Switching contact state evaluation system and method based on dynamic resistance and vibration fusion

Abstract

This invention belongs to the field of condition monitoring technology for high-voltage electrical equipment in power systems, specifically relating to a switch contact condition evaluation system and method based on dynamic resistance and vibration fusion. The evaluation system includes a data acquisition unit, a feature extraction module, a dynamic weighted fusion unit, and a status output unit. It incorporates an environmental parameter acquisition module and an interference removal step. When environmental parameters exceed the range or the signal is excessively interfered with, the system automatically determines an anomaly and re-acquires the signal or corrects the result, improving the system's environmental adaptability and reducing the risk of misjudgment. It employs a dynamic weighted fusion algorithm based on Shapley value analysis, which has simple and clear logic, low computational load, and can be embedded in an embedded monitoring device for online real-time evaluation. Based on the range of the obtained condition evaluation index (SI) value, multiple status levels are defined to match a tiered operation and maintenance strategy, achieving a closed loop from condition monitoring to operation and maintenance decision-making.

Description

Technical Field

[0001] This invention belongs to the field of condition monitoring technology for high-voltage electrical equipment in power systems, specifically relating to a switch contact condition evaluation system and method based on the fusion of dynamic resistance and vibration. Background Technology

[0002] Switch contacts are the core current-carrying components of high-voltage circuit breakers, used to connect and disconnect load currents and interrupt fault currents during line faults. During long-term operation, switch contacts are susceptible to defects such as arc erosion, mechanical wear, and environmental corrosion, leading to erosion and poor contact. These defects may not be immediately apparent, but they gradually increase contact resistance, decrease breaking capacity, and eventually cause serious faults such as weld failure. Therefore, switch contact condition evaluation is of great significance for ensuring the normal operation of the power grid.

[0003] Evaluation methods based on dynamic resistance parameters or mechanical motion parameters have been applied in the field of switch contact condition monitoring. However, traditionally, a single monitoring parameter is usually used as the basis for condition evaluation, which leads to diagnostic uncertainty. When only resistance parameters are monitored, it is difficult to distinguish whether the increased resistance is caused by contact erosion or insufficient contact pressure is caused by mechanical jamming, making it impossible to effectively identify the fault type. While monitoring only mechanical actions (such as stroke or vibration) can detect abnormal actions, it is impossible to quantify the specific degree of degradation of electrical contact performance. To overcome the above shortcomings, existing technologies have proposed a switch contact condition evaluation scheme that integrates dynamic resistance signals and mechanical motion signals. Chinese patent CN115481567A proposes a method and device for evaluating the erosion of circuit breaker contacts based on multi-dimensional information fusion. First, it acquires vibration signals, main circuit dynamic resistance, and other information during the opening and closing process of circuit breaker contacts. Then, it extracts the characteristics of various signals at different times, integrates and summarizes the characteristics of the contact segment to obtain characteristic parameters. Finally, it combines the RVM (Relevance Vector Machine) algorithm to obtain the degree of erosion of the arc contact. Chinese patent CN109164382B provides a method for diagnosing electrical erosion faults in high-voltage circuit breaker contacts. It collects dynamic contact resistance and stroke curves, uses the bat algorithm to optimize support vector machine (SVM) parameters, and establishes a nonlinear regression model to assess the contact erosion state. The aforementioned existing technical solutions evaluate the condition of circuit breaker electrical contacts by using dynamic resistance to reflect the electrical contact state (degree of erosion) and mechanical signals (vibration signals in CN115481567A and stroke signals in CN109164382B) to reflect the mechanical action state. This "electromechanical" combination eliminates the diagnostic blind spot of a single parameter and can distinguish between simple mechanical loosening (abnormal mechanical signals only) and contact erosion (synchronous abnormality in resistance and mechanical characteristics), enabling a more accurate evaluation of the erosion state of arc contacts. However, the aforementioned existing technical solutions still have three shortcomings in practical applications: (1) Poor environmental adaptability and high false alarm rate: In the actual operating environment of outdoor substations, the system is affected by strong electromagnetic interference from the substation and vibrations from equipment such as fans and transformers, which can easily lead to environmental interference being misjudged as faults. In addition, temperature and humidity conditions may exceed the system's adaptability range, which can also increase the error in signal feature extraction. The existing technical solutions CN115481567A and CN109164382B lack a targeted environmental correction mechanism, which may result in a higher false alarm rate in harsh environments.

