Intelligent detection method and system for contact stroke of tulip contact of circuit breaker

Abstract

The invention discloses a circuit breaker plum blossom contact contact stroke intelligent detection method and system, and the method comprises the steps: a configuration step: a static contact is configured with an optical fiber sensing array, an image collection module and a light supplement module, the optical fiber sensing array comprises an optical fiber sensor, the optical fiber sensor comprises an optical fiber probe, and a matched optical fiber cable and an optical fiber amplifier, the image acquisition module comprises an industrial camera; a dynamic detection step: comprehensively scheduling an optical fiber sensor and an industrial camera, and collecting displacement data of a contact piece of the moving contact and connection state data of the closed moving contact and static contact in real time; a data processing step: fusing the displacement data of the optical fiber sensor and the image data of the industrial camera, evaluating the contact state and surface defects of the static contact and the moving contact, embedding a lightweight AI reasoning chip to carry out localization real-time processing, and adopting a contact wear compensation algorithm to carry out real-time processing on the basis of the contact closing times and historical wear data. Dynamically adjusting the detection threshold value of the optical fiber probe and the parameters of the contact state evaluation model.

Description

Technical Field

[0001] The present invention relates to the field of power systems, and particularly to an intelligent detection method and system for the contact travel of a circuit breaker petal contact. Background Art

[0002] In a power system, the contact travel of a circuit breaker petal contact is a core parameter to ensure reliable circuit conduction and safe interruption. The contact travel refers to the displacement process of the moving contact from the initial contact to the fully closed state, and its accuracy directly affects the contact pressure, temperature rise characteristics, and electrical life of the contact. However, there are significant limitations in the monitoring of the contact travel of the moving contact in the prior art, which are specifically manifested as the following technical problems: Traditional detection means mostly rely on mechanical displacement sensors or simple photoelectric switches, which can only obtain the end position information of the contact closure, and cannot capture the detailed displacement trajectory (such as acceleration, synchronization deviation) of the moving contact during high-speed movement in real time. Due to the low sampling rate and large static error, it is difficult to identify the jitter, jamming, or travel deviation during the contact closing process, resulting in the inability to locate the root cause of poor contact.

[0003] The prior art generally indirectly judges the contact connection state through the conduction resistance or current signal, but such methods cannot detect the microscopic defects (such as oxidation spots, arc burns, mechanical wear) on the contact surface. When there is local oxidation or contamination on the contact surface, although the resistance value may still be within the allowable range, the actual contact area has been significantly reduced, and local overheating or even welding is likely to occur during long-term operation. In addition, there is a lack of effective monitoring means for the three-dimensional deformation matching degree (such as inclination and misalignment during closing) between the moving contact and the static contact, further increasing the risk of hidden faults.

[0004] The petal contact usually adopts a multi-contact parallel structure to improve the current-carrying capacity, but the difference in the closing synchronism of each contact will lead to uneven current distribution. The prior art lacks the ability to compare the independent displacements of multiple contacts in real time and cannot quantify the synchronism difference, resulting in long-term overload of local contacts, accelerating aging and inducing arc faults. Summary of the Invention

[0005] The purpose of the present invention is to propose an intelligent detection method and system for the contact travel of a circuit breaker petal contact to improve the technical problems of insufficient dynamic monitoring accuracy, lack of evaluation of surface state and connection quality in the prior art for the contact travel of the petal contact.

[0006] To solve the above technical problems, the present invention provides the following technical solutions: An intelligent detection method for the contact travel of a circuit breaker petal contact, comprising the following steps: Configuration steps: An optical fiber sensing array, an image acquisition module, and a supplementary lighting module are configured on the static contact. The optical fiber sensing array includes optical fiber sensors, and each optical fiber sensor includes an optical fiber probe located on the moving trajectory of the moving contact piece, as well as a supporting optical fiber cable and an optical fiber amplifier. The image acquisition module includes an industrial camera; Dynamic detection steps: Coordinating and dispatching the optical fiber sensors and the industrial camera to collect the displacement data of the moving contact piece in real time and the connection state data of the closed moving contact and static contact; Data processing steps: Fusing the displacement data of the optical fiber sensors and the image data of the industrial camera to evaluate the contact state and surface defects of the static contact and the moving contact, embedding a lightweight AI inference chip for local real-time processing, and adopting a contact wear compensation algorithm. Based on the number of contact closures and historical wear data, dynamically adjust the detection threshold of the optical fiber probe and the parameters of the contact state evaluation model.

[0007] Preferably, the contact wear compensation algorithm includes a wear quantification model construction, a dynamic adjustment mechanism, and a data fusion strategy.

[0008] The specific construction of the wear quantification model is as follows: Analyze the relationship between the number of contact closures and the contact resistance, and establish an exponential wear model, the expression of which is: R(N) = R0 + A(1 - e^{-kN}) + B·N In the formula: R0 is the initial contact resistance, A is the maximum wear increment, k is the wear rate decay factor, B is the linear wear coefficient, and N is the number of closing and opening cycles of the contact. Among them, the initial contact resistance shows a wear trend of first rising rapidly and then gradually flattening with the increase of the number of uses. The maximum wear increment reflects the ultimate loss degree of the material, and the wear rate decay factor describes the characteristic that the wear speed slows down over time; Draw a non-linear compensation function based on historical displacement data, and the non-linear compensation function adopts a piecewise polynomial: When N ≤ 500 times: C(N) = 0.05N² - 1.2N When N>500 times: C(N) = 18.7ln(N) - 112.4 In the formula: C(N) is the contact compensation amount.

