A method for improving fault collection operation and maintenance efficiency based on machine learning and RPA

Abstract

The application relates to a method for improving fault collection operation and maintenance efficiency based on machine learning and RPA, and comprises the following steps: collecting historical fault work order data as original data, and preprocessing the original data; adding data labels to the preprocessed data, wherein the data labels indicate the type of fault, the treatment scheme of the fault, and whether the work order is dispatched; placing the data with added data labels into a sample set; training a machine learning algorithm model using the sample set to obtain a trained fault diagnosis model; acquiring fault work order data in real time, inputting the fault work order data into the fault diagnosis model, and outputting the fault type corresponding to the work order, the fault treatment scheme, and whether the work order needs to be dispatched; generating a dispatched work order corresponding to the fault work order data that needs to be dispatched, and calling an RPA robot to transfer the dispatched work order to corresponding operation and maintenance personnel.

Description

Technical Field

[0001] The present invention relates to a method for improving fault collection and operation and maintenance efficiency based on machine learning and RPA, belonging to the technical field of power grid operation and maintenance automation. Background Art

[0002] In recent years, the intelligentization of power grids has made significant progress, with automated control equipment now comprising a significant portion of power supply equipment. However, the highly specialized and prone to failure characteristics of automated equipment have led to an increase in grid operation and maintenance tasks. Rapidly improving grid operation and maintenance efficiency has become a top priority.

[0003] As the application of electricity consumption information collection systems continues to deepen and strengthen, the system's operation and maintenance work has also increased, and the contradiction between urgent operation and maintenance needs and relatively low operation and maintenance efficiency has become increasingly prominent.

[0004] Analyzing from the aspects of technology, process, skills, and methods, the existing operation and maintenance cannot meet the needs. The main reasons are: the accuracy of automatic analysis of abnormal work orders in the electricity information collection system is still not high. In particular, for some equipment power outages or unstable equipment communication signals resulting in incomplete data collection, which do not require operation and maintenance orders, manual analysis and screening are still required. However, the issuance of invalid work orders cannot be completely avoided, which in turn prolongs the total work order processing time.

[0005] Manual O&M diagnostics suffer from slow analysis and diagnosis, a high number of invalid work orders, and lengthy on-site troubleshooting by maintenance personnel. This delays urgent fault repairs, leading to metering disputes and financial losses for power companies and their customers. Therefore, optimizing data collection and O&M methods is crucial to improving overall O&M efficiency and quality.

[0006] Manually dispatching work orders inadvertently increases processing time. The time spent sorting and routing work orders, and finally routing them to the next step, can take up approximately 20% of a worker's initial assessment. Optimizing this process is crucial to improving overall operational efficiency. Summary of the Invention

[0007] To address the problems existing in the above-mentioned prior art, the present invention proposes a method for improving the efficiency of fault collection and maintenance based on machine learning and RPA. This method uses a machine learning algorithm model to automatically diagnose fault work orders and uses RPA robots to complete the dispatch of work orders, thereby reducing manual work and improving the efficiency of fault collection and maintenance.

[0008] The technical solutions of the present invention are as follows:

[0009] In one aspect, the present invention provides a method for improving fault collection and maintenance efficiency based on machine learning and RPA, comprising the following steps:

[0010] Collect historical fault work order data as raw data and pre-process the raw data;

[0011] Add data labels to the preprocessed data, where the data labels indicate the type of fault, the fault handling solution, and whether the work order has been issued; and place the data with the data labels into a sample set;

[0012] Use the sample set to train the machine learning algorithm model to obtain a trained fault diagnosis model;

[0013] Acquire fault work order data in real time, input the fault work order data into the fault diagnosis model, and output the fault type, fault handling solution, and whether the work order needs to be dispatched.

[0014] Call the RPA robot to generate a dispatch work order corresponding to the fault work order data that needs to be dispatched, and transfer the dispatch work order to the corresponding operation and maintenance personnel.

[0015] As a preferred embodiment, the method for preprocessing the raw data is specifically as follows:

[0016] Organize and summarize the raw data by data type;

[0017] Perform data normalization on different types of data according to data type;

[0018] And perform one-hot encoding on discrete data.

[0019] As a preferred embodiment, the method for collecting historical fault work order data is specifically as follows:

[0020] Write a crawler script and use it to automatically capture historical fault ticket data and summarize it into a table document.

[0021] As a preferred embodiment, the data tags specifically include whether the line is disconnected, a terminal type, a power outage, a terminal metering fault, a difference reading, and an order dispatch.

