Intelligent analysis and decision-making system for high-voltage electrical equipment for electric locomotives and motor train units

Abstract

The invention discloses an intelligent analysis and decision-making system for high-voltage electrical equipment for an electric locomotive and a motor train unit, and the system comprises an equipment layer which consists of a lightning arrester, a voltage transformer and a cable terminal, and transmits the real-time state information of the equipment to a state monitoring module; the data calculation layer is composed of a state monitoring module, an edge calculation module, a data transmission module and a cloud data processing module, and is used for carrying out calculation processing on the state information of the monitoring equipment and establishing each model library representing the health state of the monitoring equipment; the analysis decision-making layer is composed of an intelligent analysis decision-making module and is used for establishing a digital twin model of the equipment, making a maintenance decision for the state of the equipment, establishing an operation and maintenance knowledge graph based on historical maintenance records and updating the operation and maintenance knowledge graph in time; and the control execution layer is composed of a control execution module and is used for receiving the operation and maintenance instruction sent by the edge calculation module and the intelligent analysis and decision module and sending operation and maintenance operation to the equipment end. According to the invention, self-state evaluation, abnormity early warning, fault analysis, life prediction and predictive state maintenance of the high-voltage electrical equipment for the electric locomotive and the motor train unit can be realized, a maintenance report is automatically formed to guide operation and maintenance personnel to operate, and autonomous thinking and intelligent operation and maintenance efficiency of the equipment can be effectively improved; and the train travel safety is prevented from being influenced by faults.

Description

Technical Field

[0001] The present invention belongs to the technical field of intelligent operation and maintenance of high-voltage electrical equipment, and more specifically relates to an intelligent analysis and decision-making system for high-voltage electrical equipment used in electric locomotives and multiple units. Background Art

[0002] The deterioration and failure of high-voltage electrical equipment are latent faults that develop slowly. The traditional regular off-line maintenance method requires removing it from the high-voltage system and conducting relatively comprehensive tests on the insulation performance, withstand voltage ability, etc. of voltage transformers, lightning arresters, cable terminals, etc. This regular maintenance method can detect the equipment performance in multiple aspects, comprehensively evaluate its operating status and remaining life, but it cannot timely grasp the real-time working status of the equipment, has a large delay, cannot timely detect sudden faults that develop rapidly, is prone to cause large-scale power outages, and seriously threatens the train operation and the personal safety of staff. At the same time, there are problems such as over-maintenance or under-maintenance in maintenance, and it is extremely easy to cause different degrees of damage to the equipment during the maintenance process, affecting its normal service life.

[0003] Therefore, how to provide an intelligent analysis and decision-making system for high-voltage electrical equipment used in electric locomotives and multiple units to solve the health management and intelligent operation and maintenance of high-voltage electrical equipment for electric locomotives and multiple units is an urgent problem to be solved by those skilled in the art. Summary of the Invention

[0004] In view of this, the present invention provides an intelligent analysis and decision-making system for high-voltage electrical equipment used in electric locomotives and multiple units, which can collect the status information of high-voltage equipment in real time, through comprehensive research and judgment by the intelligent analysis and decision-making system, propose a predictive status maintenance strategy for high-voltage electrical equipment, timely and effectively avoid equipment failures from affecting the train operation safety, and at the same time reduce the train outage time, save regular maintenance costs, improve the equipment availability and service life, and has great economic and social benefits.

[0005] To achieve the above object, the present invention adopts the following technical solutions:

[0006] An intelligent analysis and decision-making system for high-voltage electrical equipment used in electric locomotives and multiple units, comprising: an equipment layer, a data calculation layer, an intelligent analysis and decision-making layer, and a control execution layer. Among them,

[0007] The equipment layer should include lightning arresters, voltage transformers, and cable terminals, and send the real-time status data of the equipment to the status monitoring module;

[0008] The data calculation layer should be composed of a status monitoring module, an edge calculation module, a data transmission module, and a cloud data processing module, and is used for calculating and processing the status information of the monitored equipment, and establishing various model libraries representing its health status;

[0009] The analysis and decision-making layer consists of intelligent analysis and decision-making modules, which are used to establish a digital twin model of the device, propose maintenance decisions based on the operating status of the device, establish and update an operation and maintenance knowledge graph based on historical maintenance records;

[0010] The control and execution layer consists of control and execution modules, which are used to receive operation and maintenance operation instructions sent by the edge computing module and the intelligent analysis and decision-making module, and send operation and maintenance operations to the device side.

