Cloud intelligent interconnection platform and method for large hydropower station

Abstract

The invention discloses a cloud intelligent interconnection platform and method for a large hydropower station, and relates to the technical field of cloud interconnection, and the platform comprises a data collection layer, a transmission layer, a cloud decision-making layer and a feedback control layer. The data acquisition layer is provided with a sensor for acquiring key operation data of the hydropower station in real time; the transmission layer adopts the Internet of Things technology, and transmits the key operation data from the data acquisition layer to the cloud decision-making layer through edge computing equipment and a network; the cloud decision-making layer processes the key operation data based on cloud computing resources, performs data analysis by applying an AI algorithm, and generates a power generation scheduling strategy; and the feedback control layer adjusts the working state of the field equipment according to the power generation scheduling strategy generated by the cloud decision-making layer. According to the large hydropower station cloud intelligent interconnection method provided by the invention, real-time data acquisition and intelligent analysis are realized, so that the timeliness and accuracy of a scheduling decision are improved, and stable operation of the hydropower station under changing conditions is ensured.

Description

Technical Field

[0001] The present invention relates to the field of cloud interconnection technology, and in particular to a cloud intelligent interconnection platform and method for a large hydropower station. Background Art

[0002] In the operation of modern hydropower plants, data collection, transmission and processing are key links to ensure efficient power generation. Existing hydropower plant systems usually use independent sensors for data collection, transmission and analysis through traditional data management methods. These systems often lack integration, resulting in inefficient data processing and long response times. In addition, many traditional hydropower plants lack real-time monitoring and dynamic scheduling capabilities, and cannot effectively respond to complex operating conditions and environmental changes.

[0003] Current technologies have limitations in many aspects. First, data transmission channels often rely on a single communication method, making it difficult to achieve fast and reliable data transmission. Second, existing data analysis methods lack intelligence and cannot fully utilize historical data for effective trend prediction and decision support, resulting in insufficient flexibility in scheduling strategies. In addition, due to the lack of effective integration of edge computing and cloud computing, real-time data analysis and response capabilities are limited, making it difficult to adapt to rapidly changing operational needs. Summary of the invention

[0004] In view of the above-mentioned problems, the present invention is proposed.

[0005] Therefore, the technical problem solved by the present invention is that in existing hydropower stations, data interaction often has delays and information islands, resulting in insufficient decision-making basis and untimely response. In addition, the lack of effective data analysis means makes it difficult for hydropower stations to quickly adjust their operating status when facing environmental changes, thereby affecting overall efficiency and safety. Therefore, the present invention enhances the linkage between various layers by introducing Internet of Things technology and cloud computing resources, and realizes efficient and intelligent data processing and decision support.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions: A cloud-based intelligent interconnection platform for large hydropower stations, comprising:

[0007] Data collection layer, transmission layer, cloud decision layer and feedback control layer;

[0008] The data collection layer is configured with sensors to collect key operating data of the hydropower station in real time;

[0009] The transmission layer uses Internet of Things technology to transmit key operating data from the data collection layer to the cloud decision layer through edge computing devices and networks;

[0010] The cloud decision layer processes key operating data based on cloud computing resources, applies AI algorithms to perform data analysis, and generates power generation scheduling strategies;

[0011] The feedback control layer adjusts the working state of the on-site equipment according to the power generation scheduling strategy generated by the cloud decision layer.

[0012] As a preferred solution of the cloud-based intelligent interconnection platform for large hydropower stations described in the present invention, the key operating data includes water flow, water level, precipitation, temperature, wind speed and generator speed;

[0013] The edge computing device uses FPGA technology to perform data preprocessing.

[0014] As a preferred solution of the cloud-based intelligent interconnection platform for large hydropower stations described in the present invention, the cloud-based decision-making layer includes a historical data analysis module and a real-time scheduling optimization module;

[0015] The historical data analysis module models the historical operation data of the key operation data through a long short-term memory network, generates a long-term trend prediction model, and outputs the parameter range of the key operation data based on the long-term trend prediction model;

[0016] The real-time scheduling optimization module is based on a deep reinforcement learning algorithm and generates an optimal scheduling strategy and adjusts the working status of the generator set according to the key operating data and parameter range of the key operating data input in real time.

