On-line monitoring method and device for trolley circuit breaker status

Abstract

The invention discloses an online monitoring method and device for a trolley circuit breaker state, and relates to the technical field of trolley circuit breakers. The method comprises the following steps of obtaining position information of the trolley circuit breaker, wherein the position information comprises an open circuit position, a closed circuit position and an unknown position; obtaining an open circuit position state analysis model pre-trained by a video bounding box method, and using the open circuit position state analysis model to perform state monitoring on the open circuit position; obtaining a closed circuit position state analysis model pre-trained by a simulation modeling contact force analysis, and using the closed circuit position state analysis model to perform state monitoring on the closed circuit position; obtaining an open circuit position state monitoring result and a closed circuit position state monitoring result, and issuing an alarm for an abnormal state in combination with an unknown position. The present application monitors the actual state of the trolley circuit breaker at different positions by using technologies such as deep learning, so that the monitoring is more efficient and safe.

Description

Technical Field

[0001] The present invention relates to the technical field of trolley circuit breakers, and more specifically, to an online monitoring method and device for a trolley circuit breaker state. Background Art

[0002] Different faults or abnormalities may occur in different circuit breaker position states; in the open position, faults may include the trolley failing to fully open or getting stuck, resulting in the knife switch failing to separate and the circuit failing to disconnect; in the closed position, the trolley may fail to fully close or may be abnormal when closed, resulting in poor electrical contact or inability to effectively conduct current; in an unknown position, it may be due to sensor failure, monitoring system error or misjudgment of operation, resulting in the inability to accurately determine the specific position state of the knife switch, increasing operational risks and safety hazards; therefore, timely fault diagnosis and repair are essential to ensure safe and reliable operation of circuit breakers.

[0003] For example, the invention patent with announcement number CN114089175A announces an online monitoring device for mechanical characteristics of a trolley circuit breaker. Through the device body, the body is provided with a display device, a sensor, a fixing plate and a sensor controller; the sensor includes an opening time sensor, a closing time sensor, and an operation signal generating sensor, and the opening time sensor, the closing time sensor, and the operation signal generating sensor are respectively installed on the corresponding secondary circuit line; the sensor controller is electrically connected to the sensor and the display device respectively. Compared with the prior art, the present invention can directly measure the operating equipment data, is simple and quick to operate, only needs to be installed and powered off once, and no power outage is required for subsequent monitoring, which greatly improves the reliability of the equipment.

[0004] For example, the invention patent with announcement number: CN107179424B announces a wiring method for a mechanical characteristic test of a trolley circuit breaker, by selecting a trolley circuit breaker plug-in and an aviation plug-in; the trolley circuit breaker plug-in includes a plurality of antennas, the aviation plug-in includes an outlet board, an inlet board, and a hollow column located above the antenna, a hollow column is arranged above a antenna, the inlet board is connected to the open end of the hollow column, and the inlet board is provided with an inlet port matching the opening of the hollow column; step 2: determine the number of antennas to be tested, select the corresponding connecting wire, one side of the connecting wire is a probe structure, and the probe can slide up and down in the hollow column; step 3: set the outlet ports on the outlet board with the same number as the antennas to be tested, and the outlet ports are fixedly connected to the other side of the connecting wire through a wiring piece; step 4: let the probe of the connecting wire slide down through the hollow column so that the probe contacts the antenna, and then use the on-off gear of the multimeter to test whether the connecting wire is energized.

[0005] The above disclosed technical solutions have at least the following technical problems:

[0006] The existing status monitoring is only carried out through external status monitoring or manual inspection, and the actual status of the opponent vehicle circuit breaker at different positions is not well monitored.

[0007] In view of the above problems, the present invention proposes a solution. Summary of the invention

[0008] In order to overcome the above defects of the prior art, the embodiments of the present invention provide a method and device for online monitoring of the status of a trolley circuit breaker, which solve the problems raised in the above background technology by online monitoring of the status of the trolley circuit breaker.

[0009] To achieve the above object, the present invention provides the following technical solutions:

[0010] The method and device for online monitoring of the state of a trolley circuit breaker include the following steps: obtaining position information of the trolley circuit breaker, wherein the position information includes an open circuit position, a closed circuit position and an unknown position; obtaining an open circuit position state analysis model pre-trained by a video bounding box method, and using the open circuit position state analysis model to perform state monitoring on the open circuit position; obtaining a closed circuit position state analysis model pre-trained by a simulation modeling contact force analysis, and using the closed circuit position state analysis model to perform state monitoring on the closed circuit position; obtaining an open circuit position state monitoring result and a closed circuit position state monitoring result, and issuing an alarm for an abnormal state in combination with an unknown position.

[0011] In a preferred embodiment, the specific construction method of the open circuit position state analysis model is as follows: first, the original image of the trolley circuit breaker is obtained; then, a bounding box or a frame is used to limit and post-process the area for state analysis in the detection image; then, feature characters are extracted and recognized and monitored in the limited area based on optical character recognition technology; finally, the open circuit position state of the trolley circuit breaker is judged according to the difference in feature characters through a decision algorithm.

[0012] In a preferred embodiment, the specific method of using a bounding box or frame to limit and post-process the area for state analysis in the detection image is as follows: first, the range is determined: based on the preprocessed image, four coordinates are selected in the image, and the range to be analyzed is determined as the area within the four coordinates; then the line connecting the coordinates is used as a warning line to limit the image area, and finally the limited area is monitored to eliminate image data outside the limited area.

[0013] In a preferred embodiment, the feature character extraction and recognition monitoring of a limited area based on optical character recognition technology is specifically: extracting a feature map from the limited area, and activating the feature map through a ReLU activation function; establishing a sequence of multiple feature maps, and passing them through a loop layer in sequence to capture the feature character information in the sequence; and training in combination with a preset CTC loss function to obtain a set of all feature character differences.

[0014] In a preferred embodiment, the decision algorithm is used to judge the open circuit position state of the trolley circuit breaker according to the difference in feature characters, specifically: calculating the Euclidean distance between two feature characters in a set of feature character differences; comparing the Euclidean distance with a preset state evaluation threshold, and setting the states with Euclidean distance less than the preset state evaluation threshold to the same state, otherwise to different states; determining the open circuit position state of the trolley circuit breaker according to the state of the known feature characters.

[0015] In a preferred embodiment, the closed-circuit position state analysis model trained by simulation modeling contact force analysis is specifically as follows: first, three-dimensional modeling is performed on the handcart circuit breaker based on relevant modeling software through virtual prototype technology; then the part of the model that does not affect the simulation effect is simplified; then, the material and unit part grids are established for the simplified model to obtain a simulation model; and constraints are created, and the working state of the plum blossom contact in the closed-circuit position of the handcart circuit breaker is simulated by adjusting the parameters of the constraints; finally, the closed-circuit position state analysis model is established by analyzing the contact force and deformation of the handcart circuit breaker contacts based on the convolutional neural network through the simulation model.

[0016] In a preferred embodiment, the part of the model that does not affect the simulation effect is simplified, specifically: the contacts in the same group are simplified into a contact finger unit; the grooves on the contact fingers for installing the springs are removed and replaced with a simplified straight line model; a spring unit model corresponding to the actual one is used, the connection position of the spring is selected, the stiffness coefficient and the preload are set, and the radial effect of the spring on the plum blossom contact is simulated; for adjacent springs, they are simplified into a group for equivalent loading.

