Hydropower station equipment operation and maintenance method and device, computer equipment, storage medium and computer program product

Abstract

The invention relates to a hydropower station equipment operation and maintenance method and device, computer equipment, a storage medium and a computer program product. The method comprises the following steps: screening out a target hydropower station with the maximum target state prediction value from each hydropower station; according to the first defect prediction model of the target hydropower station, updating the first defect prediction models of other hydropower stations to obtain an updated defect prediction model; respectively inputting current equipment data of other hydropower stations into the updated defect prediction model to obtain second defect prediction results of the other hydropower stations; according to the first defect prediction result of the target hydropower station and the second defect prediction result of the other hydropower stations, generating a first equipment operation and maintenance instruction of the target hydropower station and second equipment operation and maintenance instructions of the other hydropower stations; and carrying out corresponding equipment operation and maintenance processing on the target hydropower station and other hydropower stations. By adopting the method, the operation and maintenance accuracy of the hydropower station equipment can be improved.

Description

Technical Field

[0001] The present application relates to the field of power grid technology, and in particular to a method and apparatus for operating and maintaining hydropower station equipment, a computer device, a computer-readable storage medium, and a computer program product. Background Art

[0002] At present, in order to ensure the normal operation of hydropower station equipment, it is very important to carry out operation and maintenance of hydropower station equipment.

[0003] Traditionally, the operation and maintenance of hydropower station equipment relies solely on the expert experience model of each hydropower station. However, this approach is subjective and limited, prone to errors, and results in low accuracy in the operation and maintenance of hydropower station equipment. Summary of the Invention

[0004] Based on this, it is necessary to provide a hydropower station equipment operation and maintenance method, device, computer equipment, computer-readable storage medium and computer program product that can improve the operation and maintenance accuracy of hydropower station equipment in response to the above technical problems.

[0005] In a first aspect, the present application provides a hydropower station equipment operation and maintenance method, which is applied to the cloud side and includes:

[0006] Obtain the target state prediction value of each hydropower station;

[0007] From each of the hydropower stations, a target hydropower station having the largest target state prediction value is selected, and a first defect prediction model of the target hydropower station is determined;

[0008] updating the first defect prediction models of other hydropower stations according to the first defect prediction model of the target hydropower station to obtain updated defect prediction models of the other hydropower stations; the other hydropower stations are used to represent the hydropower stations other than the target hydropower station in each of the hydropower stations;

[0009] Obtaining a first defect prediction result of the target hydropower station, inputting current equipment data of the other hydropower stations into the updated defect prediction models of the other hydropower stations, and obtaining second defect prediction results of the other hydropower stations;

[0010] generating a first equipment operation and maintenance instruction for the target hydropower station based on the first defect prediction result, and generating a second equipment operation and maintenance instruction for the other hydropower stations based on the second defect prediction result;

[0011] The first equipment operation and maintenance instruction and the second equipment operation and maintenance instruction are sent to the edge side, so that the edge side performs corresponding equipment operation and maintenance processing on the target hydropower station according to the first equipment operation and maintenance instruction, and performs corresponding equipment operation and maintenance processing on the other hydropower stations according to the second equipment operation and maintenance instruction.

[0012] In one embodiment, updating the first defect prediction models of other hydropower stations based on the first defect prediction model of the target hydropower station to obtain updated defect prediction models of the other hydropower stations includes:

[0013] Acquire first model architecture information of the first defect prediction model of the target hydropower station, and second model architecture information of the first defect prediction models of the other hydropower stations;

[0014] determining a similarity between the first model architecture information and the second model architecture information;

[0015] When the similarity is greater than a preset similarity, updating the model parameters of the first defect prediction model of the other hydropower station to obtain an updated defect prediction model of the other hydropower station;

[0016] When the similarity is less than or equal to the preset similarity, the first defect prediction model of the target hydropower station is used as the updated defect prediction model of the other hydropower stations.

[0017] In one embodiment, obtaining the target state prediction value of each hydropower station includes:

[0018] receiving the current equipment data and the first state prediction value for each hydropower station sent by the edge side; the edge side is configured to input the current equipment data of each hydropower station into the second defect prediction model of each hydropower station to obtain a third defect prediction result of each hydropower station, and determine the first state prediction value of each hydropower station based on the third defect prediction result of each hydropower station;

[0019] Acquire historical equipment data of each hydropower station, and obtain a second state prediction value of each hydropower station based on the current equipment data and historical equipment data of each hydropower station;

[0020] A target state prediction value of each hydropower station is obtained based on the first state prediction value and the second state prediction value respectively.

[0021] In one embodiment, obtaining the second state prediction value of each hydropower station based on the current equipment data and historical equipment data of each hydropower station includes:

[0022] Inputting the current equipment data and the historical equipment data of each hydropower station into the target defect prediction model respectively to obtain a fourth defect prediction result of each hydropower station;

[0023] Determining current weight information corresponding to the fourth defect prediction result of each hydropower station;

[0024] The second state prediction value of each hydropower station is determined according to the fourth defect prediction result of each hydropower station and the current weight information corresponding to the fourth defect prediction result of each hydropower station.

[0025] In one embodiment, determining the current weight information corresponding to the fourth defect prediction result of each hydropower station includes:

[0026] Acquire multiple candidate corresponding relationships; the multiple candidate corresponding relationships are used to represent the corresponding relationship between the defect prediction results and the weight information in different business scenarios;

[0027] Determine the current business scenario of each hydropower station respectively, and select a candidate corresponding relationship corresponding to the current business scenario from the multiple candidate corresponding relationships as the target corresponding relationship corresponding to each hydropower station;

[0028] The target correspondence relationship corresponding to each hydropower station is queried respectively to obtain current weight information corresponding to the fourth defect prediction result of each hydropower station.

[0029] In one embodiment, inputting the current equipment data of the other hydropower stations into the updated defect prediction models of the other hydropower stations to obtain second defect prediction results of the other hydropower stations includes:

[0030] Preprocessing the current equipment data of the other hydropower stations respectively to obtain the preprocessed current equipment data of the other hydropower stations;

[0031] Inputting the preprocessed current equipment data of the other hydropower stations into the updated defect prediction models of the other hydropower stations to obtain prediction probabilities corresponding to the preset defect prediction results of the other hydropower stations;

[0032] From the preset defect prediction results of the other hydropower stations, the preset defect prediction results whose corresponding prediction probabilities are greater than the preset probabilities are respectively screened out as the second defect prediction results of the other hydropower stations.

[0033] In a second aspect, the present application also provides a hydropower station equipment operation and maintenance device, which is applied to the cloud side and includes:

[0034] A data acquisition module is used to obtain the target state prediction value of each hydropower station;

[0035] a model determination module, configured to select a target hydropower station having the largest target state prediction value from each of the hydropower stations, and determine a first defect prediction model for the target hydropower station;

[0036] a model updating module, configured to update the first defect prediction models of other hydropower stations according to the first defect prediction model of the target hydropower station, to obtain updated defect prediction models of the other hydropower stations; the other hydropower stations are used to represent the hydropower stations other than the target hydropower station in each of the hydropower stations;

[0037] A result prediction module is used to obtain a first defect prediction result of the target hydropower station, input the current equipment data of the other hydropower stations into the updated defect prediction models of the other hydropower stations, and obtain a second defect prediction result of the other hydropower stations;

[0038] an instruction generation module, configured to generate a first equipment operation and maintenance instruction for the target hydropower station based on the first defect prediction result, and to generate a second equipment operation and maintenance instruction for the other hydropower stations based on the second defect prediction result;

[0039] The equipment operation and maintenance module is used to send the first equipment operation and maintenance instruction and the second equipment operation and maintenance instruction to the edge side, so that the edge side performs corresponding equipment operation and maintenance processing on the target hydropower station according to the first equipment operation and maintenance instruction, and performs corresponding equipment operation and maintenance processing on the other hydropower stations according to the second equipment operation and maintenance instruction.

[0040] In a third aspect, the present application further provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the following steps are implemented:

[0041] Obtain the target state prediction value of each hydropower station;

[0042] From each of the hydropower stations, a target hydropower station having the largest target state prediction value is selected, and a first defect prediction model of the target hydropower station is determined;

[0043] updating the first defect prediction models of other hydropower stations according to the first defect prediction model of the target hydropower station to obtain updated defect prediction models of the other hydropower stations; the other hydropower stations are used to represent the hydropower stations other than the target hydropower station in each of the hydropower stations;

[0044] Obtaining a first defect prediction result of the target hydropower station, inputting current equipment data of the other hydropower stations into the updated defect prediction models of the other hydropower stations, and obtaining second defect prediction results of the other hydropower stations;

[0045] generating a first equipment operation and maintenance instruction for the target hydropower station based on the first defect prediction result, and generating a second equipment operation and maintenance instruction for the other hydropower stations based on the second defect prediction result;

[0046] The first equipment operation and maintenance instruction and the second equipment operation and maintenance instruction are sent to the edge side, so that the edge side performs corresponding equipment operation and maintenance processing on the target hydropower station according to the first equipment operation and maintenance instruction, and performs corresponding equipment operation and maintenance processing on the other hydropower stations according to the second equipment operation and maintenance instruction.

