Static contact ablation loss and deterioration evaluation method influenced by multiple factors

Abstract

The invention provides a multi-factor influence static contact ablation loss and degradation evaluation method, which comprises the following steps: acquiring environmental condition parameters according to arc energy transfer distribution, fusing the environmental condition parameters into a simulation process, and determining a material loss initial rate; if the initial material loss rate exceeds a preset threshold value, analyzing the influence of arcing time on energy transfer through a neural network model, and judging an evaporation splashing proportion; carrying out data fusion on a material melting mechanism by adopting the evaporation splashing proportion to obtain an ablation depth quantized value; obtaining a nonlinear feature adjustment coefficient under the influence of multiple factors through the ablation depth quantized value, and determining a contact fault risk level; according to the contact fault risk level, iteratively optimizing the environmental condition variable by adopting finite element simulation to obtain a material loss accumulation curve; and if the material loss accumulation curve displays an acceleration trend, predicting the residual arcing time bearing capacity through a neural network, and judging the end point of life.

Description

Technical Field

[0001] This invention relates to the field of information technology, and in particular to a method for evaluating the erosion loss and deterioration of stationary contacts affected by multiple factors. Background Technology

[0002] High-voltage switchgear plays a crucial role in power systems, providing control, protection, and isolation circuits. Its operational reliability directly impacts the safety and stability of the entire power grid. Among these components, the stationary contacts, as the core components for conduction and switching, are subject to severe erosion from electric arcs during frequent interruptions of large currents. This leads to continuous wear and tear on the contact surface material, and in severe cases, can even cause poor contact or switch failure. Therefore, the erosion of stationary contacts has always been a key bottleneck restricting equipment lifespan and safe operation.

[0003] Current research and evaluation of static contact erosion largely rely on actual breaking tests or empirical formulas. While these methods can obtain some accurate data, they struggle to comprehensively reflect the erosion patterns under the combined effects of various factors, including different current amplitudes, arcing times, contact material properties, and the surrounding environment. Because the arc erosion process occurs within an extremely short time, with extremely high and rapidly changing temperatures, existing evaluation methods struggle to accurately capture crucial aspects such as how arc energy is distributed to the contact surface, how much energy actually causes material melting and evaporation, and the specific rate of material loss under different conditions. This often leads to significant discrepancies between predicted results and actual operating conditions.

[0004] The complexity of arc ablation lies primarily in two closely related core technical challenges: First, the arc voltage and energy transfer exhibit highly nonlinear characteristics at different current stages, meaning the relationship between energy input and material removal is not a simple proportional one. Second, under the influence of a high-temperature arc, the contact surface undergoes multiple material loss mechanisms simultaneously, including melting, evaporation, and splashing, and the proportions of these mechanisms change rapidly with the current magnitude and arc duration. For example, when interrupting a short-circuit current, the arc energy is concentrated in the first few milliseconds, potentially leading to severe splashing loss in certain areas of the contact. However, near the current zero-crossing point, the energy decreases, and the material is primarily lost through slow evaporation. The alternating action of these two mechanisms results in extremely uneven ablation morphology, with local depth differences reaching several times, directly causing an abnormally high contact resistance and a subsequent decrease in breaking capacity.

[0005] How to accurately establish the quantitative relationship between the ablation depth, mass loss and contact performance degradation of stationary contacts under the combined influence of multiple parameters such as current amplitude, arcing time, material properties and environmental conditions, so as to achieve reliable prediction of contact life, has become a key problem that urgently needs to be solved in the design and operation and maintenance of high-voltage switchgear. Summary of the Invention

[0006] This invention provides a method for evaluating the ablation loss and degradation of stationary contacts influenced by multiple factors, mainly including: By collecting current amplitude and arcing time data, the nonlinear characteristics under the influence of multiple factors are processed using the finite element simulation method to obtain the arc energy transfer distribution. Based on the distribution of electric arc energy transfer, environmental condition parameters are obtained and incorporated into the simulation process to determine the initial rate of material loss. If the initial rate of material loss exceeds a preset threshold, the impact of arcing time on energy transfer is analyzed through a neural network model to determine the evaporation and splashing ratio. Using the aforementioned evaporation-splashing ratio, data fusion is performed on the material melting mechanism to obtain a quantitative value of ablation depth; By quantifying the ablation depth, the nonlinear characteristic adjustment coefficient under the influence of multiple factors is obtained, and the contact failure risk level is determined. Based on the contact failure risk level, the environmental condition variables are iteratively optimized using finite element simulation to obtain the material loss cumulative curve; If the material loss accumulation curve shows an accelerating trend, the remaining arc time tolerance can be predicted using a neural network to determine the end of the lifespan.