[0017] Another object of the present invention is to provide a cloud-based intelligent interconnection method for large hydropower stations, which solves the problems of low data preprocessing efficiency and single analysis methods in traditional methods. Traditional methods often rely on static data analysis and fail to fully utilize the potential of historical data, resulting in insufficient flexibility in scheduling decisions. The present invention uses FPGA for data preprocessing, and combines long short-term memory networks and deep reinforcement learning algorithms for intelligent analysis, thereby achieving dynamic optimization in real-time monitoring and prediction of key operating data, significantly improving the scientificity and effectiveness of scheduling strategies.

[0018] In order to solve the above technical problems, the present invention provides the following technical solutions: A cloud-based intelligent interconnection method for a large hydropower station, comprising:

[0019] Configure sensors to collect key operating data of the hydropower station in real time and transmit the collected data to edge computing devices;

[0020] Pre-process key operation data through FPGA in edge computing devices;

[0021] Call the stored historical key operation data, analyze the historical operation data based on the long short-term memory network, generate a long-term trend prediction model, and output the parameter range of the key operation data;

[0022] Based on the deep reinforcement learning algorithm, the optimal scheduling strategy is generated according to the key operating data and the parameter range;

[0023] Adjust the working status of field equipment based on the optimal scheduling strategy.

[0024] As a preferred solution of the cloud-based intelligent interconnection method for large hydropower stations described in the present invention, wherein: the data preprocessing of key operation data includes receiving key operation data transmitted in real time from sensors;

[0025] Use digital filters to low-pass filter the received key operating data to remove high-frequency noise and retain valid signals;

[0026] Apply the mean normalization method to the filtered data to standardize the key operating data of different magnitudes to the same dimension;

[0027] Calculate the time series difference of key operating data, remove the trend of data, and smooth the data;

[0028] Store the processed data in a temporary data cache.

[0029] As a preferred solution of the cloud-based intelligent interconnection method for large hydropower stations described in the present invention, wherein: the parameter range of outputting key operating data includes extracting historical key operating data from a database and arranging the data in chronological order;

[0030] Divide the extracted historical data into a training set and a validation set;

[0031] The LSTM model was constructed, and the number of input layer nodes was set to the number of data features, including water flow, water level, precipitation, temperature, wind speed, and generator speed, to capture the time dependency of the data;

[0032] During the model training process, the mean square error is used as the loss function, and the Adam optimization algorithm is used to update the model parameters until the loss function converges;

[0033] Generate parameter ranges for key operating data, including expected minimum and maximum values ​​for each key parameter.

[0034] As a preferred solution of the cloud-based intelligent interconnection method for large hydropower stations described in the present invention, wherein: the generating of the optimal dispatching strategy according to the key operating data and the parameter range includes defining a state space, the state includes real-time key operating data and the parameter range output by LSTM, and is used to describe the current operating status of the system, and the state includes the expected range of water flow, water level, precipitation, temperature, wind speed and generator speed;

[0035] Determine the action space, which includes options such as adjusting the power output of the generator set, starting or stopping, and select the optimal operation;

[0036] Use the Q-learning algorithm to construct the value function, initialize the Q values ​​of all state-action pairs, and set the learning rate and discount factor;

[0037] The ε-greedy strategy is used to balance exploration and exploitation, and the environment feedback is received and the Q value is updated after the action is performed.

[0038] As a preferred solution of the cloud-based intelligent interconnection method for large hydropower stations described in the present invention, wherein: the adjusting the working state of the field equipment based on the optimal scheduling strategy includes converting the generated optimal scheduling strategy into a control instruction, and sending it to the control system of the field equipment through a control bus;

[0039] The field equipment receives the control instructions and performs the corresponding adjustment operations;

[0040] On-site equipment monitors its operating status in real time and collects equipment status data regularly;

[0041] The actual operation status information is sent back to the cloud decision-making layer through the data feedback channel for status assessment and deviation analysis;

[0042] If the actual operating status deviates from the expected scheduling strategy, the automatic compensation mechanism will be triggered to correct the equipment status by adjusting the control instructions to maintain the stable operation of the system.

