A data operation and maintenance management method and system for a smart power plant

Abstract

The application provides a data operation and maintenance management method and system for a smart power plant. The steps of the method include: obtaining monitoring data transmitted from a plurality of monitoring points of a generator set at a current time point; constructing a monitoring matrix based on the monitoring data transmitted from each monitoring point; inputting the real-time monitoring matrix into a pre-trained first-stage processing model, which outputs a first predicted monitoring matrix for predicting a target time point; taking the first predicted monitoring matrix as an input of a second-stage processing, which includes a plurality of sub-processing stages. In each sub-processing stage, a monitoring matrix corresponding to a time slice before the target time point is calculated by a first processing model, and the monitoring matrix is inputted into a second processing model to update the predicted monitoring matrix of the target time point. The predicted monitoring matrix outputted by the last sub-processing stage is taken as a second predicted monitoring matrix outputted by the second-stage processing.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of smart power plants, and particularly relates to a data operation and maintenance management method and system for a smart power plant. BACKGROUND

[0002] The importance of data supervision for smart power plants is self-evident. It is a key link in the intelligent transformation of the modern energy field, and plays a crucial role in improving energy utilization efficiency, optimizing power supply structure, and ensuring safe and efficient operation of power plants.

[0003] Firstly, data supervision of smart power plants can realize real-time monitoring and analysis of the operation state of the power plant. Through advanced sensors, instruments and other equipment, smart power plants can collect and store various data during the operation of the power plant, such as equipment status, energy consumption, environmental parameters, etc. These data provide a rich information base for the operation and management of the power plant. Through analysis of these data, power plant managers can timely understand the operation status of the power plant and discover potential safety hazards, so as to take appropriate measures for prevention and treatment and avoid accidents.

[0004] In addition, data supervision can also promote intelligent decision-making in power plants. Based on the results of data analysis, smart power plants can provide scientific and reasonable decision-making suggestions for the operation and management of the power plant. For example, in terms of power generation scheduling, smart power plants can predict changes in power demand in the future based on real-time data and historical data, optimize power generation plans and scheduling strategies, and improve power generation efficiency and economic benefits. In terms of equipment maintenance, smart power plants can develop maintenance plans in advance through monitoring and analysis of equipment operation data to reduce the impact of equipment failure on power plant operation.

[0005] However, the existing maintenance scheme usually needs to predict the future operation parameters of the power plant, but often needs a long length of historical data as a reference, and the processing efficiency is low.

[0006] Therefore, the present application is proposed. SUMMARY

[0007] The present application aims to provide a data operation and maintenance management method and system for a smart power plant. The present application first performs coarse-grained prediction through a first-stage processing model, and then performs granular refinement processing through multiple sub-processing stages of a second-stage processing. On the one hand, it does not require long-time window historical data, and on the other hand, it also ensures the accuracy of the prediction.

[0008] The present application provides a data operation and maintenance management method for a smart power plant. The steps of the method include:

[0009] acquiring monitoring data transmitted from a plurality of monitoring points of a generator set at a current time point;

[0010] constructing a real-time monitoring vector based on the monitoring data transmitted by each monitoring point, and constructing a monitoring matrix based on the real-time monitoring vectors of all the monitoring points;

[0011] inputting the real-time monitoring matrix into a pre-trained first-stage processing model, the first-stage processing model outputting a first predicted monitoring matrix for predicting the target time point;

[0012] taking the first predicted monitoring matrix as an input of a second-stage processing, the second-stage processing including multiple sub-processing stages, in each sub-processing stage, a monitoring matrix corresponding to a time slice before the target time point is calculated by a first processing model, and the monitoring matrix is inputted into a second processing model to update the predicted monitoring matrix of the target time point, and the predicted monitoring matrix outputted by the last sub-processing stage is taken as a second predicted monitoring matrix outputted by the second-stage processing.