[0011] Preferably, the status monitoring module includes a status data acquisition unit and a status data transmission unit; the status data acquisition unit collects device status information based on edge perception devices such as leakage current sensors and ultrasonic sensors. Among them, the leakage current sensor is used to collect leakage current signals, and the ultrasonic sensor is used to collect partial discharge signals; the status data transmission unit is used to upload the collected real-time status data to the edge computing module or upload it to the cloud data processing module through the data transmission module;

[0012] Preferably, the edge computing module includes a data processing and storage unit, a model training unit, and an operation and maintenance decision-making unit; among them,

[0013] The data processing and storage unit is used to remove anomalies and denoise the device status data uploaded by the status monitoring module, and then store the normal data;

[0014] The model training unit is used to perform preliminary training and calculation on the processed normal data in combination with deep learning algorithms such as convolutional neural networks, judge the current state of the device, predict the future state of the device, diagnose and give early warnings about possible faults of the device;

[0015] The operation and maintenance decision-making unit is used to make decisions on the judgment results of the current device state, arrange maintenance plans in advance for devices that may have faults, and send them to the control and execution module;

[0016] Preferably, the data transmission module includes a data receiving unit and a data transmission unit, which are used to upload and download relevant data through wireless communication transmission methods such as LoRa or 5G;

[0017] The data receiving unit is used to receive the device status information of the status monitoring module, the device training models of the edge computing module and the cloud data processing module;

[0018] The data transmission module is used to upload the device training model of the edge computing module to the cloud, and download the updated model from the cloud to the edge computing module;

[0019] Preferably, the cloud data processing module is used to save and update the database, and based on the updated database, perform reinforcement training on the device model to form a complete device model library; it includes a data processing unit, a model reinforcement training unit, and a data storage unit;

[0020] The data processing unit is used to remove anomalies and denoise the device status data uploaded by the data transmission module;

[0021] The model reinforcement training unit, based on the model uploaded by edge computing, combines the processed normal data with deep reinforcement learning algorithms such as DQN for reinforcement training and calculation, judges the current state of the device, predicts the future state of the device, diagnoses and gives early warnings of possible faults of the device;

[0022] The data storage unit is used to store data such as the real-time original state data of the device, historical data, historical models, and newly trained and updated models;

[0023] Preferably, the intelligent analysis and decision-making module establishes its digital twin model based on the device model library trained and updated by the cloud data processing module, determines the predictive maintenance strategy of the device to be inspected based on deep learning, and through the historical records of predictive condition-based maintenance, establishes and updates the device operation and maintenance knowledge graph in a timely manner; it includes an intelligent analysis unit, a decision execution unit, and a knowledge graph construction and update unit;

[0024] The intelligent analysis unit is used to analyze the real-time state characteristics of the device, make operation and maintenance decisions on the device with abnormal states by comparing its digital twin model, and propose the best maintenance strategy based on the latest operation and maintenance knowledge graph;

[0025] The decision execution unit is used to send the final decision instruction to the control execution module;

[0026] The knowledge graph construction and update unit is used to present and update and save the mapping relationship between the device state and the operation and maintenance decision in the form of a knowledge graph;

[0027] Preferably, the control execution module includes an operation and maintenance decision instruction receiving and sending unit and an operation and maintenance decision instruction control unit;

[0028] The operation and maintenance decision instruction receiving and sending unit is used to receive and send operation instructions and operation and maintenance decisions of the edge computing module and the intelligent analysis and decision-making module in real time;

[0029] The operation and maintenance instruction control unit is used to control the start and stop operations of the device-side operation and maintenance.

[0030] The beneficial effects of the present invention are as follows:

[0031] An intelligent analysis and decision-making system for high-voltage electrical equipment used in electric locomotives and multiple units provided by the present invention can effectively avoid the disadvantages of existing regular maintenance of high-voltage electrical equipment, such as poor timeliness, over-maintenance or under-maintenance, high operation and maintenance costs, and low safety and reliability. Aiming at the special operating conditions of high-voltage electrical equipment used in electric locomotives and multiple units, the present invention realizes long-distance, large-capacity, low-power, and highly sensitive information interaction between the device end and the cloud based on LoRa technology and wireless communication technologies such as 5G. To effectively improve the data processing efficiency, the present invention is based on the cloud-edge collaboration technology, initially trains each device model on the edge side, and conducts reinforcement training and updates the model in a timely manner on the cloud. To effectively improve the intelligent level and operation and maintenance efficiency of the device, the present invention constructs an intelligent analysis and decision-making module based on technologies such as digital twin and knowledge graph. This module can automatically evaluate the health status of the device, predict the development trend of the device status, give early warnings of possible faults of the device, and automatically form predictive maintenance strategies, with a high degree of fault self-healing ability and accident analysis and handling ability. At the same time, it can continuously update the operation and maintenance knowledge graph and improve the digital twin model of each device based on historical maintenance records, so as to provide more accurate predictive maintenance decisions and realize the intelligent operation and maintenance and health management of high-voltage electrical equipment used in electric locomotives and multiple units. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained according to the provided drawings.