[0043] A computer device comprises a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the above-mentioned large-scale hydropower station cloud-based intelligent interconnection method when executing the computer program.

[0044] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the cloud-based intelligent interconnection method for a large hydropower station as described above.

[0045] Beneficial effects of the invention: The cloud-based intelligent interconnection method for large hydropower stations provided by the invention improves the timeliness and accuracy of dispatching decisions by realizing real-time data collection and intelligent analysis, and ensures the stable operation of hydropower stations under changing conditions. The intelligent data processing mechanism not only optimizes resource allocation, reduces the need for manual intervention, but also reduces operating costs. The control system based on dynamic feedback improves the responsiveness of the equipment and enhances the safety and reliability of the overall platform. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0047] Figure 1 An overall structural diagram of a large-scale hydropower station cloud-based intelligent interconnection platform provided by an embodiment of the present invention.

[0048] Figure 2 An overall flow chart of a large-scale hydropower station cloud-based intelligent interconnection method provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0049] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0050] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0051] Example 1

[0052] Reference Figure 1 , which is an embodiment of the present invention, provides a large-scale hydropower station cloud intelligent interconnection platform, including:

[0053] Data collection layer 100, transmission layer 200, cloud decision layer 300 and feedback control layer 400;

[0054] The data collection layer 100 is configured with sensors to collect key operating data of the hydropower station in real time;

[0055] The transmission layer 200 uses Internet of Things technology to transmit key operating data from the data collection layer to the cloud decision layer through edge computing devices and networks;

[0056] The cloud decision layer 300 processes key operation data based on cloud computing resources, applies AI algorithms to perform data analysis, and generates power generation scheduling strategies;

[0057] The feedback control layer 400 adjusts the working state of the on-site equipment according to the power generation scheduling strategy generated by the cloud decision layer.

[0058] The key operating data include water flow, water level, precipitation, air temperature, wind speed and generator speed;

[0059] The edge computing device uses FPGA technology to perform data preprocessing.

[0060] The cloud decision-making layer includes a historical data analysis module and a real-time scheduling optimization module;

[0061] The historical data analysis module models the historical operation data of the key operation data through a long short-term memory network, generates a long-term trend prediction model, and outputs the parameter range of the key operation data based on the long-term trend prediction model;

[0062] The real-time scheduling optimization module is based on a deep reinforcement learning algorithm and generates an optimal scheduling strategy and adjusts the working status of the generator set according to the key operating data and parameter range of the key operating data input in real time.

[0063] Example 2

[0064] Reference Figure 2 , is an embodiment of the present invention, and provides a cloud-based intelligent interconnection method for a large hydropower station, comprising:

[0065] Configure sensors to collect key operating data of the hydropower station in real time and transmit the collected data to edge computing devices;

[0066] Pre-process key operation data through FPGA in edge computing devices;

[0067] Call the stored historical key operation data, analyze the historical operation data based on the long short-term memory network, generate a long-term trend prediction model, and output the parameter range of the key operation data;

[0068] Based on the deep reinforcement learning algorithm, the optimal scheduling strategy is generated according to the key operating data and the parameter range;

[0069] Adjust the working status of field equipment based on the optimal scheduling strategy.

[0070] The data preprocessing of the key operation data includes receiving the key operation data transmitted in real time from the sensor;

[0071] Use digital filters to low-pass filter the received key operating data to remove high-frequency noise and retain valid signals;

[0072] Apply the mean normalization method to the filtered data to standardize the key operating data of different magnitudes to the same dimension;

[0073] Calculate the time series difference of key operating data, remove the trend of data, and smooth the data;

[0074] Store the processed data in a temporary data cache.

[0075] It should be noted that the sensor collects key operating data at a frequency of ten times per second to ensure the timeliness and integrity of the data. This frequency is selected based on the dynamic response characteristics of the power station to achieve timely monitoring of rapidly changing conditions.