[0047] In a fourth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the following steps are implemented:

[0048] Obtain the target state prediction value of each hydropower station;

[0049] From each of the hydropower stations, a target hydropower station having the largest target state prediction value is selected, and a first defect prediction model of the target hydropower station is determined;

[0050] updating the first defect prediction models of other hydropower stations according to the first defect prediction model of the target hydropower station to obtain updated defect prediction models of the other hydropower stations; the other hydropower stations are used to represent the hydropower stations other than the target hydropower station in each of the hydropower stations;

[0051] Obtaining a first defect prediction result of the target hydropower station, inputting current equipment data of the other hydropower stations into the updated defect prediction models of the other hydropower stations, and obtaining second defect prediction results of the other hydropower stations;

[0052] generating a first equipment operation and maintenance instruction for the target hydropower station based on the first defect prediction result, and generating a second equipment operation and maintenance instruction for the other hydropower stations based on the second defect prediction result;

[0053] The first equipment operation and maintenance instruction and the second equipment operation and maintenance instruction are sent to the edge side, so that the edge side performs corresponding equipment operation and maintenance processing on the target hydropower station according to the first equipment operation and maintenance instruction, and performs corresponding equipment operation and maintenance processing on the other hydropower stations according to the second equipment operation and maintenance instruction.

[0054] In a fifth aspect, the present application further provides a computer program product, comprising a computer program, which, when executed by a processor, implements the following steps:

[0055] Obtain the target state prediction value of each hydropower station;

[0056] From each of the hydropower stations, a target hydropower station having the largest target state prediction value is selected, and a first defect prediction model of the target hydropower station is determined;

[0057] updating the first defect prediction models of other hydropower stations according to the first defect prediction model of the target hydropower station to obtain updated defect prediction models of the other hydropower stations; the other hydropower stations are used to represent the hydropower stations other than the target hydropower station in each of the hydropower stations;

[0058] Obtaining a first defect prediction result of the target hydropower station, inputting current equipment data of the other hydropower stations into the updated defect prediction models of the other hydropower stations, and obtaining second defect prediction results of the other hydropower stations;

[0059] generating a first equipment operation and maintenance instruction for the target hydropower station based on the first defect prediction result, and generating a second equipment operation and maintenance instruction for the other hydropower stations based on the second defect prediction result;

[0060] The first equipment operation and maintenance instruction and the second equipment operation and maintenance instruction are sent to the edge side, so that the edge side performs corresponding equipment operation and maintenance processing on the target hydropower station according to the first equipment operation and maintenance instruction, and performs corresponding equipment operation and maintenance processing on the other hydropower stations according to the second equipment operation and maintenance instruction.

[0061] The above-mentioned hydropower station equipment operation and maintenance method, apparatus, computer equipment, storage medium and computer program product first obtain the target state prediction value of each hydropower station, and screen out the target hydropower station with the largest target state prediction value from each hydropower station, and determine the first defect prediction model of the target hydropower station. Then, the hydropower stations other than the target hydropower station in each hydropower station are regarded as other hydropower stations, and the first defect prediction models of the other hydropower stations are updated according to the first defect prediction model of the target hydropower station to obtain updated defect prediction models of the other hydropower stations. Then, the first defect prediction result of the target hydropower station is obtained, and the current equipment data of the other hydropower stations are respectively input into the updated defect prediction models of the other hydropower stations to obtain second defect prediction results of the other hydropower stations. Then, based on the first defect prediction result, a first equipment operation and maintenance instruction for the target hydropower station is generated, and based on the second defect prediction result, a second equipment operation and maintenance instruction for the other hydropower stations is generated. Finally, the first equipment operation and maintenance instruction and the second equipment operation and maintenance instruction are sent to the edge side, so that the edge side performs corresponding equipment operation and maintenance processing on the target hydropower station according to the first equipment operation and maintenance instruction, and performs corresponding equipment operation and maintenance processing on the other hydropower stations according to the second equipment operation and maintenance instruction. In this way, in the process of operation and maintenance of hydropower station equipment, by screening out the target hydropower station with the largest target state prediction value, the first defect prediction model of the target hydropower station is used to update the models of other hydropower stations, and by obtaining the first defect prediction result of the target hydropower station and inputting the current equipment data of other hydropower stations into the updated defect prediction model to obtain the second defect prediction result, the defect prediction result of the hydropower station can be obtained more accurately, and then the equipment operation and maintenance instructions of the hydropower station can be generated more accurately, which is conducive to improving the operation and maintenance accuracy of the hydropower station equipment; moreover, the whole process involves model optimization and experience sharing, avoiding the subjectivity and limitations of the traditional technology of relying solely on the expert experience model of each hydropower station for operation and maintenance, which is prone to errors and leads to the defect of low operation and maintenance accuracy of hydropower station equipment, further improving the operation and maintenance accuracy of hydropower station equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the following briefly introduces the drawings required for use in the embodiments of the present application or related technical descriptions. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying any creative work.

[0063] Figure 1 This is an application environment diagram of a hydropower station equipment operation and maintenance method in one embodiment;

[0064] Figure 2A schematic flow chart of a method for operating and maintaining hydropower station equipment according to an embodiment;

[0065] Figure 3 A schematic flow chart of a method for operating and maintaining hydropower station equipment in another embodiment;

[0066] Figure 4 A schematic diagram of defect fault classification and corresponding weight identification in different business scenarios in one embodiment;

[0067] Figure 5 A schematic diagram of a ROPN (Process-Oriented Petri Net) logic diagram for analyzing and modeling oil pump operation and determining defects in one embodiment;

[0068] Figure 6 A schematic diagram of a ROPN logic diagram for oil level detection, analysis, modeling, and defect determination in one embodiment;

[0069] Figure 7 A schematic diagram of a ROPN logic diagram for oil-gas ratio analysis modeling and defect determination in one embodiment;

[0070] Figure 8 A schematic diagram of a ROPN logic diagram for weekly evaluation of governor health status in one embodiment;

[0071] Figure 9 A schematic diagram of a ROPN logic diagram for monthly evaluation of governor health status in one embodiment;

[0072] Figure 10 A schematic diagram of a ROPN logic diagram for a custom evaluation of a governor health status in one embodiment;

[0073] Figure 11 Schematic diagram of operation and maintenance health evaluation in one embodiment;

[0074] Figure 12 is a schematic diagram of a watershed expert knowledge base based on knowledge sharing in one embodiment;

[0075] Figure 13 A schematic diagram of knowledge sharing among adjacent hydropower stations in the same river basin in one embodiment;

[0076] Figure 14 A schematic diagram of the edge-cloud architecture of a cloud-collaborative device status operation and maintenance command system in one embodiment;

[0077] Figure 15 This is a structural block diagram of a hydropower station equipment operation and maintenance device in one embodiment;

[0078] Figure 16 FIG. 1 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0079] In order to make the purpose, technical solutions and advantages of this application more clear, the following further describes this application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0080] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0081] The hydropower station equipment operation and maintenance method provided in the embodiment of the present application can be applied to Figure 1 In the application environment shown in FIG. , the cloud side 102 communicates with the edge side 104 via a network. Specifically, refer to Figure 1 , the cloud side 102 obtains the target state prediction value of each hydropower station, selects the target hydropower station with the largest target state prediction value from each hydropower station, and determines the first defect prediction model of the target hydropower station; according to the first defect prediction model of the target hydropower station, updates the first defect prediction models of other hydropower stations to obtain updated defect prediction models of other hydropower stations; other hydropower stations are used to represent hydropower stations other than the target hydropower station in each hydropower station; obtains the first defect prediction result of the target hydropower station, inputs the current equipment data of other hydropower stations into the updated defect prediction models of other hydropower stations respectively, and obtains the second defect prediction results of other hydropower stations; generates a first equipment operation and maintenance instruction for the target hydropower station according to the first defect prediction result, and generates a second equipment operation and maintenance instruction for other hydropower stations according to the second defect prediction result; sends the first equipment operation and maintenance instruction and the second equipment operation and maintenance instruction to the edge side 104, so that the edge side 104 performs corresponding equipment operation and maintenance processing on the target hydropower station according to the first equipment operation and maintenance instruction, and performs corresponding equipment operation and maintenance processing on other hydropower stations according to the second equipment operation and maintenance instruction.

[0082] The cloud side 102 is a server that generates equipment operation and maintenance instructions for the hydropower station, and the edge side 104 is a server that performs equipment operation and maintenance for the hydropower station. The server can be implemented as a standalone server or a server cluster consisting of multiple servers.

[0083] In an exemplary embodiment, Figure 2 As shown, a method for operating and maintaining hydropower station equipment is provided, which is applied to Figure 1 Taking the cloud side in FIG. 1 as an example, in this embodiment, the method includes the following steps:

[0084] Step S201: Obtain the target state prediction value of each hydropower station.

[0085] Among them, a hydropower station refers to a comprehensive engineering facility that can convert water energy into electrical energy.

[0086] The target state prediction value refers to an indicator value used to evaluate the comprehensive state of the hydropower station. In actual scenarios, the target state prediction value is a score.

[0087] Exemplarily, the cloud side establishes a network path with the edge side in response to the equipment operation and maintenance request for each hydropower station; then, the cloud side receives the target state prediction value of each hydropower station sent by the edge side through the network path.

[0088] Step S202 : Filter out the target hydropower station with the largest target state prediction value from each hydropower station, and determine the first defect prediction model of the target hydropower station.

[0089] The target hydropower station refers to the hydropower station with the largest target state prediction value among all hydropower stations.

[0090] Among them, the first defect prediction model of the target hydropower station refers to the defect prediction model corresponding to the target hydropower station on the cloud side, such as a convolutional neural network model.