[0007] The technical solutions provided by the embodiments of the present invention may include the following beneficial effects: This invention discloses a method for assessing material loss and contact failure risk based on the influence of multiple factors. By collecting current amplitude and arcing time data, the method employs finite element simulation to process nonlinear characteristics, obtains the arc energy transfer distribution, and determines the initial material loss rate by combining environmental condition parameters. When the rate exceeds the limit, a neural network model is used to analyze the impact of arcing time on energy transfer, determine the evaporation and splashing ratio, and then quantifies the ablation depth through data fusion technology to derive the nonlinear characteristic adjustment coefficient, ultimately assessing the contact failure risk level. Simultaneously, this invention generates a material loss accumulation curve by iteratively optimizing environmental variables and predicts the remaining arcing time tolerance under accelerating trends to determine the end of the equipment's lifespan. The core innovation of this invention lies in combining multi-factor nonlinear characteristics with neural network prediction, achieving accurate assessment of the entire chain from energy transfer to material loss and failure risk, providing a scientific basis for equipment lifespan management and fault prevention. Attached Figure Description

[0008] Figure 1 This is a flowchart of a method for evaluating the erosion loss and degradation of stationary contacts under multiple factors according to the present invention.

[0009] Figure 2 This is a schematic diagram of a method for evaluating the erosion loss and deterioration of stationary contacts under multiple factors according to the present invention.

[0010] Figure 3 This is another schematic diagram of a method for evaluating the erosion loss and deterioration of stationary contacts under the influence of multiple factors according to the present invention. Detailed Implementation

[0011] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0012] like Figures 1-3 This embodiment of a method for evaluating the ablation loss and degradation of stationary contacts influenced by multiple factors may specifically include: S101. By collecting current amplitude and arcing time data, the nonlinear characteristics under the influence of multiple factors are processed using the finite element simulation method to obtain the arc energy transfer distribution.

[0013] Data on current amplitude and arcing time are collected. A finite element model is established based on the collected data. The finite element model is solved using the finite element simulation method to obtain the nonlinear characteristics under multiple factors. The nonlinear characteristics under multiple factors are processed using the finite element simulation method to calculate the arc energy transfer process. Data on the arc energy transfer process is obtained. The energy transfer distribution is determined based on the arc energy transfer process data. The energy transfer distribution results corresponding to the current amplitude and arcing time are extracted from the energy transfer distribution.

[0014] First, the instantaneous current value during the welding process is acquired in real time using a high-frequency current transformer (such as a Rogowski coil, bandwidth 1MHz, sampling rate 10MS / s), with data points recorded every 0.1ms. A peak detection algorithm (based on local extremum judgment with a sliding window size of 50 sampling points combined with a threshold of 0.5 × average current) is used to extract the peak arc current (typically 450A~520A) during each short-circuit transition or spray transition. Simultaneously, the arc ignition and extinguishing times are accurately marked using the voltage zero-point crossing method, and the single arc duration is calculated (typically between 0.8ms and 3.5ms). Next, the 3000 sets of arc current amplitude I_arc and corresponding arc duration t_arc data are imported into MATLAB, and a three-dimensional interpolation algorithm (scatterInterpolation) is used. The joint distribution surface of I_arc-t_arc frequencies was constructed using the LANT method (Natural Neighborhood Method), and the nonlinear mapping function of arc energy E(t,I)=k×I^2×t×(1+α×sin(π×t / T_cycle)) was obtained by KDE kernel density estimation (Gaussian kernel, bandwidth 0.12), where k=0.98 and α=0.27 are multi-factor coupling correction coefficients. Subsequently, a three-dimensional transient thermo-electro-magnetic multiphysics coupling model was established in ANSYS Workbench, with hexahedral dominant meshing (arc region refined to 0.15mm). The aforementioned energy distribution function was loaded as a volume heat source loading term (Gaussian distributed heat source radius 1.2mm, peak heat flux density dynamically changes with E(t,I), reaching a maximum of 3.8×10^7). The material properties were considered for temperature nonlinearity (the thermal conductivity of 45 steel decreased from 51 W / m·K to 28 W / m·K with temperature). A transient solution was performed for 600 time steps (total duration 2.4 s) to obtain the temperature field, current density vector field, and electromagnetic force distribution of the molten pool. Finally, the volume change curve of the isothermal surface of the molten pool (above 1350℃) was extracted with time through post-processing, and the efficiency of arc energy transfer to the workpiece (actual melting enthalpy increment / total arc input energy) was calculated. Typical results show that under the conditions of 450A and 2.1ms arcing, the energy transfer efficiency reaches 78.6%, and the energy distribution has a bimodal characteristic, with the main peak located 0.8 mm below the arc central axis, thus realizing the accurate quantification of the arc energy transfer distribution under the influence of multiple factors.

[0015] S102. Based on the distribution of electric arc energy transfer, environmental condition parameters are obtained and incorporated into the simulation process to determine the initial rate of material loss.

