Hydraulic power plant operation remote monitoring method and system based on artificial intelligence

Abstract

The invention discloses a hydraulic power plant operation remote monitoring method and system based on artificial intelligence, and relates to the technical field of hydraulic power plant operation remote monitoring, and the method comprises the steps: collecting data in real time, transmitting the collected data to an edge calculation node, carrying out the preprocessing, and carrying out the feature extraction of the preprocessed data, and carrying out data fusion by combining a multi-modal data fusion algorithm, carrying out real-time analysis on the extracted features, predicting the operation trend and detecting abnormity, adjusting a prediction strategy according to a detection result, automatically adjusting a fault early warning threshold according to historical data of the equipment and the current working condition, and adjusting equipment operation parameters according to fault diagnosis and early warning results. The long-short-term memory network is adopted for data preprocessing and feature extraction, the network training process is optimized, data fusion, running trend prediction, anomaly detection, fault diagnosis and equipment running parameter adjustment are achieved, and the intelligent level, fault early warning accuracy and remote control capacity of a monitoring system are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of remote monitoring of hydropower plant operation, and in particular to a method and system for remote monitoring of hydropower plant operation based on artificial intelligence. Background Art

[0002] In the current power system, the operation monitoring technology of hydropower plants has made significant progress after years of development. Traditional monitoring methods mainly rely on manual inspections and simple automation systems, which can ensure the stable operation of hydropower plants to a certain extent. However, with the expansion of the scale and increase in complexity of hydropower plants, traditional monitoring technologies have exposed many shortcomings in data processing, fault prediction and remote control. In recent years, the rapid development of artificial intelligence technology has provided new solutions for the operation monitoring of hydropower plants, which has significantly improved the intelligence and automation level of remote monitoring systems.

[0003] However, the existing hydropower plant operation monitoring system still has certain limitations. First, in terms of data collection, although it is possible to obtain the equipment's operating data, environmental data, and operating data, there is a lack of effective data preprocessing and feature extraction methods, which makes it difficult to ensure the accuracy and real-time performance of the monitoring data. Secondly, in terms of fault prediction and diagnosis, existing technologies often rely on simple models, which makes it difficult to accurately capture the complex laws of equipment operation, thus affecting the accuracy of fault warnings. In addition, the existing system also has shortcomings in remote control and lacks the ability to dynamically adjust equipment operating parameters based on real-time monitoring data. Summary of the invention

[0004] In view of the above existing problems, the present invention provides a method and system for remote monitoring of hydropower plant operation based on artificial intelligence, so as to solve the problem in the prior art that selecting the wrong energy storage method will cause waste of electricity.

[0005] In order to solve the above technical problems, a remote monitoring method for hydropower plant operation based on artificial intelligence is proposed, including:

[0006] Collect data in real time, transmit the collected data to the edge computing node for preprocessing, and extract features from the preprocessed data; combine multimodal data fusion algorithm for data fusion, analyze the extracted features in real time, predict operating trends and detect anomalies; adjust the prediction strategy according to the detection results, automatically adjust the fault warning threshold according to the equipment's historical data and current operating conditions, and adjust the equipment's operating parameters according to the fault diagnosis and warning results.

[0007] As a preferred solution of the artificial intelligence-based remote monitoring method for hydropower plant operation described in the present invention, the real-time data collection includes real-time collection of equipment operation data, environmental data and operating condition data.

[0008] The equipment operation data includes the rotation speed and vibration data of the rotating equipment, the output power, current, voltage and frequency of the generator set, the flow, pressure and efficiency of the turbine, the temperature, oil level and insulation status of the transformer, the operation times, contact resistance and opening and closing status of the switchgear, and the start times, operating time and load conditions of the motor.

[0009] The operating condition data includes the start and stop status of the equipment, operating mode, working cycle, fault code, alarm information, maintenance records, equipment energy consumption data, lubricant consumption, cooling system performance, safety system status and control system operation log.

[0010] As a preferred solution of the artificial intelligence-based remote monitoring method for hydropower plant operation described in the present invention, the feature extraction includes transmitting the collected data to the edge computing node, performing denoising, normalization, data completion, time synchronization and data dimension reduction, and using a long short-term memory network for feature extraction.