[0091] Exemplarily, the cloud side performs a rationality check on the target state prediction value of each hydropower station (for example, checks whether the target state prediction value of each hydropower station is within the range of 0 to 100), and obtains the verification result of the target state prediction value of each hydropower station; then, when the verification results of the target state prediction value of each hydropower station are all passed, the cloud side selects the target hydropower station with the largest target state prediction value from each hydropower station; then, the cloud side selects the defect prediction model corresponding to the target hydropower station from multiple candidate defect prediction models as the first defect prediction model.

[0092] Step S203, based on the first defect prediction model of the target hydropower station, the first defect prediction models of other hydropower stations are updated to obtain updated defect prediction models of other hydropower stations; other hydropower stations are used to represent the hydropower stations other than the target hydropower station in each hydropower station.

[0093] Among them, the first defect prediction model of other hydropower stations refers to the defect prediction models corresponding to other hydropower stations on the cloud side, such as a convolutional neural network model.

[0094] Among them, the updated defect prediction model of other hydropower stations refers to a defect prediction model obtained by updating the first defect prediction model of other hydropower stations, such as a convolutional neural network model.

[0095] Exemplarily, the cloud side regards the hydropower stations other than the target hydropower station in each hydropower station as other hydropower stations; then, the cloud side determines the update strategy information of the first defect prediction models of other hydropower stations, and according to the update strategy information, based on the first defect prediction model of the target hydropower station, updates the first defect prediction models of other hydropower stations to obtain updated defect prediction models of other hydropower stations.

[0096] Step S204: obtaining a first defect prediction result of the target hydropower station, and inputting the current equipment data of other hydropower stations into the updated defect prediction models of other hydropower stations to obtain second defect prediction results of other hydropower stations.

[0097] The first defect prediction result is used to represent the defect prediction result of the target hydropower station.

[0098] The current equipment data of other hydropower stations refers to the equipment data of other hydropower stations at the current time, including equipment oil pump data, equipment oil volume data, and equipment oil and gas data.

[0099] The second defect prediction result is used to represent the defect prediction results of other hydropower stations.

[0100] Exemplarily, the cloud side inputs the current equipment data of the target hydropower station into the updated defect prediction model of the target hydropower station to obtain the first defect prediction result of the target hydropower station; then, the cloud side performs feature extraction processing on the current equipment data of other hydropower stations respectively to obtain the feature vectors of the current equipment data of other hydropower stations; then, the cloud side inputs the feature vectors of the current equipment data of other hydropower stations into the updated defect prediction model of other hydropower stations respectively to obtain the second defect prediction results of other hydropower stations.

[0101] Step S205 : generating a first equipment operation and maintenance instruction for the target hydropower station according to the first defect prediction result, and generating a second equipment operation and maintenance instruction for other hydropower stations according to the second defect prediction result.

[0102] The first equipment operation and maintenance instruction refers to instruction information for performing equipment operation and maintenance processing on the target hydropower station.

[0103] The second equipment operation and maintenance instruction refers to instruction information for performing equipment operation and maintenance processing on other hydropower stations.

[0104] Exemplarily, the cloud side performs an integrity check on the first defect prediction result to obtain a verification result of the first defect prediction result; if the verification result of the first defect prediction result is passed, the cloud side generates an equipment operation and maintenance instruction corresponding to the first defect prediction result based on the first defect prediction result, as the first equipment operation and maintenance instruction of the target hydropower station; then, the cloud side performs an integrity check on the second defect prediction result to obtain a verification result of the second defect prediction result; if the verification result of the second defect prediction result is passed, the cloud side generates an equipment operation and maintenance instruction corresponding to the second defect prediction result based on the second defect prediction result, as the second equipment operation and maintenance instruction of other hydropower stations.

[0105] In step S206, the first equipment operation and maintenance instruction and the second equipment operation and maintenance instruction are sent to the edge side, so that the edge side performs corresponding equipment operation and maintenance processing on the target hydropower station according to the first equipment operation and maintenance instruction, and performs corresponding equipment operation and maintenance processing on other hydropower stations according to the second equipment operation and maintenance instruction.

[0106] Exemplarily, the cloud side encrypts the first device operation and maintenance instruction and the second device operation and maintenance instruction respectively to obtain the encrypted first device operation and maintenance instruction and the encrypted second device operation and maintenance instruction; then, the cloud side sends the encrypted first device operation and maintenance instruction and the encrypted second device operation and maintenance instruction to the edge side through the network path between the cloud side and the edge side; the edge side decrypts the encrypted first device operation and maintenance instruction and the encrypted second device operation and maintenance instruction respectively to obtain the first device operation and maintenance instruction and the second device operation and maintenance instruction; then, the edge side performs corresponding equipment operation and maintenance processing on the target hydropower station according to the first device operation and maintenance instruction, and performs corresponding equipment operation and maintenance processing on other hydropower stations according to the second device operation and maintenance instruction.

[0107] In the above-mentioned hydropower station equipment operation and maintenance method, the target state prediction value of each hydropower station is first obtained, and the target hydropower station with the largest target state prediction value is screened out from each hydropower station, and the first defect prediction model of the target hydropower station is determined. Then, the hydropower stations other than the target hydropower station in each hydropower station are regarded as other hydropower stations, and the first defect prediction models of the other hydropower stations are updated according to the first defect prediction model of the target hydropower station to obtain updated defect prediction models of the other hydropower stations. Then, the first defect prediction result of the target hydropower station is obtained, and the current equipment data of the other hydropower stations are respectively input into the updated defect prediction models of the other hydropower stations to obtain second defect prediction results of the other hydropower stations. Then, based on the first defect prediction result, a first equipment operation and maintenance instruction for the target hydropower station is generated, and based on the second defect prediction result, a second equipment operation and maintenance instruction for the other hydropower stations is generated. Finally, the first equipment operation and maintenance instruction and the second equipment operation and maintenance instruction are sent to the edge side, so that the edge side performs corresponding equipment operation and maintenance processing on the target hydropower station according to the first equipment operation and maintenance instruction, and performs corresponding equipment operation and maintenance processing on the other hydropower stations according to the second equipment operation and maintenance instruction. In this way, in the process of operation and maintenance of hydropower station equipment, by screening out the target hydropower station with the largest target state prediction value, the first defect prediction model of the target hydropower station is used to update the models of other hydropower stations, and by obtaining the first defect prediction result of the target hydropower station and inputting the current equipment data of other hydropower stations into the updated defect prediction model to obtain the second defect prediction result, the defect prediction result of the hydropower station can be obtained more accurately, and then the equipment operation and maintenance instructions of the hydropower station can be generated more accurately, which is conducive to improving the operation and maintenance accuracy of the hydropower station equipment; moreover, the whole process involves model optimization and experience sharing, avoiding the subjectivity and limitations of the traditional technology of relying solely on the expert experience model of each hydropower station for operation and maintenance, which is prone to errors and leads to the defect of low operation and maintenance accuracy of hydropower station equipment, further improving the operation and maintenance accuracy of hydropower station equipment.

[0108] In an exemplary embodiment, the above step S203, based on the first defect prediction model of the target hydropower station, updates the first defect prediction models of other hydropower stations to obtain updated defect prediction models of other hydropower stations, specifically including the following contents: obtaining the first model architecture information of the first defect prediction model of the target hydropower station, and the second model architecture information of the first defect prediction models of other hydropower stations; determining the similarity between the first model architecture information and the second model architecture information; when the similarity is greater than the preset similarity, updating the model parameters of the first defect prediction models of other hydropower stations to obtain updated defect prediction models of other hydropower stations; when the similarity is less than or equal to the preset similarity, using the first defect prediction model of the target hydropower station as the updated defect prediction model of other hydropower stations.

[0109] The first model architecture information refers to the model architecture information corresponding to the first defect prediction model of the target hydropower station.

[0110] Among them, the model architecture information includes hierarchical structure, number of neurons, connection method, activation function, convolution kernel size and step size.

[0111] The second model architecture information refers to the model architecture information corresponding to the first defect prediction model of other hydropower stations

[0112] Here, the similarity may refer to cosine similarity.

[0113] The preset similarity refers to a preset similarity threshold. It should be noted that the preset similarity depends on the situation.

[0114] Among them, model parameters refer to a series of adjustable values ​​used to describe the model structure and function, such as weights and biases.

[0115] Exemplarily, the cloud side obtains the model identifier of the first defect prediction model of the target hydropower station, and obtains the model architecture information corresponding to the model identifier as the first model architecture information, and obtains the model identifier of the other defect prediction model of other power stations, and obtains the model architecture information corresponding to the model identifier as the second model architecture information; then, the cloud side inputs the first model architecture information and the second model architecture information into the trained similarity prediction model, and obtains the similarity between the first model architecture information and the second model architecture information through the trained similarity prediction model; then, the cloud side judges the similarity based on the preset similarity; when the similarity is greater than the preset similarity, the cloud side updates the model parameters of the first defect prediction model of other hydropower stations according to the model parameters of the first defect prediction model of the target hydropower station, and obtains the updated defect prediction model of the other hydropower station; when the similarity is less than or equal to the preset similarity, the cloud side replaces the updated defect prediction model of the other hydropower station with the first defect prediction model of the target hydropower station.

[0116] In this embodiment, by adopting different update methods according to the similarity of the model architecture, the inefficiency caused by the unified and fixed update mode is avoided, and the blind attempt of complex parameter adjustment process is avoided, so that a matching defect prediction model can be quickly determined for other hydropower stations, which is conducive to improving the efficiency of the overall model update.