[0016] By using arc energy data and transmission distribution information, a preliminary simulation framework under environmental conditions is constructed to obtain the distribution state of energy influence. Based on the distribution state of energy influence, environmental condition parameters are incorporated, and the simulation process is dynamically adjusted to determine the energy range under changing conditions. For the energy range under changing conditions, a pre-established simulation process is used to analyze the material loss response and derive the loss assessment results. From the loss assessment results, key influencing factors of material loss are identified, and the formation mechanism of the initial rate is determined using data analysis methods. If the formation mechanism of the initial rate deviates from the distribution state of energy influence, the intermediate value of the loss assessment is recalculated through data correction for changing conditions, resulting in an adjusted rate derivation. Based on the adjusted rate derivation results and the environmental condition parameters in the simulation process, the long-term trend of material loss is analyzed to determine the final initial rate value. By correlating the data on energy influence and changing conditions with the final initial rate value, a complete record of the loss assessment is constructed, and comprehensive analytical conclusions are derived.

[0017] The gas composition, temperature, airflow velocity, and humidity parameters at the welding site were collected in real time using a multi-sensor array. The shielding gas flow rate was set to 18 L / min. An infrared thermometer was used to monitor the temperature field around the arc at a frequency of 100 Hz (range 300℃~1200℃). Simultaneously, a thermocouple array and an anemometer were used to acquire local airflow disturbances (typical values ​​0.4 m / s~1.8 m / s) and relative humidity (variing range 35%~78%) on the workpiece surface. These environmental variables were then synchronized and timestamped with previously obtained arc energy transfer distribution data. Subsequently... In a Python environment, a modified model of the effect of environmental conditions on energy transfer efficiency was constructed using multivariate regression analysis combined with principal component analysis (PCA to reduce dimensionality to the first four principal components, explaining 91.7% of the variance). The resulting environmental impact factor modification function η_env = 1 - β1 × (T_env - 25) + β2 × (v_gas - 0.8)^2 - β3 × (RH / 100), where β1 = 0.0032, β2 = 0.014, and β3 = 0.087 are the fitting coefficients. The modified arc energy function was then imported into COMSOL. Multiphysics was used to establish a two-dimensional axisymmetric transient heat conduction-fluid coupling model that includes environmental convection and radiation heat dissipation boundaries. The mesh was adaptively refined (the smallest element at the arc root was 0.08 mm). The boundary conditions were set with a convective heat transfer coefficient of h = 12 + 6.5 × v_gas W / m²·K and a piecewise linear variation of radiative emissivity according to temperature (0.42 below 800℃ and 0.68 above 1600℃). A dynamic volumetric heat source was applied, and an energy loss term caused by the environment was superimposed. Through a transient solution with a duration of 600 ms, the temperature gradient and latent heat absorption rate of the material surface in the initial stage were extracted. The initial melting loss rate of 45 steel under the current environment was calculated (typical values ​​are 0.021 g / s to 0.037 g / s under the conditions of ambient temperature 38℃, airflow 1.2 m / s, and humidity 62%). The nonlinear dependence between the loss rate and environmental factors was analyzed, thus realizing the fully automated process of integrating environmental condition parameters into the arc energy transfer simulation and quantifying the initial loss rate of the material.

[0018] S103. If the initial rate of material loss exceeds the preset threshold, the influence of arcing time on energy transfer is analyzed through a neural network model to determine the evaporation and splashing ratio.

[0019] Acquire the arc duration sequence and corresponding energy transfer ratio data. Based on the arc duration sequence, calculate the energy distribution ratio for each time period to obtain the energy distribution state sequence. For the energy distribution state sequence, obtain the current environmental medium parameters and determine the correction coefficient for energy attenuation. Adjust the energy distribution state sequence using the correction coefficient to obtain the action intensity sequence under constraint conditions. Process the action intensity sequence using a pre-configured loss response model to analyze the material surface degradation law and obtain the initial loss rate value. Separate the contributions of the evaporation mechanism and the splashing mechanism from the initial loss rate value and calculate the contribution ratio of the two mechanisms. If the contribution ratio shows that the splashing mechanism ratio exceeds a preset threshold, input the arc duration sequence and energy transfer ratio data into the neural network model and output the adjusted evaporation-splashing ratio. Based on the adjusted evaporation-splashing ratio and environmental medium parameters, apply a weighted correction to the initial loss rate value to obtain the final initial loss rate value.