[0009] The specific dynamic adjustment mechanism is as follows: When the contact resistance exceeds the preset critical value, automatically increase the displacement detection threshold according to a non-linear rule; According to the thermal effect generated by the current passing through the contact, perform dynamic reverse compensation on the contact pressure. The compensation amount is negatively correlated with the calorific value to offset the pressure increase caused by thermal expansion. The expression is: ΔF=-β·I 2 ·R·Δt Where: ΔF is the contact pressure compensation amount, β is the thermal expansion compensation coefficient, I is the effective value of the current passing through the contact, R is the current contact resistance, and Δt is the current duration time.

[0010] The specific data fusion strategy is as follows: Calculate the cumulative effect of material thermal fatigue based on the temperature change curve, analyze the contact surface image, identify the thickness of the oxide layer and the microscopic characteristics of arc ablation, calculate the comprehensive wear value through weighted fusion of mechanical wear, thermal effect wear, and surface characteristics, and update the detection threshold and evaluation parameters in real time. The expression of the temperature change curve is as follows: T(t)=T0+(T max -T0)(1-e -t / τ ) Where: T(t) is the contact temperature at time t, T0 is the initial ambient temperature, Tmax is the highest temperature of the contact in the steady state, and τ is the thermal time constant, which reflects the heat dissipation ability of the contact.

[0011] Preferably, the real-time update of the detection threshold and evaluation parameters is specifically as follows: Data normalization processing, converting indicators with different dimensions into scores from 0 to 1: Mechanical wear is linearly mapped according to the measurement value of the displacement sensor, thermal effect wear is calculated by logarithmic compression of the Joule integral, and the oxidation area and arc depth of surface characteristics are converted into percentage scores through image analysis; Assign weights based on experimental data: The weight of mechanical wear dominates early wear, the weight of thermal effect reflects long-term thermal fatigue, and the weight of surface characteristics quantifies oxidation and arc damage; Calculate the comprehensive wear value, and the expression is: Comprehensive wear value = 0.4×Mechanical wear score + 0.3×Thermal effect score + 0.3×Surface characteristics score Real-time update the detection threshold, and the threshold adjustment dynamically amplifies the initial threshold according to the comprehensive value. The expression is: New threshold = Initial threshold×(1 + 0.15×Comprehensive value) Update the evaluation parameters, and the classification strategy is: Comprehensive value < 0.3: Maintain routine detection; 0.3 ≤ Comprehensive value < 0.6: Relax the deviation to ±10%, and shorten the detection cycle to 500 ms; Comprehensive value ≥ 0.6: Relax the deviation to ±20%, trigger continuous infrared scanning, and push a maintenance work order.

[0012] Preferably, the specific dynamic detection steps are as follows: Perform zero calibration of the fiber optic sensor and white balance adjustment of the industrial camera, and complete the communication verification between modules; When the moving contact is connected to the static contact, the displacement data of the contact piece of the moving contact is collected in real time, and the contact synchronization difference is calculated. The expression is: difference degree D = (max(ti) - min(ti)) + 0.5·σ(Δt) Where: ti is the trigger time of each probe, σ is the standard deviation. When D>1ms, it is determined as asynchronous, which is the degree of dispersion of the trigger time ti of each contact piece relative to the average time; When the fiber optic probe detects the contact displacement threshold simultaneously, the industrial camera is triggered to take high-speed continuous shots of the contact surface image.

[0013] Preferably, the data processing step is specifically as follows: Fuse the stroke curve data of the fiber optic sensor and the image feature data of the industrial camera, establish the spatio-temporal mapping between the displacement points and the visual features, generate the contact stroke curve of the contact and compare it with the standard template library to evaluate the contact state; Construct a contact surface defect classification model based on deep learning. The model is trained based on multiple data samples to improve the generalization ability, output the defect type, area ratio and severity level, evaluate the surface defects of the contact, use the trained model to analyze the newly collected contact surface image, automatically classify and quantify the existing defect degree, synchronously associate the contact wear compensation algorithm, dynamically adjust the detection threshold of the fiber optic probe, and calculate the three-dimensional deformation amount when the moving contact and the static contact are closed.

[0014] Preferably, the construction of the contact surface defect classification model based on deep learning is specifically as follows: Adopt a pre-trained temporal convolutional neural network model to directly analyze the correlation between the fiber optic displacement data and the image data, and realize millisecond-level anomaly detection; the temporal convolutional neural network model adopts 8-layer convolutional stacking, each convolutional module contains normalization, activation function and spatial random inactivation operations, and adjusts the channel dimension through 1×1 convolution and then adds it to the input; downsample the high-frequency signal of the fiber optic displacement sensor to eliminate the equipment vibration noise, perform multi-scale feature extraction and three-dimensional tensor construction; adopt an improved loss function, adjust the weight coefficient and the adjustment factor, and strengthen the learning effect of the minority-class anomaly samples.