[0004] (2) The algorithm is complex and difficult to implement in engineering: CN115481567A relies on the RVM algorithm, and CN109164382B relies on the SVM algorithm. These algorithms require a large number of historical samples for offline training and have high computational requirements. It is difficult to directly deploy the evaluation system in the online monitoring device at the circuit breaker site, which increases the difficulty and cost of engineering implementation.

[0005] (3) Lack of operation and maintenance decision mapping leads to resource waste: CN115481567A outputs ambiguous numerical predictions of "ablation amount," while CN109164382B outputs a binary classification result of "abnormal / normal," lacking direct guidance for operation and maintenance decisions. When switch contacts are in a state of slight ablation, they are often simply judged as "faulty" or "needs replacement," lacking intermediate levels, which may lead to over-maintenance and waste of human and equipment resources. Summary of the Invention

[0006] To address the shortcomings of existing technologies, the technical problem to be solved by this invention is to provide a switch contact status evaluation system and method based on the fusion of dynamic resistance and vibration. It introduces an environmental correction mechanism to achieve environmental adaptation, adopts a lightweight algorithm to enable real-time operation in embedded devices, and establishes a tiered evaluation result to match specific operation and maintenance strategies.

[0007] To solve the above-mentioned technical problems, the present invention adopts the following technical solution: The first aspect of the present invention provides a switch contact state evaluation system based on the fusion of dynamic resistance and vibration, including a data acquisition unit, a feature extraction module, a dynamic weighted fusion unit and a state output unit; The data acquisition unit includes a dynamic resistance acquisition module, a vibration signal acquisition module, and an environmental parameter acquisition module; the dynamic resistance acquisition module is used to acquire resistance signals and contact travel signals during the opening process; the vibration signal acquisition module is used to acquire vibration signals during the opening process; the environmental parameter acquisition module is used to acquire temperature and humidity data, which are used for subsequent environmental correction. The feature extraction module is used to extract core differential features from the acquired dynamic resistance and vibration signals; The dynamic weighted fusion unit is used for standardizing core differentiated features, removing interference, calculating the State Evaluation Index (SI), and classifying state levels based on the SI value. The status output unit is used to generate an evaluation report, which includes core differentiated feature parameters, SI value, status level, and operation and maintenance recommendations.

[0008] A second aspect of the present invention provides a method for evaluating the state of switch contacts based on the fusion of dynamic resistance and vibration, implemented based on the above-mentioned evaluation system, and comprising the following steps: Step S1, Data Acquisition; Preferably, step S1 includes dynamic resistance acquisition, vibration signal acquisition, and environmental parameter acquisition; The dynamic resistance acquisition adopts the four-wire method, with a test current of 2000A and a sampling rate of 200kHz, and synchronously acquires the resistance signal and contact travel signal during the opening process. The vibration signal acquisition uses a piezoelectric accelerometer, which is installed in the middle of the circuit breaker's gas chamber flange to acquire vibration signals during the opening process at a sampling rate of 2kHz. The environmental parameters are collected, including temperature and humidity data.

[0009] Step S2, Feature extraction; Preferably, in step S2, based on the dynamic resistance and vibration signals acquired in step S1, core differential features are extracted. These core differential features consist of four elements: (1) Arc contact resistance increment ΔRa, where ΔRa is the difference between the average resistance of the arc contact stage during the opening process and the initial resistance of the new contact, reflecting the increase in resistance caused by ablation; (2) Arc contact overtravel difference ΔLa, where ΔLa is the difference between the new arc contact travel and the current arc contact travel, reflecting the contact shortening caused by ablation; (3) The ratio K of the 200Hz amplitude to the 300Hz amplitude in the vibration spectrum will increase during ablation due to the increased contact collision. (4) Coefficient of variation of the proportion of 100Hz frequency in the vibration spectrum S S is the proportion of the 100Hz amplitude to the total amplitude of the entire frequency range, and the coefficient of variation is... The S value is the ratio of the standard deviation to the mean of five consecutive measurements, reflecting the degree of environmental interference.