[0011] The long short-term memory network includes dividing the preprocessed data into a training set and a test set, reshaping the data into three-dimensional data, constructing an input layer, an LSTM layer and an output layer of the long short-term memory network, performing network training, and using the time series characteristics of the trained long-term memory network output data.

[0012] The output layer includes selecting an activation function according to the requirements of feature extraction. When performing a classification task, a softmax activation function is used; when performing a regression task, a linear activation function is used.

[0013] As a preferred solution of the artificial intelligence-based remote monitoring method for hydropower plant operation described in the present invention, the network training includes selecting a loss function, selecting an RMSprop optimizer to calculate the square of the gradient, and setting a learning rate, using the training set to calculate the prediction result through forward propagation, using the loss function to calculate the difference between the prediction result and the true label, obtaining the loss value, updating the network weights and biases through back propagation, and using a validation set to evaluate the model performance, and adjusting the model parameters according to the evaluation results.

[0014] The selection of the loss function includes selecting a cross entropy loss function when performing a classification task, and selecting a mean square error function when performing a regression task.

[0015] The cross entropy loss function is expressed as:

[0016]

[0017] Among them, y is the true label, is the probability distribution predicted by the model, n is the number of samples, is the loss value and i is the variable index.

[0018] The mean square error function is expressed as:

[0019]

[0020] Among them, y is the true label, is the probability distribution predicted by the model, n is the number of samples, MSE is the loss value, and i is the variable index.

[0021] As a preferred solution of the artificial intelligence-based remote monitoring method for hydropower plant operation described in the present invention, the data fusion includes performing feature-level fusion, decision-level fusion and data-level fusion in the fusion layer of the long short-term memory network design.

[0022] The predicting operation trend includes inputting the extracted features into the time series prediction model, and predicting the operation trend of the equipment according to the extracted features; the detecting anomaly includes calculating the path length of each data point to the leaf node, setting the length threshold d th , when the path length is less than d th When the data point is normal, the path length is greater than or equal to d th When the data point is abnormal, an abnormal warning is issued and the abnormal data value is eliminated.

[0023] As a preferred solution of the artificial intelligence-based remote monitoring method for hydropower plant operation described in the present invention, the fault diagnosis includes adjusting the prediction strategy according to the detection results, automatically adjusting the fault warning threshold according to the equipment historical data and current operating conditions, and using the decision function of the linear classifier to perform fault diagnosis.

[0024] The adjustment prediction strategy includes collecting anomaly detection results, adjusting long short-term memory network parameters according to the detection results, improving feature engineering, retraining the long short-term memory network, and testing the adjusted network using a validation set.

[0025] The decision function of the linear classifier is expressed as:

[0026] f ( x ) =w T x+b

[0027] Among them, f ( x) is the predicted value, w is the weight vector, b is the bias term, and x is the feature vector.

[0028] As a preferred solution of the artificial intelligence-based remote monitoring method for hydropower plant operation described in the present invention, the adjustment of equipment operating parameters includes analyzing the status of the equipment, setting a fault threshold, receiving the results of anomaly detection and fault diagnosis, analyzing the impact of the fault on the equipment operation, and adjusting the operating parameters according to the decision function calculation results of the linear classifier.

[0029] The adjustment of the operating parameters includes setting the fault threshold to f(x)=0. When f(x)>0, the equipment is judged to be operating in a normal state and the current equipment parameters are maintained; when f(x)≤0, it is judged to be a fault state, a fault warning is performed, the equipment operation mode, maintenance plan and control strategy are adjusted, and the equipment operating parameters are adjusted through remote control instructions. The adjusted equipment operating status is monitored, and the adjusted operating effect is fed back to the monitoring system. According to the feedback results, the prediction strategy and threshold adjustment are optimized.

[0030] Another object of the present invention is to provide a remote monitoring system for hydropower plant operation based on artificial intelligence. The present invention improves the operating safety, economy and environmental protection of the hydropower plant and realizes the stable operation of the hydropower plant. The system of the present invention collects equipment operation data, environmental data and operating condition data in real time, and uses deep learning technologies such as long short-term memory networks to perform feature extraction and fault diagnosis, thereby realizing real-time monitoring and early warning of hydropower plant equipment.