[0117] In an exemplary embodiment, the above-mentioned step S201, obtaining the target state prediction value of each hydropower station, specifically includes the following contents: receiving the current equipment data and the first state prediction value for each hydropower station sent by the edge side; the edge side is used to input the current equipment data of each hydropower station into the second defect prediction model of each hydropower station to obtain the third defect prediction result of each hydropower station, and determine the first state prediction value of each hydropower station based on the third defect prediction result of each hydropower station; obtaining the historical equipment data of each hydropower station, and obtaining the second state prediction value of each hydropower station based on the current equipment data and historical equipment data of each hydropower station; obtaining the target state prediction value of each hydropower station based on the first state prediction value and the second state prediction value respectively.

[0118] The current equipment data of each hydropower station refers to the equipment data of each hydropower station at the current time, including equipment oil pump data, equipment oil volume data, and equipment oil and gas data.

[0119] The first state prediction value refers to an indicator value for evaluating the state of the hydropower station on the edge side.

[0120] Among them, the second defect prediction model of each hydropower station refers to the defect prediction model corresponding to each hydropower station on the edge side, such as a convolutional neural network model.

[0121] The third defect prediction result is used to represent the defect prediction result of each hydropower station.

[0122] The historical equipment data of each hydropower station refers to the equipment data of each hydropower station in a historical time (such as the past hour, the past day, etc.), including equipment oil pump data, equipment oil volume data, and equipment oil and gas data.

[0123] The second state prediction value refers to an indicator value for evaluating the state of the hydropower station on the cloud side.

[0124] Exemplarily, the edge side inputs the current equipment data of each hydropower station into the second defect prediction model of each hydropower station, and obtains the third defect prediction result of each hydropower station through the second defect prediction model of each hydropower station; then, the edge side determines the state prediction value corresponding to the third defect prediction result of each hydropower station based on the third defect prediction result of each hydropower station, as the first state prediction value of each hydropower station; the cloud side receives the current equipment data and the first state prediction value for each hydropower station sent by the edge side through the network path between the cloud side and the edge side; then, the cloud side obtains the historical equipment data of each hydropower station from the database; then, the cloud side inputs the current equipment data and historical equipment data of each hydropower station into the target defect prediction model to obtain the second state prediction value of each hydropower station; then, the cloud side performs weighted sum processing on the first state prediction value and the second state prediction value respectively to obtain the target state prediction value of each hydropower station.

[0125] In this embodiment, the first state prediction value is obtained based on the third defect prediction result on the edge side, and the second state prediction value is obtained using current and historical equipment data. Finally, the target state prediction value is obtained by combining these two values. This can give full play to the advantages of different prediction methods, reduce the limitations and errors of a single prediction method, and further improve the accuracy of determining the target state prediction value.

[0126] In an exemplary embodiment, the second state prediction value of each hydropower station is obtained based on the current equipment data and historical equipment data of each hydropower station, which specifically includes the following contents: the current equipment data and historical equipment data of each hydropower station are input into the target defect prediction model respectively to obtain the fourth defect prediction result of each hydropower station; the current weight information corresponding to the fourth defect prediction result of each hydropower station is determined; the second state prediction value of each hydropower station is determined based on the fourth defect prediction result of each hydropower station and the current weight information corresponding to the fourth defect prediction result of each hydropower station.

[0127] Among them, the target defect prediction model refers to a network model that can determine the defect prediction results of each hydropower station, such as a convolutional neural network model.

[0128] The fourth defect prediction result is used to indicate whether there are related equipment defects in each hydropower station.

[0129] Among them, the current weight information is used to indicate the importance corresponding to the fourth defect prediction result.

[0130] Exemplarily, the cloud side inputs the current equipment data and historical equipment data of each hydropower station into the target defect prediction model respectively to obtain the defect prediction result of each hydropower station as the fourth defect prediction result; then, the cloud side determines the weight information corresponding to the fourth defect prediction result of each hydropower station as the current weight information; then, the cloud side determines the second state prediction value of each hydropower station based on the fourth defect prediction result of each hydropower station and the current weight information corresponding to the fourth defect prediction result of each hydropower station; for example, the defect prediction result of the hydropower station is "electrical interference", and the weight of "electrical interference" is 0.05. When the full score is 100 points, the score that needs to be deducted by the hydropower station is 100*0.05=5 points, and the total score obtained by the hydropower station is 95, that is, the second state prediction value of the hydropower station is 95.

[0131] In this embodiment, the current weight information corresponding to the fourth defect prediction result of each hydropower station is determined, and the second state prediction value is determined based on the prediction result and the weight information, so that the importance of each prediction result in the equipment status assessment can be measured more scientifically, thereby making the equipment status assessment more in line with the actual situation and improving the scientificity and rationality of the assessment results.

[0132] In an exemplary embodiment, the current weight information corresponding to the fourth defect prediction result of each hydropower station is determined, specifically including the following contents: obtaining multiple candidate correspondences; the multiple candidate correspondences are used to represent the correspondence between the defect prediction results and the weight information under different business scenarios; determining the current business scenario of each hydropower station respectively, and screening out the candidate correspondence corresponding to the current business scenario from the multiple candidate correspondences as the target correspondence corresponding to each hydropower station; querying the target correspondence corresponding to each hydropower station respectively to obtain the current weight information corresponding to the fourth defect prediction result of each hydropower station.

[0133] Among them, multiple candidate correspondences are used to represent the correspondence between defect prediction results and weight information in different business scenarios. For example, in business scenario A, the weight information corresponding to defect prediction result a is 0.05, and the weight information corresponding to defect prediction result b is 0.1; in business scenario B, the weight information corresponding to defect prediction result a is 0.1, and the weight information corresponding to defect prediction result b is 0.08.

[0134] The current business scenario is used to represent the business scenario of each hydropower station.

[0135] The target correspondence relationship is used to represent the correspondence relationship matching each hydropower station.

[0136] Exemplarily, the cloud side uses the correspondence between the defect prediction results and the weight information in different business scenarios as multiple candidate correspondences; then, the cloud side determines the current business scenario of each hydropower station respectively, and selects the candidate correspondence corresponding to the current business scenario from the multiple candidate correspondences as the target correspondence corresponding to each hydropower station; then, the cloud side queries the target correspondence corresponding to each hydropower station based on the fourth defect prediction result of each hydropower station, and obtains the current weight information corresponding to the fourth defect prediction result of each hydropower station.

[0137] In this embodiment, by determining the current business scenario of each hydropower station and screening out the corresponding target correspondence, the weight information can be accurately matched with the actual business scenario, avoiding the problem of unreasonable weight setting caused by the inability of general standards to adapt to the characteristics of different business scenarios, thereby more accurately reflecting the impact of defects on the equipment status and improving the accuracy of the assessment.

[0138] In an exemplary embodiment, the above-mentioned step S204, respectively inputs the current equipment data of other hydropower stations into the updated defect prediction model of other hydropower stations to obtain the second defect prediction results of other hydropower stations, specifically including the following contents: respectively preprocessing the current equipment data of other hydropower stations to obtain the preprocessed current equipment data of other hydropower stations; respectively inputting the preprocessed current equipment data of other hydropower stations into the updated defect prediction model of other hydropower stations to obtain the prediction probability corresponding to each preset defect prediction result of other hydropower stations; respectively, from each preset defect prediction result of other hydropower stations, screening out the preset defect prediction results whose corresponding prediction probability is greater than the preset probability as the second defect prediction result of other hydropower stations.

[0139] The pre-processed current equipment data of other hydropower stations refers to the current equipment data of other hydropower stations after pre-processing.

[0140] The preset defect prediction result refers to a pre-set defect prediction result. It should be noted that the preset defect prediction result depends on the situation.

[0141] The prediction probability is used to indicate the possibility that the updated defect prediction model determines that the preset defect prediction result is correct.

[0142] The preset probability refers to a pre-set probability threshold, such as 80%. It should be noted that the preset probability depends on the situation.

[0143] Exemplarily, the cloud side performs denoising processing on the current equipment data of other hydropower stations respectively to obtain the pre-processed current equipment data of other hydropower stations; then, the cloud side performs feature extraction processing on the pre-processed current equipment data of other hydropower stations respectively to obtain the feature vectors of the pre-processed current equipment data of other hydropower stations; then, the cloud side inputs the feature vectors of the pre-processed current equipment data of other hydropower stations into the updated defect prediction model of other hydropower stations, and obtains the prediction probability corresponding to each preset defect prediction result of other hydropower stations through the updated defect prediction model of other hydropower stations; then, the cloud side screens out the preset defect prediction results whose corresponding prediction probability is greater than the preset probability from each preset defect prediction result of other hydropower stations as the second defect prediction result of other hydropower stations.

[0144] In this embodiment, by preprocessing and modeling the current equipment data of other hydropower stations, and combining the prediction probabilities corresponding to each preset defect prediction result, the preset defect prediction results with a prediction probability greater than the preset probability can be accurately screened out, thereby improving the accuracy of determining the second defect prediction result, and providing an accurate data basis for subsequent data processing.

[0145] In an exemplary embodiment, Figure 3 As shown, another method for operating and maintaining hydropower station equipment is provided, which is applied to Figure 1 The cloud side in the example is used to illustrate the process, which includes the following steps:

[0146] Step S301: Obtain the target state prediction value of each hydropower station.

