Fault prediction method, device and electronic equipment for power system

Abstract

The present invention discloses a fault prediction method, device and electronic device for a power system. Among them, the method includes: obtaining target historical operation data of the power system, where the above target historical operation data at least includes first normal operation data and first abnormal operation data; based on the above target historical operation data, using a decision tree algorithm to construct a target database; based on the above target database, using the above decision tree algorithm and an adaptive boosting algorithm to construct a target fault prediction model; obtaining current operation data of a device to be tested; based on the above current operation data, using the above target fault prediction model to obtain a fault prediction result of the device to be tested. The present invention solves the technical problems of low data analysis efficiency and inaccurate fault prediction existing in the fault analysis method of the power system in the related art.

Description

Technical Field

[0001] The present invention relates to the field of fault prediction, and in particular, to a fault prediction method, device, and electronic device for a power system. Background Art

[0002] Currently, the power system has the characteristics of large data volume, fast data growth rate, and heavy analysis work. Facing the massive data accumulated during the historical operation process, it is impossible to systematically collect and index by type. In addition, in the prior art, mainly based on massive data, real-time data analysis and fault prediction of the power system are carried out manually, which requires a large amount of time cost and labor cost, with low fault prediction efficiency and poor prediction accuracy. It is difficult to efficiently complete fault prediction and avoidance at the stage where risks are controllable, which is not conducive to the stability of the power system business.

[0003] In view of the above problems, no effective solution has been proposed yet. Summary of the Invention

[0004] Embodiments of the present invention provide a fault prediction method, device, and electronic device for a power system, so as to at least solve the technical problems of low data analysis efficiency and inaccurate fault prediction existing in the fault analysis method of the power system in the related art.

[0005] According to one aspect of the embodiments of the present invention, a fault prediction method for a power system is provided, including: obtaining target historical operation data of the power system, where the target historical operation data at least includes first normal operation data and first abnormal operation data; based on the target historical operation data, using a decision tree algorithm to construct a target database; based on the target database, using the decision tree algorithm and an adaptive boosting algorithm to construct a target fault prediction model; obtaining current operation data of a device to be tested; using the target fault prediction model to predict the current operation data to obtain a fault prediction result of the device to be tested.

[0006] Optionally, the step of using a decision tree algorithm to construct a target database based on the target historical operation data includes: obtaining difference information between the first normal operation data and the first abnormal operation data; based on the difference information, using the decision tree algorithm to perform first classification processing on the target historical operation data to obtain a first data type corresponding to the target historical operation data; establishing an index relationship according to the first data type and the target historical operation data; constructing the target database according to the index relationship.

[0007] Optionally, establishing an index relationship between the first data type and the target historical operation data includes: classifying the first data type to obtain a second data type and a first label corresponding to the second data type; establishing an index relationship between the first label, the second data type, and the target historical operation data.

[0008] Optionally, constructing a target fault prediction model based on the target database using the decision tree algorithm and the adaptive boosting algorithm includes: based on the data in the target database, using the adaptive boosting algorithm to obtain a first feature set and a second feature set, where the first feature set is used to characterize fault features and the second feature set is used to optimize a first decision tree; based on the first feature set, using the decision tree algorithm to obtain the first decision tree; using the second feature set to optimize the first decision tree to obtain a first optimization result; determining the target fault prediction model according to the first optimization result.

[0009] Optionally, based on the data in the target database, using the adaptive boosting algorithm to determine a first feature set and a second feature set, where the first feature set is used to characterize fault features and the second feature set is used to optimize a first decision tree, includes: using the adaptive boosting algorithm to classify the data in the target database to obtain second normal operation data and second abnormal operation data; based on the first abnormal operation data and the second abnormal operation data, obtaining the first feature set; based on the first normal operation data and the second normal operation data, obtaining the second feature set.

[0010] Optionally, using the adaptive boosting algorithm to classify the data in the target database to obtain hidden danger feature data; based on the hidden danger feature data and the first normal operation data, obtaining a third feature set; using the second feature set and the third feature set to optimize the first decision tree to obtain a second optimization result; updating the target fault prediction model according to the second optimization result.