[0031] As a preferred solution of the artificial intelligence-based remote monitoring system for hydropower plant operation described in the present invention, it is characterized by including a data acquisition and preprocessing module, a feature extraction module, a data fusion module and a fault diagnosis and early warning module.

[0032] The data acquisition and preprocessing module is used to collect equipment operation data, environmental data and working condition data in real time, transmit the collected data to the edge computing node, and perform preprocessing.

[0033] The feature extraction module is used to extract features from the preprocessed data using a long short-term memory network to extract time series features of the device status.

[0034] The data fusion module is used to perform feature-level fusion, decision-level fusion and data-level fusion in the fusion layer of the long short-term memory network design, and fuse the data of different sensors.

[0035] The fault diagnosis and early warning module is used to adjust the prediction strategy according to the detection results, automatically adjust the fault warning threshold according to the equipment historical data and the current working conditions, perform fault diagnosis using the decision function of the linear classifier, adjust the equipment operating parameters, adjust the equipment operating parameters through remote control instructions, monitor the adjusted equipment operating status, and feed back the adjusted operating effect to the monitoring system, and optimize the prediction strategy and threshold adjustment based on the feedback results.

[0036] A computer device comprises a memory and a processor, wherein the memory stores a computer program, and is characterized in that when the processor executes the computer program, the steps of a method described in a method for remote monitoring of hydropower plant operation based on artificial intelligence are implemented.

[0037] 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 a method described in a method for remote monitoring of hydropower plant operation based on artificial intelligence are implemented.

[0038] Beneficial effects of the present invention: The present invention collects and transmits data to edge computing nodes in real time for preprocessing and feature extraction, combines multimodal data fusion algorithms, uses long short-term memory networks to deeply mine the time series characteristics of equipment status, and optimizes the network training process, thereby achieving comprehensive monitoring and accurate prediction of the operating status of the hydropower plant. By automatically adjusting the fault warning threshold and remotely controlling the operating parameters of the equipment, the system self-optimizes and responds to state changes in real time, which improves the intelligence level of the monitoring system, reduces operation and maintenance costs, and improves production efficiency and safety. The use of computer equipment and storage media provides a solution that is easy to deploy and upgrade, ensuring the continuous updating of monitoring technology and the long-term stable monitoring service of the hydropower plant. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative work, among which:

[0040] Figure 1 An overall flow chart of an artificial intelligence-based remote monitoring method for hydropower plant operation provided for one embodiment of the present invention.

[0041] Figure 2 A system solution flow chart of a remote monitoring system for hydropower plant operation based on artificial intelligence provided by one embodiment of the present invention. DETAILED DESCRIPTION

[0042] 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.

[0043] 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.

[0044] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor is it an embodiment that is mutually exclusive with other embodiments, either individually or selectively.

[0045] The present invention is described in detail with reference to schematic diagrams. When describing the embodiments of the present invention, for the sake of convenience, the cross-sectional diagrams showing the device structure will not be partially enlarged according to the general scale, and the schematic diagrams are only examples, which should not limit the scope of protection of the present invention. In addition, in actual production, the three-dimensional dimensions of length, width and depth should be included.

[0046] At the same time, in the description of the present invention, it should be noted that the directions or positional relationships indicated by the terms "upper, lower, inner and outer" are based on the directions or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific direction, be constructed and operated in a specific direction, and therefore cannot be understood as limiting the present invention. In addition, the terms "first, second or third" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance.

[0047] In the present invention, unless otherwise clearly specified and limited, the terms "install, connect, connect" should be understood in a broad sense, for example: it can be a fixed connection, a detachable connection or an integral connection; it can also be a mechanical connection, an electrical connection or a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0048] Example 1, reference Figure 1 , which is the first embodiment of the present invention, provides a method for remote monitoring of hydropower plant operation based on artificial intelligence, comprising:

[0049] S1: Collect data in real time, transmit the collected data to the edge computing node for preprocessing, and perform feature extraction on the preprocessed data.

[0050] The real-time data collection includes real-time collection of equipment operation data, environmental data and operating condition data; the equipment operation data includes the speed and vibration data of rotating equipment, the output power, current, voltage and frequency of the generator set, the flow, pressure and efficiency of the turbine, the temperature, oil level and insulation status of the transformer, the number of operations, contact resistance and opening and closing status of the switchgear, and the number of starts, operating time and load conditions of the motor.