[0022] As a preferred embodiment, the machine learning algorithm model is specifically a LightGBM decision tree algorithm model, and the step of training the machine learning algorithm model using the sample set is specifically as follows:

[0023] Divide the sample set into a test set and a training set;

[0024] Set label columns according to data labels and label descriptions;

[0025] Save the label column data into the LightGBM binary file;

[0026] Set the parameters of the LightGBM decision tree algorithm model;

[0027] Use the training set to train the LightGBM decision tree algorithm model;

[0028] Save the model parameters and use the test set to test the accuracy of the model prediction. If the accuracy reaches the threshold, the trained model is output. If the accuracy is lower than the threshold, the parameters of the LightGBM decision tree algorithm model are reset.

[0029] As a preferred embodiment, during the training process of the LightGBM decision tree algorithm model, the model parameters are adjusted by finding the optimal parameter script;

[0030] In each iteration, the script for finding the optimal parameters determines whether the accuracy of the model prediction reaches a threshold;

[0031] If the standard is not met, the parameters of the LightGBM decision tree algorithm model are updated through the optimization algorithm, or the model structure of the LightGBM decision tree algorithm model is adjusted until the accuracy of the model prediction reaches the threshold.

[0032] As a preferred embodiment, it is characterized in that: the RPA robot specifically adopts the Uibot RPA robot.

[0033] As a preferred embodiment, the steps of calling the RPA robot to generate a dispatch work order corresponding to the fault work order data to be dispatched and forwarding the dispatch work order to the corresponding operation and maintenance personnel are as follows:

[0034] RPA robots include console management robots, executor robots, and production robots;

[0035] The console management robot receives the fault work order data that needs to be dispatched output by the fault diagnosis model, and distributes the fault work order data to the actuator robot;

[0036] The actuator robot sends the fault work order data to the corresponding production robot according to the fault type;

[0037] The production robot prepares a dispatch work order based on the fault type and fault handling solution, and uploads the prepared dispatch work order to the console management robot;

[0038] The console management robot transfers the dispatched work order to the corresponding operation and maintenance personnel through the intranet server.

[0039] On the other hand, the present invention provides an electronic device, including a memory, a processor, and a computer program stored in the memory and runnable on the processor. When the processor executes the program, it implements the method for improving fault collection and operation efficiency based on machine learning and RPA as described in any embodiment of the present invention.

[0040] On the other hand, the present invention provides a computer-readable storage medium having a computer program stored thereon, characterized in that when the program is executed by a processor, it implements the method for improving fault collection and operation efficiency based on machine learning and RPA as described in any embodiment of the present invention.

[0041] The present invention has the following beneficial effects:

[0042] The present invention provides a method for improving the efficiency of fault collection and operation and maintenance based on machine learning and RPA. It uses a machine learning algorithm model to automatically diagnose fault work orders and uses RPA robots to complete the dispatch of work orders, reducing manual work and improving the efficiency of fault collection and operation and maintenance. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 is a flow chart of a method according to an embodiment of the present invention;

[0044] Figure 2 This is an example diagram of data labels in an embodiment of the present invention. DETAILED DESCRIPTION

[0045] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0046] It should be understood that the step numbers used herein are only for convenience of description and are not intended to limit the order in which the steps are executed.

[0047] It should be understood that the terms used in the present specification are only for the purpose of describing specific embodiments and are not intended to limit the present invention. As used in the present specification and the appended claims, the singular forms "a", "an" and "the" are intended to include the plural forms unless the context clearly indicates otherwise.

[0048] The terms “include” and “comprising” indicate the presence of described 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.

[0049] The term "and / or" means any combination of one or more of the associated listed items and all possible combinations, and includes these combinations.

[0050] Embodiment one:

[0051] Referring to Figure 1 The embodiment provides a method for improving fault collection and operation and maintenance efficiency based on machine learning and RPA, comprising the following steps:

[0052] Collect historical fault work order data as original data, and pre-process the original data into unified data segments;

[0053] Adding data labels to the pre-processed data, the data labels indicating the type of fault, the processing scheme of the fault, and whether the work order is dispatched or not; and putting the data with added data labels into a sample set;

[0054] Training a machine learning algorithm model by using the sample set to obtain a trained fault diagnosis model; the machine learning algorithm can be realized by using methods such as decision tree, K-nearest neighbor algorithm, and neural network;

[0055] Real-time acquisition of fault work order data, input of the fault work order data into the trained fault diagnosis model, and prediction of the fault type, the fault processing scheme, and whether the work order needs to be dispatched or not corresponding to the fault work order data by using the fault diagnosis model;

[0056] Calling an RPA robot to generate a dispatch work order corresponding to the fault work order data that needs to be dispatched, and transferring the dispatch work order to corresponding operation and maintenance personnel.