[0011] Optionally, obtaining the target historical operation data of the power system includes: obtaining the initial historical operation data of the power system; preprocessing the initial historical operation data to obtain the target historical operation data.

[0012] According to another aspect of the embodiments of the present invention, there is provided a fault prediction device for a power system, including: a first acquisition module configured to acquire target historical operation data of the power system, where the target historical operation data includes at least first normal operation data and first abnormal operation data; a first construction module configured to construct a target database based on the target historical operation data by using a decision tree algorithm; a second construction module configured to construct a target fault prediction model based on the target database by using the decision tree algorithm and an adaptive boosting algorithm; a second acquisition module configured to acquire current operation data of a device to be tested; and a prediction module configured to predict the current operation data by using the target fault prediction model to obtain a fault prediction result of the device to be tested.

[0013] According to still another aspect of the embodiments of the present invention, there is provided a non-volatile storage medium storing multiple instructions, and the instructions are adapted to be loaded and executed by a processor to perform any one of the fault prediction methods for a power system.

[0014] According to yet another aspect of the embodiments of the present invention, there is provided an electronic device including: one or more processors and a memory, where the memory is used to store one or more programs, and when the one or more programs are executed by the one or more processors, the one or more processors are caused to implement any one of the fault prediction methods for a power system.

[0015] In the embodiments of the present invention, a method of establishing a fault prediction model is adopted. By acquiring target historical operation data of the power system, where the target historical operation data includes at least first normal operation data and first abnormal operation data; constructing a target database based on the target historical operation data by using a decision tree algorithm; constructing a target fault prediction model based on the target database by using the decision tree algorithm and an adaptive boosting algorithm; acquiring current operation data of a device to be tested; and predicting the current operation data by using the target fault prediction model to obtain a fault prediction result of the device to be tested. The purpose of introducing an algorithm to analyze data and construct a fault prediction model, and accurately performing fault prediction based on the model is achieved, and the technical effect of efficiently indexing data and predicting potential faults is realized, thereby solving the technical problem of low data analysis efficiency and inaccurate fault prediction existing in the fault analysis method of the power system in the related art. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The drawings described herein are used to provide a further understanding of the present invention, and constitute a part of this application. The illustrative embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation to the present invention. In the drawings:

[0017] Figure 1Flowchart of the fault prediction method for the power system provided by the embodiment of the present invention;

[0018] Figure 2 Schematic diagram of the fault prediction device for the power system provided by the embodiment of the present invention;

[0019] Figure 3 Schematic diagram of an electronic device provided by the embodiment of the present invention. Detailed implementation manners

[0020] In order to enable those skilled in the art to better understand the solution of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0021] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and do not necessarily need to be used to describe a specific order or sequence. It should be understood that such data can be interchanged under appropriate circumstances so that the embodiments of the present invention described here can be implemented in an order other than those illustrated or described here. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device including a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0022] Term description

[0023] Decision Tree, the decision tree is a very commonly used classification method and a supervised learning method based on given samples. The so-called supervised learning means that each sample has a pre-determined attribute and category, and a classifier is obtained through learning, and this classifier can give the correct classification to newly emerging objects.

[0024] Adaboost (Adaptive Boosting), the adaptive boosting algorithm is an iterative algorithm, and its core idea is to train different classifiers (weak classifiers) for the same sample set, and then combine these weak classifiers to form a stronger final classifier (strong classifier).

[0025] According to an embodiment of the present invention, a method embodiment of a fault prediction method for a power system is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in a different order than here.

[0026] Figure 1 is a fault prediction method for a power system according to an embodiment of the present invention. As Figure 1 shown, the method includes the following steps:

[0027] Step S102, obtain the target historical operation data of the power system, where the above target historical operation data includes at least first normal operation data and first abnormal operation data;

[0028] Step S104, based on the above target historical operation data, use the decision tree algorithm to construct a target database;

[0029] Step S106, based on the above target database, use the above decision tree algorithm and the adaptive boosting algorithm to construct a target fault prediction model;

[0030] Step S108, obtain the current operation data of the device to be tested;

[0031] Step S110, use the above target fault prediction model to predict the above current operation data to obtain the fault prediction result of the device to be tested.