[0017] In a preferred embodiment, the constraint conditions are created by adjusting the parameters of the constraint conditions to simulate the working state of the plum blossom contact in the closed position of the circuit breaker of the opponent vehicle, specifically: the working state of the plum blossom contact includes a normal state, an abnormal lateral displacement state, an abnormal longitudinal displacement state, and an abnormal angular displacement state; appropriate constraint conditions are created in the simulation software, and the constraint conditions include the position of the static contact and the contact posture and the stiffness coefficient of the spring; the contact posture includes the displacement in the Y-axis direction, the displacement in the X-axis direction, the angle around the X-axis, and the angle around the Y-axis; the parameters of each constraint condition are set and adjusted separately to obtain the change in the parameters of other constraint conditions, and the corresponding constraint condition-contact force relationship curve is established in combination with the working state of the plum blossom contact.

[0018] In a preferred embodiment, the simulation model analyzes the contact force and deformation of the trolley circuit breaker contacts based on a convolutional neural network to establish a closed-circuit position state analysis model, specifically: obtaining historical data of the contacts of the trolley circuit breaker in the closed-circuit position, and simulating the data through a model to obtain a constraint condition-contact force relationship data set; the convolutional neural network includes a first convolutional layer, a second convolutional layer, and a third convolutional layer; adding noise to the constraint condition-contact force relationship data set for data enhancement;

[0019] The enhanced constraint-contact force relationship dataset is reduced in dimensionality through the first convolutional layer;

[0020] The complexity of the enhanced constraint-contact force relationship dataset is reduced through the second convolutional layer;

[0021] The closed-loop position state analysis model is activated and constructed based on the processed constraint-contact force relationship dataset through the third convolutional layer.

[0022] The device for online monitoring the state of a trolley circuit breaker comprises a position monitoring module, a closed-circuit monitoring module, an open-circuit monitoring module and an alarm module; the position monitoring module is used to obtain the position information of the trolley circuit breaker, wherein the position information comprises an open-circuit position, a closed-circuit position and an unknown position; the open-circuit monitoring module is used to obtain an open-circuit position state analysis model pre-trained by a video bounding box method, and use the open-circuit position state analysis model to perform state monitoring on the open-circuit position; the closed-circuit monitoring module is used to obtain a closed-circuit position state analysis model pre-trained by a simulation modeling contact force analysis, and use the closed-circuit position state analysis model to perform state monitoring on the closed-circuit position; the alarm module is used to obtain an open-circuit position state monitoring result and a closed-circuit position state monitoring result, and in combination with the unknown position, issue an alarm for an abnormal state.

[0023] The technical effects and advantages of the on-line monitoring method and device for the state of a handcart circuit breaker of the present invention are as follows:

[0024] 1. The present invention trains the open circuit position state analysis model through the video bounding box method, which can effectively extract key information from the original image and realize accurate monitoring of the open circuit position state of the opponent vehicle circuit breaker; this method not only improves the accuracy and efficiency of monitoring, but also can effectively cope with complex scenes and changing operating environments, providing strong support for the safe operation and maintenance of industrial equipment.

[0025] 2. The present invention uses simulation modeling technology and convolutional neural networks to analyze the working state of the plum blossom contacts in the closed-circuit position of the trolley circuit breaker. By establishing an accurate three-dimensional model and simplifying the parts that do not affect the simulation effect, including the simplification of the contact finger unit and the replacement of the groove with a straight line model. The setting of constraint conditions simulates the behavior of the contacts in normal and abnormal working conditions by adjusting parameters. These states include normal, lateral offset, longitudinal offset and angular offset. The contact force and deformation of the contacts are analyzed using convolutional neural networks, and a closed-circuit position state analysis model is established. This technology can accurately evaluate and predict the safety and performance of circuit breakers, and provide important technical support and guarantee for the long-term reliability and safety of power equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 It is a structural schematic diagram of the on-line monitoring method of the trolley circuit breaker status of the present invention.

[0027] Figure 2 It is a structural schematic diagram of the on-line monitoring device for the status of a trolley circuit breaker of the present invention. DETAILED DESCRIPTION

[0028] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0029] Embodiment 1, Figure 1 The present invention provides a method for online monitoring of the state of a trolley circuit breaker, comprising the following steps:

[0030] S1, obtaining the position information of the trolley circuit breaker, wherein the position information includes an open position, a closed position and an unknown position;

[0031] S2, obtaining an open circuit position state analysis model pre-trained by a video bounding box method, and using the open circuit position state analysis model to perform state monitoring on the open circuit position;

[0032] S3, obtaining a closed-circuit position state analysis model that is pre-trained through simulation modeling contact force analysis, and using the closed-circuit position state analysis model to perform state monitoring on the closed-circuit position;

[0033] S4, obtaining the monitoring results of the open circuit position state and the closed circuit position state, and issuing an alarm for the abnormal state in combination with the unknown position;

[0034] S5, when the alarm of the abnormal state belongs to the abnormality of the closed-circuit position, maintenance is performed to maintain the contact force within the preset normal range, and a fixed contact force and a fixed maximum displacement are set based on the constraint condition-contact force relationship curve;

[0035] S6, setting different contact angle changes, obtaining corresponding spring stiffness influence data and component force data, and performing data analysis to obtain an angle change evaluation value;

[0036] S7, constructing an angle-effect curve according to different contact angle changes and corresponding angle change evaluation values, and performing data analysis on the angle-effect curve to obtain an optimal contact angle change, and maintaining the abnormality of the closed circuit position through the optimal contact angle change.

[0037] In this embodiment, S1, obtain the position information of the trolley circuit breaker, and the position information includes an open position, a closed position and an unknown position.

[0038] Specifically, the position information of the trolley circuit breaker refers to the indicator or display device used to display the position status of the circuit breaker switch; it is mainly used to inform the operator whether the circuit breaker is currently in the open position, closed position or unknown position; the following is a detailed introduction:

[0039] Open circuit position: The open circuit position refers to the state where the knife switch of the trolley circuit breaker is fully opened. In the open circuit position, the knife switch of the circuit breaker is in a completely separated state, the electrical contact is in a disconnected state, and the current cannot flow through the knife switch. It is displayed as "open" on a visual indicator (such as a display screen or indicator light), a mechanical indicator or a computer display interface, and an obvious position mark or indicator pointer points to the open position;

[0040] Closed circuit position: refers to the state where the circuit breaker is completely closed, the two ends of the circuit breaker are tightly connected, the electrical contacts are in contact, and current is allowed to pass through the circuit breaker. It is also displayed as "closed" on the visual indicator, mechanical indicator or computer display interface, accompanied by an obvious position mark or indicator pointer pointing to the closed position;

[0041] Unknown position: The unknown position means that the computer vision system cannot accurately determine the current position status of the circuit breaker. This may be because the visual recognition algorithm cannot clearly determine the position of the knife switch, or insufficient system data makes it impossible to determine the exact position status of the knife switch. The unknown position will cause the operator to be uncertain about the circuit status, which may increase operational risks and safety hazards.