[0033] Figure 1 It is the overall structure diagram of the present invention.

[0034] Figure 2 It is a schematic diagram of device data calculation and processing based on cloud-edge collaboration.

[0035] Figure 3 It is the construction process of the device digital twin model.

[0036] Figure 4 The formation process of the device operation and maintenance knowledge graph.

[0037] Figure 5 It is the device operation and maintenance decision flow chart.

[0038] Figure 6 The formation process of the operation and maintenance decision table. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0039] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described 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 the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0040] Please refer to the attached Figure 1 , the present invention provides an intelligent analysis and decision-making system for high-voltage electrical equipment used in electric locomotives and multiple units, including: an equipment layer, a data calculation layer, an intelligent analysis and decision-making layer, and a control execution layer. Among them,

[0041] The equipment layer is composed of arresters, voltage transformers, and cable terminals, and sends real-time device status information to the status monitoring module;

[0042] The data calculation layer is composed of a status monitoring module, an edge computing module, a data transmission module, and a cloud data processing module, and is used to calculate and process the status information of the monitored equipment, and establish various model libraries representing its health status;

[0043] The analysis and decision-making layer is composed of an intelligent analysis and decision-making module, which is used to establish a digital twin model of the equipment, propose maintenance decisions on the status of the equipment, establish an operation and maintenance knowledge graph based on historical maintenance records, and update it in a timely manner;

[0044] The control execution layer is composed of a control execution module, which is used to receive the operation and maintenance instructions sent by the edge computing module and the intelligent analysis and decision-making module, and send operation and maintenance operations to the device end.

[0045] In the present invention, the data calculation layer mainly realizes the calculation and processing of device status information through edge computing and edge-cloud collaboration technologies, as Figure 2 shown. Among them, the edge computing module can process part of the device status information, initially train the device model, determine the operation and maintenance decision based on operation and maintenance experience, and make timely responses. The cloud data processing module can process most of the device status information, and strengthen the training of the device model uploaded by the edge computing module based on the latest data to form a complete device model library. The two realize information interaction and sharing through the data transmission module, which mainly ensures the collaborative operation of "edge + cloud" through a combination of wireless communication technologies such as LoRa and 5G, and realizes the intelligent perception, processing, application, and storage of the status information of high-voltage electrical equipment used in electric locomotives and multiple units.

[0046] In the present invention, the intelligent analysis and decision-making layer mainly realizes the self-evaluation of the equipment health status, self-diagnosis of faults, self-prediction of remaining life, and self-generation of operation and maintenance decisions through digital twin technology and knowledge graph technology, so as to guide the operation and maintenance operations and improve the self-healing ability of equipment faults. At the same time, the equipment model is updated through real-time status information, and the equipment operation and maintenance knowledge graph is updated in a timely manner through accident analysis and handling situations, continuously self-improving and optimizing the operation and maintenance decisions to achieve high-precision, high-reliability, and high-efficiency operation and maintenance.

[0047] In the present invention, the construction process of the digital twin model of the equipment is as Figure 3 shown. It is mainly divided into the construction, inspection, and management of the model. The construction of the model is mainly based on the physical characteristics and operating characteristics of the equipment. Among them, the model based on physical characteristics is mainly the output characteristic model of the equipment under multi-physical fields obtained through finite element simulation or experiments, while the model based on operating characteristics is mainly the health management models such as the equipment health status evaluation model, life prediction model, and fault diagnosis model trained in the cloud; the inspection of the model is mainly to compare the detected performance parameters of the equipment with the results calculated and predicted by the digital twin body. When the accuracy is met, it can be used as the final digital twin body. When the accuracy is not met, the model library needs to be disassembled and retrained and updated until the accuracy requirements are met; the model management is mainly based on the digital twin bodies of each equipment, and an independent and complete model library for each equipment is established through a hierarchical concept and updated and saved in a timely manner.