[0013] With the above scheme, the present scheme only needs the monitoring data of the current time point as the input data, constructs a monitoring matrix based on the monitoring data of the current time point, in the prediction process, firstly performs coarse-grained prediction by the first-stage processing model to obtain a first predicted monitoring matrix of the preliminary prediction, and then performs granular refinement processing by multiple sub-processing stages of the second-stage processing, in the multiple sub-processing stages, the monitoring matrix of the time point before the target time point is continuously predicted, and the monitoring matrix of the target time point is generated again based on the monitoring matrix, which can continuously refine the prediction result and ensure the accuracy of the prediction.

[0014] In some embodiments of the present application, in the processing step of inputting the real-time monitoring matrix into a pre-trained first-stage processing model, the first-stage processing model outputting a first predicted monitoring matrix for predicting the target time point:

[0015] determining the number of loop processing based on the number of time slices between the current time point and the target time point;

[0016] in each loop processing, the monitoring matrix of the next time point is predicted by the first-stage processing model.

[0017] In some embodiments of the present application, in the loop processing of inputting the real-time monitoring matrix into a pre-trained first-stage processing model, the first-stage processing model outputting a first predicted monitoring matrix for predicting the target time point, each loop processing corresponds to a monitoring matrix of a time slice, each loop processing takes the monitoring matrix outputted by the previous loop processing as an input, and the monitoring matrix outputted by the last loop processing is taken as the first predicted monitoring matrix.

[0018] With the above scheme, the first stage processing of the scheme is cyclically processed by the first stage processing model, and the monitoring matrix of the next time slice is predicted in each processing. The granularity refinement degree of the first stage processing can be improved through multiple cyclic processing, and the processing precision is ensured.

[0019] In some embodiments of the present application, the second stage processing is composed of multiple sub-processing stages, and each processing stage is completed by the first processing model and the second processing model.

[0020] In the processing process of the first processing model, the real-time monitoring matrix and the predicted monitoring matrix of the current updated target time point are combined and input into the first processing model. The first processing model outputs the transition monitoring matrix predicted for the time slice between the current time point and the target time point.

[0021] The transition monitoring matrix is input into the second processing model, and the second processing model outputs the predicted monitoring matrix of the updated target time point.

[0022] In some embodiments of the present application, the number of sub-processing stages of the second stage processing corresponds to the number of cycles of the first processing stage. The order of the sub-processing stage based on the second stage processing corresponds to the cyclic processing order of the first processing stage. The first processing model outputs the transition monitoring matrix of the time slice corresponding to the cyclic processing order of the first processing stage, and the second processing model outputs the predicted monitoring matrix of the updated target time point based on the transition monitoring matrix of the time slice.

[0023] With the above scheme, the number of sub-processing stages of the second stage processing of the scheme corresponds to the number of cycles of the first stage processing. Since each cycle corresponds to a time slice, the scheme makes the sub-processing stage one-to-one correspond to the cycle processing, so that the first processing model of the sub-processing stage outputs the transition monitoring matrix of the time slice corresponding to the cycle processing, which can represent the smooth generation at different time steps, thereby capturing the time dynamics and propagation characteristics, and generating the predicted monitoring matrix of the target time point for each time slice. The problem of relying on a long observation window in the prior art is effectively solved.

[0024] In some embodiments of the present application, in the step of constructing a real-time monitoring vector based on the monitoring data transmitted by each monitoring point and constructing a monitoring matrix based on the real-time monitoring vectors of all monitoring points, the monitoring data transmitted by each monitoring point is encoded to obtain a real-time monitoring vector corresponding to the monitoring point. The real-time monitoring vectors are sequentially arranged based on the numbers of the monitoring points to obtain the monitoring matrix.

[0025] In some embodiments of the present application, the method further comprises a health degree determination, and the health degree determination comprises:

[0026] inputting the second prediction monitoring matrix into a pre-trained health degree determination model, outputting a health degree value based on the health degree determination model, and determining an overall health degree of the generator set based on the health degree value;

[0027] comparing the overall health degree with a health degree threshold value, and determining whether the generator set is in an overall healthy state.

[0028] By using the above scheme, the second prediction monitoring matrix obtained by continuous fine processing is used to determine the overall health condition of the generator set, and the overall health condition of the generator set at a future target time point can be determined through monitoring data at a time point, thereby ensuring the accuracy of health determination.