[0042] Different faults or abnormalities may occur in different circuit breaker position states; in the open position, faults may include the trolley failing to fully open or getting stuck, resulting in the knife switch failing to separate and the circuit failing to disconnect; in the closed position, the trolley may fail to fully close or may be abnormal when closed, resulting in poor electrical contact or inability to effectively conduct current; in an unknown position, it may be due to sensor failure, monitoring system error or misjudgment of operation, resulting in the inability to accurately determine the specific position state of the knife switch, increasing operational risks and safety hazards; therefore, timely fault diagnosis and repair are essential to ensure safe and reliable operation of circuit breakers.

[0043] A trolley circuit breaker in an open circuit state may encounter a variety of faults. For example, the trolley cannot be fully opened or is stuck in the middle position, causing the knife switch to be unable to be completely separated; mechanical parts such as transmission chains or gears are damaged or broken, affecting the normal movement of the trolley; motor failure causes the trolley operation to be blocked or the knife switch cannot be fully opened; sensors or circuits in the electrical control system fail and cannot accurately detect the knife switch position or provide correct operating signals; environmental factors such as high temperature or humidity may cause the mechanical parts of the circuit breaker to operate unstably or rust, thereby affecting the operating performance of the trolley. These problems may cause the trolley circuit breaker to fail to cut off the circuit normally or fail to operate reliably, increasing the complexity and risk of equipment operation and maintenance;

[0044] The types of faults that may occur in a trolley circuit breaker in a closed circuit state are different from those in an open circuit state. Common problems include the trolley encountering resistance or getting stuck when closing, causing the knife switch to not be completely and tightly closed; electrical contact parts such as electrical contacts corroding or oxidizing, causing unstable current conduction; mechanical transmission systems such as chains or gears being damaged or broken, affecting the trolley's closing operation; motor or operating mechanism failures making it impossible for the trolley to drive the knife switch to close smoothly; and environmental conditions such as high temperature or humidity causing performance degradation of electrical or mechanical components; these problems may affect the normal operation of the trolley circuit breaker in a closed circuit state, requiring timely maintenance and repair to ensure equipment safety and power system reliability;

[0045] There are many types of failures that may occur in the trolley circuit breaker in an unknown or abnormal position. For example, the system may not be able to accurately display or judge the actual position status of the trolley due to sensor failure or monitoring system error, resulting in the operator being unable to accurately understand the opening and closing status of the circuit breaker; or misjudgment of operation may cause the trolley to be in an unsafe position, increasing the risk of circuit operation; in addition, the performance of electrical or mechanical components may deteriorate due to environmental conditions such as high temperature or humidity, affecting the normal operation of the trolley; these problems need to be diagnosed and repaired in a timely manner to ensure that the trolley circuit breaker can reliably cut off the circuit or restore circuit conduction when necessary, ensuring the stable operation and safety of the power system.

[0046] Since the types of faults that may occur in the trolley circuit breaker in open circuit, closed circuit and unknown positions vary greatly, different state assessment models are needed for online state monitoring; for the open circuit state, the monitoring focus may be on whether the trolley can be fully opened and the normal operation of the mechanical transmission system; for the closed circuit state, the focus is on ensuring that the trolley can be fully closed and the electrical contact components are reliable; and for the unknown position, the monitoring model needs to focus on the reliability of the detection sensor and control system to ensure that the actual position state of the circuit breaker can be accurately judged; this monitoring method for different position states can effectively improve the comprehensive evaluation of the operating status of the trolley circuit breaker, thereby ensuring the safe operation of the power system and the long-term reliability of the equipment.

[0047] In this embodiment, S2, obtain an open circuit position state analysis model pre-trained by a video bounding box method, and use the open circuit position state analysis model to perform state monitoring on the open circuit position.

[0048] In this embodiment, the specific construction method of the open circuit position state analysis model is as follows:

[0049] First, the original image of the trolley circuit breaker is obtained, and then the area for state analysis in the detection image is limited and post-processed using a bounding box or frame. Next, feature characters are extracted and recognized and monitored in the limited area based on optical character recognition technology. Finally, the open circuit position state of the trolley circuit breaker is judged according to the difference in feature characters through a decision algorithm.

[0050] The video bounding box method is a technology used for video analysis and target detection. It helps computer vision systems automatically identify and track specific objects or targets by drawing bounding boxes in video frames to represent targets or areas of interest. This method usually involves the following steps: first, obtain consecutive image frames from the video stream; second, apply object detection algorithms (such as deep learning-based object detectors) to identify the target objects in each frame; then, draw rectangular or polygonal bounding boxes around the identified targets to indicate the location and boundaries of the targets; finally, by tracking the movement and changes of the bounding boxes between consecutive frames, real-time tracking and dynamic analysis of the targets are achieved. The video bounding box method is widely used in video surveillance, autonomous driving, behavior analysis and other fields, providing an effective solution for real-time target monitoring and tracking, and promoting the development and application of intelligent vision systems.

[0051] Optical character recognition (OCR) is a technology that can automatically extract text information from images or scanned documents. It recognizes words, numbers or symbols in images and converts them into editable text formats, thereby realizing automatic processing and information extraction of digital documents. OCR technology usually includes the following steps: first, image preprocessing, including adjusting image contrast, removing noise and smoothing to improve recognition accuracy; then text area detection and segmentation to determine the area in the image containing text; then feature extraction and classification to identify the shape, size and position of characters; finally, character recognition and output to convert the recognized characters into text data that can be processed by computers. OCR technology is widely used in various fields, including document management, automated data entry, identity authentication systems, license plate recognition, etc., which greatly improves the efficiency and accuracy of information processing, and provides important support for the digital transformation and intelligent applications of modern society.

[0052] The specific method for obtaining the original image of the trolley circuit breaker is as follows:

[0053] Image acquisition: Use a dedicated surveillance camera or camera to acquire images of the trolley circuit breaker; these devices are usually installed in a fixed position to capture different positions and states of the trolley;

[0054] Image transmission and storage: The collected images are transmitted to the processing unit or storage device through the network. The stability and real-time performance of image transmission are guaranteed so that subsequent processing and analysis can be carried out in a timely manner;

[0055] Image preprocessing: Before performing status analysis, it is necessary to preprocess the acquired original image; the purpose of preprocessing includes eliminating noise and interference in the image to improve the accuracy of subsequent processing; adjusting the contrast, brightness, etc. of the image to optimize the image quality; correcting the distortion or distortion in the image to ensure the accurate shape and position of the trolley circuit breaker.

[0056] The specific method of using a bounding box or a framework to limit and post-process the area for state analysis in the detection image is as follows:

[0057] First, the range is determined: based on the preprocessed image, four coordinates are selected in the image, and the range to be analyzed is determined as the area within the four coordinates; then the line connecting the coordinates is used as a warning line to limit the image area, and finally the limited area is monitored and the image data outside the limited area is eliminated.

[0058] Among them, feature character extraction and recognition monitoring are performed on the limited area based on optical character recognition technology, specifically:

[0059] Extract feature maps from limited areas and activate them through ReLU activation function; create a sequence of multiple feature maps and pass them through the recurrent layer in turn to capture the feature character information in the sequence, and train them with the preset CTC loss function to obtain a set of all feature character differences;

[0060] The specific calculation formula of the feature map is as follows:

[0061]

[0062] In the formula, is the value of the fth channel of the output feature map TX at position (i, j); is the pixel value of the trolley circuit breaker image at the coordinate (i+k, j+l); is the weight, is the bias term, and there is one bias for each output channel f.