[0032] Through the above steps, the purpose of introducing algorithm analysis data and constructing a fault prediction model can be achieved, and the technical effects of efficiently indexing data and predicting potential faults can be realized, thereby solving the technical problems of low data analysis efficiency and inaccurate fault prediction existing in the fault analysis method of the power system in the related technology.

[0033] In the fault prediction method for a power system provided by the embodiment of the present invention, first, obtain the target historical operation data of the power system, and use the decision tree algorithm to construct a target database according to the above target historical operation data for systematic data query. Secondly, combine the decision tree and the adaptive boosting algorithm to process the data in the above target database to obtain a target fault detection model. The above target fault detection model performs fault prediction on the device to be tested in the power system. After that, obtain the current operation data of the device to be tested for prediction to obtain the corresponding fault prediction result.

[0034] Optionally, the above target historical operation data includes but is not limited to: operation data generated by actual devices, process data generated by power communication systems, and operation data generated by power transmission and transformation systems.

[0035] Optionally, the device to be tested described above is at least a device that has obtained target historical operation data or the same type of device, where the same type of device may be a device with a similar application method or a similar structure.

[0036] Optionally, there may be multiple types of the above-mentioned target databases. For example, relational databases such as Oracle and MySql.

[0037] In an optional embodiment, based on the above-mentioned target historical operation data, using a decision tree algorithm to construct a target database, including: obtaining the difference information between the above-mentioned first normal operation data and the above-mentioned first abnormal operation data; based on the above-mentioned difference information, using the above-mentioned decision tree algorithm to perform a first classification process on the above-mentioned target historical operation data to obtain the first data type corresponding to the above-mentioned target historical operation data; establishing an index relationship according to the above-mentioned first data type and the above-mentioned target historical operation data; and constructing the above-mentioned target database according to the above-mentioned index relationship.

[0038] It can be understood that the above-mentioned first normal operation data and the above-mentioned first abnormal operation data included in the above-mentioned target historical operation data are obtained, and classified to obtain difference information. The decision tree algorithm is used to perform classification processing according to the difference information, and the first data type is obtained through the above-mentioned processing. An index relationship is established according to the first data type and the above-mentioned target historical operation data, and a target database is established according to the above-mentioned index relationship. Different from simple storage, the above-mentioned target database can quickly obtain multiple data of the same data type according to the established index relationship.

[0039] Optionally, there may be multiple types of the above-mentioned difference information. For example, the difference between the normal operation state and the abnormal operation state, the difference between different working modes in the normal state, the difference between different device operation information in the normal state of the power system, the difference between different device operation information in the abnormal state of the power system, and so on.

[0040] Optionally, the index relationship between the above-mentioned first data type and the above-mentioned target historical data may be multiple types. For example, one first data type corresponds to at least one of the above-mentioned target historical data.

[0041] In an optional embodiment, establishing the index relationship according to the above-mentioned first data type and the above-mentioned target historical operation data includes: performing a classification process on the above-mentioned first data type to obtain a second data type and the first label corresponding to the above-mentioned second data type; and establishing an index relationship among the above-mentioned first label, the above-mentioned second data type, and the above-mentioned target historical operation data.

[0042] It can be understood that the above-mentioned second data is obtained by classifying multiple above-mentioned first data types, and the first label corresponding to the above-mentioned second data type is obtained. An index relationship is established among the above-mentioned first label, the above-mentioned second data type, and the above-mentioned target historical operation data. By classifying to obtain the first label, the data classification is further strengthened, facilitating type-indexing of the above-mentioned target historical data.

[0043] Optionally, there can be multiple index relationships between the above-mentioned second data type and the above-mentioned first label. For example, there is an index relationship between multiple second data types and the same first label, and there is an index relationship between multiple second data types and multiple first labels.

[0044] Optionally, there can be multiple index relationships between the above-mentioned second data type and the above-mentioned first data type. For example, one second data type can correspond to at least one first data type. Further, based on the index relationship between the first data type and the target historical data, for another example, one second data type corresponds to multiple target historical data, one second data type corresponds to one target historical data, and so on.