[0010] coefficient of variation This reflects the temporal instability of signal components—true environmental electromagnetic interference typically manifests as intermittent, non-steady-state intrusions, causing the 100Hz proportion to fluctuate drastically over time, with a high coefficient of variation. The vibration will increase significantly; however, the vibration characteristics caused by faults such as mechanical loosening are relatively stable, with a lower coefficient of variation. Lower.

[0011] Step S3: Feature standardization and interference removal; Preferably, in step S3, the feature standardization process includes: Normalize ΔRa, ΔLa, and K to the interval [0,1], as shown in the following formula: , in, The denominator 120 in the formula represents 120µΩ. The denominator 6 in the formula represents 6mm; the 120µΩ, 6mm and The denominator 0.4 in the formula is set as the severe abnormality threshold for ΔRa, ΔLa, and K, respectively.

[0012] Preferably, in step S3, the interference removal includes: When the When the value exceeds 10%, it is judged as environmental interference, and the vibration signal is re-acquired; when the temperature and humidity parameters exceed the range of "-40~80℃" and / or "0~95% RH", it is judged as an environmental anomaly, and the evaluation results are subsequently corrected.

[0013] Experimental tests were conducted, using the coefficient of variation. With an interference rejection threshold exceeding 10%, the system can effectively distinguish between "random environmental noise" and "persistent mechanical fault characteristics," reducing false alarm rates. Combined with an effective working environment limitation and correction mechanism that includes a temperature range of -40 to 80°C and relative humidity below 95%, the system can ensure stable operation in complex industrial environments such as those with strong electromagnetic fields and signal-to-noise ratios as low as 5dB. It is suitable for substations in different climate zones, such as the low-temperature and dry Northwest, the high-temperature and high-humidity South, and the high-temperature and high-humidity plateau.

[0014] Step S4: Calculate the state evaluation index and classify the state levels; Preferably, in step S4, the contribution of each core differentiated feature to the state evaluation is quantified through Shapley value analysis, and weights are assigned accordingly to calculate the state evaluation index SI, as shown in the following formula: .

[0015] The weighting coefficients of the algorithm were determined using the Shapley value analysis method. By calculating the marginal contribution of each core differentiating feature to the evaluation accuracy after being added to the fusion model, the normalized contribution weights of each core differentiating feature were obtained (ΔRa 40%, ΔLa 30%, K 20%). (Accounting for 10%), while directly reflecting the contact state of the contacts, it effectively corrects for environmental interference by utilizing vibration characteristics.

[0016] Preferably, in step S4, the switch contact status is divided into four levels according to the SI value: "Normal", "Caution", "Abnormal", and "Critical", according to the following rules: (1) Normal: SI ≤ 0.25; (2) Note: 0.25 < SI ≤ 0.5; (3) Abnormal: 0.5 < SI ≤ 0.75; (4) Severe: SI > 0.75.

[0017] The typical core differentiating characteristics for "normal" are: ΔRa≤40µΩ, ΔLa≤2mm, K≤1.0. ≤5%; Typical values ​​for the core differentiating characteristics corresponding to "Note": 40<ΔRa≤80µΩ, 2<ΔLa≤4mm, 1.0<K≤1.2, ≤8%; Typical core differentiating characteristics corresponding to "abnormal" are: 80<ΔRa≤120µΩ, 4<ΔLa≤6mm, 1.2<K≤1.4, ≤10%; Typical core differentiating characteristics for "severe" are: ΔRa>120µΩ, ΔLa>6mm, K>1.4. ≤10%.