[0147] Step S302 : Filter out the target hydropower station with the largest target state prediction value from each hydropower station, and determine the first defect prediction model of the target hydropower station.

[0148] Step S303, obtaining the first model architecture information of the first defect prediction model of the target hydropower station and the second model architecture information of the first defect prediction models of other hydropower stations; other hydropower stations are used to represent the hydropower stations other than the target hydropower station in each hydropower station.

[0149] Step S304: Determine the similarity between the first model architecture information and the second model architecture information.

[0150] Step S305: When the similarity is greater than the preset similarity, the model parameters of the first defect prediction model of other hydropower stations are updated to obtain the updated defect prediction model of other hydropower stations; when the similarity is less than or equal to the preset similarity, the first defect prediction model of the target hydropower station is used as the updated defect prediction model of other hydropower stations.

[0151] Step S306: Obtain a first defect prediction result of the target hydropower station.

[0152] Step S307 , pre-processing the current equipment data of other hydropower stations respectively to obtain the pre-processed current equipment data of other hydropower stations.

[0153] In step S308 , the pre-processed current equipment data of the other hydropower stations are respectively input into the updated defect prediction models of the other hydropower stations to obtain the prediction probabilities corresponding to the preset defect prediction results of the other hydropower stations.

[0154] Step S309 , respectively screening out preset defect prediction results of other hydropower stations whose corresponding prediction probabilities are greater than the preset probabilities as second defect prediction results of other hydropower stations.

[0155] Step S310: generating a first equipment operation and maintenance instruction for the target hydropower station according to the first defect prediction result, and generating a second equipment operation and maintenance instruction for other hydropower stations according to the second defect prediction result.

[0156] In step S311, the first equipment operation and maintenance instruction and the second equipment operation and maintenance instruction are sent to the edge side, so that the edge side performs corresponding equipment operation and maintenance processing on the target hydropower station according to the first equipment operation and maintenance instruction, and performs corresponding equipment operation and maintenance processing on other hydropower stations according to the second equipment operation and maintenance instruction.

[0157] In the above-mentioned hydropower station equipment operation and maintenance method, in the process of operation and maintenance of hydropower station equipment, the target hydropower station with the largest target state prediction value is screened out, and the models of other hydropower stations are updated using the first defect prediction model of the target hydropower station. The first defect prediction result of the target hydropower station is obtained, and the current equipment data of other hydropower stations is input into the updated defect prediction model to obtain the second defect prediction result, so that the defect prediction result of the hydropower station can be obtained more accurately, and then the equipment operation and maintenance instructions of the hydropower station can be generated more accurately, which is conducive to improving the operation and maintenance accuracy of the hydropower station equipment. Moreover, the whole process involves model optimization and experience sharing, which avoids the subjectivity and limitations of the traditional technology of relying solely on the expert experience model of each hydropower station for operation and maintenance, which is prone to errors and leads to low operation and maintenance accuracy of the hydropower station equipment, thereby further improving the operation and maintenance accuracy of the hydropower station equipment.

[0158] In an exemplary embodiment, in order to more clearly illustrate the hydropower station equipment operation and maintenance method provided by the embodiment of the present application, the hydropower station equipment operation and maintenance method is specifically described below with a specific embodiment. In one embodiment, the present application also provides a hydropower station basin end-edge cloud collaborative equipment status operation and maintenance command system based on ROPN+AI (Artificial Intelligence) knowledge mining technology. In the process of operating and maintaining the hydropower station equipment, the target state prediction value of each hydropower station is first obtained, and the target hydropower station with the largest target state prediction value is screened out from each hydropower station, and the first defect prediction model of the target hydropower station is determined. Then, the hydropower stations other than the target hydropower station in each hydropower station are regarded as other hydropower stations, and the first defect prediction models of other hydropower stations are updated according to the first defect prediction model of the target hydropower station to obtain the updated defect prediction models of other hydropower stations. Then, the target hydropower station is obtained. The first defect prediction result of the target hydropower station is used to input the current equipment data of other hydropower stations into the updated defect prediction model of other hydropower stations to obtain the second defect prediction results of other hydropower stations. Then, based on the first defect prediction result, the first equipment operation and maintenance instruction of the target hydropower station is generated, and based on the second defect prediction result, the second equipment operation and maintenance instruction of other hydropower stations is generated. Finally, the first equipment operation and maintenance instruction and the second equipment operation and maintenance instruction are sent to the edge side, so that the edge side performs corresponding equipment operation and maintenance processing on the target hydropower station according to the first equipment operation and maintenance instruction, and performs corresponding equipment operation and maintenance processing on other hydropower stations according to the second equipment operation and maintenance instruction. Specifically including the following contents:

[0159] The hydropower station basin end-edge-cloud collaborative equipment status operation and maintenance command system, based on ROPN+AI knowledge mining technology, enables defect analysis and knowledge sharing for all hydropower station equipment under the cloud. It can also conduct health status assessments and expert knowledge sharing for all equipment under the cloud.

[0160] 1. Intelligent status evaluation system at the edge of hydropower station:

[0161] The ROPN+AI-based intelligent equipment status evaluation system and defect diagnosis method require a very sophisticated design throughout the entire process. To implement the intelligent equipment evaluation system, the most important thing is to divide the entire equipment evaluation system. The following uses the speed regulator as an example to illustrate:

[0162] (1) The entire runtime evaluation system of the speed regulator adopts a 100-point scoring standard and implements a business scenario division method for the speed regulator. This business scenario division includes the main defect risks of the speed regulator and sets weights according to importance;

[0163] (2) Based on the divided business scenarios, classify the defect types that need to be diagnosed for each business scenario and determine the weights according to their importance;

[0164] (3) Based on the above classification of defects, the most important thing left is to build a model of the state gradual change process according to the scenarios of these defects. For example, what states and data changes have the entire defect state of the guide vane relay experienced in history to lead to such a defect state? The entire business scenario state change process is modeled using ROPN+AI.

[0165] (4) The specific process of implementing the overall intelligent evaluation system and defect diagnosis is as follows:

[0166] Step 1: Use a 100-point system to evaluate the entire governor's operation. This system is divided into three business scenarios: oil pump operation analysis, oil volume detection analysis, and oil-gas ratio analysis. From the perspective of the governor's impact, the impact weights are α, β, and γ, where α + β + γ = 1.

[0167] Step 2: Classify the defects and faults in different business scenarios and their corresponding weights as follows: Figure 4 As shown in the figure, the weight mark for reminding the operating personnel to pay more attention is α1, the weight mark for oil receiver oil leakage is α2, the weight mark for oil cross-contamination of the guide vane relay is α3, the weight mark for oil cross-contamination of the emergency pressure distribution valve is α4, the weight mark for electrical interference is α5, the weight mark for the zeroing lever of the main pressure distribution valve is α6, the weight mark for the set value drift of the safety valve or unloading valve is α7, the weight mark for the oil discharge valve not closing tightly is α8, the weight mark for the check valve of a pump not closing tightly is α9, the weight mark for the check valve of the leaking oil pump not closing tightly is α10, the weight mark for water ingress into the unit rotor is β1, the weight mark for water ingress into the unit leakage tank is β2, the weight mark for oil leakage in the unit pressure system is β3, the weight mark for oil leakage in the oil pressure device is γ1, and the weight mark for the air supply valve not closing tightly with air leakage is γ2.

[0168] in, The governor deducts corresponding points based on different defect weights when a defect occurs.

[0169] Step 3: Build a ROPN+AI knowledge representation model for different business scenarios and corresponding defects to reflect the state change process of the equipment in the business scenario. Based on this model, the current equipment state is quantitatively characterized. When the corresponding defect occurs, the corresponding score is deducted. The modeling of each business scenario is explained below:

[0170] Oil pump operation analysis: For the oil pump operation analysis scenario, it involves reminding the operating personnel to pay more attention to defects such as large oil leakage from the oil receiver, oil cross-contamination of the guide vane relay, oil cross-contamination of the emergency pressure distribution valve, electrical interference, zero adjustment rod of the main pressure distribution valve, drift of the set value of the safety valve or unloading valve, loose closure of the oil discharge valve, loose closure of the pump check valve, loose closure of the leaking oil pump check valve, etc. The corresponding weights are: [α1, α2, ..., α10 ]. When a defect occurs, different scores are deducted according to different weights. Two figures are mainly used to show it, one is the logic diagram, and the other is the ROPN+AI modeling relationship diagram. The specific oil pump operation analysis and the entire modeling and judgment of the defect ROPN logic diagram are as follows Figure 5 It should be noted that, in the figure, Q9 and Q10 are threshold values ​​for judging whether the number of active power adjustments of the unit is frequent and large within the statistical period of oil pump operation; Q13 and Q14 are threshold values ​​for judging whether the number of active power adjustments of the unit is infrequent and small within the statistical period of oil pump operation; Q16 and Q17 are threshold values ​​for judging whether the number of active power adjustments of the unit is infrequent and small within the operating time of the oil pump; Q11 is the threshold value for the maximum number of operations allowed by the oil pump under normal operating mode; Q12 is the threshold value for the maximum temperature rise allowed by the return oil tank under normal operating mode; Q15 is the threshold value for the oil pump The maximum operating time threshold allowed for a single operation under normal operating mode; Q18 is the threshold value for the maximum difference allowed for the oil volume of the unit's oil pressure system under normal operating mode; unit active power P: represents the power change amplitude of the actual active power output of the unit; ΔK refers to the degree of power change; the number of power changes n refers to the number of times the unit's active power changes; T1, T2, and T3 refer to the time it takes to record the operation of different pumps; S1 refers to the frequency of changes in the oil pump's on-off operation; M is the preset threshold; Y refers to the actual oil volume currently detected in the oil pressure system; and Y' is the total oil volume in the oil pressure system.