[0063] The ReLU activation function is specifically:

[0064]

[0065] The specific calculation formula of the CTC loss function is:

[0066]

[0067] In the formula, is the set of characteristic character differences, For all possible different paths of L, is the total number of sequences, t is the sequence number, y is the activated feature map, is the activated feature map of the t sequence number, z is the known feature character, is the probability that y and z are different.

[0068] Among them, the decision algorithm is used to judge the open circuit position state of the opponent vehicle circuit breaker according to the difference in characteristic characters, specifically:

[0069] Calculate the Euclidean distance between two feature characters in the set of feature character differences, compare the Euclidean distance with a preset state evaluation threshold, set the states whose Euclidean distance is less than the preset state evaluation threshold to the same state, otherwise to different states, and determine the open position state of the trolley circuit breaker based on the state of the known feature characters.

[0070] It should be noted that ReLU is a nonlinear activation function that allows neural networks to learn nonlinear relationships. The output of ReLU is linear in the positive interval, which makes the network more stable and reduces the possibility of gradient explosion. During backpropagation, the ReLU function sets the gradient of negative numbers to zero, which helps reduce the gradient vanishing problem and accelerates the convergence process. The calculation of the ReLU function is simple and fast, and only requires comparing whether the input is greater than zero.

[0071] It should be noted that the original image is the actual data obtained from the field, which contains the real status information of the trolley circuit breaker; by using the original image, model training and monitoring can be carried out based on actual scenarios to improve the practical applicability and accuracy of the model.

[0072] It should be noted that the bounding box or frame can help accurately locate and segment the trolley circuit breaker in the image; limitation and post-processing can effectively reduce the interference of irrelevant information, focus on the area where the trolley circuit breaker is located, and improve the efficiency and accuracy of subsequent feature extraction and recognition; it helps to improve the efficiency of subsequent processing and reduce the waste of computing resources; through precise area limitation, it can ensure that the subsequent steps can carry out targeted feature extraction and analysis to avoid interference from invalid data.

[0073] It should be noted that optical character recognition technology can effectively extract text information from images, including identifiers on trolley circuit breakers; the extraction and recognition of characteristic characters can provide necessary data support for subsequent status judgment; the benefits are that key information can be automatically extracted from images, reducing the need for manual intervention and improving the real-time and efficiency of monitoring; based on optical character recognition technology, the status identifiers on trolley circuit breakers can be accurately obtained, providing reliable data basis for subsequent decision-making.

[0074] The decision algorithm can judge the specific status of the trolley circuit breaker (such as the open circuit position) based on the difference or similarity of the characteristic characters; based on the difference judgment, it can realize automatic status monitoring and alarm, and timely discover and handle abnormal situations; it helps to improve the accuracy and real-time performance of monitoring, so that operators can respond to problems quickly; the automated decision-making process reduces the subjectivity and error rate of human judgment, and enhances the stability and reliability of the system.

[0075] Combining the above steps, the open circuit position state analysis model is trained through the video bounding box method, which can effectively extract key information from the original image and realize accurate monitoring of the open circuit position state of the opponent vehicle circuit breaker. This method not only improves the accuracy and efficiency of monitoring, but also can effectively cope with complex scenarios and changing operating environments, providing strong support for the safe operation and maintenance of industrial equipment.

[0076] This embodiment uses image processing technology and state analysis models to monitor the open circuit position state of the trolley circuit breaker. The circuit breaker is accurately located and segmented by the video bounding box method, and the characteristic characters are extracted by combining the optical character recognition technology. Then, the state is judged by the decision algorithm, thereby realizing automatic state monitoring and alarm. This technology not only improves the accuracy and efficiency of monitoring, but also effectively copes with complex environments and operating conditions, providing strong support for ensuring the safe operation of industrial equipment.

[0077] In embodiment 2, a closed-circuit position state analysis model pre-trained by simulation modeling contact force analysis is obtained, and the closed-circuit position state analysis model is used to perform state monitoring on the closed-circuit position.

[0078] It should be noted that the hidden dangers generated in the closed circuit position are usually caused by the plum blossom contacts of the trolley circuit breaker trolley, so it is particularly important to monitor the status of the plum blossom contacts.

[0079] The plum blossom contact is a key electrical contact component in the trolley circuit breaker. Its status directly affects the closed-circuit operation and current conduction of the circuit breaker. If the plum blossom contact has problems or is in a bad state, it may cause the following problems:

[0080] Poor electrical contact: If the plum blossom contact is corroded, oxidized or physically damaged, it will cause poor electrical contact with the knife switch; this will prevent the current from passing stably when the knife switch is closed, causing unreliable or incomplete conduction of the circuit, thus affecting the normal power supply of the equipment;

[0081] Unstable conduction: If the surface of the plum blossom contact is rough or contains impurities, it may cause unstable current conduction; in this case, the current may be intermittently interrupted or cause arcing in the circuit breaker when it is closed, increasing the risk of circuit operation and wear of the equipment;

[0082] Reduced operational reliability: Damage or poor condition of the plum blossom contacts will affect the normal operation of the trolley circuit breaker; for example, the trolley may become stuck or difficult to operate when closing, or even fail to fully close, resulting in the circuit breaker being unable to effectively cut off the circuit or provide a reliable power connection;

[0083] Increased safety hazards: Poor plum blossom contact status may cause arcing or overheating, further leading to equipment damage or even serious accidents such as fire; this poses a direct threat to the safety of industrial equipment and power systems, and requires timely detection and repair to prevent potential catastrophic consequences;

[0084] Therefore, regular monitoring and maintenance of the status of the plum blossom contacts is essential to ensure the safe operation of the trolley circuit breaker; through an effective status monitoring and fault warning system, possible problems with the plum blossom contacts can be discovered and resolved in a timely manner, thereby ensuring the long-term reliability and safety of the equipment.

[0085] In this embodiment, the closed-circuit position state analysis model trained by simulation modeling contact force analysis is specifically:

[0086] Firstly, the virtual prototype technology is used to conduct 3D modeling with the opponent’s circuit breaker based on relevant modeling software;

[0087] Then simplify the parts of the model that do not affect the simulation results;

[0088] Then, the meshes of materials and unit parts are established for the simplified model to obtain the simulation model;

[0089] And create constraint conditions, and simulate the working state of the plum blossom contact in the closed position of the circuit breaker of the hand car by adjusting the parameters of the constraint conditions;

[0090] Finally, the contact force and deformation of the trolley circuit breaker contacts are analyzed through the simulation model based on the convolutional neural network to establish a closed-circuit position state analysis model.

