=[a'1,...,a' j" ,...,a' I ], where a' j" For the updated j"th action, j" is the action indicator, and I is the total number of dimensions in the action space;

[0097] S5-5-3: Input the updated state space into the adjusted reinforcement learning module, and use the adjusted deep Q network to generate the Q value of each possible action in the updated action space based on the updated state space.

[0098] The formula is:

[0099] Q(s' p' ,a' p' )=(1-α")·Q(s p' ,a p' )+α"·(R(s p' ,a p' ,s' p' )+γ·Q max (s p' ,a p' ))

[0100] In the formula, Q(s' p' ,a' p' ) represents the updated state s' p' and the update action a' p' The corresponding updated Q value; Q(s) p' ,a p' ) represents state s p' and action a p' The corresponding predicted Q value; α" is the learning rate; Q max (s p' ,a p') is the highest predicted Q-value; p' is the composite indicator; g is the update parameter; R(s p' ,a p' ,s' p' ) is the reward value based on action a p' transformation of state s p' to updated state s' p' .

[0101] S5-5-4: Repeat the above Q-value update step until the number of iterations reaches the iteration threshold, use the greedy strategy to select the possible action with the highest updated Q-value as the execution action, and obtain the real-time hydropower station intelligent management strategy according to the execution action;

[0102] The real-time hydropower station intelligent management strategy includes a control scheme of the target physical entity and a control scheme of the target auxiliary control device, etc.

[0103] S5-6: According to the real-time hydropower station intelligent management strategy, generate real-time control instructions of the target physical entity and the target auxiliary control device, and send the real-time control instructions to the target physical entity and the target auxiliary control device;

[0104] S6: The intelligent management base station uploads the real-time OPC UA information data, the real-time data analysis result, and the real-time hydropower station intelligent management strategy to the cloud data center for distributed storage after encryption, including the following steps:

[0105] S6-1: The intelligent management base station compresses and encrypts the real-time OPC UA information data of the target physical entity, the real-time data analysis result, and the real-time hydropower station intelligent management strategy using an asymmetric encryption algorithm according to the encryption key to obtain a real-time encrypted compression package, and uploads it to the cloud data center;

[0106] S6-2: The cloud data center decrypts and decompresses the real-time encrypted compression package using an asymmetric decryption algorithm according to the decryption key to obtain the decrypted real-time OPC UA information data of the target physical entity, the decrypted real-time data analysis result, and the decrypted real-time hydropower station intelligent management strategy;

[0107] S6-3: The decrypted real-time OPC UA information data of the target physical entity, the decrypted real-time data analysis result, and the decrypted real-time hydropower station intelligent management strategy are stored in a distributed manner using a blockchain network.

[0108] Embodiment 2:

[0109] As Figure 2As shown, the embodiment provides a hydropower station intelligent management system for realizing a hydropower station intelligent management method, the system comprising a cloud data center and a plurality of intelligent management base stations, the cloud data center being in communication connection with the plurality of intelligent management base stations respectively, each intelligent management base station being correspondingly arranged in a hydropower station, the intelligent management base station being in communication connection with a plurality of data acquisition devices and a plurality of target physical entities and target auxiliary control equipment of the corresponding hydropower station, and the intelligent management base station being provided with an OPC UA server;

[0110] The cloud data center is used for constructing a data analysis model and a hydropower station intelligent management model using an artificial intelligence algorithm, and deploying the model into the intelligent management base station of each hydropower station connected with the cloud data center; the real-time OPC UA information data, real-time data analysis results and real-time hydropower station intelligent management strategies sent by the intelligent management base station are stored in a distributed manner;

[0111] The intelligent management base station is used for constructing a corresponding OPC UA instance in the pre-set OPC UA server according to the physical entity data of the target physical entity; real-time OPC UA information data of the OPC UA instance is extracted, and the data analysis model is used to perform data analysis on the real-time OPC UA information data to obtain real-time data analysis results; the hydropower station intelligent management model is used to perform hydropower station intelligent management according to the real-time data analysis results, to obtain and execute real-time hydropower station intelligent management strategies; the real-time OPC UA information data, real-time data analysis results and real-time hydropower station intelligent management strategies are encrypted and uploaded to the cloud data center;

[0112] The data acquisition device is used for acquiring real-time monitoring data of the target physical entity of the hydropower station, and writing the real-time monitoring data into the corresponding OPC UA instance in the OPC UA server using the OPC UA protocol.