[0076] Kalman filter is used for data preprocessing. Kalman filter can effectively reduce high-frequency noise and improve signal quality, thus ensuring the accuracy of subsequent data analysis. The advantage of this filter is that it can update the estimated value in real time, which is suitable for the processing of dynamic systems.

[0077] The received key operation data is normalized. The specific steps include calculating the mean and standard deviation of the data and converting all the data into a standard normal distribution. This processing method can effectively eliminate the impact of data of different magnitudes on the analysis model and improve the training effect of the model.

[0078] First-order difference processing is performed to remove the trend in the data and smooth the data fluctuations. This step can reduce the impact of non-stationarity on model prediction and improve the stability of subsequent analysis.

[0079] The parameter range of outputting the key operating data includes extracting the historical key operating data from the database and arranging the data in chronological order;

[0080] Divide the extracted historical data into a training set and a validation set;

[0081] The LSTM model was constructed, and the number of input layer nodes was set to the number of data features, including water flow, water level, precipitation, temperature, wind speed, and generator speed, to capture the time dependency of the data;

[0082] During the model training process, the mean square error is used as the loss function, and the Adam optimization algorithm is used to update the model parameters until the loss function converges;

[0083] Generate parameter ranges for key operating data, including expected minimum and maximum values ​​for each key parameter.

[0084] It should be noted that when extracting historical key operation data from the database, the ETL (extraction, transformation, loading) process is used to remove duplicates and process missing values ​​to ensure the accuracy and reliability of the data. Each data record must contain a timestamp and corresponding environment variables.

[0085] Input layer design: The number of input layer nodes is set according to the number of data features, such as water flow, water level, precipitation, air temperature, wind speed, and generator speed, to ensure that the model fully captures the time dependency of the data.

[0086] Model training: The mean square error is used as the loss function, and the Adam optimization algorithm is used to update the model parameters. During the training process, cross-validation is used to prevent overfitting and ensure the generalization ability of the model.

[0087] Parameter range output: After model training is completed, the expected minimum and maximum values ​​of each key parameter are calculated based on the output long-term trend prediction model. This process can effectively provide a reference for subsequent scheduling decisions.

[0088] Generating the optimal scheduling strategy according to the key operating data and the parameter range includes defining a state space, the state includes the real-time key operating data and the parameter range output by the LSTM, and is used to describe the current operating status of the system, and the state includes the expected range of water flow, water level, precipitation, temperature, wind speed and generator speed;

[0089] Determine the action space, which includes options such as adjusting the power output of the generator set, starting or stopping, and select the optimal operation;

[0090] Use the Q-learning algorithm to construct the value function, initialize the Q values ​​of all state-action pairs, and set the learning rate and discount factor;

[0091] The ε-greedy strategy is used to balance exploration and exploitation, and the environment feedback is received and the Q value is updated after the action is performed.

[0092] It should be noted that the real-time key operating data (such as water flow, water level, etc.) and the parameter range of LSTM output are combined to form a state space description. The state is defined as the expected range of each key variable in order to reflect the current operating status of the system.

[0093] Identify action options, including adjusting the power output of the generator set, starting or stopping it, and define the conditions and expected effects of each action.

[0094] Initialize the Q-values ​​of all state-action pairs to zero, and set the learning rate (e.g., 0.1) and discount factor (e.g., 0.9) to balance short-term and long-term rewards.

[0095] Use the ε-greedy strategy to explore the action space and randomly choose to explore or exploit. After the action is executed, the Q value is updated according to the feedback from the environment to gradually optimize the strategy.

[0096] The adjusting the working state of the field device based on the optimal scheduling strategy includes converting the generated optimal scheduling strategy into a control instruction and sending it to the control system of the field device through a control bus;

[0097] The field equipment receives the control instructions and performs the corresponding adjustment operations;

[0098] On-site equipment monitors its operating status in real time and collects equipment status data regularly;

[0099] The actual operation status information is sent back to the cloud decision-making layer through the data feedback channel for status assessment and deviation analysis;

[0100] If the actual operating status deviates from the expected scheduling strategy, the automatic compensation mechanism will be triggered to correct the equipment status by adjusting the control instructions to maintain the stable operation of the system.