[0045] In an optional embodiment, based on the above-mentioned target database, the above-mentioned decision tree algorithm and the adaptive boosting algorithm are used to construct a target fault prediction model, including: based on the data in the above-mentioned target database, the above-mentioned adaptive boosting algorithm is used to obtain a first feature set and a second feature set, where the above-mentioned first feature set is used to characterize fault features, and the above-mentioned second feature set is used to optimize the first decision tree; based on the above-mentioned first feature set, the above-mentioned decision tree algorithm is used to obtain the above-mentioned first decision tree; the above-mentioned first decision tree is optimized by using the above-mentioned second feature set to obtain a first optimization result; the above-mentioned target fault prediction model is determined according to the above-mentioned first optimization result.

[0046] It can be understood that by using the adaptive boosting algorithm and based on the data in the above-mentioned target database, two different feature sets are obtained: a first feature set and a second feature set. Among them, the above-mentioned first feature set is used to characterize fault features, and the above-mentioned second feature set is used to optimize the first decision tree. Based on the above-mentioned first feature set, the decision tree algorithm constructs the first decision tree according to the characterized fault features. The data features in the above-mentioned second feature set are used to optimize the above-mentioned first decision tree, increasing the judgment ability of the first decision tree to obtain a first optimization result. The above-mentioned target fault model is determined based on the above-mentioned first optimization result.

[0047] Optionally, the above-mentioned first decision tree can be optimized using the above-mentioned second feature set. For example, the data features in the above-mentioned second feature set are used to prune the above-mentioned first decision tree to remove incorrect classification nodes in the algorithm. The above-mentioned pruning is a process of optimizing the algorithm of the first decision tree, which improves the judgment ability of the first decision tree.

[0048] In an alternative embodiment, based on the data in the above-mentioned target database, using the above-mentioned adaptive boosting algorithm, the first feature set and the second feature set are determined, including: classifying the data in the above-mentioned target database using the above-mentioned adaptive boosting algorithm to obtain second normal operation data and second abnormal operation data; obtaining the above-mentioned first feature set based on the above-mentioned first abnormal operation data and the above-mentioned second abnormal operation data; obtaining the above-mentioned second feature set based on the above-mentioned first normal operation data and the above-mentioned second normal operation data.

[0049] It can be understood that first, the above-mentioned adaptive boosting algorithm is used to classify and process to obtain second normal operation data and second abnormal operation data. The second normal operation data and the second abnormal operation data are the results after algorithm processing, and the first normal operation data and the first abnormal operation data are real data. Based on the above-mentioned first abnormal operation data and the above-mentioned second abnormal operation data, the above-mentioned first feature set is obtained, which is used to characterize fault features; based on the above-mentioned first normal operation data and the above-mentioned second normal operation data, the above-mentioned second feature set is obtained, which is used to optimize the above-mentioned first decision tree.

[0050] In an alternative embodiment, the data in the above-mentioned target database is classified using the above-mentioned adaptive boosting algorithm to obtain potential hazard feature data, where the potential hazard feature data is first normal operation data with a preset abnormal trend; the third feature set is obtained based on the potential hazard feature data and the above-mentioned first normal operation data; the above-mentioned first decision tree is optimized using the above-mentioned second feature set and the above-mentioned third feature set to obtain a second optimization result; the above-mentioned target fault prediction model is updated according to the above-mentioned second optimization result.

[0051] It can be understood that in order to further optimize the above-mentioned target fault model, a new feature set is added to improve the judgment ability of the first decision tree. The above-mentioned adaptive boosting algorithm is used for classification processing to obtain first normal operation data with a preset abnormal trend as potential hazard feature data. The potential hazard feature data is the data after algorithm processing, and the first normal operation data is real data. The third feature set is obtained based on the potential hazard feature data and the above-mentioned first normal operation data. Combining the above-mentioned second feature set and the above-mentioned third feature set, the above-mentioned first decision tree is further optimized to obtain a second optimization result. The target fault prediction model is updated based on the above-mentioned second optimization result to obtain a target fault prediction model with stronger prediction ability.