[0113] The application discloses a hydropower station intelligent management method and system, a data analysis model and a hydropower station intelligent management model constructed using an artificial intelligence algorithm, which can process and analyze massive data in real time and efficiently, thereby improving the accuracy and timeliness of data analysis; the hydropower station intelligent management model can automatically adjust the management strategy of the hydropower station according to the operation state and specific conditions of the hydropower station, thereby improving the adaptability and management efficiency of different hydropower stations; the data analysis model can mine deep features of monitoring data, thereby significantly improving the accuracy and reliability of fault prediction; by constructing a unified OPC UA information model, standardized conversion between different data sources is realized, the data processing process is simplified, and the processing efficiency is improved; by enhancing the automation and intelligent level, manual intervention is significantly reduced, the operation cost and error rate are reduced, and the decision-making process is more scientific and objective.

[0114] The above detailed description of the specific embodiments of the present application has been given to understand the purpose, technical solutions and beneficial effects of the present application. It should be understood that the above description is only a specific embodiment of the present application and is not used to limit the protection scope of the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A method for intelligent management of hydropower stations, characterized in that: Includes the following steps: The cloud data center uses artificial intelligence algorithms to build data analysis models and intelligent management models for hydropower stations, and deploys them to the intelligent management base stations of each hydropower station connected to the cloud data center, including the following steps: The cloud data center collects historical monitoring data of several historical physical entities and preprocesses it to obtain several preprocessed historical monitoring data. Using the OPC UA protocol, several preprocessed historical monitoring data were transformed into graph structures to obtain several historical OPC UA information graphs. Based on several historical OPC UA infographics, a data analysis model is constructed using deep learning algorithms, several historical data analysis results are generated, and the first model metadata of the data analysis model is extracted. The data analysis model is constructed based on the GCN-LSTM algorithm, and the data analysis model includes a graph feature extraction module constructed based on the GCN algorithm and a data analysis module constructed based on the LSTM algorithm, which are connected in sequence. Based on the analysis results of several historical data, an enhanced reinforcement learning algorithm is used to construct a smart management model for hydropower stations, and the second model metadata of the smart management model for hydropower stations is extracted. The intelligent management model for hydropower stations is constructed based on the MPO-DQN algorithm. The intelligent management model for hydropower stations includes a meta-policy optimization module constructed based on the MPO algorithm and a reinforcement learning module constructed based on the DQN algorithm. The reinforcement learning module is equipped with an agent, a deep Q network and an experience replay pool. The agent is connected to the deep Q network, the experience replay pool and the meta-policy optimization module respectively. The meta-policy optimization module is connected to the experience replay pool. Send the first model metadata and the second model metadata to the smart management base station of each hydropower station connected to the cloud data center; The intelligent management base station reconstructs the model based on the metadata of the first model and the metadata of the second model, resulting in a reconstructed data analysis model and a reconstructed intelligent management model for the hydropower station. The intelligent management base station also includes building a corresponding OPC UA instance in a pre-configured OPC UA server based on the physical entity data of the target physical entity; The data acquisition device collects real-time monitoring data of the target physical entities of the hydropower station and uses the OPC UA protocol to write the real-time monitoring data into the corresponding OPC UA instance in the OPC UA server. The intelligent management base station also includes extracting real-time OPC UA information data from OPC UA instances, using a data analysis model to analyze the real-time OPC UA information data, obtaining real-time data analysis results, using a hydropower station intelligent management model to perform intelligent management of the hydropower station based on the real-time data analysis results, obtaining and executing real-time hydropower station intelligent management strategies, and encrypting and uploading the real-time OPC UA information data, real-time data analysis results, and real-time hydropower station intelligent management strategies to a cloud data center for distributed storage.