[0018] Since SI is obtained by weighted summation of four core differentiated features, there may be complementary values ​​between features. When a single feature value does not conform to the above typical value range, the SI value is used as the standard for level determination to avoid false alarms triggered by instantaneous fluctuations of a certain feature parameter.

[0019] Step S5: Generate an evaluation report.

[0020] Preferably, in step S5, the evaluation report includes core differentiated characteristic parameters, SI value, status level, and operation and maintenance recommendations; the operation and maintenance recommendations are as follows: for "normal" status, routine inspection is recommended; for "attention" status, retesting is recommended within 1 month; for "abnormal" status, contact replacement is recommended within 1 week; and for "serious" status, immediate shutdown is recommended.

[0021] By meticulously classifying the contact condition into four levels and providing corresponding tiered maintenance recommendations, we can effectively avoid over-maintenance caused by "immediate replacement under critical conditions" and eliminate the potential failure risks of "continuing to operate under severe conditions," thus meeting the actual needs of the project.

[0022] In summary, the condition of switch contacts is reflected not only in their electrical performance but also in their mechanical characteristics, and is also influenced by environmental factors. This invention evaluates the condition of switch contacts based on dynamic weighted fusion of multi-dimensional features. The system simultaneously collects dynamic resistance, vibration signals, and environmental parameters, extracts core differential features of dynamic resistance and vibration signals (such as arc contact resistance increment, arc contact overtravel difference, vibration spectrum ratio, etc.), and uses the Shapley value analysis method to determine the contribution (weight) of each core differential feature to the condition evaluation, thereby calculating a comprehensive condition evaluation index (SI). By comparing the SI value with a preset four-level threshold, the system ultimately achieves accurate classification of the contact condition (normal-attention-abnormal-severe) and provides corresponding operation and maintenance suggestions.

[0023] Compared with the prior art, the present invention has the following beneficial effects: (1) This invention introduces an environmental correction mechanism. An environmental parameter acquisition module is set in the data acquisition unit to synchronously collect on-site temperature and humidity data. Clear environmental tolerance indicators are set to ensure that the system can work normally in a wide temperature range of -40 to 80℃ and an environment with relative humidity below 95%. An interference elimination step is also set in the algorithm flow. If the environmental parameters exceed the range or the vibration signal is severely interfered with (such as excessive amplitude fluctuation of 100Hz), the system will automatically determine the abnormality and re-acquire the signal or correct it, instead of directly outputting the wrong result. As a result, the anti-interference capability of the evaluation system is significantly enhanced. Even under strong electromagnetic interference in substations or extreme climatic conditions, the evaluation error can be guaranteed to be no more than 5%, which solves the problem of misjudgment caused by poor environmental adaptability of existing technical methods.

[0024] (2) This invention adopts a dynamic weighted fusion algorithm based on Shapley value analysis, which does not require offline training with massive amounts of data. It only needs to normalize and weight the extracted core differential features (such as arc contact resistance increment, vibration spectrum ratio, etc.). The algorithm logic is simple and clear, with low computational load, and it is very easy to implement in resource-limited embedded monitoring devices to achieve online real-time evaluation. It has strong engineering applicability, greatly reduces the difficulty of on-site debugging and deployment, and has the feasibility of large-scale promotion.

[0025] (3) Based on the range of the obtained State Evaluation Index (SI) values, this invention defines four levels: "Normal," "Attention," "Abnormal," and "Severe" to match a tiered maintenance strategy. Specific characteristic thresholds and corresponding maintenance recommendations are established for each level, avoiding over-maintenance. For example, when the equipment is in a state of slight ablation (e.g., Ra is 40-80µΩ, corresponding to the "Attention" level), the system no longer requires immediate replacement but recommends continued monitoring. This aligns with the actual operating patterns of the equipment (which can typically operate safely for 1-3 months), thus avoiding resource waste caused by a single threshold judgment in existing technologies. Furthermore, maintenance personnel can directly execute the corresponding strategy based on the level output by the system, achieving a closed loop from state evaluation to maintenance decision-making. Attached Figure Description

[0026] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, those skilled in the art can obtain other drawings based on the structures shown in these drawings without creative effort.