[0052] Optionally, there may be multiple types of the above-mentioned potential hazard feature data. Since, based on a preset abnormal trend, data in the first normal operation data is selected as the potential hazard feature data. The method of setting the above-mentioned preset abnormal trend determines different ranges for selecting the above-mentioned potential hazard feature data. For example, the occurrence of some abnormalities in the power system is predictable. Before the actual abnormality occurs, the corresponding trend law can be found from the operation data. Although the above-mentioned operation data is still within the normal range, it will enter the abnormal range within a certain period according to the above-mentioned corresponding trend law. The above-mentioned preset abnormal trend includes, but is not limited to: the difference from the limit value is less than the preset threshold, and the change amplitude exceeds the preset amplitude range within the preset time period.

[0053] In an alternative embodiment, the above-mentioned obtaining of the target historical operation data of the power system includes: obtaining the initial historical operation data of the above-mentioned power system; preprocessing the above-mentioned initial historical operation data to obtain the above-mentioned target historical operation data.

[0054] Optionally, there can be multiple types of the above-mentioned preprocessing methods. There are multiple data sources for the initial historical operation data in the power system, resulting in differences in data storage formats, file types, and naming methods, which are not convenient for processing. For the above reasons, preprocessing is performed on the initial historical data. Multiple processing methods can be, for example, adopting the same naming method and the same storage format, which is convenient for subsequent processing on the same basis.

[0055] Based on the above embodiments and alternative embodiments, the present invention proposes an alternative implementation method, which specifically includes the following steps:

[0056] Step S1, obtain the initial historical operation data from multiple data sources of the power system. By converting it into files of the same format and adopting the same naming method, the target historical operation data is obtained. Among them, the above-mentioned target historical operation data includes at least the first normal operation data and the first abnormal operation data.

[0057] Step S2: Based on the difference information between the above-mentioned first normal operation data and the first abnormal operation data. For example, the difference information can be the difference between the normal operation state and the abnormal operation state, the difference between the operation information of devices in different power systems in the normal state, and the difference between the operation information of devices in different power systems in the abnormal state. Use the above decision tree algorithm to perform the first classification process on the above target historical operation data to obtain the first data type corresponding to the above target historical operation data. For example, based on the difference between the normal operation state and the abnormal operation state in the difference information, the first data type A is obtained; based on the difference between the operation information of devices in different power systems in the normal state, the first data type B is obtained; based on the difference between the operation information of devices in different power systems in the abnormal state, the first data type C is obtained. The first data type A corresponds to the target historical operation data A, the first data type B corresponds to the target historical operation data B, and the first data type C corresponds to the target historical operation data C. Classify the first data type A, the first data type B, and the first data type C to obtain the second data type and the corresponding first data label. The first data type A and the first data type B correspond to the second data type D, and the first data type C corresponds to the second data type E. Both the second data type D and the second data type E correspond to the first data label. Based on the above corresponding index relationship, construct the target database. In the above target database, the target historical operation data A, the target historical operation data B, and the target historical operation data C can be indexed through the above first data label.

[0058] Step S3: Based on the above target database, use the adaptive boosting algorithm to perform classification processing to obtain the second normal operation data, the second abnormal operation data, and the hidden danger feature data. Based on the above first abnormal operation data and the second abnormal operation data, obtain the above first feature set; based on the above first normal operation data and the second normal operation data, obtain the above second feature set; based on the first normal operation data and the hidden danger feature data, obtain the above third feature set. Based on the above first feature set, use the above decision tree algorithm to obtain the above first decision tree; use the above second feature set to optimize the above first decision tree to obtain the first optimization result; determine the above target fault prediction model according to the above first optimization result. Use the above second feature set and the above third feature set to optimize the above first decision tree to obtain the second optimization result; update the above target fault prediction model according to the above second optimization result.

[0059] Step S4: Obtain the current operation data of the device to be tested;

[0060] Step S5: Use the above-mentioned target fault prediction model to predict the above-mentioned current operation data, and obtain the fault prediction result of the device to be tested. Based on the above-mentioned fault prediction result, a prompt message is sent to relevant staff to play a role in risk avoidance.