2. The intelligent management method for a hydropower station according to claim 1, characterized in that: The target physical entities include turbines, generators, transformers, switchyards and power distribution equipment, gates and flood discharge systems, as well as dams and hydraulic structures.

3. The intelligent management method for a hydropower station according to claim 1, characterized in that: The physical entity data includes the basic information of the target physical entity, monitoring indicators, indicator attributes, and indicator relationships; The real-time monitoring data includes real-time electrical signal monitoring data, real-time environmental monitoring data, real-time mechanical monitoring data, and real-time communication monitoring data.

4. The intelligent management method for a hydropower station according to claim 3, characterized in that: The process of constructing a corresponding OPC UA instance in a pre-configured OPC UA server based on the physical entity data of the target physical entity includes the following steps: Intelligent management base stations define several object nodes based on the basic information and monitoring indicators in the physical entity data of any target physical entity; Based on the indicator attributes in the physical entity data of the same target physical entity, define the attribute node of each node, and define the node relationship between object nodes based on the indicator relationship in the physical entity data. Traverse all target physical entities managed by the intelligent management base station to obtain several object nodes, several attribute nodes, and several node relationships for each target physical entity; Integrate several object nodes, several attribute nodes, and several node relationships of all target physical entities, use the OPCUA protocol to construct an OPC UA information model, and extract the corresponding third-party model metadata; The third model metadata is sent to the OPC UA server of the intelligent management base station. Based on the third model metadata, the corresponding OPC UA instance is constructed in the address space of the OPC UA server.

5. The intelligent management method for a hydropower station according to claim 1, characterized in that: The process of extracting real-time OPC UA information data from OPC UA instances and using a data analysis model to analyze the real-time OPC UA information data to obtain real-time data analysis results includes the following steps: The intelligent management base station extracts real-time OPC UA information data from OPC UA instances and performs graph structure transformation on the real-time OPC UA information data to obtain a real-time OPC UA information graph. The graph feature extraction module of the data analysis model is used to extract real-time graph features of the real-time OPC UA information graph, and the real-time graph features are input into the data analysis module. Using the data analysis module, data analysis is performed based on the characteristics of the real-time graph to obtain real-time data analysis results.

6. The intelligent management method for a hydropower station according to claim 1, characterized in that: The process of using a hydropower station intelligent management model based on real-time data analysis results to obtain and execute real-time hydropower station intelligent management strategies includes the following steps: The intelligent management base station extracts the real-time meta-policy optimization experience of the meta-policy optimization module from the experience playback pool of the intelligent management model of the hydropower station based on the real-time data analysis results. Based on real-time meta-policy optimization experience, the meta-policy optimization module is used to generate real-time network parameters for the deep Q-network of the reinforcement learning module. Based on real-time network parameters, the deep Q-network of the reinforcement learning module is adjusted to obtain the adjusted reinforcement learning module. Based on the real-time data analysis results, the real-time reinforcement learning experience of the reinforcement learning module is extracted from the experience playback pool of the intelligent management model of the hydropower station. Based on real-time data analysis results and real-time reinforcement learning experience, an adjusted reinforcement learning module is used to carry out intelligent management of hydropower stations, resulting in a real-time intelligent management strategy for hydropower stations. Based on the real-time intelligent management strategy of the hydropower station, real-time control commands are generated for the target physical entity and the target auxiliary control equipment, and then sent to the target physical entity and the target auxiliary control equipment.

7. A hydropower station intelligent management system, used to implement the hydropower station intelligent management method as described in any one of claims 1-6, characterized in that: The system includes a cloud data center and several intelligent management base stations. The cloud data center is communicatively connected to several intelligent management base stations. Each intelligent management base station is set up in a corresponding hydropower station. The intelligent management base station is communicatively connected to several data acquisition devices, several target physical entities, and target auxiliary control equipment of the corresponding hydropower station. The intelligent management base station is equipped with an OPC UA server.

Images