[0015] To solve the above technical problems, the present invention also provides the following technical solutions: An intelligent detection system for the contact stroke of the plum blossom contact of a circuit breaker, comprising: The fiber optic sensing array is arranged on the detection plate vertically arranged on the outer periphery of the static contact near the connection contact of the static contact and the moving contact. The fiber optic sensing array includes fiber optic sensors. The fiber optic sensors include three fiber optic probes embedded in the outer edge of the detection plate and evenly distributed at 120°, located on the movement track of the moving contact blade, and the supporting fiber optic cables and fiber optic amplifiers. The fiber optic probes are configured to detect the approaching distance and contact state of the moving contact blade in real time. The fiber optic probes are connected to the fiber optic amplifiers through the fiber optic cables; The image acquisition module is arranged directly above the static contact. The image acquisition module uses an industrial camera with more than 20 million pixels and is equipped with a macro lens. The image acquisition module is configured to acquire the surface image of the moving contact and the static contact in the connected state; The supplementary light module is arranged around the lens of the industrial camera to form a coaxial illumination system. The supplementary light module integrates a ring-shaped LED light source with a color temperature adjustable range of 5000 - 6500K, providing light sources for the fiber optic probes and the industrial camera; The dynamic detection module is used to overall dispatch the fiber optic sensors and the industrial camera, and collect the displacement data of the moving contact blade and the connection state data of the closed moving contact and static contact in real time. The dynamic detection module is also used to perform zero calibration of the fiber optic sensors and white balance adjustment of the industrial camera, and complete the communication verification between modules; The data processing module is used to fuse the displacement data of the fiber optic sensors and the image data of the industrial camera, and evaluate the contact state and surface defects of the static contact and the moving contact. The data processing module embeds a lightweight AI inference chip for local real-time processing. The data processing module is also configured with a contact wear compensation algorithm, which dynamically adjusts the detection threshold of the fiber optic probes and the parameter of the contact state evaluation model based on the number of contact closures and historical wear data; The control module is used to provide data interaction between the fiber optic sensors, the industrial camera, and the fiber optic amplifiers, and trigger the breaker protection action. The control module supports the Modbus TCP protocol. The control module is also used for hierarchical alarm of abnormal states. When the contact delay of the contact blade > 1ms, the surface oxidation area > 15%, or the deformation amount exceeds 80% of the material yield limit, it triggers an audible and visual alarm. At the same time, a linkage control interface is set up to trigger breaker locking, start the ventilation system in the cabinet, or push maintenance reminders.

[0016] Compared with the prior art, the beneficial technical effects of the present invention are: 1. The present invention integrates multi-modal sensors of a fiber optic probe array and an industrial camera. The fiber optic probes capture the details of the displacement trajectory of the moving contact in real time, solving the problems of low sampling rate and large error in traditional methods. The industrial camera combined with ring-shaped supplementary light directly detects microscopic defects such as oxidation spots and arc burns, avoiding the indirect judgment blind area that only relies on resistance values.

[0017] 2. This invention calculates the contact closure delay and displacement deviation through real-time comparison of three optical fiber signals, and locates contacts with asynchronous contact. When the three probes reach the displacement threshold at the same time, the industrial camera is triggered to take high-speed continuous shots, accurately capturing the instantaneous state of multiple contact closures, avoiding the omission of actions by traditional methods. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 This is a schematic diagram of the moving contact not being connected; Figure 2 This is a schematic diagram of the moving contact connected state; Figure 3 Schematic diagram of the static contact port; Figure 4 It is a schematic diagram of the module structure; Figure 5 The figure is a flow chart of the intelligent detection method for the contact stroke of the circuit breaker plum blossom contacts.

[0019] The following are marked in the figure: 1. static contact; 2. detection board; 3. fiber optic probe; 4. fiber optic cable; 5. moving contact. DETAILED DESCRIPTION

[0020] In the description of the present application, the meaning of "multiple" is at least two, such as two, three, etc., unless otherwise clearly and specifically limited. In the embodiments of the present application, all directional indications (such as up, down, left, right, front, back, top, bottom ...) are only used to explain the relative position relationship, movement, etc. between the components under a certain specific posture (as shown in the accompanying drawings). If the specific posture changes, the directional indication also changes accordingly. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions. For example, a process, method, system, product or Internet of Things terminal that includes a series of steps or units is not limited to the listed steps or units, but optionally also includes steps or units that are not listed, or optionally also includes other steps or units inherent to these processes, methods, products or Internet of Things terminals.

[0021] In addition, references to "embodiments" herein mean that a particular feature, structure, or characteristic described in connection with the embodiments may be included in at least one embodiment of the present application. The appearance of such phrases in various places in the specification does not necessarily refer to the same embodiment, nor does it necessarily refer to independent or alternative embodiments that are mutually exclusive of other embodiments. It is understood, both explicitly and implicitly, by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0022] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without making creative efforts belong to the scope of protection of the present application. It should be noted that the numerical values described in the present application should not be fixed values. The numerical values are only example values and should not limit the scope of protection of the present application. The relevant numerical data in the present application can be determined according to the specific situation of the power grid.

[0023] The purpose of the present invention is to propose an intelligent detection method and system for the contact stroke of the breaker petal contact, so as to improve the technical problems of insufficient dynamic monitoring accuracy of the contact stroke of the petal contact, lack of evaluation of surface state and connection quality in the prior art. Through multi-dimensional sensing fusion, intelligent analysis algorithms and active control mechanisms, the core problems in traditional technologies such as rough dynamic monitoring, missed detection of surface defects, and synchronization blind spots are accurately solved, significantly improving the accuracy, real-time performance and intelligent level of the breaker contact state monitoring, and providing a reliable guarantee for the safe operation of the power system.