[0027] Figure 1 This is a schematic diagram of the vibration signal acquisition device according to an embodiment of the present invention.

[0028] Figure 2 This is a frequency domain diagram of the vibration of the switch contact under different contact states in an embodiment of the present invention.

[0029] Figure 3 This is a schematic diagram of the switch contact condition evaluation method based on the fusion of dynamic resistance and vibration according to the present invention. Detailed Implementation

[0030] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0031] The present invention provides a switch contact status evaluation system based on the fusion of dynamic resistance and vibration, which consists of a data acquisition unit, a feature extraction module, a dynamic weighted fusion unit, and a status output unit.

[0032] The data acquisition unit includes a dynamic resistance acquisition module, a vibration signal acquisition module, and an environmental parameter acquisition module. The dynamic resistance acquisition module uses a four-wire method, a test current of 2000A, and a sampling rate of 200kHz, simultaneously acquiring resistance and contact travel signals during the opening process. The vibration signal acquisition module uses a piezoelectric accelerometer, installed in the middle of the circuit breaker's gas chamber flange, to acquire vibration signals during the opening process, with a sampling rate of 2kHz. The device structure is as follows: Figure 1 As shown; the environmental parameter acquisition module synchronously collects temperature and humidity data for subsequent environmental calibration.

[0033] The feature extraction module extracts four core differential features from the collected dynamic resistance and vibration signals: First, the arc contact resistance increment ΔRa, which is the difference between the average resistance of the arc contact during the opening process and the initial resistance of the new contact, reflecting the resistance increase caused by ablation; second, the arc contact overtravel difference ΔLa, which is the difference between the arc contact stroke of the new contact and the current arc contact stroke, reflecting the contact shortening caused by ablation; and third, the ratio K of the 200Hz amplitude to the 300Hz amplitude in the vibration spectrum. Increased K value occurs during ablation due to intensified contact collisions. The vibration frequency domain diagrams under different contact states are shown below. Figure 2 As shown; fourth is the coefficient of variation of the proportion S of the 100Hz frequency in the vibration spectrum. S is the proportion of the 100Hz amplitude to the total amplitude of the entire frequency range, and the coefficient of variation is... The S value is the ratio of the standard deviation to the mean of five consecutive measurements, reflecting the degree of environmental interference.

[0034] In strong electromagnetic interference scenarios at outdoor substations, directly judging interference based on the magnitude of the S-value can easily lead to misjudgments due to load fluctuations; while the coefficient of variation... This reflects the temporal instability of signal components—real environmental electromagnetic interference typically manifests as intermittent, non-steady-state intrusions, causing the 100Hz frequency proportion to fluctuate drastically over time, with a high coefficient of variation. The vibration will increase significantly; however, the vibration characteristics caused by faults such as mechanical loosening are relatively stable, with a lower coefficient of variation. Lower.

[0035] Based on the aforementioned core differentiating features, this invention can construct a highly discriminative combination of fused features for use in distinguishing and determining the state of switch contacts: The combination of arc contact resistance increment ΔRa and arc contact overtravel difference ΔLa is used to distinguish between ablation and insufficient contact pressure. If ΔRa increases while ΔLa remains unchanged, it is judged as insufficient pressure. If both increase simultaneously, it is judged as ablation. Using the amplitude ratio K of 200Hz / 300Hz and the coefficient of variation of the frequency proportion of 100Hz The combination of these elements is used to distinguish between ablation and environmental disturbance. An increase in K without significant fluctuations indicates environmental interference. The simultaneous increase in K indicates ablation.

[0036] The specificity of the above feature combination is no less than 90%, which improves the accuracy of status assessment from the source.

[0037] The workflow of the dynamic weighted fusion unit is as follows: First, feature standardization is performed, normalizing ΔRa, ΔLa, and K to the [0,1] interval, as shown in the following formula: , Among them, 120µΩ, 6mm, and 0.4 are the severe anomaly thresholds for ΔRa, ΔLa, and K, respectively.