[0048] In the present invention, the equipment status evaluation and life prediction are realized by using a multi-level evaluation method, and its algorithm flow is as follows:

[0049] ① The equipment status evaluation model is divided into three levels. Among them, the most important indicators for the equipment are selected in the first level, which can directly reflect the equipment health status and the equipment degradation amount, mainly including the equipment factory time, design life, service life, etc. The evaluation result is reduced to a health value H within the range of 0-10. 1 . When this value is in the range of 0-3, it indicates that the equipment status is good; when it is in the range of 3.0-6.5, it indicates that the equipment has shown obvious aging phenomena, and the aging process has started to rise significantly; when H 1 > 6.5, it indicates that the equipment has shown serious aging phenomena. In this state, the probability of failure increases significantly. The specific calculation formula is as follows:

[0050]

[0051] In the formula, H 0 is the initial health index of the equipment; B is the aging coefficient; T 1 is the equipment commissioning year; T 2 is the current year or future year. The calculation of B is as follows:

[0052] The initial health index H of the equipment0 Generally taken as 0.5, assuming the expected operation life of the device is T = T 2 -T 1 , at this time H 1 = 6.5, so

[0054] ② The second-level evaluation mainly extracts features from data such as the leakage current and partial discharge of the device, assigns weights to them through the variable weight theory, and then evaluates the device status based on this. By comparing the experimental data with the standard data, the evaluation result can be reduced to a health value H within the range of 0-10 2 , and this value has a non-linear relationship with conditions such as the usage of the device.

[0055] ③ The third-level evaluation mainly proofreads the health values obtained from the first- and second-level evaluations, and its proofreading coefficient is I. Through proofreading, the final health index of the device is obtained as H = H 1 ×H 2 ×I; the value of I can be comprehensively obtained according to the defects, failures, repairs, etc. of the device in the early stage.

[0056] In the present invention, the formation process of the device operation and maintenance knowledge graph is as Figure 4 shown, which is mainly divided into the initial construction link of the operation and maintenance knowledge graph and the update and improvement link of the operation and maintenance knowledge graph. In the initial construction link of the operation and maintenance knowledge graph, based on means such as deep learning, structured data such as device operation data is fused with unstructured data such as maintenance records to establish a preliminary device operation and maintenance knowledge graph, realizing functions such as device accident analysis, anomaly handling, and maintenance decision-making; while in the update link of the operation and maintenance knowledge graph, the operation and maintenance knowledge of the device is complemented, the method is complemented, the case is complemented, and the decision is complemented mainly through the relationship reasoning between the device health state and the characteristic value, knowledge reasoning, logical reasoning, and case reasoning after the accident, so as to establish a more complete device operation and maintenance knowledge graph.

[0057] In the present invention, the flowchart of the device operation and maintenance decision-making is as Figure 5 shown. First, the device operation status information uploaded to the cloud is preprocessed and transmitted to the intelligent analysis unit, and its health state is evaluated through the model library inside the device digital twin; then it is judged whether the evaluation result meets the maintenance condition (whether it exceeds the threshold). When the condition is met, a maintenance decision table is formed by searching the knowledge graph module corresponding to the evaluated status and sent to the decision execution unit, while when the condition is not met, the maintenance task is ended.

[0058] In the present invention, the formation process of the maintenance decision table is as Figure 6As shown in the figure, it is mainly divided into several steps: information analysis, information search and positioning, and operation and maintenance decision generation. In the information analysis stage, it mainly judges the abnormal situation based on the operation status information of the device, and then generates abnormal text information; in the information search and positioning stage, it mainly queries the abnormal situation graph based on the abnormal text information, and then matches the operation and maintenance knowledge graph; in the operation and maintenance decision generation stage, it mainly determines the operation and maintenance case library based on the matched operation and maintenance knowledge graph through logical knowledge reasoning, and analyzes and locates the abnormal situation in this case library, and then determines the optimal operation and maintenance decision and generates a decision table.

Claims

1. An intelligent analysis and decision-making system for high-voltage electrical equipment for electric locomotives and EMUs, characterized in that: include: Equipment layer, data computing layer, intelligent analysis and decision-making layer, and control execution layer. The equipment layer should include lightning arresters, voltage transformers and cable terminals, and send real-time equipment status data to the status monitoring module; The data computing layer should include a status monitoring module, an edge computing module, a data transmission module and a cloud data processing module, which are used to calculate and process the status information of the monitoring equipment and establish various model libraries that characterize its health status; The analysis and decision-making layer is composed of intelligent analysis and decision-making modules, which are used to establish a digital twin model of the equipment, make maintenance decisions based on the operating status of the equipment, and establish and update the equipment operation and maintenance knowledge map based on historical maintenance records; The control execution layer is composed of a control execution module, which is used to receive operation and maintenance instructions sent by the edge computing module and the intelligent analysis and decision-making module, and send operation and maintenance operations to the device end.