[0061] From the above optional implementation manners, a large amount of data in the power system can be utilized to construct the above-mentioned target database, which provides an automated data processing means and a classification indexing function, reducing the manual periodic and repetitive workload. Introducing an algorithm for fault prediction increases the prediction accuracy, warns of potential risk hazards, improves the possibility of fault avoidance during the abnormal controllable period, and ensures the stability of the power system business.

[0062] It should be noted that the steps shown in the flowchart of the drawings can be executed in a computer system such as a set of computer-executable instructions, and although the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in a different order than here.

[0063] In this embodiment, a fault prediction device for a power system is also provided. This device is used to implement the above-mentioned embodiments and preferred implementation manners, and those that have been described will not be repeated. As used below, the terms "module" and "device" can be a combination of software and / or hardware that can achieve a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, implementation in hardware, or a combination of software and hardware is also possible and contemplated.

[0064] According to an embodiment of the present invention, an apparatus embodiment for implementing the fault prediction method of a power system is also provided. Figure 2 It is a schematic diagram of a fault prediction device for a power system according to an embodiment of the present invention, as Figure 2 shown. The above-mentioned fault prediction device for a power system includes a first acquisition module 202, a first construction module 204, a second construction module 206, a second acquisition module 208, and a prediction module 210, and an explanation of this device will be given.

[0065] The first acquisition module 202 is used to acquire the target historical operation data of the power system, where the above-mentioned target historical operation data includes at least first normal operation data and first abnormal operation data;

[0066] The first construction module 204 is connected to the first acquisition module 202 and is used to construct a target database based on the above-mentioned target historical operation data by using a decision tree algorithm;

[0067] The second construction module 206 is connected to the first construction module 204 and is used to construct a target fault prediction model based on the above-mentioned target database by using the above-mentioned decision tree algorithm and an adaptive boosting algorithm for the target database;

[0068] A second acquisition module 208, connected to the second construction module 206, is configured to acquire the current operation data of the device to be tested.

[0069] A prediction module 210, connected to the second acquisition module 208, is configured to predict the current operation data by using the target fault prediction model to obtain a fault prediction result of the device to be tested.

[0070] In a fault prediction device for a power system provided by an embodiment of the present invention, by setting a first acquisition module for acquiring target historical operation data of the power system, where the target historical operation data at least includes first normal operation data and first abnormal operation data; a first construction module for constructing a target database based on the target historical operation data by using a decision tree algorithm; a second construction module for constructing a target fault prediction model based on the target database by using the decision tree algorithm and an adaptive boosting algorithm; a second acquisition module for acquiring the current operation data of the device to be tested; and a prediction module for obtaining a fault prediction result of the device to be tested based on the current operation data by using the target fault prediction model. The purpose of introducing an algorithm to analyze data and construct a fault prediction model, and accurately performing fault prediction based on the model is achieved, and the technical effect of efficiently indexing data and predicting potential faults is realized. Furthermore, the technical problem of low data analysis efficiency and inaccurate fault prediction existing in the fault analysis method of the power system in the related art is solved.

[0071] It should be noted that the above-mentioned respective modules can be implemented by software or hardware. For example, for the latter, it can be implemented in the following manner: the above-mentioned respective modules can be located in the same processor; or, the above-mentioned respective modules are located in different processors in any combination.

[0072] It should be noted here that the first acquisition module 202, the first construction module 204, the second construction module 206, the second acquisition module 208, and the prediction module 210 correspond to steps S102 to S110 in the embodiment. The examples and application scenarios implemented by the above-mentioned modules and the corresponding steps are the same, but are not limited to the content disclosed in the above-mentioned embodiment. It should be noted that the above-mentioned modules, as a part of the device, can run in a computer terminal.

[0073] It should be noted that the optional or preferred implementation manners of this embodiment can refer to the relevant descriptions in the embodiment, and will not be repeated here.

[0074] The above virtual learning scenario construction device based on the power system may further include a processor and a memory. The first acquisition module 202, the first construction module 204, the second construction module 206, the second acquisition module 208, the prediction module 210, etc. are all stored in the memory as program units, and the corresponding functions are implemented by the processor executing the above program units stored in the memory.