[0091] It should be noted that the virtual prototype modeling of the circuit breaker trolley is a process of three-dimensional modeling and analysis of the trolley circuit breaker in a digital environment through computer-aided design and simulation technology; the main advantage of this method is that it can accurately simulate and evaluate the performance of the circuit breaker under various working conditions, while avoiding the high cost and time consumption that may be caused by traditional experimental methods;

[0092] First, virtual prototype modeling uses modern modeling software such as CAD or professional simulation software to accurately capture the geometric shape and structural features of the trolley circuit breaker, including key components such as plum blossom contacts. The step of simplifying the model is to ensure the efficiency and accuracy of the simulation, usually by removing details that have little effect on the simulation results, such as surface texture or non-critical components;

[0093] Secondly, the mesh of materials and unit parts is established to accurately map the physical characteristics and material properties into the simulation model for mechanical and thermal analysis; this includes determining the elastic modulus, thermal conductivity and other parameters of the material and applying them to various parts of the model to ensure the authenticity and reliability of the simulation results;

[0094] One of the biggest benefits of virtual prototype modeling is the ability to simulate the performance of the trolley circuit breaker under a variety of operating conditions, including normal operation, abnormal conditions, and emergencies. By adjusting constraints and simulation parameters, the operating state of the circuit breaker in the closed position, such as the contact force and deformation of the contacts, can be simulated and predicted, which is crucial for evaluating the safety, stability, and reliability of the equipment.

[0095] Finally, convolutional neural network-based analysis further improves the accuracy and complexity of the simulation model; this technology can effectively process large amounts of simulation data, extract and analyze complex mechanical features, and help engineers and researchers better understand and optimize the design and operation of circuit breaker carts;

[0096] Therefore, through the virtual prototype modeling of the circuit breaker cart, the efficiency and accuracy of the engineering design process can be significantly improved, while reducing the development cycle and cost, providing important technical support and guarantee for safe and reliable power equipment.

[0097] It should be noted that convolutional neural networks play an important role in plum blossom contact analysis. They are a deep learning model specifically used to process and analyze complex data such as images and sequence data. Their application in the mechanical behavior and contact force analysis of plum blossom contacts can help us deeply understand the working status and performance of circuit breakers.

[0098] First, the convolutional neural network can effectively capture the mechanical characteristics of the plum blossom contact under different working conditions by learning the patterns and features in a large amount of simulation data; for example, it can analyze the contact surface shape, contact force size and change trend of the contact under different environmental conditions. This information is of great significance for predicting contact life, optimizing design and preventing failures.

[0099] Secondly, through the combination of multiple convolution and pooling layers, the convolutional neural network can automatically extract and learn key features in the data, such as nonlinear patterns in the force-displacement curve or mutations under abnormal conditions. These features are crucial for building a closed-circuit position state analysis model, which can help engineers more accurately evaluate and predict various situations that circuit breakers may encounter in actual operation;

[0100] Finally, the analysis based on convolutional neural networks is not limited to mechanical properties, but can also be combined with other sensor data or experimental results to further improve the accuracy and universality of the model. By training and optimizing the convolutional neural network model, a comprehensive analysis of circuit breaker trolleys under complex working conditions can be achieved, providing scientific basis and technical support for the safety and reliability of the equipment.

[0101] Therefore, the plum blossom contact analysis based on convolutional neural networks not only expands the limitations of traditional simulation methods, but also plays an important role in understanding and optimizing circuit breaker design, providing new ideas and solutions for the development and application of power equipment.

[0102] In this embodiment, the part of the model that does not affect the simulation effect is simplified, specifically:

[0103] Simplify the touch pieces in the same group into one touch finger unit;

[0104] The groove on the contact finger for installing the spring is removed and replaced by a simplified straight line model;

[0105] Use the spring unit model corresponding to the actual situation, select the connection position of the spring, set the stiffness coefficient and preload force, and simulate the radial effect of the spring on the plum blossom contact;

[0106] For adjacent springs, simplify them into a group for equivalent loading;

[0107] It should be noted that the contacts in the same group are simplified into a contact finger unit; this can reduce the number of contact fingers in the model and reduce the computational complexity without affecting the model's accurate description of the overall dynamics of the contact;

[0108] It should be noted that there are usually grooves on the stylus for installing springs. In the modeling process, it is no longer necessary to reproduce the details of each groove in detail, but to use a simplified straight line model instead; this method can avoid complex meshing and reduce the details in the model, thereby improving simulation efficiency.

[0109] It should be noted that a suitable spring unit model is used, such as the Comb14 spring unit provided in Workbench; by selecting the connection position of the spring, setting the appropriate stiffness coefficient and preload, the radial effect of the spring on the plum contact is simulated; this non-solid unit can effectively simplify the model and ensure that the simulation results are consistent with the actual working conditions.

[0110] It should be noted that adjacent springs can be simplified into a group for equivalent loading; this can reduce the number of spring units in the model while maintaining an accurate description of the mechanical properties of the contact.

[0111] Through the above simplification steps, the complexity of the plum blossom contact model can be effectively reduced and the efficiency and accuracy of the simulation can be improved; these simplification methods are not only applicable to specific examples, but can also be widely used in the modeling and simulation of various types of plum blossom contacts.

[0112] In this embodiment, a constraint condition is created, and the working state of the plum blossom contact in the closed position of the circuit breaker of the hand vehicle is simulated by adjusting the parameters of the constraint condition, specifically:

[0113] The plum blossom contact working state includes a normal state, a lateral deviation abnormal state, a longitudinal deviation abnormal state, and an angular deviation abnormal state;

[0114] Creating appropriate constraint conditions in the simulation software, wherein the constraint conditions include the position and contact posture of the stationary contact and the stiffness coefficient of the spring; the contact posture includes the displacement in the Y-axis direction, the displacement in the X-axis direction, the angle around the X-axis and the angle around the Y-axis;

[0115] The parameters of each constraint condition are set and adjusted individually to obtain the variation of other constraint condition parameters, and the corresponding constraint condition-contact force relationship curve is established in combination with the working state of the plum blossom contact.

[0116] The following is a specific calculation formula for the feasible constraint relationship curve:

[0117]

[0118] In the formula, is the contact force of the plum blossom contact, is the constraint parameter of contact displacement, is the change in the spring stiffness coefficient, is the angle change of the contact, , is the working status adjustment factor, and d is the correction term.

[0119] In this embodiment, the simulation model analyzes the contact force and deformation of the trolley circuit breaker contacts based on the convolutional neural network to establish a closed circuit position state analysis model, specifically:

[0120] Obtain the historical data of contacts in the closed position of the trolley circuit breaker, and obtain the constraint condition-contact force relationship data set through model simulation;

[0121] The convolutional neural network includes a first convolutional layer, a second convolutional layer, and a third convolutional layer;

[0122] Add noise to the constraint-contact force relationship dataset for data enhancement;

[0123] The enhanced constraint-contact force relationship dataset is reduced in dimensionality through the first convolutional layer;

[0124] The complexity of the enhanced constraint-contact force relationship dataset is reduced through the second convolutional layer;

[0125] The closed-loop position state analysis model is activated and constructed based on the processed constraint condition-contact force relationship dataset through the third convolutional layer;

[0126] The specific calculation formula for the data enhancement is:

[0127]

[0128] In the formula, The enhanced constraint condition-contact force relationship data; is the i-th constraint condition-contact force relationship data in the constraint condition-contact force relationship data set, is the noise factor.

[0129] It should be noted that Gaussian noise is added to the overall sample to enhance the sample quantity and robustness.