[0024] Embodiment 1, As Figure 5 shown in the step flow chart of the present application, the present invention provides an intelligent detection method for the contact stroke of the breaker petal contact, including the following steps: Configuration step: An optical fiber sensing array, an image acquisition module and a supplementary light module are configured on the static contact 1. The optical fiber sensing array includes optical fiber sensors. The optical fiber sensors include an optical fiber probe 3 located on the movement track of the moving contact piece, and a supporting optical fiber cable 4 and an optical fiber amplifier. The image acquisition module includes an industrial camera; Further, as Figures 1-3 shown, an optical fiber sensing array is arranged on the detection plate vertically arranged on the outer periphery of the connection contact between the static contact and the moving contact near the static contact. The optical fiber sensing array is composed of high-precision optical fiber sensors. The optical fiber sensors include three optical fiber probes 3 and a supporting optical fiber cable 4 and an optical fiber amplifier. The optical fiber probes 3 are embedded at the outer edge of the detection plate 2 and are evenly distributed at 120°. And the optical fiber probe 3 is located on the movement track of the moving contact piece 5 of the moving contact. The probe can be installed at a certain inclination angle. The axis of the probe can be set at an angle of 15° with the plane of the movement track of the contact piece to avoid physical occlusion. The optical fiber probe 3 is configured to detect the approaching distance and contact state of the contact piece of the moving contact 5 in real time. The optical fiber probe 3 is connected to the optical fiber amplifier through the optical fiber cable 4. The optical fiber amplifier is mainly used to enhance the signal collected by the optical fiber probe 3 to ensure the accuracy and reliability of the data; When the moving contact 5 moves, the change in the distance between the optical fiber probe 3 and the contact piece causes the intensity of the reflected light to change. Within a specific distance range, the intensity of the reflected light has a non-linear relationship with the distance between the optical fiber probe and the contact piece, which is approximately Gaussian distribution; The change in light intensity is mapped to a displacement value through a calibration curve. The received optical signal is converted into a current signal by a detector. By comparing the synchronism of the three signals, it is judged whether the moving contact 5 is aligned with the static contact 1 when it is closed. When the displacements of the three channels reach the contact threshold simultaneously, it is determined as fully closed; otherwise, it is incompletely closed. The threshold can be set to 0.5 mm. A gap of 3 ± 0.5 mm is preset between the detection end face of the optical fiber probe 3 and the top surface of the contact piece of the moving contact 5 to ensure that the optical fiber probe 3 does not come into direct contact with the moving contact 5, avoiding wear or damage caused by contact. At the same time, it also ensures that during the approach of the moving contact 5, the optical fiber probe 3 can perform precise measurements within its optimal working range. If the distance is too close or too far, it will affect the performance and accuracy of the optical fiber sensor; The sampling rate of the optical fiber probe 3 is not less than 10 kHz, which is convenient for capturing rapid or subtle changes during the movement of the moving contact 5. More data points can be obtained per unit time, which helps to improve the resolution of the displacement trajectory of the moving contact 5. The static detection accuracy is ±0.15 mm, ensuring high-precision measurement; The image acquisition module uses an industrial camera with more than 20 million pixels and is equipped with a macro lens. It is installed directly above the static contact, and the installation height can be about 150 mm from the end face of the static contact. It is configured to collect the surface images of the moving contact and the static contact in the connected state, providing intuitive visual information as a supplement to the detection results of the optical fiber probe 3; The supplementary light module is integrated around the camera lens to form a coaxial lighting system. An integrated ring LED light source is used, and the adjustable range of the color temperature is 5000 - 6500K, providing light sources for the optical fiber probe and the industrial camera; Furthermore, it also includes a control module that supports the Modbus TCP protocol and is configured to provide data interaction between the optical fiber sensor, the industrial camera, and the optical fiber amplifier, as well as trigger the breaker protection action; The control module is also used for hierarchical alarm of abnormal states. When the contact delay of the contact piece > 1 ms, the surface oxidation area > 15%, or the deformation amount exceeds 80% of the material yield limit, an audible and visual alarm is triggered. At the same time, a linkage control interface is set up to trigger breaker locking, start the ventilation system in the cabinet, or push maintenance reminders.

[0025] Dynamic detection steps: Overall scheduling of the optical fiber sensor and the industrial camera, real-time collection of the displacement data of the contact piece of the moving contact and the connection state data of the moving contact and the static contact after closing; Further, the dynamic detection step also performs zero calibration of the fiber optic sensor and white balance adjustment of the industrial camera, and completes the communication verification between modules; When the moving contact 5 is connected to the static contact 1, the fiber optic probe 3 is adjusted to collect the displacement data of the contact piece of the moving contact 5 in real time, and the contact synchronization difference is calculated. The expression is: Difference degree D = (max(ti) - min(ti)) + 0.5·σ(Δt) In the formula, ti: trigger time of each probe (μs), σ: standard deviation. When D>1ms, it is determined as asynchronous, which is the degree of dispersion of the trigger time ti of each contact piece relative to the average time; When the three fiber optic probes 3 simultaneously detect the displacement threshold, the industrial camera is triggered to take high-speed continuous shots.

[0026] The data processing step fuses the displacement data of the fiber optic sensor and the image data of the industrial camera, evaluates the contact state and surface defects of the static contact 1 and the moving contact 5, embeds a lightweight AI inference chip for local real-time processing, adopts a contact wear compensation algorithm, and based on the contact closing times and historical wear data, dynamically adjusts the detection threshold of the fiber optic probe 3 and the parameters of the contact state evaluation model to avoid misjudgment caused by long-term wear.