[0075] The processor contains a kernel, and the kernel retrieves the corresponding program units from the memory. One or more kernels can be set. The memory may include non-permanent memory in the computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of, for example, read-only memory (ROM) or flash memory (flash RAM), and the memory includes at least one memory chip.

[0076] An embodiment of the present invention provides a non-volatile storage medium, on which a program is stored, and when the program is executed by a processor, a fault prediction method for a power system is implemented.

[0077] As Figure 3 shown, an embodiment of the present invention provides an electronic device. The electronic device 10 includes a processor, a memory, and a program stored on the memory and executable on the processor. When the processor executes the program, the following steps are implemented: obtaining target historical operation data of the power system, where the target historical operation data at least includes first normal operation data and first abnormal operation data; based on the target historical operation data, using a decision tree algorithm to construct a target database; based on the target database, using the decision tree algorithm and an adaptive boosting algorithm to construct a target fault prediction model; obtaining current operation data of the device to be tested; using the target fault prediction model to predict the current operation data to obtain a fault prediction result of the device to be tested. The device in this article can be a server, a PC, etc.

[0078] The present invention also provides a computer program product, which is suitable for executing a program initialized with the following method steps when executed on a data processing device: obtaining target historical operation data of the power system, where the target historical operation data at least includes first normal operation data and first abnormal operation data; based on the target historical operation data, using a decision tree algorithm to construct a target database; based on the target database, using the decision tree algorithm and an adaptive boosting algorithm to construct a target fault prediction model; obtaining current operation data of the device to be tested; using the target fault prediction model to predict the current operation data to obtain a fault prediction result of the device to be tested.

[0079] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a system, or a computer program product. Therefore, the present invention can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0080] The present invention is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to embodiments of the present invention. It should be understood that each flow and / or block in the flowchart and / or block diagram, and the combination of flows and / or blocks in the flowchart and / or block diagram, can be realized by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate means for realizing the functions specified in Figure 1 one or more of the processes Figure 1 or a plurality of processes and / or blocks

[0081] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory generate a manufactured article including instruction means, and the instruction means realizes the functions specified in Figure 1 one or more of the processes Figure 1 or a plurality of processes and / or blocks

[0082] These computer program instructions can also be loaded onto a computer or other programmable data processing device, so that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process. Thus, the instructions executed on the computer or other programmable device provide steps for realizing the functions specified in Figure 1 one or more of the processes Figure 1 or a plurality of processes and / or blocks

[0083] In a typical configuration, a computing device includes one or more processors (CPUs), an input / output interface, a network interface, and a memory.

[0084] The memory may include non-permanent memory in a computer-readable medium, in the form of random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash memory (flash RAM). The memory is an example of a computer-readable medium.

[0085] A computer-readable medium includes permanent and non-permanent, removable and non-removable media that can implement information storage by any method or technology. The information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassette tapes, magnetic disk storage or other magnetic storage devices, or any other non-transitory medium that can be used to store information accessible by a computing device. As defined herein, a computer-readable medium does not include transitory computer-readable media, such as modulated data signals and carrier waves.

[0086] It should also be noted that the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, such that a process, method, article or apparatus comprising a series of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article or apparatus. Without further limitation, an element defined by the statement "comprising an..." does not exclude the presence of additional identical elements in the process, method, article or apparatus comprising the element.

[0087] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a system or a computer program product. Therefore, the present invention can take the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk memory, CD-ROM, optical memory, etc.) containing computer-usable program code.

[0088] The above are only embodiments of the present invention and are not used to limit the present invention. For those skilled in the art, the present invention can have various changes and modifications. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included within the scope of the claims of the present invention.