[0130] The specific calculation formula of the first convolutional layer is:

[0131]

[0132] In the formula, is the i-th constraint condition-contact force relationship data of the h+1-th layer, is the bias and weight term connecting the h layer and the h+1 layer, The constraint condition-contact force relationship data for training, For correction items.

[0133] This embodiment uses simulation modeling technology and convolutional neural networks to analyze the working state of the plum blossom contact in the closed-circuit position of the trolley circuit breaker. By establishing an accurate three-dimensional model and simplifying the parts that do not affect the simulation effect, including the simplification of the contact finger unit and the replacement of the groove with a straight line model. The setting of constraints simulates the behavior of the contact in normal and abnormal working conditions by adjusting parameters. These states include normal, lateral offset, longitudinal offset and angular offset. The contact force and deformation of the contact are analyzed using a convolutional neural network, and a closed-circuit position state analysis model is established. This technology can accurately evaluate and predict the safety and performance of the circuit breaker, and provide important technical support and guarantee for the long-term reliability and safety of power equipment.

[0134] Embodiment 3, S4, obtains the open circuit position state monitoring result and the closed circuit position state monitoring result, and in combination with the unknown position, issues an alarm for the abnormal state.

[0135] Specifically, when the open circuit position status monitor detects that the trolley circuit breaker is in a pushed-in state, it is determined to be an abnormal state;

[0136] When the closed-circuit position status monitoring finds that the trolley circuit breaker contact is in an abnormal state of lateral displacement, abnormal state of longitudinal displacement or abnormal state of angular displacement, it is determined to be an abnormal state;

[0137] When the position information of the trolley circuit breaker is unknown, it is determined to be an abnormal state;

[0138] Issue alerts for abnormal conditions.

[0139] S5, when the alarm of the abnormal state belongs to the abnormality of the closed-circuit position, maintenance is performed to maintain the contact force within the preset normal range, and a fixed contact force and a fixed maximum displacement are set based on the constraint condition-contact force relationship curve.

[0140] It should be noted that when the system issues an alarm for abnormal closed-circuit position, maintenance is required to restore normal status; at this time, it is necessary to ensure that the contact force remains within the set normal range, and at the same time, set a fixed contact force and maximum displacement based on the constraint condition and contact force relationship curve to ensure that the equipment operates stably within the normal operating range; this process involves precise adjustments to ensure that the equipment can still maintain optimal performance under different working conditions.

[0141] S6, setting different contact angle changes, obtaining corresponding spring stiffness influence data and component force data, and performing data analysis to obtain an angle change evaluation value;

[0142] It should be noted that the change in contact angle has the following effects on the stiffness of the spring:

[0143] Too high contact angle variation leads to mismatched stiffness: When the contact angle variation is too high, the contact angle between the contact and the spring increases; this may cause the stiffness of the spring to no longer align with the actual position of the contact, making it impossible for the spring to effectively provide the required support force; this mismatch may cause the circuit breaker to be unable to maintain a stable contact force during the switching operation, affecting its opening and closing performance;

[0144] Poor contact: Excessive angle variation may cause uneven contact between the contact and the spring. The contact may generate extra friction or irregular contact surface at the contact point, resulting in poor electrical contact or increased contact resistance; this will affect the conductivity and reliability of the circuit breaker and may cause overheating or failure;

[0145] Deterioration of mechanical properties: Excessive angle changes will change the force distribution of the spring, subjecting it to additional stress; the nonlinear stress distribution of the spring may cause its stiffness characteristics to decrease, thereby reducing the stability and operating accuracy of the circuit breaker. In the long run, this may lead to fatigue and shorten the life of the spring,

[0146] The spring rate influence data includes a spring rate influence coefficient.

[0147] The spring stiffness influence coefficient is an important parameter used to evaluate the impact of different contact angle changes on the performance of the opponent vehicle circuit breaker. It is used to reflect the specific impact of excessive contact angle changes on the spring stiffness. This coefficient helps analyze and optimize the performance of the circuit breaker in actual applications by quantifying the change in spring stiffness caused by the angle change. A high spring stiffness influence coefficient indicates that the angle change has a greater negative impact on the spring performance, which may cause poor contact or unstable operation, while a lower coefficient indicates that this impact is smaller, and the spring can still maintain good stiffness characteristics under a larger angle change. Through this coefficient, engineers can optimize the contact design and adjust the spring parameters to ensure the stability and reliability of the circuit breaker.

[0148] Analyzing the influence coefficient of spring stiffness has the following advantages for evaluating the application effect of contact angle change on opponent vehicle circuit breakers:

[0149] Accurately evaluate performance changes: The spring stiffness influence coefficient can quantify the specific impact of the contact angle change on the spring stiffness, providing an accurate evaluation of the circuit breaker performance change; this helps to understand how the spring stiffness under different angle changes affects the switching operation and stability of the circuit breaker, so as to optimize the design and adjust the parameters to ensure reliable operation;

[0150] Optimize design parameters: By analyzing the spring stiffness influence coefficient, it is possible to identify which angle changes have a significant impact on the spring stiffness, thereby guiding designers to take measures in the design of springs and contacts to optimize design parameters and ensure that the circuit breaker can maintain good performance under various working conditions;

[0151] Improve reliability and stability: Understanding and applying the spring stiffness influence coefficient can effectively avoid performance problems caused by excessive contact angle variation, such as poor contact or spring fatigue; by optimizing the contact angle variation, ensuring that the spring stiffness during operation remains within an appropriate range, thereby improving the reliability and stability of the circuit breaker;

[0152] Fault prediction and prevention: The spring stiffness influence coefficient can help predict possible faults caused by improper contact angle changes and take preventive measures in advance; this forward-looking analysis can reduce the failure rate of circuit breakers, reduce maintenance costs, and improve the overall safety of the equipment;

[0153] Data-driven decision support: This coefficient provides engineers with a data-driven tool for evaluating and optimizing the design and operating parameters of circuit breakers. By systematically analyzing the spring stiffness influence coefficient, the contact angle change can be scientifically adjusted to improve the performance and application effect of the circuit breaker.

[0154] The specific method for obtaining the spring stiffness influence coefficient is as follows:

[0155] Obtain spring stiffness data at different contact angle changes;

[0156] According to the spring stiffness data, the influence rate of the contact angle change on the spring stiffness is calculated based on the finite difference method;

[0157] The influence rate of the contact angle change on the spring stiffness is used as the gradient direction of the preset optimization objective function to optimize the spring stiffness influence coefficient;

[0158] The specific calculation formula for the influence rate of the contact angle change on the spring stiffness is as follows:

[0159]

[0160] The specific calculation formula of the optimization objective function is as follows:

[0161]

[0162] In the formula, k ( ) is the angle change The spring stiffness under is the minimum value of the angle change, CY is the difference between the actual stiffness value and the predicted stiffness value during the optimization process, The actual stiffness value, is the base stiffness; is the spring stiffness influence coefficient to be optimized, x is the number of the actual stiffness value, and z is the total number of the actual stiffness values.