[0020] If the initial rate of material melting loss exceeds a preset threshold of 0.040 g / s, a time-series analysis process based on a long short-term memory network is automatically triggered. First, current and voltage waveform data and corresponding total energy input are extracted from the welding database at 10 ms intervals within 3000 consecutive arc cycles. This data, along with the synchronously acquired arc spectral radiation intensity sequence, forms a multi-channel input feature. Then, a neural network model containing two LSTM units (128 and 64 hidden nodes respectively) is constructed, using the Adam optimizer (learning rate 0.001, decay rate 0.95), with a batch size of 32, training epochs limited to 150 epochs, and an early stopping mechanism enabled (patience value 12). The loss function is a combination of mean squared error and L2 regularization (regularization coefficient 0.0005). After model training, the current arc duration sequence (typically within the range of 0.2) is input. The model outputs predicted values ​​of the energy distribution ratio among melting, evaporation, and splashing mechanisms at each time point (s~2.8s). For 45 steel material under the conditions of peak current 220A and arcing time 1.4s, the model calculates that the evaporation ratio is approximately 27.3%~34.6%, the splashing ratio is approximately 18.9%~26.2%, and the remaining part is mainly used for molten pool heating and heat conduction. Furthermore, the contribution of key time steps to the evaporation and splashing ratio is visualized through a gradient weighted activation mapping algorithm. It is found that the influence weight of the splashing ratio in the 80ms~160ms interval after the current rise edge is as high as 0.41. Subsequently, the ratio results are correlated with the previous loss rate threshold judgment results to form a closed loop feedback. If the total evaporation and splashing ratio exceeds 48%, process parameter adjustment suggestions are automatically generated and recorded in the database, thereby realizing the quantitative judgment and optimization of the impact of arcing time on the energy transfer mechanism.

[0021] S104. Using the evaporation and splashing ratio, data fusion is performed on the material melting mechanism to obtain a quantitative value of ablation depth.

[0022] The process involves acquiring the melt layer thickness sequence and the evaporation / splashing ratio sequence during the arcing process. Based on the melt layer thickness sequence, the rate of change of the molten pool volume at each time interval is calculated, yielding the dynamic state sequence of the molten pool. The current arc power density distribution is obtained from the dynamic state sequence, and the amplification factor of power density on the molten pool volume is determined. This amplification factor is used to correct the dynamic state sequence, resulting in a corrected melt volume sequence. The corrected melt volume sequence is processed using a depth accumulation model to analyze the layer-by-layer material removal pattern, obtaining the initial ablation depth value. The contributions of melting removal and evaporation removal are separated from the initial ablation depth value, and their proportions are calculated. If the evaporation removal proportion exceeds a preset threshold, the melt layer thickness sequence and the evaporation / splashing ratio sequence are input into a neural network model, which outputs a corrected removal proportion. Based on the corrected removal proportion and the arc power density distribution, the initial ablation depth value is updated with weights to obtain the final quantified ablation depth value.

[0023] If the initial melting loss rate of the material exceeds a preset threshold of 0.040 g / s, a quantitative analysis process for ablation depth targeting the melting mechanism is initiated. First, real-time data on the surface temperature distribution of the molten pool, the dynamic change sequence of the molten pool width, and the synchronous real-time feedback value of the laser power are acquired every 5 ms after the start of arc combustion from a multi-sensor fusion system. This data is then matched with the temperature-dependent curves of the specific heat capacity, thermal conductivity, and density of 45 steel in the material thermophysical parameter library to form a three-dimensional feature tensor containing temperature gradient, molten pool volume estimation, and energy accumulation. Subsequently, a one-dimensional convolutional neural network is used... A hybrid model combining network and attention mechanism is used for deep feature extraction. The first convolutional kernel size is 7×1 with 64 channels, and the second convolutional kernel size is 5×1 with 128 channels. Then, a multi-head self-attention layer (4 heads, 256 dimensions) is connected. The contribution of each time step to the ablation depth is fused by softmax weighting. The model uses the RMSprop optimizer (learning rate 0.0008, weight decay 0.0002), the batch size is set to 16, and a dropout ratio of 0.15 is applied during training. The validation loss is allowed to remain unchanged for 8 consecutive rounds. Early termination is triggered, and the Huber loss function (δ=1.0) is used to improve robustness to abnormal depths. During the model inference phase, the complete temporal characteristics of the current arcing process are input, and the instantaneous ablation depth increments from the start to the end of the arc are output and accumulated to obtain the final quantified ablation depth value. For 45 steel under typical conditions of an average power of 3800W and an arcing duration of 1.6s, the predicted ablation depth range is 1.12mm to 1.47mm. Furthermore, the SHAP value analysis method is used to quantify the contribution of each input feature to the depth prediction, revealing that the highest temperature at the center of the molten pool... The duration of heat exceeding 2150K has the greatest impact on ablation depth, contributing 0.37, followed by the peak range of melt pool width expansion rate (approximately 0.9s to 1.3s), contributing 0.29. Finally, the quantified ablation depth results are jointly verified with the previously predicted evaporation and splashing ratio. If the ablation depth exceeds 1.35mm and the total evaporation and splashing ratio is higher than 45%, the power attenuation coefficient for the next cycle (recommended range 0.92 to 0.97) is automatically calculated and updated, and written into the process optimization log database to achieve precise closed-loop control of energy distribution in the melting mechanism.