Claims

1. A fault prediction method for a power system, characterized in that, it includes: Obtain the target historical operation data of the power system, where the target historical operation data at least includes first normal operation data and first abnormal operation data; Based on the target historical operation data, use the decision tree algorithm to construct a target database; Based on the target database, use the decision tree algorithm and the adaptive boosting algorithm to construct a target fault prediction model; Obtain the current operation data of the device to be tested; Use the target fault prediction model to predict the current operation data to obtain the fault prediction result of the device to be tested; Among them, the step of constructing the target database based on the target historical operation data by using the decision tree algorithm includes: obtaining the difference information between the first normal operation data and the first abnormal operation data; based on the difference information, using the decision tree algorithm to perform a first classification process on the target historical operation data to obtain the first data type corresponding to the target historical operation data; establishing an index relationship according to the first data type and the target historical operation data; constructing the target database according to the index relationship; Among them, the step of constructing the target fault prediction model based on the target database by using the decision tree algorithm and the adaptive boosting algorithm includes: based on the data in the target database, using the adaptive boosting algorithm to obtain a first feature set and a second feature set, where the first feature set is used to characterize fault features, and the second feature set is used to optimize the first decision tree; based on the first feature set, using the decision tree algorithm to obtain the first decision tree; using the second feature set to perform an optimization process on the first decision tree to obtain a first optimization result; determining the target fault prediction model according to the first optimization result.

2. The method according to claim 1, characterized in that, the step of establishing an index relationship according to the first data type and the target historical operation data includes; Classify the first data type to obtain a second data type and a first label corresponding to the second data type; Establish an index relationship among the first label, the second data type, and the target historical operation data.

3. The method according to claim 1, characterized in that, based on the data in the target database, using the adaptive boosting algorithm to determine the first feature set and the second feature set includes: Using the adaptive boosting algorithm to classify the data in the target database to obtain second normal operation data and second abnormal operation data; Based on the first abnormal operation data and the second abnormal operation data, obtain the first feature set; Based on the first normal operation data and the second normal operation data, obtain the second feature set.

4. The method according to claim 1, characterized in that, the method further includes: Classify the data in the target database using the adaptive enhancement algorithm to obtain hidden danger feature data, where the hidden danger feature data is the first normal operation data with a preset abnormal trend; Based on the hidden danger feature data and the first normal operation data, obtain a third feature set; Use the second feature set and the third feature set to optimize the first decision tree to obtain a second optimization result; Update the target fault prediction model according to the second optimization result.

5. The method according to any one of claims 1 to 4, wherein, the obtaining of the target historical operation data of the power system includes: obtain the initial historical operation data of the power system; preprocess the initial historical operation data to obtain the target historical operation data.

6. A fault prediction device for a power system, wherein, it includes a first obtaining module, configured to obtain target historical operation data of a power system, where the target historical operation data includes at least first normal operation data and first abnormal operation data; a first constructing module, configured to construct a target database based on the target historical operation data using a decision tree algorithm; a second constructing module, configured to construct a target fault prediction model based on the target database using the decision tree algorithm and an adaptive enhancement algorithm for the target database; a second obtaining module, configured to obtain current operation data of a device to be measured; a prediction module, configured to predict the current operation data using the target fault prediction model to obtain a fault prediction result of the device to be measured; wherein, the first constructing module is further configured to: obtain the difference information between the first normal operation data and the first abnormal operation data; based on the difference information, perform a first classification process on the target historical operation data using the decision tree algorithm to obtain the first data type corresponding to the target historical operation data; establish an index relationship according to the first data type and the target historical operation data; construct the target database according to the index relationship; the second constructing module is further configured to: based on the data in the target database, use the adaptive enhancement algorithm to obtain a first feature set and a second feature set, where the first feature set is used to characterize fault features, and the second feature set is used to optimize a first decision tree; based on the first feature set, use the decision tree algorithm to obtain the first decision tree; use the second feature set to optimize the first decision tree to obtain a first optimization result; determine the target fault prediction model according to the first optimization result.

7. A non-volatile storage medium, wherein, the non-volatile storage medium stores multiple instructions, and the instructions are suitable for being loaded and executed by a processor to perform the fault prediction method for a power system according to any one of claims 1 to 5.

8. An electronic device, wherein, it includes: One or more processors and a memory, the memory being configured to store one or more programs, wherein, when the one or more programs are executed by the one or more processors, cause the one or more processors to implement the fault prediction method of the power system according to any one of claims 1 to 5.

Images