2. According to claim 1, the intelligent analysis and decision-making system for high-voltage electrical equipment for electric locomotives and EMUs is characterized in that: The status monitoring module includes a status data acquisition unit and a status data transmission unit; the status data acquisition unit collects device status information based on edge sensing devices such as leakage current sensors and ultrasonic sensors, wherein the leakage current sensor is used to collect leakage current signals and the ultrasonic sensor is used to collect partial discharge signals; the status data transmission unit is used to upload the collected real-time status data to the edge computing module or to the cloud data processing module through the data transmission module.

3. The intelligent analysis and decision-making system for high-voltage electrical equipment for electric locomotives and EMUs according to claim 1 is characterized in that: The edge computing module includes a data processing and storage unit, a model training unit and an operation and maintenance decision unit; wherein, The data processing and storage unit is used to perform abnormal elimination, denoising and feature extraction on the device status data uploaded by the status monitoring module, and then store normal data and feature values; The model training unit is used to combine the processed normal data and feature values ​​with deep learning algorithms such as convolutional neural networks for preliminary training and calculation, to judge the current state of the equipment, to predict the future state of the equipment, to diagnose possible equipment failures and to issue early warnings; The operation and maintenance decision unit is used to make a decision on the evaluation result of the current equipment status, arrange a maintenance plan for the equipment that may fail in advance, and send a control execution module.

4. The intelligent analysis and decision-making system for high-voltage electrical equipment for electric locomotives and EMUs according to claim 1 is characterized in that: The data transmission module includes a data receiving unit and a data transmission unit, which are used to upload and send relevant data through wireless communication transmission methods such as LoRa or 5G; The data receiving unit is used to receive the device status information of the status monitoring module, the device training model of the edge computing module and the cloud data processing module; The data transmission module is used to upload the device training model of the edge computing module to the cloud, and send the updated model in the cloud to the edge computing module.

5. The intelligent analysis and decision-making system for high-voltage electrical equipment for electric locomotives and EMUs according to claim 1 is characterized in that: The cloud data processing module is used to save and update the database, and to perform enhanced training on the device model based on the updated database to form a complete device model library; it includes a data processing unit, a model enhanced training unit, and a data storage unit; The data processing unit is used to perform abnormal elimination, noise removal and feature extraction on the device status data uploaded by the data transmission module; The model reinforcement training unit, based on the model uploaded by edge computing, combines the processed normal data and feature values ​​with deep reinforcement learning algorithms such as DQN for reinforcement training and calculation, and judges the current state of the device, predicts the future state of the device, diagnoses possible faults of the device and provides early warning; The data storage unit is used to store data such as real-time original status data of the device, historical data and historical models, newly trained and updated models, etc.

6. The intelligent analysis and decision-making system for high-voltage electrical equipment for electric locomotives and EMUs according to claim 1 is characterized in that: The intelligent analysis and decision-making module establishes its digital twin model based on the equipment model library updated by the cloud data processing module training, determines the predictive maintenance strategy of the equipment to be inspected based on deep learning, and establishes and timely updates the equipment operation and maintenance knowledge graph through the predictive status maintenance history record; it includes an intelligent analysis unit, a decision execution unit and a knowledge graph construction and update unit; The intelligent analysis unit is used to analyze the real-time status characteristics of the equipment, make operation and maintenance decisions for the equipment with abnormal status by comparing its digital twin model, and propose the best maintenance strategy based on the latest operation and maintenance knowledge graph; The decision execution unit is used to send the final decision instruction to the control execution module; The knowledge graph construction and updating unit is used to present the mapping relationship between equipment status and operation and maintenance decisions in the form of a knowledge graph and update and save it in a timely manner.

7. The intelligent analysis and decision-making system for high-voltage electrical equipment for electric locomotives and EMUs according to claim 1 is characterized in that: The control execution module includes an operation and maintenance decision instruction transceiver unit and an operation and maintenance decision instruction control unit; The operation and maintenance decision instruction transceiver unit is used to receive and send operation instructions and operation and maintenance decisions of the edge computing module and the intelligent analysis and decision module in real time; The operation and maintenance command control unit is used to control the operation and maintenance start and stop operations on the device side.

Images