[0024] S105. By quantifying the ablation depth, obtain the nonlinear characteristic adjustment coefficient under the influence of multiple factors to determine the contact failure risk level.

[0025] The process involves acquiring the material surface temperature sequence and material removal rate sequence during the electric arc process. The cumulative heat input for each time period is calculated based on the temperature sequence to obtain a cumulative heat sequence. Material thermophysical parameters are obtained from the cumulative heat sequence, and correction factors for these parameters on the removal rate are determined. The material removal rate sequence is adjusted using these correction factors to obtain an adjusted removal rate sequence. The adjusted removal rate sequence is processed using a layer-by-layer cumulative calculation model to obtain a preliminary removal depth value. The thermal melting removal portion and the vaporization removal portion are separated from the preliminary removal depth value to obtain the proportion of vaporization removal. If the proportion of vaporization removal exceeds a preset ratio, the temperature sequence and removal rate sequence are input into a support vector regression model to obtain an adjusted removal contribution ratio. Based on the adjusted removal contribution ratio and the arc heat flux density distribution, a proportional weighted correction is applied to the preliminary removal depth value to obtain a final quantified ablation depth value. A nonlinear correction coefficient under multi-factor coupling is calculated based on the final quantified ablation depth value to obtain the contact failure risk level.

[0026] After obtaining the multi-factor nonlinear influence characteristics through the ablation depth quantification value, the system automatically extracts key variables such as power fluctuation amplitude, estimated value of molten pool convection intensity, and material surface tension temperature coefficient during the arcing process, forming a comprehensive influence factor vector with 12 dimensions. Subsequently, a gradient boosting decision tree algorithm combined with a target encoding method is used to perform nonlinear mapping on each factor. The tree depth is limited to 7, the learning rate is set to 0.065, the minimum number of leaf node samples per tree is 9, the subsampling ratio is 0.78, and the regularization parameters lambda and gamma are 1.2 and 0.3, respectively. The model uses the ablation depth quantification value as the target and employs 5-fold cross-validation training with squared error loss, obtaining adjustment coefficient prediction values ​​typically between 0.84 and 1.19. During inference, the real-time feature vector of the current operating condition is input, and the nonlinear feature adjustment coefficient is output, for example, the instantaneous peak value of the laser power. Under the conditions of reaching 4350W and the molten pool convection velocity exceeding 0.38m / s, the adjustment coefficient is 1.07. Then, based on this coefficient, the original contact resistance model is corrected, and the percentage deviation between the adjusted contact resistance value and the standard value is calculated. Combined with the failure probability distribution of similar operating conditions in the historical failure case library, the contact failure risk level is calculated using logistic regression. The risk score formula is 1 / (1+exp(-(β0+β1×adjustment coefficient+β2×deviation percentage+β3×cumulative energy density))), where β0=-4.82, β1=3.41, β2=0.096, and β3=0.00074. When the risk score exceeds 0.62, it is judged as high risk. The system automatically marks this operating condition and generates a warning signal. Simultaneously, the adjustment coefficient and risk score are recorded in the time-series database for adaptive iterative optimization of process parameters in subsequent batches.

[0027] S106. Based on the contact failure risk level, the environmental condition variables are iteratively optimized using finite element simulation to obtain the material loss cumulative curve.

[0028] Obtain the contact failure risk level and material loss rate sequence. Calculate the cumulative loss for each time period based on the material loss rate sequence to obtain the cumulative loss sequence. Extract the corresponding environmental condition sequence from the cumulative loss sequence to obtain the environmental condition sequence. Use finite element simulation to input the cumulative loss sequence and environmental condition sequence to calculate the material loss distribution under the current conditions, obtaining a preliminary loss distribution result. Determine whether the location of the maximum loss in the preliminary loss distribution result exceeds a preset threshold. If it does, extract the current key control variable and adjust its value range to obtain the adjusted key control variable. Update the environmental condition sequence using the adjusted key control variable to obtain the updated environmental condition sequence. Use finite element simulation again to input the updated environmental condition sequence and cumulative loss sequence to recalculate the material loss distribution, obtaining the optimized material loss cumulative curve.