[0051] It should be noted that the environmental data include the water level, water temperature, water quality and flow of the reservoir, atmospheric temperature, humidity, wind speed and direction, rainfall, water level changes and geological activities in the hydropower plant area, as well as noise level, light intensity and air pollution index.

[0052] The operating condition data includes the start and stop status of the equipment, operating mode, working cycle, fault code, alarm information, maintenance records, equipment energy consumption data, lubricant consumption, cooling system performance, safety system status and control system operation log.

[0053] It should be further explained that the feature extraction includes transmitting the collected data to the edge computing node, performing denoising, normalization, data completion, time synchronization and data dimension reduction, and using a long short-term memory network for feature extraction.

[0054] The long short-term memory network includes dividing the preprocessed data into a training set and a test set, reshaping the data into three-dimensional data, constructing an input layer, an LSTM layer and an output layer of the long short-term memory network, performing network training, and using the time series characteristics of the trained long-term memory network output data.

[0055] The output layer includes selecting an activation function according to the requirements of feature extraction. When performing a classification task, a softmax activation function is used; when performing a regression task, a linear activation function is used.

[0056] S2: Combine multimodal data fusion algorithms to perform data fusion, analyze the extracted features in real time, predict operating trends and detect anomalies.

[0057] Furthermore, the network training includes selecting a loss function, selecting an RMSprop optimizer to calculate the square of the gradient, and setting a learning rate, using the training set to calculate the prediction result through forward propagation, using the loss function to calculate the difference between the prediction result and the true label, obtaining the loss value, updating the network weights and biases through back propagation, and using the validation set to evaluate the model performance, and adjusting the model parameters according to the evaluation results.

[0058] The selection of the loss function includes selecting a cross entropy loss function when performing a classification task, and selecting a mean square error function when performing a regression task.

[0059] It should be noted that the cross entropy loss function is expressed as:

[0060]

[0061] Among them, y is the true label, is the probability distribution predicted by the model, n is the number of samples, is the loss value and i is the variable index.

[0062] The mean square error function is expressed as:

[0063]

[0064] Among them, y is the true label, is the probability distribution predicted by the model, n is the number of samples, MSE is the loss value, and i is the variable index.

[0065] Furthermore, the data fusion includes performing feature-level fusion, decision-level fusion and data-level fusion in a fusion layer of a long short-term memory network.

[0066] The predicting operation trend includes inputting the extracted features into the time series prediction model, and predicting the operation trend of the equipment according to the extracted features; the detecting anomaly includes randomly selecting a data point from the data set as a root node, and randomly selecting a feature and a threshold value to segment the data at the current node, and stopping the segmentation when only one data point remains at the current node after the data point is segmented, and continuing until there are two or more data points remaining at the current node after the data point is segmented, until a tree is established, and calculating the path length of each data point to reach the leaf node, and setting the length threshold d th , when the path length is less than d th When the data point is normal, the path length is greater than or equal to d th When the data point is abnormal, an abnormal warning is issued and the abnormal data value is eliminated.

[0067] S3: Adjust the prediction strategy according to the detection results, automatically adjust the fault warning threshold according to the equipment historical data and current working conditions, and adjust the equipment operating parameters according to the fault diagnosis and warning results.

[0068] Furthermore, the fault diagnosis includes adjusting the prediction strategy according to the detection results, automatically adjusting the fault warning threshold according to the equipment historical data and the current working condition, and performing fault diagnosis using the decision function of the linear classifier.

[0069] The adjustment prediction strategy includes collecting anomaly detection results, adjusting long short-term memory network parameters according to the detection results, improving feature engineering, retraining the long short-term memory network, and testing the adjusted network using a validation set.

[0070] It should be noted that the decision function of the linear classifier is expressed as:

[0071] f(x)=w T x+b

[0072] Among them, f(x(is the predicted value, w is the weight vector, b is the bias term, and x is the feature vector.

[0073] Furthermore, the adjustment of equipment operating parameters includes, based on the decision function calculation results of the linear classifier, analyzing the status of the equipment, setting a fault threshold, receiving the results of anomaly detection and fault diagnosis, analyzing the impact of the fault on equipment operation, and adjusting the operating parameters.