[0029] In some embodiments of the present application, the step of health degree determination further comprises:

[0030] If the generator set is in an overall unhealthy state, the second prediction monitoring matrix corresponding to each monitoring point is split to obtain a prediction monitoring vector corresponding to each monitoring point;

[0031] determining whether the corresponding monitoring point is in a state to be inspected based on the prediction monitoring vector.

[0032] In some embodiments of the present application, in the steps of splitting the second prediction monitoring matrix corresponding to each monitoring point to obtain a prediction monitoring vector corresponding to each monitoring point, and determining whether the corresponding monitoring point is in a state to be inspected based on the prediction monitoring vector:

[0033] obtaining power generation data and working frequency data of the generator set at a plurality of historical time points;

[0034] constructing a power prediction vector and a frequency prediction vector based on the power generation data and working frequency data at a plurality of historical time points, predicting a power prediction value and a frequency prediction value at a target time point based on the power prediction vector and the frequency prediction vector;

[0035] constructing the power prediction value and the frequency prediction value into a supplementary data set, splicing the supplementary data set with the prediction monitoring vector split from the second prediction monitoring matrix, inputting the spliced vector into a determination model corresponding to each monitoring point, and determining whether the monitoring point is in a state to be inspected through the determination model.

[0036] With the above scheme, when it is determined that the current generator set is in a non-integral health state, the scheme splits the prediction monitoring vectors corresponding to each monitoring point in the second prediction monitoring matrix, but individual environmental monitoring data often lacks comparability, making it difficult to ensure accuracy of the determination, and the scheme further predicts the power generation and working frequency of the generator set through historical power generation and working frequency data, splices the predicted power generation and working frequency into each prediction monitoring vector, provides a comparison of the load of the generator set, and determines whether the monitoring point is in a state to be inspected through a neural network model, thereby improving the accuracy of the determination.

[0037] Another aspect of the present application also relates to a data operation and maintenance management system for a smart power plant, which comprises a computer device including a processor and a memory, the memory storing computer instructions, and the processor is configured to execute the computer instructions stored in the memory, and the system realizes the steps implemented by the method.

[0038] In summary, the present application has the following advantages:

[0039] 1. The present scheme only needs monitoring data at the current time point as input data, and constructs a monitoring matrix from the monitoring data at the current time point. In the prediction process, a first-stage processing model is used for coarse-grained prediction to obtain a first prediction monitoring matrix, and a plurality of sub-processing stages of a second-stage processing are used for granular refinement processing. In the plurality of sub-processing stages, the monitoring matrix of the time point before the target time point is continuously predicted, and the monitoring matrix of the target time point is generated again according to the monitoring matrix, which can continuously refine the prediction result and ensure the accuracy of the prediction.

[0040] 2. The first-stage processing of the present scheme is processed by a first-stage processing model, and the monitoring matrix of the next time slice is predicted in each processing, which can improve the granular refinement degree of the first-stage processing through multiple loop processing, and ensure the processing fineness.

[0041] 3. The sub-processing stages of the second-stage processing of the present scheme correspond to the number of loop processing of the first-stage processing. Since each loop processing corresponds to a time slice, the present scheme makes the sub-processing stages correspond to the loop processing one by one, so that the first processing model of the sub-processing stage outputs the transition monitoring matrix of the time slice corresponding to the loop processing, which can represent the smooth generation at different time steps, thereby capturing the time dynamics and propagation characteristics, and generating the prediction monitoring matrix of the target time point piece by piece, effectively solving the problem of relying on a long observation window in the prior art.

[0042] 4. When it is determined that the current generator set is in a non-integrated healthy state, the scheme splits the prediction monitoring vectors corresponding to each monitoring point in the second prediction monitoring matrix, but the individual environmental monitoring data often lacks comparability, making it difficult to ensure the accuracy of the determination, so the scheme further predicts the power generation power and working frequency of the generator set through historical power generation power data and working frequency data, splices the predicted power generation power and working frequency into each prediction monitoring vector, provides a comparison of the load of the generator set, and then determines whether the monitoring point is in a state to be inspected through a neural network model, thereby improving the accuracy of the determination. BRIEF DESCRIPTION OF DRAWINGS

[0043] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings in the following description are only some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained from these drawings without creative labor.