[0029] Based on the high-risk assessment result of a contact failure risk level exceeding 0.62, the system automatically triggers the finite element simulation module, using the current operating parameters as initial boundary conditions. These include an electrical contact pressure set to 12.5N, a constant ambient temperature of 85℃, a peak current density of 8.7×10^6 A / m², and a calculated risk score of 0.68. The finite element model employs a thermo-electrical-structural multiphysics coupling method, using tetrahedral elements for mesh generation, with the total number of elements controlled to approximately 420,000. Local mesh refinement in the contact area is reduced to a minimum mesh size of 0.008mm. Material properties exhibit nonlinear temperature variations; the thermal conductivity of the copper alloy is assigned a piecewise function within the 300K to 1200K range, and the resistivity temperature coefficient is [value missing]. The iterative optimization process takes minimizing material volume loss as the objective function and employs a sequential quadratic programming algorithm. Environmental variables are updated in each iteration: First, the current density is fixed, and 36 simulations are performed by adjusting the relative humidity from 35% to 92% in a step size of 0.8%. The cumulative ablation volume is recorded within 1000 discharge cycles after each simulation. Then, the optimal humidity is fixed at 92%, and the ambient air pressure is optimized in a step size of 0.02 atm within the range of 0.65 atm to 1.15 atm, resulting in the lowest loss at an air pressure of 0.89 atm. Finally, the results from the two stages were combined to generate a material loss cumulative curve. This curve shows that the loss increases approximately linearly before the cumulative power-on time reaches 4200 hours, with a slope of 0.00147 mm³ / h. After that, it enters an accelerated degradation stage, and the cumulative loss volume reaches 0.875 mm³ at 4500 hours, corresponding to a decrease in contact surface height of about 17.3 μm. At this point, the system automatically outputs optimized environmental control recommendations: maintain relative humidity of 90%±2% and air pressure of 0.88atm±0.03atm. The parameterized equation of this cumulative curve is stored in the knowledge base for life prediction correction in the next batch of contact component design stage.

[0030] S107. If the material loss accumulation curve shows an accelerating trend, the remaining arc time tolerance is predicted by a neural network to determine the end of the lifespan.

[0031] Obtain the material loss rate sequence and the arcing time sequence. Calculate the cumulative value segmented by the material loss rate sequence to obtain the material loss cumulative sequence. Extract the corresponding arcing time sequence from the material loss cumulative sequence to obtain the arcing time sequence. Use a neural network to input the material loss cumulative sequence and the arcing time sequence to obtain the predicted remaining arcing time value. If the predicted remaining arcing time value is lower than the lifetime threshold, extract the positions in the material loss cumulative sequence where the rate of increase increases to obtain the acceleration segment position sequence. Locate the corresponding part in the material loss cumulative sequence using the acceleration segment position sequence to obtain the updated material loss cumulative sequence. Use a neural network to input the updated material loss cumulative sequence and the arcing time sequence to obtain the optimized lifetime end time.

[0032] Based on a warning threshold of 0.71 for contact fault risk level, the system automatically activates the neural network prediction module. The input feature vector uses the current cumulative energization time of 3780 hours and the loss volume of 0.592 mm³ calculated in the previous finite element method. Simultaneously, it incorporates the real-time monitored peak arc energy density of 6.4 × 10⁵ J / m², average arc duration of 4.2 ms, and the contact resistance increment trend slope of 0.018 mΩ / h. A hybrid model combining a Long Short-Term Memory (LSTM) network and an attention mechanism is employed, with 128 hidden layer units and 8 attention heads. During training, 3200 sets of accelerated aging experimental data from previous batches are used as the dataset. The loss function is a combination of mean squared error and L1 regularization. The optimizer uses AdamW with an initial learning rate of 0.0008, decaying using a cosine annealing strategy. After 120 epochs of convergence, the model predicts a remaining arcing time of approximately 820 hours, corresponding to a predicted total lifespan end of 4600 hours. At this point, the accumulated loss volume will reach 1.128 mm³, and the contact surface height will decrease by approximately 22.6 μm. The system further extracts the second derivative change points in the latter part of the prediction curve, revealing a significant increase in degradation acceleration within the remaining arcing time range of less than 280 hours. The slope jumps from 0.0021 mm³ / h to 0.0078 mm³ / h, indicating the onset of irreversible accelerated failure. Subsequently, a lifespan end warning report is automatically generated, storing the parameterized results of the predicted remaining arcing time of 820 hours and the confidence interval ±65 hours in the database. This data is then linked and updated to the full lifecycle health records of the same model of contact components for subsequent production batch reliability design iterations and preventative maintenance strategy adjustments.

[0033] If the technical solution of this application involves the collection, storage, use, processing, transmission, provision, disclosure, or deletion of personal information, the products using this technical solution have clearly and understandably informed the users of the personal information processing rules before processing personal information, and have obtained the individuals' voluntary consent in accordance with the law. If the technical solution of this application involves sensitive personal information (such as biometrics, religious beliefs, specific identities, medical and health information, financial accounts, and location tracking), the products using this solution have obtained the individuals' separate consent before processing sensitive personal information, and have also met the requirement of "express consent," ensuring that individuals make authorization decisions voluntarily based on full knowledge.

[0034] Specific implementation methods include, but are not limited to, the following: setting up clear and prominent signs at personal information collection devices such as cameras and sensors to inform relevant personnel that they have entered the scope of personal information collection and that their personal information will be collected and processed. If an individual voluntarily enters the collection scope after being informed, it is deemed that they have agreed to the collection of their personal information; or using obvious icons, text descriptions, or other means on the terminal device or system interface for personal information processing to inform them of the rules for personal information processing, and obtaining the individual's explicit authorization through interactive methods such as pop-up prompts, check confirmation boxes, or asking the individual to upload their personal information themselves.