[0027] Further, by fusing the stroke curve data of the fiber optic sensor and the image feature data of the industrial camera, the industrial camera and the fiber optic sensor share the same clock source, and each frame of image corresponds to an accurate timestamp. When the fiber optic detects the contact displacement threshold, the camera is triggered to capture the contact surface image, establish the spatio-temporal mapping between the displacement point and the visual feature, generate the contact stroke curve of the contact and compare it with the standard template library. The standard template library contains the stroke curve in the ideal contact state. By comparing, it is judged whether the current contact state meets the expectation, establish a dynamic model of the contact state between the moving contact 5 and the static contact 1, and evaluate the contact state. The horizontal axis of the coordinate of the contact stroke curve: time (ms), from the start of the contact to the complete closing, and the vertical axis: displacement (mm), which is the real-time position of the moving contact relative to the static contact; Build a contact surface defect classification model based on deep learning and evaluate the surface defects of the contact, which helps to identify various problems that may affect the performance of the contact. The data set includes working conditions such as oxidation spots, arc burns and mechanical wear. The contact surface defect classification model collects high-resolution images of the contact surface and uses a lightweight convolutional neural network to build a defect classification model to realize the automatic identification of oxidation spots, arc burns and mechanical wear. The model is trained based on multiple data samples to improve the generalization ability, and outputs the defect type, area ratio and severity level, and synchronously associates the contact wear compensation algorithm to dynamically adjust the detection threshold of the fiber optic probe 3; Further, use the trained model to analyze the newly collected contact surface images, automatically classify and quantify the existing defect degree; Embed a lightweight AI inference chip in the data processing module to support local real-time processing. Through a pre-trained temporal convolutional neural network model, directly analyze the correlation between fiber optic displacement data and image data to achieve millisecond-level anomaly detection and avoid cloud delays; The temporal convolutional neural network model uses 8 layers of convolutional stacking, and the temporal receptive field of each convolutional kernel expands exponentially, enabling the network model to capture both short-term fluctuations and long-term trends simultaneously. Each layer contains 28 convolutional kernels with a width of 7 to ensure coverage of the temporal features within a 1024-millisecond time window; Each convolutional module includes normalization, activation function, and spatial dropout operations, and adjusts the channel dimension through 1×1 convolution and then adds it to the input. This design alleviates the problem of gradient disappearance in deep networks and improves the ability to extract weak anomaly features; Downsample the 10kHz high-frequency signal of the fiber optic displacement sensor to 1024Hz, eliminate device vibration noise, and perform multi-scale feature extraction and three-dimensional tensor construction; adopt an improved loss function, and by adjusting the weight coefficients (normal state 0.2, minor anomaly 0.3, severe anomaly 0.5) and adjustment factors, strengthen the learning effect on minority-class anomaly samples; Calculate the three-dimensional deformation amount when the moving contact and the static contact are closed, specifically to quantify the displacement deviation of the moving contact and the static contact in the X / Y / Z axis directions, which is used to detect potential structural problems or uneven pressure distributions.

[0028] Furthermore, the contact state assessment includes: Matching score based on the slope of the travel curve and the standard template; Output the comprehensive rating and adjustment and maintenance suggestions. Among them, the comprehensive rating categories are divided into four levels: excellent, good, poor, and dangerous.

[0029] Furthermore, the contact wear compensation algorithm includes wear quantification model construction, dynamic adjustment mechanism, and data fusion strategy; The wear quantification model construction is configured to: by analyzing the relationship between the contact closing times and the contact resistance, establish an exponential wear model, and the expression is: R(N) = R0 + A(1 - e^{-kN}) + B·N In the formula: R0: initial contact resistance (μΩ), A = 120: maximum wear increment, k = 0.007: wear rate decay factor, B = 0.3: linear wear coefficient, N: number of closing and opening cycles of the contact; among them, the initial contact resistance shows a wear trend of first rising rapidly and then gradually flattening with the increase of the use times, where the maximum wear increment reflects the material limit loss degree, and the wear rate decay factor describes the characteristic that the wear speed slows down over time; A non - linear compensation function is plotted based on historical displacement data, comprehensively considering the dual effects of linear wear and abnormal wear. The non - linear compensation function adopts a piece - wise polynomial: When N ≤ 500 times: C(N) = 0.05N² - 1.2N When N > 500 times: C(N) = 18.7ln(N) - 112.4 In the formula, C(N) is the contact compensation amount.

[0030] The dynamic adjustment mechanism is configured as follows: when the contact resistance exceeds the preset critical value, the displacement detection threshold is automatically increased according to the non - linear law to avoid false alarms caused by material aging; according to the thermal effect generated by the current passing through the contact, dynamic reverse compensation is performed on the contact pressure. Among them, the compensation amount is negatively correlated with the calorific value, offsetting the pressure increase caused by thermal expansion. The expression is: ΔF=-β·I 2 ·R·Δt In the formula, ΔF: contact pressure compensation amount (unit: N), β: thermal expansion compensation coefficient (typical value: 0.0035 N / (A²·Ω·s)), I: effective value of the current passing through the contact (A), R: current contact resistance (Ω), Δt: current duration (s).

[0031] The data fusion strategy is configured as follows: calculate the cumulative effect of material thermal fatigue based on the temperature change curve, analyze the contact surface image, identify microscopic features such as the thickness of the oxide layer and arc ablation, calculate the comprehensive wear value through weighted fusion of mechanical wear, thermal - effect wear and surface features, and update the detection threshold and evaluation parameters in real - time; Furthermore, the expression of the temperature change curve is: T(t)=T0+(T max -T0)(1 - e -t / τ ) In the formula, T(t): contact temperature at time t, T0: initial ambient temperature, T max : highest temperature of the contact in the stable state, τ: thermal time constant, reflecting the heat dissipation ability of the contact; Furthermore, the real - time update of the detection threshold and evaluation parameters is specifically as follows: Data normalization processing, converting indicators with different dimensions into scores from 0 to 1 uniformly: Mechanical wear: linearly mapped according to the measurement value of the displacement sensor (such as 0 - 2 mm corresponding to 0 - 1 point), Thermal - effect wear: calculated by logarithmic compression of the Joule integral (such as 0 - 1.5×10 6 J corresponding to 0 - 1 point), Surface features: the oxidation area and arc depth are converted into percentage scores through image analysis (such as 0% - 30% oxidation corresponding to 0 - 1 point); Weight assignment, assigning weights based on experimental data: Weight of mechanical wear is 0.4: Dominating early wear (such as displacement deviation), Weight of thermal effect is 0.3: Reflecting long-term thermal fatigue (such as cumulative Joule value), Weight of surface characteristics is 0.3: Quantifying oxidation and arc damage (such as oxidation area × arc depth); Calculation of comprehensive wear value, the expression is: Comprehensive wear value = 0.4 × mechanical wear score + 0.3 × thermal effect score + 0.3 × surface characteristics score For example: mechanical wear score is 0.6, thermal effect score is 0.5, surface score is 0.8 → comprehensive value = 0.4×0.6 + 0.3×0.5 + 0.3×0.8 = 0.63; Real-time update of detection threshold, the threshold is adjusted to dynamically amplify the initial threshold according to the comprehensive value, The expression is: new threshold = initial threshold × (1 + 0.15 × comprehensive value) Example: initial threshold is 0.5mm, comprehensive value is 0.6 → new threshold = 0.5×1.09 = 0.545mm Update of evaluation parameters, grading strategy: Comprehensive value < 0.3: Maintain routine detection (such as contact resistance deviation ±5%), 0.3 ≤ comprehensive value < 0.6: Relax the deviation to ±10% and shorten the detection cycle to 500ms, Comprehensive value ≥ 0.6: Relax the deviation to ±20%, trigger continuous infrared scanning and push a maintenance work order.