[0044] Figure 1 a schematic diagram of a first embodiment of the data operation and maintenance management method for a smart power plant of the present application;

[0045] Figure 2 a schematic diagram of a second embodiment of the data operation and maintenance management method for a smart power plant of the present application;

[0046] Figure 3 a schematic diagram of a third embodiment of the data operation and maintenance management method for a smart power plant of the present application;

[0047] Figure 4 a schematic diagram of a fourth embodiment of the data operation and maintenance management method for a smart power plant of the present application. DETAILED DESCRIPTION

[0048] The exemplary embodiments will be described in detail herein with reference to the drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present application. Instead, they are merely examples of systems and methods consistent with some aspects of the present application as detailed in the appended claims.

[0049] The terminology used in the present application is for the purpose of describing particular embodiments only and is not intended to be limiting of the present application. As used in the present application and the accompanying claims, the singular form "a", "an" and "the" include plural references unless the context clearly dictates otherwise. It will be further understood that the terms "comprises" and / or "comprising", when used in this specification, specify the presence of stated features, integers, steps, operations, elements, and / or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof.

[0050] As shown in Figure 1 The present application provides a data operation and maintenance management method for a smart power plant, and the steps of the method include:

[0051] In step s100, monitoring data transmitted from a plurality of monitoring points of a generator set at a current time point is acquired.

[0052] In the specific implementation process, the monitoring data includes temperature data, pressure data and sound data of the monitoring points.

[0053] In step S200, a real-time monitoring vector is constructed based on the monitoring data transmitted from each monitoring point, and a monitoring matrix is constructed based on the real-time monitoring vectors of all monitoring points.

[0054] In the specific implementation process, in the step of constructing a real-time monitoring vector based on the monitoring data transmitted from each monitoring point and constructing a monitoring matrix based on the real-time monitoring vectors of all monitoring points, if the monitoring data includes temperature data, pressure data and sound data of the monitoring points, the temperature data, pressure data and sound data of each monitoring point are first constructed into a triple, and the temperature data, pressure data and sound data in the triple are respectively encoded into vectors of a preset length, and the three vectors are connected as a real-time monitoring vector of a monitoring point.

[0055] In step S300, the real-time monitoring matrix is input into a pre-trained first-stage processing model, and the first-stage processing model outputs a first predicted monitoring matrix for predicting a target time point.

[0056] In the specific implementation process, the network structure of the first-stage processing model can be sequentially connected convolutional layers, attention layers, embedding layers, attention layers, embedding layers, attention layers and convolutional layers; the structure of the second processing model of each sub-stage processing in the second-stage processing is the same as the network structure of the first-stage processing model; both models are trained using a pre-set training data set, specifically, the training process is completed by calculating the loss value of the loss function and performing back propagation.

[0057] In some embodiments of the present application, the loss function can use a mean square error loss function:

[0058] Specifically, the mean square error loss function is calculated by the following formula:

[0059]

[0060] wherein, MSE represents the value of the mean square error, N is the sample quantity, y i represents the true value of the i-th sample, represents the predicted value of the i-th sample.

[0061] In step S400, the first predicted monitoring matrix is taken as the input of the second stage processing, the second stage processing includes multiple sub-processing stages, in each sub-processing stage, a monitoring matrix corresponding to a time slice before the target time point is calculated by a first processing model, and the monitoring matrix is input to a second processing model to update the predicted monitoring matrix of the target time point, and the predicted monitoring matrix output by the last sub-processing stage is taken as the second predicted monitoring matrix output by the second stage processing.

[0062] In the specific implementation process, the first processing model can adopt a convolutional neural network model, and is also completed by using a pre-set training data set. The second processing model can also adopt a convolutional neural network model.