[0101] It should be noted that based on the optimal scheduling strategy, it is converted into specific control instructions through algorithms. This process includes formatting and parameter setting of instructions to ensure the accuracy and executability of the instructions.

[0102] The generated control instructions are sent to the field device control system through the safety control bus to ensure the integrity and security of the data during transmission. After receiving the control instructions, the field device performs the corresponding adjustment operation.

[0103] On-site equipment needs to monitor its operating status in real time, collect equipment status data regularly, and send the actual operating status information back to the cloud decision-making layer through the data feedback channel. If the actual status deviates from the expected scheduling strategy, the automatic compensation mechanism is triggered to correct the equipment status by adjusting the control instructions to maintain the stable operation of the system.

[0104] Furthermore, the adjusting the working state of the field equipment based on the optimal scheduling strategy includes converting the generated optimal scheduling strategy into a control instruction. The specific steps are:

[0105] Receive the optimal dispatch strategy from the cloud decision layer, which includes the target power output, operation mode and coordinated operation requirements of the generator set.

[0106] These policy information are analyzed and key parameters are extracted through the algorithm to generate corresponding control instructions, for example:

[0107] Generator power adjustment command: Based on the specific power increase or decrease value, the command format is "adjust the generator power to XXXMW".

[0108] Device start / stop command: The format is "start / stop device Y".

[0109] The generated control instructions should adopt a standardized encoding method to ensure the compatibility and readability of the instructions among various devices.

[0110] The instructions are sent to the control system of the field device via the control bus. The control bus uses industrial standard communication protocols, such as CAN, MODBUS or EtherCAT, to ensure efficient and reliable instruction transmission. Specific implementation details include:

[0111] Redundant design: To ensure the reliability of instructions during transmission, a dual channel redundancy mechanism is set up to ensure that when the main channel fails, the backup channel can immediately take over the communication.

[0112] Real-time guarantee: Using time-triggered or event-triggered communication modes, instructions can be delivered to the target device within milliseconds to meet the needs of rapid response.

[0113] The on-site equipment is equipped with a variety of sensors to monitor its operating status in real time, including but not limited to:

[0114] Temperature sensor: monitors the internal temperature of the device to avoid overheating risks.

[0115] Pressure sensor: Ensures that the equipment operates within a safe pressure range.

[0116] Vibration sensor: monitors the vibration of equipment during operation and detects potential mechanical failures in a timely manner.

[0117] The monitoring data is fed back to the cloud decision-making layer in real time through wireless communication modules (such as Zigbee or LoRa), ensuring that the data is updated once per second to facilitate real-time analysis and adjustment.

[0118] The device status data received by the cloud decision layer will be evaluated through the following algorithms:

[0119] Statistical analysis: Use statistical methods (such as mean and standard deviation) to conduct real-time analysis of monitoring data to determine whether the equipment operation is within the expected range.

[0120] Machine learning model: Use trained machine learning models to perform deviation detection based on historical data and real-time data to automatically identify potential faults.

[0121] Evaluation indicators: Set clear evaluation indicators, such as equipment operation stability, response time, and deviation amplitude, and use these indicators to determine the deviation between the actual status and the expected scheduling strategy.

[0122] When the actual operating status is detected to deviate from the expected scheduling strategy, the system will trigger the automatic compensation mechanism. The specific operation process is as follows:

[0123] Trigger condition: Set the deviation threshold. When the deviation exceeds the preset range (such as ±5%), the compensation process is automatically started.

[0124] Compensation strategy: Different compensation strategies are adopted according to the deviation type:

[0125] PID control algorithm: Automatically adjusts control instructions based on deviation and optimizes device output for rapid recovery.

[0126] Fuzzy logic control: In situations with high uncertainty, a fuzzy controller is used to adjust the device state to balance system stability and response speed.

[0127] Execution feedback: The adjusted control instructions are sent to the field devices again through the control bus to ensure that the system maintains the best operating state.

[0128] Example 3

[0129] An embodiment of the present invention is different from the first two embodiments in that:

[0130] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc., which can store program codes.

[0131] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.