[0163] It should be noted that the angle change of the contact has an important influence on the force of the component; specifically, too small an angle change will lead to the following situations:

[0164] Poor contact: Too small a change in the contact angle may result in poor contact between the contacts. In this case, the contact may not be able to fully contact the contact, resulting in increased contact resistance, which affects the normal operation of the circuit breaker; poor contact can lead to unstable current transmission, increased heat generation, and may even cause electrical failures;

[0165] Uneven contact force: Too little angular variation may result in uneven pressure distribution of the contacts at the contact points; if the contacts cannot move within the designed angular range, the contact force will be concentrated at certain points, which may cause local wear, premature failure or accelerated component aging;

[0166] Increased mechanical friction: Too little change in contact angle may lead to increased friction during contact operation; this friction will cause additional mechanical stress on the components, which will affect the long-term stability and durability of the components;

[0167] Unstable switch action: If the contact angle change is too small, the switch action may be unstable; when the circuit breaker is working, the contact may not move to the predetermined position correctly, which may cause the switch to act slowly or fail to close or open completely, thus affecting the reliability and safety of the equipment;

[0168] In short, the proper setting of the contact angle variation is crucial to ensure the normal operation and service life of the circuit breaker. Too small an angle variation can lead to poor contact, uneven contact force, inappropriate spring stiffness, increased mechanical friction, and unstable switching action, thus affecting the performance and reliability of the circuit breaker; in the design and maintenance process, the contact angle variation needs to be fully considered to optimize its performance.

[0169] The component force data includes a component force uniformity coefficient.

[0170] The component force uniformity coefficient is an indicator used to quantify the abnormal force exerted on the component under different contact angle changes. It measures and analyzes the force exerted on the component to evaluate whether the component is subjected to force beyond the normal working range under specific angle change conditions. This coefficient can reflect the impact of contact angle change on component force, help identify and optimize potential problems in the design, and ensure that the force exerted on the component in actual operation remains within a safe and effective range, thereby improving the reliability and stability of the system.

[0171] Analyzing the uniformity coefficient of component force has the following advantages for selecting the optimal contact angle variation:

[0172] Improve system stability: By evaluating the force uniformity coefficient of components under different angle changes, the optimal contact angle can be identified, thereby reducing uneven force and mechanical friction, ensuring the stability of the system during operation;

[0173] Extend equipment life: Optimizing the contact angle change can reduce abnormal component stress, excessive wear and mechanical fatigue, thereby extending the service life of the equipment and reducing maintenance and replacement frequency;

[0174] Improve operational reliability: Selecting the optimal contact angle variation can ensure good contact between the contacts and the contact points, reduce poor contact and unstable current transmission, and thus improve the operational reliability and safety of the equipment;

[0175] Optimize energy efficiency: By optimizing the contact angle change, the contact resistance and energy loss can be reduced, the energy efficiency and overall performance of the equipment can be improved, and energy consumption and operating costs can be reduced;

[0176] Reduce failure rate: Analyzing the uniformity coefficient of component stress helps identify potential design defects and operational problems, reducing failures caused by abnormal stress, thereby improving the overall reliability and safety of the system;

[0177] Improve maintenance management: By optimizing the change in contact angle, the stress condition of components can be better controlled, which facilitates the formulation of scientific maintenance plans and improves maintenance efficiency and accuracy.

[0178] The specific method for obtaining the force uniformity coefficient of the component is as follows:

[0179] A plurality of measuring points are set on the plum blossom contact, and the forces at the plurality of measuring points under standard conditions are obtained respectively;

[0180] Obtain the actual forces at several measuring points under the corresponding contact variation, calculate the difference between the forces under the corresponding standard conditions, and construct a difference set;

[0181] Calculate the standard deviation and average value within the difference set, and use the ratio of the standard deviation to the average value as the component force uniformity coefficient.

[0182] The method for obtaining the angle variation evaluation value is as follows:

[0183] The component force uniformity coefficient and spring stiffness influence coefficient are used based on the BP neural network algorithm to construct an angle change evaluation model and generate an angle change evaluation value.

[0184] The specific calculation formula of the angle change evaluation value is as follows:

[0185]

[0186] In the formula, is the evaluation value of the angle change, is the spring stiffness influence coefficient, is the component force uniformity coefficient, is the weight coefficient obtained from the training of the spring stiffness influence coefficient, It is the weight coefficient obtained from the training of component force uniformity coefficient.

[0187] S7, constructing an angle-effect curve according to different contact angle changes and corresponding angle change evaluation values, and performing data analysis on the angle-effect curve to obtain an optimal contact angle change, and maintaining the abnormality of the closed circuit position through the optimal contact angle change.

[0188] The specific method of constructing the angle-effect curve is as follows:

[0189] Obtaining output angle variation evaluation values ​​under different contact angle variations from the angle variation evaluation model;

[0190] A rectangular coordinate system is established, with the horizontal axis representing the contact angle change and the vertical axis representing the angle change evaluation value;

[0191] Mark the angle change evaluation value corresponding to each contact angle change on the rectangular coordinate system to form a data point;

[0192] Connect data points to form an angle-effect curve;

[0193] The angle-effect curve is analyzed to obtain the optimal contact angle change, which is specifically:

[0194] Obtain the peak values ​​of several angle-effect curves, perform secondary derivative on each peak value, and obtain the peak derivative;

[0195] The contact angle variation corresponding to the peak value whose peak derivative is closest to zero is taken as the optimal contact angle variation, and is used as the contact angle variation for maintaining the abnormality of the closed circuit position.

[0196] Embodiment 4, Figure 2 Provided is a trolley circuit breaker state online monitoring device, comprising a position monitoring module, a closed-circuit monitoring module, an open-circuit monitoring module, an alarm module, a feedback module, a feedback evaluation module, and a maintenance module.

[0197] A position monitoring module is used to obtain the position information of the trolley circuit breaker, wherein the position information includes an open circuit position, a closed circuit position and an unknown position;

[0198] An open circuit monitoring module, used to obtain an open circuit position state analysis model pre-trained by a video bounding box method, and use the open circuit position state analysis model to perform state monitoring on the open circuit position;

[0199] Closed-circuit monitoring module, used to analyze the contact force and deformation of the trolley circuit breaker contacts based on convolutional neural network, construct constraint condition-contact force relationship curve, and establish a closed-circuit position state analysis model;

[0200] The alarm module is used to obtain the open circuit position status monitoring results and the closed circuit position status monitoring results, and in combination with the unknown position, issue an alarm for abnormal status.

[0201] The feedback module is used to perform maintenance when the alarm of abnormal state belongs to the abnormality of closed-circuit position, maintain the contact force within the preset normal range, and set the fixed contact force and the fixed maximum displacement based on the constraint condition-contact force relationship curve;

[0202] A feedback evaluation module is used to set different contact angle changes, obtain corresponding spring stiffness impact data and component force data, and perform data analysis to obtain an angle change evaluation value;

[0203] The maintenance module is used to construct an angle-effect curve according to different contact angle changes and corresponding angle change evaluation values, and perform data analysis on the angle-effect curve to obtain the optimal contact angle change, and maintain the abnormality of the closed circuit position through the optimal contact angle change.

[0204] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0205] The above embodiments may be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, the above embodiments may be implemented in whole or in part in the form of a computer program product.

[0206] Those of ordinary skill in the art will appreciate that the modules and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of 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. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0207] In addition, each functional module in each embodiment of the present application may be integrated into one processing module, or each module may exist physically separately, or two or more modules may be integrated into one module.