[0063] In some embodiments of the present application, in the step of calculating the monitoring matrix corresponding to the time slice before the target time point by the first processing model, the time slice before the target time point can be any time slice between the target time point and the current time point.

[0064] In the specific implementation process, the time slice is a time period uniformly divided in the time period between the target time point and the current time point.

[0065] By using the above scheme, the present scheme only needs the monitoring data of the current time point as the input data, constructs the monitoring matrix from the monitoring data of the current time point, in the prediction process, firstly performs coarse-grained prediction by the first stage processing model to obtain the first predicted monitoring matrix of the preliminary prediction, and then performs granular refinement processing by multiple sub-processing stages of the second stage processing, in multiple sub-processing stages, the monitoring matrix of the time point before the target time point is continuously predicted, and the monitoring matrix of the target time point is generated again according to the monitoring matrix, which can continuously refine the prediction result and ensure the accuracy of the prediction.

[0066] As Figure 2 shown, in some embodiments of the present application, in the processing step of inputting the real-time monitoring matrix into the pre-trained first stage processing model and outputting the first predicted monitoring matrix for predicting the target time point:

[0067] Step S310, determining the number of loop processing based on the number of time slices between the current time point and the target time point;

[0068] Step S320, in each loop processing, the monitoring matrix of the next time point is predicted by the first stage processing model.

[0069] In the specific implementation process, the time slices can be finely divided or coarsely divided, and the finer the division granularity is, the higher the processing accuracy is, but the more the number of loop processing is, and the slower the processing speed is. The specific division granularity can be set according to the actual situation.

[0070] In some embodiments of the present application, in the input of the real-time monitoring matrix into the pre-trained first stage processing model, the first stage processing model outputs the loop processing of the first predicted monitoring matrix of the target time point, each loop processing corresponds to a monitoring matrix of a time slice, and each loop processing takes the monitoring matrix output by the previous loop processing as input and takes the monitoring matrix output by the last loop processing as the first predicted monitoring matrix.

[0071] In the specific implementation process, since the length of each time slice is equal, the length of each backward prediction is the same, and the first stage processing of the present application can be loop processed by a first stage processing model, without the need to additionally increase the model, thereby reducing the training burden of the model.

[0072] By using the above scheme, the first stage processing of the present application is loop processed by the first stage processing model, and the monitoring matrix of the next time slice is predicted in each processing, so that the granularity refinement degree of the first stage processing can be improved through multiple loop processing, and the processing precision is ensured.

[0073] In some embodiments of the present application, the second stage processing is composed of multiple sub-processing stages, and each processing stage is completed by a first processing model and a second processing model:

[0074] In the processing process of the first processing model, the real-time monitoring matrix and the predicted monitoring matrix of the current updated target time point are combined and input into the first processing model, and the first processing model outputs the transition monitoring matrix predicted for the time slice between the current time point and the target time point;

[0075] The transition monitoring matrix is input into the second processing model, and the second processing model outputs the predicted monitoring matrix of the updated target time point.

[0076] In the specific implementation process, the real-time monitoring matrix and the currently updated target time point prediction monitoring matrix are combined and input into the first processing model, and then the real-time monitoring matrix and the currently updated target time point prediction monitoring matrix are spliced ​​together.

[0077] In the specific implementation process, this solution first generates a monitoring matrix of an intermediate time slice in each sub-processing stage, and then uses the monitoring matrix of the time slice to predict the prediction monitoring matrix of the target time point. By continuously performing intermediate segmentation, it continuously achieves refined processing, ensuring prediction accuracy without requiring real data of a long window.

[0078] In some embodiments of the present invention, the number of sub-processing stages in the second stage corresponds to the number of cycles in the first stage. The sub-processing stages correspond to the cyclic processing order in the first stage based on their order in the second stage. A transition monitoring matrix for a time slice corresponding to the cyclic processing order in the first stage is output by a first processing model, and an updated prediction monitoring matrix for the target time point is output by a second processing model based on the transition monitoring matrix of the time slice.