[0074] The adjustment of the operating parameters includes setting the fault threshold to f(x)=0. When f(x)>0, the equipment is judged to be operating in a normal state and the current equipment parameters are maintained; when f(x)≤0, it is judged to be a fault state, a fault warning is performed, the equipment operation mode, maintenance plan and control strategy are adjusted, and the equipment operating parameters are adjusted through remote control instructions. The adjusted equipment operating status is monitored, and the adjusted operating effect is fed back to the monitoring system. According to the feedback results, the prediction strategy and threshold adjustment are optimized.

[0075] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. 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.

[0076] Example 2, reference Figure 2, which is the second embodiment of the present invention, provides an artificial intelligence-based remote monitoring system for hydropower plant operation, including a data acquisition and preprocessing module 100, a feature extraction module 200, a data fusion module 300, and a fault diagnosis and early warning module 400.

[0077] The data acquisition and preprocessing module 100 is used to collect equipment operation data, environmental data and working condition data in real time, transmit the collected data to the edge computing node, and perform preprocessing.

[0078] The feature extraction module 200 is used to extract features from the preprocessed data using a long short-term memory network to extract time series features of the device status.

[0079] The data fusion module 300 is used to perform feature-level fusion, decision-level fusion and data-level fusion in the fusion layer of the long short-term memory network design, and fuse the data of different sensors.

[0080] The fault diagnosis and warning module 400 is used to adjust the prediction strategy according to the detection results, automatically adjust the fault warning threshold according to the equipment historical data and the current operating conditions, perform fault diagnosis using the decision function of the linear classifier, adjust the equipment operating parameters, adjust the equipment operating parameters through remote control instructions, monitor the adjusted equipment operating status, and feed back the adjusted operating effect to the monitoring system, and optimize the prediction strategy and threshold adjustment based on the feedback results.

[0081] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. 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.

[0082] Embodiment 3, the third embodiment of the present invention, is different from the first two embodiments in that:

[0083] 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, and other media that can store program codes.

[0084] 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.

[0085] 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.

[0086] 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.

Claims

1. A remote monitoring method for hydropower plant operation based on artificial intelligence, characterized in that: include, Collect data in real time, transmit the collected data to the edge computing node for preprocessing, and perform feature extraction on the preprocessed data; Combine multimodal data fusion algorithms to perform data fusion, analyze extracted features in real time, predict operating trends, and detect anomalies; Adjust the prediction strategy based on the detection results, automatically adjust the fault warning threshold based on the equipment's historical data and current operating conditions, and adjust the equipment's operating parameters based on the fault diagnosis and warning results.

2. The method for remote monitoring of hydropower plant operation based on artificial intelligence according to claim 1, characterized in that: The real-time data collection includes real-time collection of equipment operation data, environmental data and working condition data; The equipment operation data includes the speed and vibration data of the rotating equipment, the output power, current, voltage and frequency of the generator set, the flow, pressure and efficiency of the turbine, the temperature, oil level and insulation status of the transformer, the operation times, contact resistance and opening and closing status of the switchgear, and the start times, operation time and load conditions of the motor; The operating condition data includes the start and stop status of the equipment, operating mode, working cycle, fault code, alarm information, maintenance records, equipment energy consumption data, lubricant consumption, cooling system performance, safety system status and control system operation log.

3. The method for remote monitoring of hydropower plant operation based on artificial intelligence as claimed in claim 2, characterized in that: The feature extraction includes transmitting the collected data to the edge computing node, performing denoising, normalization, data completion, time synchronization and data dimension reduction, and performing feature extraction using a long short-term memory network; The long short-term memory network includes dividing the preprocessed data into a training set and a test set, reshaping the data into three-dimensional data, constructing an input layer, an LSTM layer and an output layer of the long short-term memory network, performing network training, and using the time series characteristics of the trained long-term memory network output data; The output layer includes selecting an activation function according to the requirements of feature extraction. When performing a classification task, a softmax activation function is used; when performing a regression task, a linear activation function is used.