[0171] Oil quantity detection and analysis: For the oil quantity detection and analysis business scenario, it involves defects such as water ingress into the unit's rotor body, water ingress into the unit's oil tank, and oil leakage in the unit's oil pressure system. The corresponding weights are: [β1, β2, β3]. When a defect occurs, different scores are deducted according to different weights. Specifically, two diagrams are used for display: one is a logic diagram, and the other is a ROPN+AI modeling relationship diagram. The specific oil quantity detection and analysis of the entire modeling and judgment of the defect ROPN logic diagram is as follows Figure 6 As shown. It should be noted that ag is the cross-sectional area of ​​the corresponding oil tank or oil tank; h and j are the oil storage coefficients for different openings of the guide and impeller blades; C and D are the functional relationship coefficients between the theoretical pressure tank oil level and the oil levels of the return tank and leakage tank; Y1 is the change in the actual pressure tank oil level; Y2 is the change in the actual return tank and storage tank oil levels converted to the pressure tank oil level change; Q1 and Q2 are the upper and lower threshold values ​​for the allowable change in actual oil volume; Q3 and Q4 are the upper and lower threshold values ​​for the allowable change in the pressure tank oil level, the actual return tank oil level, and the leakage tank oil level converted to the pressure tank oil level.

[0172] Oil and gas ratio analysis: For oil and gas ratio analysis business scenarios, defects such as oil leakage in hydraulic devices and lax closure of air supply valves are involved, and the corresponding weights are: [γ1,γ2]. When a defect occurs, different scores are deducted according to different weights. Two main diagrams are used to display it: one is a logic diagram, and the other is a ROPN+AI modeling relationship diagram. The specific oil and gas ratio analysis modeling and defect judgment ROPN logic diagram is as follows: Figure 7 It should be noted that c and d are the functional relationship coefficients for converting the theoretical pressure tank oil pressure to the pressure tank oil level; Q7-Q8 are the upper and lower threshold values ​​allowed under corresponding actual and theoretical conditions; Q5 and Q6 are the maximum number of actuations and actuation time thresholds allowed under normal operation of the air supply valve; and X3' is the actual pressure tank oil pressure converted to the pressure tank oil level.

[0173] Step 4: After the relevant ROPN+AI model is built, the overall work of the entire intelligent evaluation system is established. The ROPN+AI model built based on the business scenario can complete defect diagnosis. Defect diagnosis is also a process of process execution. For example, in the oil and gas ratio analysis scenario, the defect reasoning process is realized from left to right. When a defect occurs, the corresponding score is immediately deducted.

[0174] Step 5. Based on the above content, determine the overall evaluation system results. The scoring rules are as follows: 90-100 points are excellent; 75-90 points are average, requiring attention; 60-75 points are for increased attention; and below 60 points are dangerous.

[0175] 2. Hydropower Station Cloud Service Equipment Status Operation and Maintenance Command System:

[0176] Weekly and Monthly Equipment Health Assessment Methods Based on Defect Analysis: This weekly and monthly equipment health assessment method, based on ROPN+AI equipment defect analysis, evaluates the operational health of equipment over a specific time period. This method not only considers the statistical patterns of defect data but also considers the impact and causal relationships of operational status data, including operating condition data, analog data, and switch correlations. The time periods used are primarily categorized into three types: weekly, monthly, and custom time windows. This weekly and monthly equipment health assessment method, based on ROPN+AI equipment defect analysis, not only uses forward reasoning but also utilizes intermediate event judgment conditions to complete health reasoning. The following uses a speed regulator as an example to illustrate the specific process.

[0177] Step 1: Analyze governor system defects within weekly, monthly, or user-defined cycles. Specific defects include: the number of Class I defects within the unit's ballast system cycle, the number of Class II defects within the unit's oil pressure system cycle, the number of Class III defects within the unit's oil pressure system cycle, the number of oil leaks within the unit's oil pressure system cycle, the number of oil seepage points within the unit's oil pressure system cycle, the number of air leaks within the unit's oil pressure system cycle, the number of other problems within the unit's oil pressure system cycle, the number of equipment gaps within the unit's oil pressure system cycle, water ingress into the unit's rotor, water ingress into the unit's oil leakage tank, oil leakage in the unit's ballast system, air leakage in the unit's turbid pressure device, excessive air supply to the unit's oil pressure device, lax oil drain valve closure, lax check valve closure, pump safety valve or unloading valve setting drift, and internal leaks in the unit's oil pressure system. This alerts operators to pay close attention. Time periods represent: weekly, monthly, or user-defined time intervals.

[0178] Step 2: Statistics of historical equipment operation data within the weekly, monthly, and user-defined time periods of the speed regulator. The statistical contents are as follows: unit active power, pump operating time, pump operation recovery time, air supply valve operation time, air supply valve operation recovery time, oil leakage pump operating time, and oil leakage pump operation recovery time.

[0179] Step 3: Perform time series modeling and analysis on the governor's oil tank level time series data within weekly, monthly, and custom time periods. The specific contents are as follows: return tank oil level, leakage tank oil level, pressure tank oil level, emergency tank berth, guide vane opening, and blade opening.

[0180] Step 4: Based on the above data, a ROPN logic diagram for speed governor health status evaluation is constructed. The ROPN logic diagram for weekly evaluation of speed governor health status is as follows: Figure 8 As shown, the ROPN logic diagram of the monthly evaluation of the governor health status is as follows Figure 9 As shown, the ROPN logic diagram of the custom evaluation of the governor health status is as follows Figure 10 shown.

[0181] It should be noted that Figure 8D19 and D20 are the operating time and number of times of the oil pump under standard operating conditions within a week; D21 and D22 are the operating time and number of times of the air supply valve under standard operating conditions within a week; D23 and D24 are the operating time and number of times of the leakage pump under standard operating conditions within a week; D25 is the number of defects allowed in the unit oil pressure system within a week; D40 is the percentage of deduction allowed for one-time leakage of the pressure device in the unit oil pressure system within a week; D41 is the percentage of deduction allowed for one-time excessive air supply of the pressure device in the unit oil pressure system within a week; D42 is the percentage of deduction allowed for one-time lax closure of the injection device discharge valve and check valve in the unit oil pressure system within a week; D43 is the percentage of deduction allowed for one-time drift of the pump safety valve or unloading set value in the unit oil pressure system within a week; D44 is the percentage of deduction allowed for one-time internal leakage in the pressure system in the unit oil pressure system within a week; D45 is the percentage of deduction allowed for one-time internal leakage in the pressure system in the unit oil pressure system The percentage deducted for each reminder to strengthen attention; D58 is the ratio of the unit's oil pressure pump operating time to the amplitude of the unit's active power change under theoretical conditions within a week; D61 is the score deducted for the existence and past existence of Class I defects in the unit's oil pressure system within a week; D62 is the score deducted for each existence and past existence of Class I1 defects in the unit's oil pressure system within a week; D63 is the score deducted for each existence and past existence of Class III defects in the unit's oil pressure system within a week; D64 is the score deducted for the existence and past existence of an oil leak in the unit's oil pressure system within a week; D65 is the score deducted for the removal of each oil seepage point in the unit's oil pressure system within a week; D66 is the score deducted for the existence and past existence of an oil leak in the unit's oil pressure system within a week; D67 is the score deducted for the existence and past existence of other problems in the unit's oil pressure system within a week; D68 is the score deducted for each existence and past existence of various hidden dangers in the unit's oil pressure system within a week.

[0182] It should be noted that Figure 9D26 and D27 are the operating time and number of times of the mooring pump under standard operating conditions within a month; D28 and D29 are the operating time and number of times of the air supply valve under standard operating conditions within a month; D30 and D31 are the operating time and number of times of the pump under standard operating conditions within a month; D32 is the number of defects allowed in the unit's oil pressure system within a month; D46 is the percentage deducted for one oil pressure device leak allowed in the unit's oil pressure system within a month; D47 is the percentage deducted for one oil pressure device leak allowed in the unit's oil pressure system within a month The percentage of deduction allowed for the occurrence of excessive air supply to the pressure starter in the oil pressure system; D48 is the percentage of deduction allowed for the occurrence of lax closure of the pressure relief valve and check valve in the unit oil pressure system within a month; D49 is the percentage of deduction allowed for the occurrence of drift of the set value of the pump safety valve or load valve in the unit oil pressure system within a month; D50 is the percentage of deduction allowed for internal leakage in the unit oil pressure system within a month; D51 is the percentage of deduction allowed for leakage in the unit oil pressure system within a month We now remind you to pay close attention to the percentage of one-time deduction; D59 is the ratio of the running time of the unit's oil pressure pump to the degree of change in the unit's active power under theoretical conditions within a month; D69 is the score for the one-time deduction of Class I defects that exist or have existed in the unit's oil pressure system within a month; D70 is the score for the one-time deduction of Class II defects that exist or have existed in the unit's oil pressure system within a month; D71 is the score for the one-time deduction of Class II defects that exist or have existed in the unit's oil pressure system within a month; D72 is the score for the existence or existence of one turbidity point and defect in the unit's oil pressure system within a month; D73 is the score for the existence or existence of one turbidity point and defect in the unit's oil pressure system within a month; D74 is the score for the existence or existence of one turbidity point and defect in the unit's oil pressure system within a month; D75 is the score for the existence or existence of other problems in the unit's oil pressure system within a month; D76 is the score for the existence or existence of one equipment hidden danger in the unit's oil pressure system within a month.