[0079] In the specific implementation process, this scheme corresponds the number of sub-processing stages in the second stage to the number of cycles in the first stage, so as to achieve granularity alignment between the two processes. Through granularity alignment, the prediction for each time slice can be refined twice, further ensuring the processing accuracy.

[0080] Using the above scheme, the sub-processing stage of the second stage of the scheme corresponds to the number of loop processing in the first stage. Since each loop processing corresponds to a time slice, this scheme makes the sub-processing stage correspond one-to-one with the loop processing. The first processing model of the sub-processing stage outputs the transition monitoring matrix of the time slice of the corresponding loop processing, which can represent the smooth generation at different time steps, thereby capturing the temporal dynamics and propagation characteristics, and generating the prediction monitoring matrix of the target time point for each time slice. This effectively solves the problem that the existing technology relies on a long observation window.

[0081] In some embodiments of the present invention, in the steps of constructing a real-time monitoring vector based on the monitoring data input from each monitoring point and constructing a monitoring matrix based on the real-time monitoring vectors of all monitoring points, the monitoring data input from each monitoring point is encoded to obtain the real-time monitoring vector of the corresponding monitoring point, and the real-time monitoring vectors are arranged sequentially based on the numbers of each monitoring point to obtain the monitoring matrix.

[0082] like Figure 3 As shown, in some embodiments of the present invention, the method further includes a health determination step, the health determination step including:

[0083] Step S510, inputting the second prediction monitoring matrix into a pre-trained health degree determination model, outputting a health degree value based on the health degree determination model, and determining the overall health degree of the generator set based on the health degree value;

[0084] Step S520, comparing the overall health degree with a health degree threshold value, and determining whether the generator set is in an overall healthy state.

[0085] In specific implementation, the health degree determination model can be a convolutional neural network model or a multi-layer perception model.

[0086] In specific implementation, if the overall health degree is lower than the health degree threshold value, it is determined that the generator set is in an overall unhealthy state.

[0087] By using the above scheme, the second prediction monitoring matrix obtained through continuous fine processing is used to determine the overall health condition of the generator set, the overall health condition of the generator set at a target time point in the future can be determined through monitoring data at one time point, and the health determination accuracy is ensured.

[0088] As shown in Figure 4 In some embodiments of the present application, the step of determining the health degree further comprises:

[0089] Step S530, if the generator set is in an overall unhealthy state, the second prediction monitoring matrix corresponding to each monitoring point is split to obtain a prediction monitoring vector corresponding to each monitoring point;

[0090] Step S540, determining whether the corresponding monitoring point is in a to-be-inspected state based on the prediction monitoring vector.

[0091] In specific implementation, each row of the second prediction monitoring matrix corresponds to one monitoring point, and the data of the rows corresponding to each monitoring point is split to obtain a prediction monitoring vector corresponding to each monitoring point.

[0092] In some embodiments of the present application, in the steps of splitting the second prediction monitoring matrix corresponding to each monitoring point to obtain a prediction monitoring vector corresponding to each monitoring point, and determining whether the corresponding monitoring point is in a to-be-inspected state based on the prediction monitoring vector:

[0093] Obtaining power generation data and working frequency data of the generator set at a plurality of historical time points;

[0094] The power prediction vector and the frequency prediction vector are constructed based on power generation power data and working frequency data of a plurality of historical time points, and the power prediction value and the frequency prediction value of a target time point are predicted based on the power prediction vector and the frequency prediction vector.

[0095] The power prediction value and the frequency prediction value are constructed into a supplementary data set, and are spliced with a prediction monitoring vector split from a second prediction monitoring matrix, and the spliced vector is input into a determination model corresponding to each monitoring point, and whether the monitoring point is in a state to be inspected is determined through the determination model.

[0096] In the specific implementation process, in the step of constructing the power prediction vector and the frequency prediction vector based on power generation power data and working frequency data of a plurality of historical time points, and predicting the power prediction value and the frequency prediction value of a target time point based on the power prediction vector and the frequency prediction vector, the power generation power data and the working frequency data of each time point are sequentially arranged based on the time stamps in the power generation power data and the working frequency data, to obtain the power prediction vector and the frequency prediction vector; and the prediction is performed through a pre-trained LSTM model.