[0208] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0209] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. The method for online monitoring of the state of a trolley circuit breaker is characterized in that: The steps include: Acquire an open circuit position state analysis model, wherein the open circuit position state analysis model is obtained by video bounding box method training, and use the open circuit position state analysis model to perform state monitoring on the open circuit position to obtain an open circuit position state monitoring result; The contact force and corresponding deformation of the trolley circuit breaker contacts were analyzed, the constraint condition-contact force relationship curve was constructed, and a closed-circuit position state analysis model was established based on the convolutional neural network to obtain the closed-circuit position state monitoring results; Obtain the open circuit position status monitoring results and the closed circuit position status monitoring results, and combine with the unknown position to issue an alarm for abnormal status; When the alarm of abnormal state belongs to the abnormality of closed-circuit position, maintenance is performed to maintain the contact force within the preset normal range, and a fixed contact force and a fixed maximum displacement are set based on the constraint condition-contact force relationship curve; Set different contact angle changes, obtain corresponding spring stiffness impact data and component force data, and perform data analysis to obtain angle change evaluation values; An angle-effect curve is constructed based on different contact angle changes and corresponding angle change evaluation values, and data analysis is performed on the angle-effect curve to obtain the optimal contact angle change. The abnormality of the closed-circuit position is maintained through the optimal contact angle change; The specific construction method of the open circuit position state analysis model is as follows: Get the original image of the trolley circuit breaker; Using bounding boxes or frames to define and post-process the region for state analysis in the inspection image; Extract characteristic characters and identify and monitor the limited area based on optical character recognition technology; Finally, the decision algorithm is used to judge the open circuit position status of the opponent's circuit breaker according to the difference in characteristic characters; The closed-circuit position state analysis model trained by simulation modeling contact force analysis is specifically: Use virtual prototype technology to conduct 3D modeling with the opponent's circuit breaker based on relevant modeling software; Simplify the parts of the model that do not affect the simulation results; The material and unit part meshes are established for the simplified model to obtain a simulation model; And create constraint conditions, and simulate the working state of the plum blossom contact in the closed position of the circuit breaker of the hand car by adjusting the parameters of the constraint conditions; Through the simulation model, the contact force and deformation of the trolley circuit breaker contacts are analyzed based on the convolutional neural network to establish a closed-circuit position state analysis model.

2. The method for online monitoring of the state of a trolley circuit breaker according to claim 1 is characterized in that: The constraint conditions are created by adjusting the parameters of the constraint conditions to simulate the working state of the plum blossom contacts in the closed position of the circuit breaker of the hand vehicle, specifically: The plum blossom contact working state includes a normal state, a lateral deviation abnormal state, a longitudinal deviation abnormal state, and an angular deviation abnormal state; Creating appropriate constraint conditions in the simulation software, wherein the constraint conditions include the position and contact posture of the stationary contact and the stiffness coefficient of the spring; the contact posture includes the displacement in the Y-axis direction, the displacement in the X-axis direction, the angle around the X-axis and the angle around the Y-axis; The parameters of each constraint condition are set and adjusted individually to obtain the variation of other constraint condition parameters, and the corresponding constraint condition-contact force relationship curve is established in combination with the working state of the plum blossom contact.

3. The method for online monitoring of the state of a trolley circuit breaker according to claim 2 is characterized in that: The specific method of constructing the angle-effect curve is as follows: Obtaining output angle variation evaluation values ​​under different contact angle variations from the angle variation evaluation model; A rectangular coordinate system is established, with the horizontal axis representing the contact angle change and the vertical axis representing the angle change evaluation value; Mark the angle change evaluation value corresponding to each contact angle change on the rectangular coordinate system to form a data point; Connect the data points to form an angle-effect curve.

4. The method for online monitoring of the state of a trolley circuit breaker according to claim 3 is characterized in that: The specific calculation formula of the constraint condition-contact force relationship curve is: In the formula, is the contact force of the plum blossom contact, is the constraint parameter of contact displacement, is the change in the spring stiffness coefficient, is the angle change of the contact, , is the working status adjustment factor, and d is the correction term.

5. The method for online monitoring of the state of a trolley circuit breaker according to claim 4 is characterized in that: The spring stiffness influence data includes a spring stiffness influence coefficient, and a specific method for obtaining the spring stiffness influence coefficient is as follows: Obtain spring stiffness data at different contact angle changes; According to the spring stiffness data and based on the finite difference method, the influence rate of the contact angle change on the spring stiffness is calculated; The influence rate of the contact angle change on the spring stiffness is used as the gradient direction of the preset optimization objective function to optimize the spring stiffness influence coefficient.

6. The method for online monitoring of the state of a trolley circuit breaker according to claim 5, characterized in that: The component force data includes a component force uniformity coefficient; The specific method for obtaining the force uniformity coefficient of the component is as follows: A plurality of measuring points are set on the plum blossom contact, and the forces at the plurality of measuring points under standard conditions are obtained respectively; Obtain the actual forces at several measuring points under the corresponding contact variation, calculate the difference between the forces under the corresponding standard conditions, and construct a difference set; Calculate the standard deviation and average value within the difference set, and use the ratio of the standard deviation to the average value as the component force uniformity coefficient.

7. The method for online monitoring of the state of a trolley circuit breaker according to claim 6, characterized in that: The angle-effect curve is analyzed to obtain the optimal contact angle change, which is specifically: Obtain the peak values ​​of several angle-effect curves, perform secondary derivative on each peak value, and obtain the peak derivative; The contact angle variation corresponding to the peak value whose peak derivative is closest to zero is taken as the optimal contact angle variation, and is used as the contact angle variation for maintaining the abnormality of the closed circuit position.

8. A trolley circuit breaker status online monitoring device, used to implement the trolley circuit breaker status online monitoring method according to any one of claims 1 to 7, characterized in that: Including position monitoring module, closed-circuit monitoring module, open-circuit monitoring module and alarm module; A position monitoring module is used to obtain the position information of the trolley circuit breaker, wherein the position information includes an open circuit position, a closed circuit position and an unknown position; An open circuit monitoring module, used to obtain an open circuit position state analysis model pre-trained by a video bounding box method, and use the open circuit position state analysis model to perform state monitoring on the open circuit position; Closed-circuit monitoring module, used to analyze the contact force and deformation of the trolley circuit breaker contacts based on convolutional neural network, construct constraint condition-contact force relationship curve, and establish a closed-circuit position state analysis model; An alarm module is used to obtain the monitoring results of the open circuit position status and the closed circuit position status, and to issue an alarm for abnormal status in combination with the unknown position; The feedback module is used to perform maintenance when the alarm of abnormal state belongs to the abnormality of closed-circuit position, maintain the contact force within the preset normal range, and set the fixed contact force and the fixed maximum displacement based on the constraint condition-contact force relationship curve; A feedback evaluation module is used to set different contact angle changes, obtain corresponding spring stiffness impact data and component force data, and perform data analysis to obtain an angle change evaluation value; The maintenance module is used to construct an angle-effect curve according to different contact angle changes and corresponding angle change evaluation values, and perform data analysis on the angle-effect curve to obtain the optimal contact angle change, and maintain the abnormality of the closed circuit position through the optimal contact angle change.

Images