[0038] Next, interference removal is performed. If the vibration rate exceeds 10%, it is considered environmental interference, and the vibration signal is re-acquired. If the temperature and humidity parameters exceed the range of "-40~80℃" or "0~95% RH", it is considered an environmental anomaly.

[0039] Based on laboratory and field measurements, the coefficient of variation was used to determine the optimal values. Using a threshold of over 10% for interference rejection can effectively distinguish between "random environmental noise" and "persistent mechanical fault characteristics," thereby reducing the false alarm rate.

[0040] Subsequently, feature weights were determined based on Shapley value analysis, and the state evaluation index was calculated.

[0041] The algorithm quantifies the contribution of each feature to the state evaluation through Shapley value analysis and assigns weights accordingly. Table 1 shows the quantification results of the contribution of each feature based on Shapley value analysis: Table 1. Quantitative data on the contribution of each feature based on Shapley value.

[0042] As shown in the table above, Shapley value analysis objectively quantifies the independent contribution and synergistic gain of each feature to the accuracy of condition assessment. The resistance increment ΔRa contributes the most (approximately 39.2%), which is consistent with the physical mechanism that contact erosion directly leads to an increase in contact resistance; the arc contact overtravel difference ΔLa is the second most significant (approximately 30.5%), reflecting the independent judgment value of mechanical wear; the two vibration frequency domain features contribute a total of approximately 30%, mainly playing a role in "electromechanical" cross-verification and anti-interference correction.

[0043] Finally, the status levels are divided according to the SI value as follows:

[0044] Since SI is obtained by weighted summation of four core differentiated features, there may be complementary values ​​between features. When a single feature value does not conform to the above typical value range, the SI value is used as the standard for level determination to avoid false alarms triggered by instantaneous fluctuations of a certain parameter.

[0045] The status output unit generates an evaluation report, which includes characteristic parameters, SI values, status levels, and maintenance recommendations. For a "normal" status, routine inspections are recommended; for a "caution" status, retesting within one month is recommended; for an "abnormal" status, contact replacement within one week is recommended; and for a "critical" status, immediate shutdown is recommended.

[0046] The flowchart of the switch contact condition evaluation method based on the fusion of dynamic resistance and vibration of this invention is as follows: Figure 3 As shown.

[0047] To verify the effectiveness of this invention, a simulation verification environment was constructed based on typical circuit breaker parameters and a contact erosion model. This environment encompassed 217 tripping samples, involving 15 circuit breakers with minor erosion faults. Some diagnostic results using the evaluation system and method of this invention are shown in Table 2.

[0048] Table 2. Details of Verification Diagnostic Results

[0049] Of the 217 simulation diagnostic results, 3 were triggered by strong electromagnetic interference. When values ​​exceeded the standard, the system automatically resampled according to the interference rejection mechanism, and all results returned normal, with a false alarm rate of 0. All 15 known faults were correctly identified as "attention" or "abnormal," with a missed alarm rate of 0. This indicates that the present invention has excellent diagnostic accuracy and anti-interference performance in the environment of outdoor substations with strong electromagnetic interference, and possesses good potential for engineering applications.

Claims

1. A switch contact condition evaluation system based on the fusion of dynamic resistance and vibration, characterized in that, It includes a data acquisition unit, a feature extraction module, a dynamic weighted fusion unit, and a status output unit; The data acquisition unit includes a dynamic resistance acquisition module, a vibration signal acquisition module, and an environmental parameter acquisition module; the dynamic resistance acquisition module is used to acquire resistance signals and contact travel signals during the opening process; the vibration signal acquisition module is used to acquire vibration signals during the opening process; and the environmental parameter acquisition module is used to acquire temperature and humidity data. The feature extraction module is used to extract core differential features from the acquired dynamic resistance and vibration signals; The dynamic weighted fusion unit is used for standardization of core differentiated features, interference removal, calculation of the State Evaluation Index (SI), and classification of state levels based on the SI value. The status output unit is used to generate an evaluation report, which includes core differentiated feature parameters, SI value, status level, and operation and maintenance recommendations.