[0097] LSTM (Long Short-Term Memory, Long Short-Term Memory) is a special recursive neural network (RNN).

[0098] With the above scheme, when it is determined that the current generator set is in a non-integral health state, the prediction monitoring vector corresponding to each monitoring point in the second prediction monitoring matrix is split in the scheme, but the separate environmental monitoring data often lacks controllability, and it is difficult to ensure the accuracy of the determination, and the scheme further predicts the power generation power and the working frequency of the generator set through the historical power generation power data and the working frequency data, splices the predicted power generation power and the working frequency into each prediction monitoring vector, provides a control of the load of the generator set, and then determines whether the monitoring point is in a state to be inspected through a neural network model, thereby improving the accuracy of the determination.

[0099] Another aspect of the present application also relates to a data operation and maintenance management system for a smart power plant, which comprises a computer device, the computer device comprising a processor and a memory, the memory storing computer instructions, and the processor being configured to execute the computer instructions stored in the memory, so that the system implements the steps implemented by the method.

[0100] The embodiment of the present application further provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to implement the foregoing data operation and maintenance management method for a smart power plant. The computer readable storage medium can be a tangible storage medium, such as a random access memory (RAM), a memory, a read only memory (ROM), an electrically programmable ROM, an electrically erasable programmable ROM, a register, a floppy disk, a hard disk, a removable storage disk, a CD-ROM, or any other form of storage medium known in the technical field.

[0101] Those of ordinary skill in the art will appreciate that the various illustrative components, systems and methods described in connection with the embodiments disclosed herein can be implemented as hardware, software, or a combination thereof. The choice of hardware or software implementation is a matter of design choice and will depend on the particular application and constraints of the design. Those of skill in the art will be able to make such design choices without departing from the scope of the present application. Aspects of the application can be implemented in hardware, for example, as an electronic circuit, as an application specific integrated circuit (ASIC), as a firmware, as a plug-in, as a functional card, etc. Aspects of the application can also be implemented in software, for example, as a program or code segment. The program or code segment can be stored in a machine readable medium, or transmitted through a carrier wave in a transmission medium or communication link.

[0102] It is to be understood that the application is not limited to the particular configurations and processes described herein and shown in the drawings, which are provided by way of example only. Detailed descriptions of known methods are omitted so as not to obscure the description of the present application. In the above embodiments, several specific steps are described and illustrated as examples. However, the method processes of the present application are not limited to the specific steps described and illustrated, and one skilled in the art will be able to make various changes, modifications and additions, or change the order of the steps, without departing from the spirit of the present application.

[0103] In the present application, features described and / or illustrated in relation to one embodiment can be used in the same or a similar way in one or more other embodiments, and / or combined with or instead of features of other embodiments.

[0104] The above description is merely illustrative of the application, and is not intended to limit the application. The embodiments of the application can be variously changed and modified without departing from the spirit and scope of the application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the application shall be included in the scope of the application.

Claims

1. A data operation and maintenance management method for a smart power plant, characterized in that, The steps of the method include: acquiring monitoring data transmitted from a plurality of monitoring points of a generator set at a current time point; constructing a real-time monitoring vector based on the monitoring data transmitted from each monitoring point, and constructing a monitoring matrix based on the real-time monitoring vectors of all monitoring points; inputting the real-time monitoring matrix into a pre-trained first-stage processing model, and the first-stage processing model outputs a first predicted monitoring matrix for predicting a target time point; inputting the first predicted monitoring matrix as an input of a second-stage processing, and the second-stage processing includes a plurality of sub-processing stages, in each sub-processing stage, a transition monitoring matrix corresponding to a time slice before the target time point is calculated by a first processing model, in the processing process of the first processing model, the real-time monitoring matrix and the predicted monitoring matrix of the current updated target time point are combined and input into the first processing model, the first processing model outputs a transition monitoring matrix predicted for a time slice between the current time point and the target time point, and the transition monitoring matrix is input into a second processing model to update the predicted monitoring matrix of the target time point, and the predicted monitoring matrix output by the last sub-processing stage is taken as the second predicted monitoring matrix output by the second-stage processing. 2.The data operation and maintenance management method for a smart power plant according to claim 1, wherein, In the processing step of inputting the real-time monitoring matrix into a pre-trained first-stage processing model, and the first-stage processing model outputs a first predicted monitoring matrix for predicting a target time point: determine the number of loop processing based on the number of time slices between the current time point and the target time point; in each loop processing, the monitoring matrix of the next time point is predicted by the first-stage processing model. 3.The data operation and maintenance management method for a smart power plant according to claim 1, wherein, In the loop processing of inputting the real-time monitoring matrix into a pre-trained first-stage processing model, and the first-stage processing model outputs a first predicted monitoring matrix for predicting a target time point, each loop processing corresponds to a monitoring matrix of a time slice, and each loop processing takes the monitoring matrix output by the previous loop processing as an input, and the monitoring matrix output by the last loop processing is taken as the first predicted monitoring matrix.