[0132] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or, if necessary, processing in another suitable manner, and then stored in a computer memory.

[0133] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, a plurality of steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0134] Example 4

[0135] As an embodiment of the present invention, a cloud-based intelligent interconnection method for a large hydropower station is provided. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through simulation experiments.

[0136] A large hydropower station was selected and a traditional power generation dispatching system was set up. The system was based on timed collection and manual dispatching, using standard sensors to monitor water flow, water level and other key parameters.

[0137] During the one-week experiment, key operating data (such as water flow, water level, air temperature, etc.) were collected every hour, and the power output and dispatch response time of the generator sets were recorded.

[0138] Based on the collected data, dispatchers manually analyze the data, develop hourly power generation plans, and adjust the power output of generator sets through traditional methods.

[0139] Record any failures that occurred during the week, as well as the response and repair times.

[0140] In the same hydropower station, the cloud-based intelligent interconnection platform of the present invention is deployed, real-time sensors are configured to collect key operating data, and edge computing equipment is built to perform data preprocessing.

[0141] During the one-week experiment, key operating data was collected in real time and preprocessed using edge computing devices, including low-pass filtering, mean normalization, and time series differencing.

[0142] Based on the processed data, the long short-term memory network (LSTM) model is used to generate long-term trend forecasts, and the deep reinforcement learning algorithm is used to generate the optimal power generation scheduling strategy in real time to automatically adjust the power output of the generator set.

[0143] The actual operating status of the on-site equipment is monitored through the feedback control layer, and the information is transmitted back to the cloud in real time for evaluation and optimization. If the actual operating status deviates from the scheduling strategy, the system will automatically compensate.

[0144] The fault conditions that occurred during the experiment were recorded, including fault identification time, response time, and repair time. The experimental results are shown in Table 1.

[0145] Table 1 Comparison of experimental results

[0146] index Experimental group (intelligent interconnection platform) Control group (traditional method) Dispatch response time (minutes) 5 15 Energy consumption per kilowatt-hour (kWh) 0.5 0.8 Equipment failure rate (times / hour) 0.1 0.3 Power generation efficiency (%) 92 85

[0147] Traditional methods rely on manual data collection and meetings, which results in a slow decision-making process and is affected by subjective factors, and cannot quickly adapt to the dynamically changing operating environment. However, the present invention can generate scheduling strategies in a short time through real-time data collection and intelligent decision-making based on cloud computing, thus significantly improving the response speed.

[0148] The present invention effectively reduces the energy consumption per kilowatt-hour, can monitor and optimize the operating status of the equipment in real time, and avoids the excessive energy consumption in traditional methods. Traditional methods usually fail to adjust the power output of the generator set in time, resulting in unnecessary energy waste. Through intelligent scheduling, the present invention achieves efficient use of resources and reduces operating costs.

[0149] Traditional methods lack real-time monitoring and cannot identify potential equipment failures in a timely manner, resulting in frequent failures. However, the present invention can quickly detect anomalies and issue early warnings through real-time data feedback and status evaluation, greatly reducing the incidence of equipment failures.

[0150] Traditional dispatching methods often make decisions based on historical data, which is difficult to reflect the current operating status. However, the intelligent algorithm of the present invention analyzes based on real-time data, which can optimize the power generation process and ensure that the equipment operates in the best state, thereby improving the overall power generation efficiency.

[0151] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A cloud-based intelligent interconnection platform for large hydropower stations, characterized in that: include: Data collection layer (100), transmission layer (200), cloud decision layer (300) and feedback control layer (400); The data collection layer (100) is configured with sensors to collect key operating data of the hydropower station in real time; The transmission layer (200) uses Internet of Things technology to transmit key operating data from the data collection layer to the cloud decision layer through edge computing devices and networks; The cloud decision layer (300) processes key operation data based on cloud computing resources, applies AI algorithms to perform data analysis, and generates power generation scheduling strategies; The feedback control layer (400) adjusts the working state of the on-site equipment according to the power generation dispatching strategy generated by the cloud decision layer.