2. A method for evaluating the condition of switch contacts based on the fusion of dynamic resistance and vibration, implemented based on the switch contact condition evaluation system of claim 1, characterized in that, Includes the following steps: Step S1, Data Acquisition; Step S2, Feature extraction; Step S3: Feature standardization and interference removal; Step S4: Calculate the state evaluation index and classify the state levels; Step S5: Generate an evaluation report.

3. The method for evaluating the state of switch contacts according to claim 2, characterized in that, Step S1 includes dynamic resistance acquisition, vibration signal acquisition, and environmental parameter acquisition. The dynamic resistance acquisition adopts the four-wire method, with a test current of 2000A and a sampling rate of 200kHz, and synchronously acquires the resistance signal and contact travel signal during the opening process. The vibration signal acquisition uses a piezoelectric accelerometer, which is installed in the middle of the circuit breaker's gas chamber flange to acquire vibration signals during the opening process at a sampling rate of 2kHz. The environmental parameters are collected, including temperature and humidity data.

4. The method for evaluating the state of switch contacts according to claim 3, characterized in that, In step S2, based on the dynamic resistance and vibration signals acquired in step S1, core differential features are extracted. These core differential features consist of four elements: (1) Arc contact resistance increment ΔRa, where ΔRa is the difference between the average resistance of the arc contact during the opening process and the initial resistance of the new contact; (2) Arc contact overtravel difference ΔLa, where ΔLa is the difference between the new arc contact travel and the current arc contact travel; (3) The ratio K of the 200Hz amplitude to the 300Hz amplitude in the vibration spectrum; (4) Coefficient of variation of the proportion of 100Hz frequency in the vibration spectrum S S is the proportion of the 100Hz amplitude to the total amplitude of the entire frequency range, and the coefficient of variation is... It is the ratio of the standard deviation of the S value obtained from 5 consecutive measurements to its mean.

5. The method for evaluating the state of switch contacts according to claim 4, characterized in that, In step S3, the feature standardization process includes: Normalize ΔRa, ΔLa, and K to the interval [0,1], as shown in the following formula: ， in, The denominator 120 in the formula represents 120µΩ. The denominator 6 in the formula represents 6mm; the 120µΩ, 6mm and The denominator 0.4 in the formula is set as the severe abnormality threshold for ΔRa, ΔLa, and K, respectively.

6. The method for evaluating the state of switch contacts according to claim 5, characterized in that, In step S3, the interference removal includes: When the When the value exceeds 10%, it is judged as environmental interference, and the vibration signal is re-acquired; when the temperature and humidity parameters exceed the range of "-40~80℃" and / or "0~95% RH", it is judged as an environmental anomaly, and the evaluation results are subsequently corrected.

7. The method for evaluating the state of switch contacts according to claim 6, characterized in that, In step S4, the contribution of each core differentiated feature to the state evaluation is quantified through Shapley value analysis, and weights are assigned accordingly. The state evaluation index SI is then calculated using the following formula: 。 8. The method for evaluating the state of switch contacts according to claim 7, characterized in that, In step S4, the switch contact status is divided into four levels based on the SI value: "Normal", "Caution", "Abnormal", and "Critical", according to the following rules: (1) Normal: SI ≤ 0.25; (2) Note: 0.25 < SI ≤ 0.5; (3) Abnormal: 0.5 < SI ≤ 0.75; (4) Severe: SI > 0.

75.

9. The method for evaluating the state of switch contacts according to claim 8, characterized in that, In step S5, the evaluation report includes core differentiated characteristic parameters, SI value, status level, and operation and maintenance recommendations. The operation and maintenance recommendations are as follows: for the "normal" status, routine inspection is recommended; for the "attention" status, retesting is recommended within one month; for the "abnormal" status, contact replacement is recommended within one week; and for the "serious" status, immediate shutdown is recommended.

Images