4. The data operation and maintenance management method for a smart power plant according to claim 3, characterized in that, The number of sub-processing stages of the second-stage processing corresponds to the number of loops of the first processing stage, the order of the sub-processing stages in the second-stage processing corresponds to the loop processing order of the first processing stage, the transition monitoring matrix of the time slice corresponding to the loop processing order of the first processing stage is output by the first processing model, and the updated predicted monitoring matrix of the target time point is output by the second processing model based on the transition monitoring matrix of the time slice.

5. The data operation and maintenance management method for a smart power plant according to claim 1, characterized in that, In the step of constructing a real-time monitoring vector based on the monitoring data transmitted from each monitoring point, and constructing a monitoring matrix based on the real-time monitoring vectors of all monitoring points, the monitoring data transmitted from each monitoring point is encoded to obtain a real-time monitoring vector corresponding to the monitoring point, and the real-time monitoring vectors are sequentially arranged based on the numbers of the monitoring points to obtain the monitoring matrix. 6.The data operation and maintenance management method for a smart power plant according to any one of claims 1-5, characterized in that, The steps of the method further include health degree determination, and the steps of the health degree determination include: inputting the second prediction monitoring matrix into a pre-trained health degree determination model, outputting a health degree value based on the health degree determination model, and determining an overall health degree of the generator set based on the health degree value; comparing the overall health degree with a health degree threshold value, and determining whether the generator set is in an overall healthy state.

7. The data operation and maintenance management method for a smart power plant according to claim 6, characterized in that, The step of health degree determination further comprises: if the generator set is in a non-overall healthy state, splitting the second prediction monitoring matrix corresponding to each monitoring point to obtain a prediction monitoring vector corresponding to each monitoring point; determining whether the corresponding monitoring point is in a state to be inspected based on the prediction monitoring vector.

8. The data operation and maintenance management method for a smart power plant according to claim 7, characterized in that, In the steps of splitting the second prediction monitoring matrix corresponding to each monitoring point to obtain a prediction monitoring vector corresponding to each monitoring point, and determining whether the corresponding monitoring point is in a state to be inspected based on the prediction monitoring vector: obtaining power generation data and working frequency data of the generator set at a plurality of historical time points; constructing a power prediction vector and a frequency prediction vector based on the power generation data and working frequency data at the plurality of historical time points, and predicting a power prediction value and a frequency prediction value at a target time point based on the power prediction vector and the frequency prediction vector; constructing the power prediction value and the frequency prediction value into a supplementary data set, splicing the supplementary data set with the prediction monitoring vector split from the second prediction monitoring matrix, inputting the spliced vector into a determination model corresponding to each monitoring point, and determining whether the monitoring point is in a state to be inspected through the determination model.

9. A data operation and maintenance management system for a smart power plant, characterized in that: The system comprises a computer device, the computer device comprising a processor and a memory, the memory storing computer instructions, and the processor being configured to execute the computer instructions stored in the memory, and when the computer instructions are executed by the processor, the system implements the steps of the method according to any one of claims 1-8.

Images