4. The method for remote monitoring of hydropower plant operation based on artificial intelligence as claimed in claim 3, characterized in that: The network training includes selecting a loss function, selecting an RMSprop optimizer to calculate the square of the gradient, and setting a learning rate, using the training set to calculate the prediction result by forward propagation, using the loss function to calculate the difference between the prediction result and the true label, obtaining the loss value, updating the weight and bias of the network by back propagation, and using the validation set to evaluate the model performance, and adjusting the model parameters according to the evaluation results; The selection of the loss function includes selecting a cross entropy loss function when performing a classification task, and selecting a mean square error function when performing a regression task; The cross entropy loss function is expressed as: Among them, y is the true label, is the probability distribution predicted by the model, n is the number of samples, is the loss value, i is the variable index; The mean square error function is expressed as: Among them, y is the true label, is the probability distribution predicted by the model, n is the number of samples, MSE is the loss value, and i is the variable index.

5. The method for remote monitoring of hydropower plant operation based on artificial intelligence as claimed in claim 4, characterized in that: The data fusion includes performing feature-level fusion, decision-level fusion and data-level fusion in a fusion layer designed in a long short-term memory network; The predicting operation trend includes inputting the extracted features into the time series prediction model, and predicting the operation trend of the equipment according to the extracted features; the detecting anomaly includes calculating the path length of each data point to the leaf node, setting the length threshold d th , when the path length is less than d th When the data point is normal, the path length is greater than or equal to d th When the data point is abnormal, an abnormal warning is issued and the abnormal data value is eliminated.

6. The method for remote monitoring of hydropower plant operation based on artificial intelligence as claimed in claim 5, characterized in that: The fault diagnosis includes adjusting the prediction strategy according to the detection results, automatically adjusting the fault warning threshold according to the equipment historical data and the current working condition, and performing fault diagnosis using the decision function of the linear classifier; The adjustment prediction strategy includes collecting anomaly detection results, adjusting long short-term memory network parameters according to the detection results, improving feature engineering, retraining the long short-term memory network, and testing the adjusted network using a validation set; The decision function of the linear classifier is expressed as: f ( x ) =w T x+b Among them, f(x) is the predicted value, w is the weight vector, b is the bias term, and x is the feature vector.

7. The method for remote monitoring of hydropower plant operation based on artificial intelligence as claimed in claim 6, characterized in that: The adjusting of the equipment operating parameters includes, according to the decision function calculation result of the linear classifier, analyzing the state of the equipment, setting the fault threshold, receiving the results of the abnormality detection and fault diagnosis, analyzing the impact of the fault on the equipment operation, and adjusting the operating parameters; The adjustment of the operating parameters includes setting the fault threshold to f(x)=0. When f(x)>0, the equipment is judged to be operating in a normal state and the current equipment parameters are maintained; when f(x)≤0, it is judged to be a fault state, a fault warning is performed, the equipment operation mode, maintenance plan and control strategy are adjusted, and the equipment operating parameters are adjusted through remote control instructions. The adjusted equipment operating status is monitored, and the adjusted operating effect is fed back to the monitoring system. According to the feedback results, the prediction strategy and threshold adjustment are optimized.

8. A system using the artificial intelligence-based remote monitoring method for hydropower plant operation as claimed in any one of claims 1 to 7, characterized in that: It includes data acquisition and preprocessing module, feature extraction module, data fusion module and fault diagnosis and early warning module; The data acquisition and preprocessing module is used to collect equipment operation data, environmental data and working condition data in real time, transmit the collected data to the edge computing node, and perform preprocessing; The feature extraction module is used to extract features from the preprocessed data using a long short-term memory network to extract time series features of the device status; The data fusion module is used to perform feature-level fusion, decision-level fusion and data-level fusion in the fusion layer of the long short-term memory network design, and fuse the data of different sensors; The fault diagnosis and early warning module is used to adjust the prediction strategy according to the detection results, automatically adjust the fault warning threshold according to the equipment historical data and the current working conditions, perform fault diagnosis using the decision function of the linear classifier, adjust the equipment operating parameters, adjust the equipment operating parameters through remote control instructions, monitor the adjusted equipment operating status, and feed back the adjusted operating effect to the monitoring system, and optimize the prediction strategy and threshold adjustment based on the feedback results.

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 a method for remote monitoring of hydropower plant operation based on artificial intelligence as described in any one of claims 1 to 7 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 a method for remote monitoring of hydropower plant operation based on artificial intelligence described in any one of claims 1 to 7 are implemented.

Images