[0035] The aforementioned personal information processing rules should include, but are not limited to, the name and contact information of the personal information processor, the specific purpose of personal information processing, the processing method, the types of personal information processed, the retention period, and the methods and procedures for individuals to exercise their relevant rights.

[0036] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A method for evaluating the erosion loss and deterioration of stationary contacts influenced by multiple factors, characterized in that, The method includes: By collecting current amplitude and arcing time data, the nonlinear characteristics under the influence of multiple factors are processed using the finite element simulation method to obtain the arc energy transfer distribution. Based on the distribution of electric arc energy transfer, environmental condition parameters are obtained and incorporated into the simulation process to determine the initial rate of material loss. If the initial rate of material loss exceeds a preset threshold, the impact of arcing time on energy transfer is analyzed through a neural network model to determine the evaporation and splashing ratio. Using the aforementioned evaporation-splashing ratio, data fusion is performed on the material melting mechanism to obtain a quantitative value of ablation depth; By quantifying the ablation depth, the nonlinear characteristic adjustment coefficient under the influence of multiple factors is obtained, and the contact failure risk level is determined. Based on the contact failure risk level, the environmental condition variables are iteratively optimized using finite element simulation to obtain the material loss cumulative curve; If the material loss accumulation curve shows an accelerating trend, the remaining arc time tolerance can be predicted using a neural network to determine the end of the lifespan.

2. The method for evaluating the ablation loss and deterioration of stationary contacts under multiple factors according to claim 1, characterized in that, The process involves collecting current amplitude and arcing time data, and using the finite element method to process the nonlinear characteristics under the influence of multiple factors to obtain the arc energy transfer distribution, including: Collect current amplitude and arcing time data; A finite element model was established based on the collected current amplitude and arcing time data; The finite element model was solved using the finite element simulation method to obtain the nonlinear characteristics under the influence of multiple factors; The nonlinear characteristics under multiple factors obtained by finite element simulation are used to calculate the energy transfer process of electric arc. Acquire data on the energy transfer process of electric arc; Determine the energy transfer distribution based on data from the electric arc energy transfer process; Extract the energy transfer distribution results corresponding to the current amplitude and arcing time from the energy transfer distribution.

3. The method for evaluating the erosion loss and deterioration of stationary contacts under multiple factors according to claim 1, characterized in that, The step of obtaining environmental condition parameters based on the arc energy transfer distribution and incorporating them into the simulation process to determine the initial rate of material loss includes: By combining electric arc energy data with transmission distribution information, a preliminary simulation framework under environmental conditions is constructed to obtain the distribution state of energy influence. Based on the distribution of energy impact, environmental condition parameters are incorporated to dynamically adjust the simulation process and determine the energy impact range under changing conditions. To analyze the response of material loss under varying conditions and the range of energy action, a pre-established simulation process was used to derive the loss assessment results. From the loss assessment results, we can identify the key influencing factors of material loss and, combined with data analysis methods, determine the formation mechanism of the initial rate. If there is a discrepancy between the formation mechanism of the initial rate and the distribution of energy influence, the intermediate value of the loss assessment is recalculated through data correction of condition changes to obtain the adjusted rate derivation result. Based on the adjusted rate derivation results and combined with the environmental condition parameters in the simulation process, the long-term trend of material loss is analyzed to determine the final initial rate value. By using the final initial rate value, correlated with data on energy effects and condition changes, a complete record of loss assessment is constructed, leading to comprehensive analytical conclusions.

4. The method for evaluating the ablation loss and deterioration of stationary contacts under multiple factors according to claim 1, characterized in that, If the initial rate of material loss exceeds a preset threshold, the influence of arcing time on energy transfer is analyzed using a neural network model to determine the evaporation and splashing ratio, including: Obtain the arc burning duration sequence and the corresponding energy transfer ratio data; Based on the arc duration sequence, the energy distribution ratio of each time period is calculated to obtain the energy distribution state sequence; For the energy distribution state sequence, obtain the current environmental medium parameters and determine the correction coefficient of the medium for energy attenuation; The energy distribution state sequence is adjusted using a correction coefficient to obtain the action intensity sequence under constraint conditions; By processing the action intensity sequence through a pre-configured loss response model, the degradation law of the material surface is analyzed, and a preliminary loss rate value is obtained. Separate the contributions of the evaporation mechanism and the splashing mechanism from the initial loss rate values, and calculate the contribution ratio of the two mechanisms; If the contribution ratio shows that the proportion of the splashing mechanism exceeds the preset threshold, the arc duration sequence and energy transfer ratio data are input into the neural network model, and the adjusted evaporation splashing ratio is output. Based on the adjusted evaporation-splash ratio and environmental medium parameters, the initial loss rate value is weighted and corrected to obtain the final initial loss rate value.