[0032] The active control of this application is reflected in: automatically correcting the trigger threshold of the fiber optic probe according to the contact wear quantification model. When detecting the out-of-sync deviation of the contact piece, a pulse signal is sent to the operating mechanism through the control module to drive the servo motor for stroke compensation, and the energy storage motor power supply is synchronously cut off when the circuit breaker is locked.

[0033] Embodiment 2, Such as Figure 4 shown, this application also provides an intelligent detection system for the contact stroke of the circuit breaker's plum blossom contact, including: An optical fiber sensing array is arranged on a detection plate (2) vertically disposed near the connection contact of the static contact (1) and around the outer periphery of the connection contact of the static contact and the moving contact (5). The optical fiber sensing array includes optical fiber sensors. Each optical fiber sensor includes three optical fiber probes (3) embedded in the outer edge of the detection plate (2) and evenly distributed at 120°, located on the movement track of the moving contact blade, and the supporting optical fiber cables (4) and optical fiber amplifiers. The optical fiber probes (3) are configured to detect the approaching distance and contact state of the moving contact blade in real time. The optical fiber probes (3) are connected to the optical fiber amplifiers through the optical fiber cables (4); An image acquisition module is arranged directly above the static contact (1). The image acquisition module uses an industrial camera with a pixel count greater than 20 million and is equipped with a macro lens. The image acquisition module is configured to acquire the surface image of the moving contact and the static contact in the connected state; A supplementary lighting module is arranged around the lens of the industrial camera to form a coaxial lighting system. The supplementary lighting module integrates a ring-shaped LED light source with a color temperature adjustable range of 5000 - 6500K, providing light sources for the optical fiber probes and the industrial camera; A dynamic detection module is used to overall coordinate and dispatch the optical fiber sensors and the industrial camera, and to collect the displacement data of the moving contact blade and the connection state data of the closed moving contact and static contact in real time. The dynamic detection module is also used to perform zero calibration of the optical fiber sensors and white balance adjustment of the industrial camera, and to complete communication verification between various modules; A data processing module is used to fuse the displacement data of the optical fiber sensors and the image data of the industrial camera, and to evaluate the contact state and surface defects of the static contact and the moving contact. The data processing module embeds a lightweight AI inference chip for local real-time processing. The data processing module is also configured with a contact wear compensation algorithm, which dynamically adjusts the detection threshold of the optical fiber probes (3) and the parameters of the contact state evaluation model based on the number of contact closures and historical wear data; A control module is used to provide data interaction between the optical fiber sensors, the industrial camera, and the optical fiber amplifiers, and to trigger the protection action of the circuit breaker. The control module supports the Modbus TCP protocol. The control module is also used to perform hierarchical alarm for abnormal states. When the contact delay of the blade > 1ms, the surface oxidation area > 15%, or the deformation amount exceeds 80% of the material yield limit, an audible and visual alarm is triggered. At the same time, a linkage control interface is set up to trigger circuit breaker locking, start the ventilation system in the cabinet, or push a maintenance reminder.

[0034] Furthermore, this solution provides an intelligent detection system for the contact travel of the circuit breaker's plum blossom contacts. A detection board 2 is arranged on the static contact 1, and three fiber optic probes 3 are evenly distributed on the detection board at 120°. They continuously monitor the approaching distance and contact state of the contact piece of the moving contact 5, enhance the signal through a fiber optic amplifier to ensure data accuracy, and preset a gap of 3 ± 0.5 mm to avoid wear. An industrial camera is installed directly above the static contact, with a pixel count of more than 20 million and equipped with a macro lens to capture the surface image in the state of contact between the contacts. A ring-shaped LED light source provides suitable illumination; The dynamic detection module is responsible for scheduling the fiber optic sensors and industrial cameras, collecting displacement and connection state data, performing calibration and communication verification. When the moving contact closes, it calculates the synchronization difference and triggers high-speed continuous shooting when the threshold is reached. The data processing module fuses the data from the fiber optic sensors and industrial cameras to generate a travel curve, establishes a contact state model to evaluate the contact situation and surface defects, and uses deep learning to identify oxidation spots, arc burns, etc. The control module supports the Modbus TCP protocol, processes data interaction and triggers protection actions, such as hierarchical alarm, starting the ventilation system, etc., to ensure the safe operation of electrical equipment. Through the cooperation of these components, intelligent monitoring of the contact travel of the contacts and output of maintenance suggestions are achieved.

[0035] The above detailed description is the same as that of the method in Embodiment 1 and is omitted here.