[0057] As a preferred embodiment, the method for pre-processing the original data is specifically:

[0058] The original data is sorted and summarized according to data types;

[0059] Different types of data are normalized according to data types;

[0060] And one-hot encoding is performed on discrete data.

[0061] As a preferred embodiment of the embodiment, the method for collecting historical fault work order data is specifically:

[0062] A crawler script is written by using Python, and historical fault work order data is automatically grabbed by using the crawler script and summarized into a total table document.

[0063] As a preferred embodiment of the embodiment, specifically referring to Figure 2The data tags specifically include a disconnection tag, a terminal type tag, a power failure tag, a terminal metering fault tag, a copy error tag, and a dispatch tag. Figure 2 The specific values and meanings of the tags are shown.

[0064] As a preferred embodiment of the present embodiment, the machine learning algorithm model is specifically a LightGBM decision tree algorithm model. The LightGBM decision tree algorithm is based on a histogram algorithm (first discretize continuous floating-point feature values into k integers, and then construct a histogram with a width of k), supplemented by a leaf-wise leaf growth strategy, a GOSS algorithm, and an EFB algorithm.

[0065] The step of training the machine learning algorithm model using the sample set specifically includes:

[0066] Dividing the sample set into a test set and a training set;

[0067] According to the data tags shown in the table, the tag columns are set, i.e., the tag column setting is performed according to whether the line is disconnected, the terminal type, whether the power is off, whether there is a terminal metering fault, whether there is a copy error, whether there is a dispatch, and a sentence pattern description; Figure 2

[0068] Saving the set tag column data into a LightGBM binary file;

[0069] Setting the parameters of the LightGBM decision tree algorithm model;

[0070] Training the LightGBM decision tree algorithm model using the training set;

[0071] Saving the model parameters and testing the accuracy of the model prediction using the test set. If the accuracy reaches a threshold value, the trained model is output. If the accuracy is lower than the threshold value, the parameters of the LightGBM decision tree algorithm model are reset.

[0072] The LightGBM decision tree algorithm has a faster training speed, consumes less memory, has a high recognition success rate, and supports distributed processing of massive data.

[0073] As a preferred embodiment of the present embodiment, in the training process of the LightGBM decision tree algorithm model, the model parameters are adjusted by a script that finds optimal parameters;

[0074] In each iteration, the script that finds optimal parameters judges whether the accuracy of the model prediction reaches a threshold value;

[0075] ​If the standard is not met, the parameters of the LightGBM decision tree algorithm model are updated through the optimization algorithm. The optimization algorithm can use the steepest gradient descent method, Newton descent method, conjugate gradient descent method, etc., or adjust the model structure of the LightGBM decision tree algorithm model, such as increasing the number of hidden layers, until the accuracy of the model prediction reaches the threshold.

[0076] As a preferred implementation of this embodiment, the RPA robot specifically adopts the Uibot RPA robot.

[0077] As a preferred implementation of this embodiment, the steps of calling the RPA robot to generate a dispatch work order corresponding to the fault work order data to be dispatched and forwarding the dispatch work order to the corresponding operation and maintenance personnel are as follows:

[0078] RPA robots include console management robots, executor robots, and production robots;

[0079] The console management robot receives the fault work order data that needs to be dispatched output by the fault diagnosis model, and distributes the fault work order data to the actuator robot;

[0080] The actuator robot sends the fault work order data to the corresponding production robot according to the fault type;

[0081] The production robot prepares a dispatch work order based on the fault type and fault handling solution, and uploads the prepared dispatch work order to the console management robot;

[0082] The console management robot transfers the dispatched work order to the corresponding operation and maintenance personnel through the intranet server.

[0083] Example 2:

[0084] This embodiment provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, the method for improving fault collection and maintenance efficiency based on machine learning and RPA as described in any embodiment of the present invention is implemented.

[0085] Example 3:

[0086] This embodiment provides a computer-readable storage medium having a computer program stored thereon, characterized in that when the program is executed by a processor, the method for improving fault collection and operation and maintenance efficiency based on machine learning and RPA as described in any embodiment of the present invention is implemented.