[0183] It should be noted that Figure 10D33 and D34 are the operating time and number of times of the oil pump under standard operating conditions within the user-defined time; D35 and D36 are the operating time and number of times of the air supply valve under standard operating conditions within the user-defined time; D37 and D38 are the operating time and number of times of the leakage pump under standard operating conditions within the user-defined time; D39 is the number of defects allowed in the unit's oil pressure system within the user-defined time; D52 is the percentage of deduction allowed for one-time oil pressure leakage in the unit's oil pressure system within the user-defined time; D53 is the percentage of deduction allowed for one-time oil pressure device overfill within the user-defined time; D54 is the percentage of deduction allowed for one-time oil pressure device oil discharge valve and check valve not closing tightly within the user-defined time; D55 is the percentage of deduction allowed for one-time oil pump safety valve or unloading valve setting drift within the user-defined time; D56 is the percentage of deduction allowed for one-time oil pressure system internal leakage within the user-defined time; D57 is The percentage of points that can be deducted if a reminder or enhanced attention is given to the unit's oil pressure system within a custom time period; D60 is the ratio of the unit's oil pressure pump operating time to the amplitude of the unit's active power change under theoretical conditions within the custom time period; D77 is the point of deduction for any Class I defects that exist or have existed in the unit's oil pressure system within a week; D78 is the point of deduction for any Class II defects that exist or have existed in the unit's oil pressure system within a week; D79 is the point of deduction for any Class III defects that exist or have existed in the unit's oil pressure system within a week; D80 is the point of deduction for any oil leak that exists or has existed in the unit's oil pressure system within a week; D81 is the point of deduction for any oil seepage that exists or has existed in the unit's oil pressure system within a week; D82 is the point of deduction for any oil and gas leak that exists or has existed in the unit's oil pressure system within a week; D83 is the point of deduction for any other problems that exist or have existed in the unit's oil pressure system within a week; D84 is the point of deduction for any equipment hidden danger that exists or has existed in the unit's oil pressure system within a week.

[0184] Step 5: Complete the evaluation of the speed regulator equipment status according to the process model of speed regulator status evaluation.

[0185] According to the governor status evaluation score, the historical status and the historical data relationship experienced by the historical status, as well as the historical status data correlation relationship are determined, and the time series visualization method is used to display and intuitively express the data change trend.

[0186] According to the above modeling method, the health evaluation of the operating status of hydropower station equipment in weeks, months and custom periods and the intelligent defect analysis of equipment defect status are constructed, such as Figure 11 shown.

[0187] Build a basin expert knowledge base for knowledge sharing, such as Figure 12As shown. Build a unified end-edge-cloud collaboration, establish a shared knowledge system for all hydropower stations, and combine ROPN + AI knowledge mining to conduct in-depth knowledge mining, making this knowledge available to all hydropower stations within the basin, thereby achieving comprehensive O&M control of equipment status. The following describes knowledge sharing methods, using adjacent hydropower stations in the same basin as an example. As shown in Figure 13, Hydropower Stations A and B are adjacent stations in the same basin. Cloud analysis shows that Hydropower Station A's equipment is well-evaluated and maintained, while that of Hydropower Station B is poorly evaluated and maintained. Hydropower Station B wants to learn from Hydropower Station A's experience. By directly copying the excellent model of Hydropower Station A (for example, model component 5) and applying it to Hydropower Station B, a direct comparison can be performed, enabling experience sharing and improving the efficiency of Hydropower Station B.

[0188] Cloud Collaborative Equipment Status Operation and Maintenance Command System: Based on the health status evaluation of the edge side intelligent status evaluation and cloud side defect analysis, with the knowledge sharing basin expert knowledge base as the core, a cloud collaborative equipment status operation and maintenance command system is built through ROPN+AI. To achieve the requirements of cluster management, comprehensive monitoring, and centralized operation and maintenance management of hydropower stations in the basin, the edge-cloud architecture of the cloud collaborative equipment status operation and maintenance command system is as follows: Figure 14 shown.

[0189] In the above embodiment, during the operation and maintenance of hydropower station equipment, the target hydropower station with the largest target state prediction value is screened out, and the models of other hydropower stations are updated using the first defect prediction model of the target hydropower station. The first defect prediction result of the target hydropower station is obtained, and the current equipment data of other hydropower stations is input into the updated defect prediction model to obtain the second defect prediction result, so that the defect prediction result of the hydropower station can be obtained more accurately, and then the equipment operation and maintenance instructions of the hydropower station can be generated more accurately, which is conducive to improving the operation and maintenance accuracy of the hydropower station equipment. Moreover, the whole process involves model optimization and experience sharing, which avoids the subjectivity and limitations of the traditional technology of relying solely on the expert experience model of each hydropower station for operation and maintenance processing, which is prone to errors and leads to low operation and maintenance accuracy of the hydropower station equipment, thereby further improving the operation and maintenance accuracy of the hydropower station equipment. At the same time, a complete intelligent defect analysis system has been built on the site side, addressing the current shortcomings in analyzing equipment anomalies and faults. Through ROPN technology, the operation and maintenance experience of multiple parties has been integrated into the intelligent defect analysis system, forming an expert knowledge base. This is optimized through AI, thereby improving and enriching the site's judgment of equipment defects. Through cloud-edge deployment, the site's expert knowledge base is shared, and combined with ROPN+AI technology, the knowledge base is fully utilized and mined to form knowledge. This improves the intelligent defect analysis capabilities of all hydropower stations under the cloud. A complete cloud-edge equipment status operation and maintenance command system has been built to monitor equipment anomalies, analyze operating and maintenance status, and provide real-time understanding of equipment working status and changing trends.

[0190] It should be understood that, although the various steps in the flowcharts involved in the various embodiments described above are displayed in sequence according to the instructions of the arrows, these steps are not necessarily executed in sequence in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction on the execution of these steps, and these steps can be executed in other orders. Moreover, at least a portion of the steps in the flowcharts involved in the various embodiments described above can include multiple steps or multiple stages, and these steps or stages are not necessarily executed and completed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a portion of steps or stages in other steps.

[0191] Based on the same inventive concept, embodiments of the present application also provide a hydropower station equipment operation and maintenance device for implementing the aforementioned hydropower station equipment operation and maintenance method. The solution provided by this device is similar to the solution described in the aforementioned method. Therefore, the specific limitations of one or more embodiments of the hydropower station equipment operation and maintenance device provided below can be found in the above-mentioned limitations of the hydropower station equipment operation and maintenance method, and will not be repeated here.

[0192] In an exemplary embodiment, Figure 15 As shown, a hydropower station equipment operation and maintenance device is provided, including: a data acquisition module 1501, a model determination module 1502, a model update module 1503, a result prediction module 1504, an instruction generation module 1505 and an equipment operation and maintenance module 1506, wherein:

[0193] The data acquisition module 1501 is used to obtain the target state prediction value of each hydropower station.

[0194] The model determination module 1502 is used to select a target hydropower station with the largest target state prediction value from each hydropower station, and determine a first defect prediction model for the target hydropower station.

[0195] The model updating module 1503 is used to update the first defect prediction models of other hydropower stations according to the first defect prediction model of the target hydropower station to obtain updated defect prediction models of other hydropower stations; other hydropower stations are used to represent the hydropower stations other than the target hydropower station in each hydropower station.

[0196] The result prediction module 1504 is used to obtain the first defect prediction result of the target hydropower station, input the current equipment data of other hydropower stations into the updated defect prediction models of other hydropower stations, and obtain the second defect prediction results of other hydropower stations.

[0197] The instruction generation module 1505 is used to generate a first equipment operation and maintenance instruction for the target hydropower station according to the first defect prediction result, and to generate a second equipment operation and maintenance instruction for other hydropower stations according to the second defect prediction result.

[0198] The equipment operation and maintenance module 1506 is used to send the first equipment operation and maintenance instruction and the second equipment operation and maintenance instruction to the edge side, so that the edge side performs corresponding equipment operation and maintenance processing on the target hydropower station according to the first equipment operation and maintenance instruction, and performs corresponding equipment operation and maintenance processing on other hydropower stations according to the second equipment operation and maintenance instruction.

[0199] In an exemplary embodiment, the model updating module 1503 is also used to obtain the first model architecture information of the first defect prediction model of the target hydropower station and the second model architecture information of the first defect prediction model of other hydropower stations; determine the similarity between the first model architecture information and the second model architecture information; when the similarity is greater than the preset similarity, update the model parameters of the first defect prediction model of other hydropower stations to obtain the updated defect prediction model of other hydropower stations; when the similarity is less than or equal to the preset similarity, use the first defect prediction model of the target hydropower station as the updated defect prediction model of other hydropower stations.

[0200] In an exemplary embodiment, the data acquisition module 1501 is also used to receive the current equipment data and the first state prediction value for each hydropower station sent by the edge side; the edge side is used to input the current equipment data of each hydropower station into the second defect prediction model of each hydropower station to obtain the third defect prediction result of each hydropower station, and determine the first state prediction value of each hydropower station based on the third defect prediction result of each hydropower station; obtain the historical equipment data of each hydropower station, and obtain the second state prediction value of each hydropower station based on the current equipment data and historical equipment data of each hydropower station; obtain the target state prediction value of each hydropower station based on the first state prediction value and the second state prediction value respectively.