2. The large-scale hydropower station cloud intelligent interconnection platform as claimed in claim 1, characterized in that: The key operating data include water flow, water level, precipitation, air temperature, wind speed and generator speed; The edge computing device uses FPGA technology to perform data preprocessing.

3. The large-scale hydropower station cloud intelligent interconnection platform as claimed in claim 2, characterized in that: The cloud decision-making layer includes a historical data analysis module and a real-time scheduling optimization module; The historical data analysis module models the historical operation data of the key operation data through a long short-term memory network, generates a long-term trend prediction model, and outputs the parameter range of the key operation data based on the long-term trend prediction model; The real-time scheduling optimization module is based on a deep reinforcement learning algorithm and generates an optimal scheduling strategy and adjusts the working status of the generator set according to the key operating data and parameter range of the key operating data input in real time.

4. A method using the cloud-based intelligent interconnection platform for large hydropower stations as described in any one of claims 1 to 3, characterized in that: include: Configure sensors to collect key operating data of the hydropower station in real time and transmit the collected data to edge computing devices; Pre-process key operation data through FPGA in edge computing devices; Call the stored historical key operation data, analyze the historical operation data based on the long short-term memory network, generate a long-term trend prediction model, and output the parameter range of the key operation data; Based on the deep reinforcement learning algorithm, the optimal scheduling strategy is generated according to the key operating data and the parameter range; Adjust the working status of field equipment based on the optimal scheduling strategy.

5. The cloud-based intelligent interconnection method for large hydropower stations according to claim 3, characterized in that: The data preprocessing of the key operation data includes receiving the key operation data transmitted in real time from the sensor; Use digital filters to low-pass filter the received key operating data to remove high-frequency noise and retain valid signals; Apply the mean normalization method to the filtered data to standardize the key operating data of different magnitudes to the same dimension; Calculate the time series difference of key operating data, remove the trend of the data, and smooth the data; Store the processed data in a temporary data cache.

6. The cloud-based intelligent interconnection method for a large hydropower station according to claim 4, characterized in that: The parameter range of outputting the key operating data includes extracting the historical key operating data from the database and arranging the data in chronological order; Divide the extracted historical data into a training set and a validation set; The LSTM model was constructed, and the number of input layer nodes was set to the number of data features, including water flow, water level, precipitation, temperature, wind speed, and generator speed, to capture the time dependency of the data; During the model training process, the mean square error is used as the loss function, and the Adam optimization algorithm is used to update the model parameters until the loss function converges; Generate parameter ranges for key operating data, including expected minimum and maximum values ​​for each key parameter.

7. The cloud-based intelligent interconnection method for a large hydropower station according to claim 5, characterized in that: Generating the optimal scheduling strategy according to the key operating data and the parameter range includes defining a state space, the state includes the real-time key operating data and the parameter range output by the LSTM, and is used to describe the current operating status of the system, and the state includes the expected range of water flow, water level, precipitation, temperature, wind speed and generator speed; Determine the action space, which includes options such as adjusting the power output of the generator set, starting or stopping, and select the optimal operation; Use the Q-learning algorithm to construct the value function, initialize the Q values ​​of all state-action pairs, and set the learning rate and discount factor; The ε-greedy strategy is used to balance exploration and exploitation, and the environment feedback is received and the Q value is updated after the action is performed.

8. The cloud-based intelligent interconnection method for a large hydropower station according to claim 6, characterized in that: The adjusting the working state of the field device based on the optimal scheduling strategy includes converting the generated optimal scheduling strategy into a control instruction and sending it to the control system of the field device through a control bus; The field equipment receives the control instructions and performs the corresponding adjustment operations; On-site equipment monitors its operating status in real time and collects equipment status data regularly; The actual operation status information is sent back to the cloud decision-making layer through the data feedback channel for status assessment and deviation analysis; If the actual operating status deviates from the expected scheduling strategy, the automatic compensation mechanism will be triggered to correct the equipment status by adjusting the control instructions to maintain the stable operation of the system.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the large hydropower station cloud-based intelligent interconnection method described in any one of claims 4 to 8 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the large hydropower station cloud-based intelligent interconnection method described in any one of claims 4 to 8 are implemented.

Images