[0087] In the embodiments of the present application, "at least one" refers to one or more, and "more" refers to two or more. "And / or" describes the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B can represent the existence of A alone, the existence of A and B at the same time, and the existence of B alone. Among them, A and B can be singular or plural. The character " / " generally indicates that the previous and next associated objects are in an "or" relationship. "At least one of the following" and similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one of a, b and c can represent: a, b, c, a and b, a and c, b and c or a and b and c, where a, b, c can be single or multiple.

[0088] Those skilled in the art will appreciate that the various units and algorithm steps described in the embodiments disclosed herein can be implemented using a combination of electronic hardware, computer software, and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0089] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0090] In the several embodiments provided in this application, if any function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of this application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (Read-Only Memory; hereinafter referred to as: ROM), random access memory (Random Access Memory; hereinafter referred to as: RAM), magnetic disk or optical disk, and other media that can store program code.

[0091] The above descriptions are merely embodiments of the present invention and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention's description and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A method for improving fault collection and maintenance efficiency based on machine learning and RPA, characterized in that: The following steps are involved: Collect historical fault work order data as raw data and pre-process the raw data; Add data labels to the preprocessed data, where the data labels indicate the type of fault, the fault handling solution, and whether a work order has been issued; and place the data with the data labels into a sample set; Use the sample set to train the machine learning algorithm model to obtain a trained fault diagnosis model; Acquire fault work order data in real time, input the fault work order data into the fault diagnosis model, and output the fault type, fault handling solution, and whether the work order needs to be dispatched. Invoke the RPA robot to generate a dispatch work order based on the fault work order data that needs to be dispatched, and then forward the dispatch work order to the corresponding operation and maintenance personnel; The method for preprocessing the original data is specifically as follows: Organize and summarize the raw data by data type; Perform data normalization on different types of data according to data type; And perform one-hot encoding on discrete data; The machine learning algorithm model is specifically a LightGBM decision tree algorithm model, and the steps of training the machine learning algorithm model using the sample set are specifically as follows: Divide the sample set into a test set and a training set; Set label columns according to data labels and label descriptions; Save the label column data into the LightGBM binary file; Set the parameters of the LightGBM decision tree algorithm model; Use the training set to train the LightGBM decision tree algorithm model; Save the model parameters and use the test set to test the accuracy of the model prediction. If the accuracy reaches the threshold, the trained model is output. If the accuracy is lower than the threshold, the parameters of the LightGBM decision tree algorithm model are reset. Among them, during the training process of the LightGBM decision tree algorithm model, the model parameters are adjusted by finding the script of the optimal parameters; In each iteration, the script for finding the optimal parameters determines whether the accuracy of the model prediction reaches a threshold; If the standard is not met, the parameters of the LightGBM decision tree algorithm model are updated through the optimization algorithm, or the model structure of the LightGBM decision tree algorithm model is adjusted until the accuracy of the model prediction reaches the threshold; The steps of calling the RPA robot to generate a dispatch work order based on the fault work order data that needs to be dispatched and forwarding the dispatch work order to the corresponding operation and maintenance personnel are as follows: RPA robots include console management robots, executor robots, and production robots; The console management robot receives the fault work order data that needs to be dispatched output by the fault diagnosis model, and distributes the fault work order data to the actuator robot; The actuator robot sends the fault work order data to the corresponding production robot according to the fault type; The production robot prepares a dispatch work order based on the fault type and fault handling solution, and uploads the prepared dispatch work order to the console management robot; The console management robot transfers the dispatched work order to the corresponding operation and maintenance personnel through the intranet server.

2. The method for improving fault collection and operation efficiency based on machine learning and RPA according to claim 1 is characterized in that: The method for collecting historical fault work order data is specifically as follows: Write a crawler script and use it to automatically capture historical fault ticket data and summarize it into a table document.

3. The method for improving fault collection and maintenance efficiency based on machine learning and RPA according to claim 1 is characterized by: The data tags specifically include whether the line is disconnected, the terminal type, whether there is a power outage, whether there is a terminal metering fault, whether a difference reading is performed, and whether an order is dispatched.

4. The method for improving fault collection and maintenance efficiency based on machine learning and RPA according to claim 1 is characterized by: The RPA robot specifically adopts the Uibot RPA robot.

5. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the method for improving fault collection and operation and maintenance efficiency based on machine learning and RPA as described in any one of claims 1 to 4 is implemented.

6. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the method for improving fault collection and operation and maintenance efficiency based on machine learning and RPA as described in any one of claims 1 to 4 is implemented.

Images