[0201] In an exemplary embodiment, the data acquisition module 1501 is also used to input the current equipment data and historical equipment data of each hydropower station into the target defect prediction model respectively to obtain the fourth defect prediction result of each hydropower station; determine the current weight information corresponding to the fourth defect prediction result of each hydropower station; and determine the second state prediction value of each hydropower station based on the fourth defect prediction result of each hydropower station and the current weight information corresponding to the fourth defect prediction result of each hydropower station.

[0202] In an exemplary embodiment, the data acquisition module 1501 is also used to obtain multiple candidate correspondences; the multiple candidate correspondences are used to represent the correspondences between defect prediction results and weight information under different business scenarios; the current business scenario of each hydropower station is determined respectively, and the candidate correspondences corresponding to the current business scenario are screened out from the multiple candidate correspondences as the target correspondences corresponding to each hydropower station; the target correspondences corresponding to each hydropower station are queried respectively to obtain the current weight information corresponding to the fourth defect prediction result of each hydropower station.

[0203] In an exemplary embodiment, the result prediction module 1504 is also used to preprocess the current equipment data of other hydropower stations respectively to obtain the preprocessed current equipment data of other hydropower stations; input the preprocessed current equipment data of other hydropower stations into the updated defect prediction model of other hydropower stations respectively to obtain the prediction probability corresponding to each preset defect prediction result of other hydropower stations; and screen out the preset defect prediction results whose corresponding prediction probability is greater than the preset probability from each preset defect prediction result of other hydropower stations as the second defect prediction result of other hydropower stations.

[0204] Each module in the aforementioned hydropower station equipment operation and maintenance device can be implemented in whole or in part through software, hardware, or a combination thereof. Each module can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in the computer device in the form of software, so that the processor can call and execute the corresponding operations of each module.

[0205] In an exemplary embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as shown in FIG. Figure 16 As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, abbreviated as I / O) and a communication interface. The processor, memory and input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and computer program in the non-volatile storage medium. The database of the computer device is used to store data such as current device data and target state prediction values. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a method for operating and maintaining hydropower station equipment is implemented.

[0206] Those skilled in the art will understand that Figure 16 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0207] In an exemplary embodiment, a computer device is further provided, including a memory and a processor. The memory stores a computer program, and the processor implements the steps in the above method embodiments when executing the computer program.

[0208] In an exemplary embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments are implemented.

[0209] In an exemplary embodiment, a computer program product is provided, including a computer program. When the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0210] Those skilled in the art will appreciate that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, database or other media used in the embodiments provided in this application may include at least one of non-volatile and volatile memory. Non-volatile memory may include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory may include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in the various embodiments provided herein may include at least one of a relational database and a non-relational database. Non-relational databases may include, but are not limited to, distributed databases based on blockchains. The processor involved in the various embodiments provided herein may be, but are not limited to, a general-purpose processor, a central processing unit, a graphics processing unit, a digital signal processor, a programmable logic unit, a data processing logic unit based on quantum computing, and the like.

[0211] The technical features of the above embodiments can be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0212] The above-described embodiments merely represent several implementation methods of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present application. It should be noted that a person of ordinary skill in the art may make various modifications and improvements without departing from the spirit of the present application, and these modifications and improvements fall within the scope of protection of the present application. Therefore, the scope of protection of the present application shall be determined by the appended claims.

Claims

1. A method for operating and maintaining hydropower station equipment, characterized in that: Applied to the cloud side, the method includes: Obtain the target state prediction value of each hydropower station; From each of the hydropower stations, a target hydropower station having the largest target state prediction value is selected, and a first defect prediction model of the target hydropower station is determined; updating the first defect prediction models of other hydropower stations according to the first defect prediction model of the target hydropower station to obtain updated defect prediction models of the other hydropower stations; the other hydropower stations are used to represent the hydropower stations other than the target hydropower station in each of the hydropower stations; Obtaining a first defect prediction result of the target hydropower station, and inputting current equipment data of the other hydropower stations into the updated defect prediction models of the other hydropower stations to obtain second defect prediction results of the other hydropower stations; generating a first equipment operation and maintenance instruction for the target hydropower station based on the first defect prediction result, and generating a second equipment operation and maintenance instruction for the other hydropower stations based on the second defect prediction result; The first equipment operation and maintenance instruction and the second equipment operation and maintenance instruction are sent to the edge side, so that the edge side performs corresponding equipment operation and maintenance processing on the target hydropower station according to the first equipment operation and maintenance instruction, and performs corresponding equipment operation and maintenance processing on the other hydropower stations according to the second equipment operation and maintenance instruction.

2. The method according to claim 1, characterized in that The updating of the first defect prediction models of other hydropower stations according to the first defect prediction model of the target hydropower station to obtain updated defect prediction models of the other hydropower stations includes: Acquire first model architecture information of the first defect prediction model of the target hydropower station, and second model architecture information of the first defect prediction models of the other hydropower stations; determining a similarity between the first model architecture information and the second model architecture information; When the similarity is greater than a preset similarity, updating the model parameters of the first defect prediction model of the other hydropower station to obtain an updated defect prediction model of the other hydropower station; When the similarity is less than or equal to the preset similarity, the first defect prediction model of the target hydropower station is used as the updated defect prediction model of the other hydropower stations.

3. The method according to claim 1, characterized in that The step of obtaining the target state prediction value of each hydropower station includes: receiving the current equipment data and the first state prediction value for each hydropower station sent by the edge side; the edge side is configured to input the current equipment data of each hydropower station into the second defect prediction model of each hydropower station to obtain a third defect prediction result of each hydropower station, and determine the first state prediction value of each hydropower station based on the third defect prediction result of each hydropower station; Acquire historical equipment data of each hydropower station, and obtain a second state prediction value of each hydropower station based on the current equipment data and historical equipment data of each hydropower station; A target state prediction value of each hydropower station is obtained based on the first state prediction value and the second state prediction value respectively.

4. The method according to claim 3, characterized in that The step of obtaining the second state prediction value of each hydropower station based on the current equipment data and historical equipment data of each hydropower station includes: Inputting the current equipment data and the historical equipment data of each hydropower station into the target defect prediction model respectively to obtain a fourth defect prediction result of each hydropower station; Determining current weight information corresponding to the fourth defect prediction result of each hydropower station; The second state prediction value of each hydropower station is determined according to the fourth defect prediction result of each hydropower station and the current weight information corresponding to the fourth defect prediction result of each hydropower station.

5. The method according to claim 4, characterized in that The determining of the current weight information corresponding to the fourth defect prediction result of each hydropower station includes: Acquire multiple candidate corresponding relationships; the multiple candidate corresponding relationships are used to represent the corresponding relationship between the defect prediction results and the weight information in different business scenarios; Determine the current business scenario of each hydropower station respectively, and select a candidate corresponding relationship corresponding to the current business scenario from the multiple candidate corresponding relationships as the target corresponding relationship corresponding to each hydropower station; The target correspondence relationship corresponding to each hydropower station is queried respectively to obtain current weight information corresponding to the fourth defect prediction result of each hydropower station.

6. The method according to any one of claims 1 to 5, characterized in that The step of inputting the current equipment data of the other hydropower stations into the updated defect prediction models of the other hydropower stations to obtain second defect prediction results of the other hydropower stations includes: Preprocessing the current equipment data of the other hydropower stations respectively to obtain the preprocessed current equipment data of the other hydropower stations; Inputting the preprocessed current equipment data of the other hydropower stations into the updated defect prediction models of the other hydropower stations to obtain prediction probabilities corresponding to the preset defect prediction results of the other hydropower stations; From the preset defect prediction results of the other hydropower stations, the preset defect prediction results whose corresponding prediction probabilities are greater than the preset probabilities are respectively screened out as the second defect prediction results of the other hydropower stations.

7. A hydropower station equipment operation and maintenance device, characterized in that: Applied to the cloud side, the device includes: A data acquisition module is used to obtain the target state prediction value of each hydropower station; a model determination module, configured to select a target hydropower station having the largest target state prediction value from each of the hydropower stations, and determine a first defect prediction model for the target hydropower station; a model updating module, configured to update the first defect prediction models of other hydropower stations according to the first defect prediction model of the target hydropower station, to obtain updated defect prediction models of the other hydropower stations; the other hydropower stations are used to represent the hydropower stations other than the target hydropower station in each of the hydropower stations; A result prediction module is used to obtain a first defect prediction result of the target hydropower station, input the current equipment data of the other hydropower stations into the updated defect prediction models of the other hydropower stations, and obtain a second defect prediction result of the other hydropower stations; an instruction generation module, configured to generate a first equipment operation and maintenance instruction for the target hydropower station based on the first defect prediction result, and to generate a second equipment operation and maintenance instruction for the other hydropower stations based on the second defect prediction result; The equipment operation and maintenance module is used to send the first equipment operation and maintenance instruction and the second equipment operation and maintenance instruction to the edge side, so that the edge side performs corresponding equipment operation and maintenance processing on the target hydropower station according to the first equipment operation and maintenance instruction, and performs corresponding equipment operation and maintenance processing on the other hydropower stations according to the second equipment operation and maintenance instruction.

8. 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 method according to any one of claims 1 to 6 are implemented.

9. 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 method according to any one of claims 1 to 6 are implemented.

10. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images