[0036] The basic principles of this application have been described above in conjunction with specific embodiments. However, it should be noted that the advantages, benefits, effects, etc. mentioned in this application are only examples and not limitations. It cannot be considered that these advantages, benefits, effects, etc. are essential for each embodiment of this application. Additionally, the above-disclosed specific details are only for the purposes of illustration and facilitating understanding, and are not limitations. These details do not limit this application to necessarily adopt the above specific details for implementation. The above description is provided to enable any person skilled in the art to make or use this application. Various modifications to these aspects are very obvious to those skilled in the art, and the general principles defined herein can be applied to other aspects without departing from the scope of this application. Therefore, this application is not intended to be limited to the aspects shown herein, but rather to the broadest scope consistent with the principles and novel features invented herein.

[0037] The above are only the preferred embodiments of this application's creation and are not used to limit this application's creation. Any modifications, equivalent replacements, etc. made within the spirit and principles of this application's creation shall be included within the protection scope of this application's creation.

[0038] Although embodiments of the present invention have been shown and described, it will be understood by those of ordinary skill in the art that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the present invention, and the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. An intelligent detection method for the contact stroke of the breaker plum blossom contact, characterized in that, It includes the following steps: Configuration step: An optical fiber sensing array, an image acquisition module, and a supplementary lighting module are configured on the static contact (1). The optical fiber sensing array includes optical fiber sensors, and each optical fiber sensor includes an optical fiber probe (3) located on the movement trajectory of the moving contact piece, a supporting optical fiber cable (4), and an optical fiber amplifier. The image acquisition module includes an industrial camera; Dynamic detection step: Coordinating and dispatching the optical fiber sensors and the industrial camera to collect the displacement data of the moving contact piece and the connection state data of the closed moving contact and static contact in real time; Data processing step: Fusing the displacement data of the optical fiber sensors and the image data of the industrial camera to evaluate the contact state and surface defects of the static contact and the moving contact, embedding a lightweight AI inference chip for local real-time processing, and adopting a contact wear compensation algorithm. Based on the contact closing times and historical wear data, dynamically adjust the detection threshold of the optical fiber probe (3) and the parameters of the contact state evaluation model.

2. The intelligent detection method for the contact stroke of the breaker plum blossom contact according to claim 1, characterized in that, The contact wear compensation algorithm includes wear quantification model construction, dynamic adjustment mechanism, and data fusion strategy.

3. The intelligent detection method for the contact stroke of the breaker plum blossom contact according to claim 2, characterized in that, The specific construction of the wear quantification model is as follows: Analyze the relationship between the contact closing times and the contact resistance, and establish an exponential wear model, the expression of which is: R(N) = R0 + A(1 - e^{-kN}) + B·N In the formula: R0 is the initial contact resistance, A is the maximum wear increment, k is the wear rate decay factor, B is the linear wear coefficient, and N is the number of closing and opening cycles of the contact. Among them, the initial contact resistance shows a wear trend of first rising rapidly and then gradually flattening with the increase of the usage times. The maximum wear increment reflects the ultimate loss degree of the material, and the wear rate decay factor describes the characteristic that the wear speed slows down over time; Draw a non-linear compensation function based on the historical displacement data. The non-linear compensation function uses a piecewise polynomial: When N ≤ 500 times: C(N) = 0.05N² - 1.2N When N > 500 times: C(N) = 18.7ln(N) - 112.4 In the formula: C(N) is the contact compensation amount.

4. The intelligent detection method for the contact travel of the plum blossom contact of a circuit breaker according to claim 3, characterized in that, The specific dynamic adjustment mechanism is as follows: when the contact resistance exceeds the preset critical value, the displacement detection threshold is automatically increased according to a non-linear law; according to the thermal effect generated by the current passing through the contact, dynamic reverse compensation is performed on the contact pressure, and the compensation amount is negatively correlated with the heat generation amount to offset the pressure increase caused by thermal expansion. The expression is: ΔF = -β·I 2 ·R·Δt In the formula: ΔF is the contact pressure compensation amount, β is the thermal expansion compensation coefficient, I is the effective value of the current passing through the contact, R is the current contact resistance, and Δt is the current duration.

5. A method for intelligently detecting the contact stroke of a circuit breaker's plum blossom contact, according to claim 4, characterized in that, The specific data fusion strategy is as follows: Calculate the cumulative effect of material thermal fatigue based on the temperature change curve, analyze the image of the contact surface, identify the thickness of the oxide layer and the microscopic characteristics of arc ablation, and calculate the comprehensive wear value through weighted fusion of mechanical wear, thermal effect wear, and surface characteristics, and update the detection threshold and evaluation parameters in real time. The expression of the temperature change curve is: T(t)=T0+(T max -T0)(1-e -t / τ ) In the formula: T(t) is the contact temperature at time t, T0 is the initial ambient temperature, Tmax is the highest temperature of the contact in the stable state, and τ is the thermal time constant, which reflects the heat dissipation ability of the contact.

6. The intelligent detection method for the contact stroke of the breaker plum blossom contact according to claim 5, characterized in that, The specific real-time update of the detection threshold and evaluation parameters is as follows: Data normalization is performed to uniformly convert indicators with different dimensions into scores ranging from 0 to 1: Mechanical wear is linearly mapped according to the measurement values of displacement sensors, thermal effect wear is calculated through logarithmic compression of the Joule integral, and the oxidation area and arc depth of surface features are converted into percentage scores through image analysis; Weights are assigned based on experimental data: The weight of mechanical wear dominates early wear, the weight of thermal effect reflects long-term thermal fatigue, and the weight of surface features quantifies oxidation and arc damage; Calculate the comprehensive wear value, and the expression is: Comprehensive wear value = 0.4×Mechanical wear score + 0.3×Thermal effect score + 0.3×Surface feature score Update the detection threshold in real time. The threshold adjustment dynamically amplifies the initial threshold according to the comprehensive value, and the expression is: New threshold = Initial threshold×(1 + 0.15×Comprehensive value) Update the evaluation parameters, and the grading strategy is: Comprehensive value < 0.3: Maintain regular detection; 0.3 ≤ Comprehensive value < 0.6: Relax the deviation to ±10%, and shorten the detection cycle to 500 ms; 7. An intelligent detection method for the contact stroke of the plum blossom contact of a circuit breaker according to claim 1, characterized in that, Comprehensive value ≥ 0.6: Relax the deviation to ±20%, trigger continuous infrared scanning and push a maintenance work order. The specific dynamic detection steps are as follows: Perform zero calibration of the fiber optic sensor and white balance adjustment of the industrial camera, and complete the communication verification between modules; When the moving contact connects with the static contact, collect the displacement data of the moving contact blade in real time, and calculate the contact synchronization difference. The expression is: Difference degree D = (max(ti) - min(ti)) + 0.5·σ(Δt) In the formula: ti is the trigger time of each probe, σ is the standard deviation. When D > 1 ms, it is determined as asynchronous, which is the degree of dispersion of the trigger time ti of each contact blade relative to the average time; 8. A method for intelligent detection of the contact stroke of a circuit breaker's plum blossom contact, according to claim 1, characterized in that When the fiber optic probe detects the contact displacement threshold at the same time, trigger the industrial camera to take high-speed continuous shots of the contact surface image. The specific data processing steps are as follows: Fuse the stroke curve data of the fiber optic sensor and the image feature data of the industrial camera, establish a spatio-temporal mapping between displacement points and visual features, generate a contact stroke curve of the contact and compare it with the standard template library to evaluate the contact state; 9. The intelligent detection method for the contact travel of the plum blossom contact of a circuit breaker according to claim 8, characterized in that, Construct a contact surface defect classification model based on deep learning. The model is trained based on multiple data samples to improve the generalization ability, output the defect type, area ratio and severity level, evaluate the surface defects of the contact, use the trained model to analyze the newly acquired contact surface image, automatically classify and quantify the existing defect degree, synchronously associate the contact wear compensation algorithm, dynamically adjust the detection threshold of the fiber optic probe (3), and calculate the three-dimensional deformation amount when the moving contact and the static contact are closed. The specific construction of the contact surface defect classification model based on deep learning is as follows: Using a pre-trained temporal convolutional neural network model, directly analyze the correlation between fiber optic displacement data and image data to achieve millisecond-level anomaly detection; the temporal convolutional neural network model uses 8 layers of convolutional stacking, each convolutional module contains normalization, activation function and spatial random inactivation operations, and adjusts the channel dimension through 1×1 convolution and then adds it to the input; downsample the high-frequency signal of the fiber optic displacement sensor to eliminate equipment vibration noise, perform multi-scale feature extraction and three-dimensional tensor construction; use an improved loss function, adjust the weight coefficient and adjustment factor, and strengthen the learning effect of minority-class anomaly samples.

10. An intelligent detection system for the contact travel of the plum blossom contact of a circuit breaker, characterized in that, Including: A fiber optic sensing array is arranged on a detection plate (2) vertically arranged near the connection contact of the static contact (1) and the moving contact (5). The fiber optic sensing array includes fiber optic sensors. The fiber optic sensors include three fiber optic probes (3) embedded in the outer edge of the detection plate (2) and evenly distributed at 120°, located on the moving track of the moving contact blade, and the supporting fiber optic cable (4) and fiber optic amplifier. The fiber optic probe (3) is configured to detect the approaching distance and contact state of the moving contact blade in real time. The fiber optic probe (3) is connected to the fiber optic amplifier through the fiber optic cable (4); An image acquisition module is arranged directly above the static contact (1). The image acquisition module uses an industrial camera with more than 20 million pixels and is equipped with a macro lens. The image acquisition module is configured to acquire the surface image of the moving contact and the static contact in the connected state; A supplementary light module is arranged around the industrial camera lens to form a coaxial lighting system. The supplementary light module integrates a ring-shaped LED light source with a color temperature adjustable range of 5000 - 6500K to provide light sources for the fiber optic probe and the industrial camera; A dynamic detection module is used to overall dispatch the fiber optic sensors and the industrial camera, and collect the displacement data of the moving contact blade and the connection state data of the closed moving contact and static contact in real time. The dynamic detection module is also used to perform zero calibration of the fiber optic sensors and white balance adjustment of the industrial camera, and complete the communication verification between each module; A data processing module is used to fuse the fiber optic sensor displacement data and the industrial camera image data, evaluate the contact state and surface defects of the static contact and the moving contact. The data processing module embeds a lightweight AI inference chip for local real-time processing. The data processing module is also configured with a contact wear compensation algorithm, which dynamically adjusts the detection threshold of the fiber optic probe (3) and the parameters of the contact state evaluation model based on the number of contact closures and historical wear data; A control module is used to provide data interaction between the fiber optic sensors, the industrial camera, and the fiber optic amplifier and trigger the breaker protection action. The control module supports the Modbus TCP protocol. The control module is also used to perform hierarchical alarm for abnormal states. When the contact delay of the blade > 1ms, the surface oxidation area > 15%, or the deformation amount exceeds 80% of the material yield limit, it triggers an audible and visual alarm. At the same time, a linkage control interface is set to trigger breaker locking, start the ventilation system in the cabinet, or push a maintenance reminder.

Images