5. The method for evaluating the erosion loss and deterioration of stationary contacts under multiple factors according to claim 1, characterized in that, The process of using the evaporation-splashing ratio to perform data fusion on the material melting mechanism to obtain a quantitative value of ablation depth includes: Obtain the sequence of melt layer thickness and the sequence of evaporation and splashing ratio during the electric arc process; The rate of change of molten pool volume in each time period is calculated based on the molten layer thickness sequence to obtain the dynamic state sequence of the molten pool; The current arc power density distribution is obtained from the dynamic state sequence of the molten pool, and the amplification factor of the power density on the molten pool volume is determined. The dynamic state sequence of the molten pool is corrected by using an amplification factor to obtain the corrected molten volume sequence; By processing the corrected melt volume sequence using a depth accumulation model, the layer-by-layer removal pattern of the material is analyzed to obtain the preliminary ablation depth value. Separate the contributions of melting removal and evaporation removal from the initial ablation depth values, and calculate the proportion of the two removal methods; If the evaporation removal ratio exceeds the preset threshold, the neural network model is input with the melt layer thickness sequence and the evaporation splash ratio sequence, and the corrected removal ratio is output. Based on the corrected removal ratio and arc power density distribution, the initial ablation depth value is updated with weights to obtain the final ablation depth quantification value.

6. The method for evaluating the ablation loss and deterioration of stationary contacts under multiple factors according to claim 1, characterized in that, The process of obtaining nonlinear characteristic adjustment coefficients under the influence of multiple factors through ablation depth quantification to determine the contact failure risk level includes: Obtain the material surface temperature sequence and material removal rate sequence during the electric arc process; The cumulative heat input for each time period is calculated based on the temperature sequence to obtain the cumulative heat input sequence. The thermal properties of the material are obtained from the thermal accumulation sequence, and the correction factor of the thermal properties on the removal rate is determined. The material removal rate sequence was adjusted using a correction factor to obtain the adjusted removal rate sequence; The adjusted removal rate sequence is processed by a layer-by-layer cumulative calculation model to obtain the initial removal depth value; The hot melt removal portion and the vaporization removal portion are separated from the initial removal depth value to obtain the proportion of the vaporization removal portion; If the proportion of vaporization removal exceeds the preset proportion, the temperature sequence and removal rate sequence are input into the support vector regression model to obtain the adjusted removal contribution ratio. Based on the adjusted removal contribution ratio and arc heat flux density distribution, the initial removal depth value is proportionally weighted and corrected to obtain the final ablation depth quantification value. The nonlinear correction coefficient under multi-factor coupling is calculated based on the final ablation depth quantification value to obtain the contact failure risk level.

7. The method for evaluating the ablation loss and deterioration of stationary contacts under multiple factors according to claim 1, characterized in that, The method involves using finite element simulation to iteratively optimize environmental condition variables based on the contact failure risk level, resulting in a material loss accumulation curve, including: Obtain the sequence of contact failure risk level and material loss rate; The cumulative loss for each time period is calculated based on the material loss rate sequence to obtain the cumulative loss sequence; The environmental condition sequence is obtained by extracting the environmental condition sequence at the corresponding time from the loss accumulation sequence; By using finite element simulation to input the cumulative loss sequence and the environmental condition sequence, the material loss distribution under the current conditions is calculated, and preliminary loss distribution results are obtained. Determine whether the location of the maximum loss in the preliminary loss distribution results exceeds the preset threshold. If it does, extract the current key control variable and adjust its value range to obtain the adjusted key control variable. The updated environmental condition sequence is obtained by updating the key control variables after adjustment. By re-inputting the updated environmental condition sequence and loss accumulation sequence using finite element simulation, the material loss distribution is recalculated to obtain the optimized material loss accumulation curve.

8. The method for evaluating the ablation loss and deterioration of stationary contacts under multiple factors according to claim 1, characterized in that, If the material loss accumulation curve shows an accelerating trend, the remaining arcing time tolerance is predicted using a neural network to determine the end of the lifespan, including: Obtain material loss rate sequences and arcing time sequences; The cumulative value is calculated segment by segment for the material loss rate sequence to obtain the material loss cumulative sequence; The arcing time series is obtained by extracting the corresponding arcing time series from the material loss accumulation sequence; The predicted remaining arcing time is obtained by using a neural network to input the material loss accumulation sequence and the arcing time sequence. If the predicted remaining arc time is lower than the lifetime threshold, the positions where the rate of increase increases in the material loss accumulation sequence are extracted to obtain the acceleration segment position sequence. The updated material loss accumulation sequence is obtained by locating the corresponding part in the material loss accumulation sequence using the acceleration section position sequence; The optimized lifespan endpoint time is obtained by using the updated material loss accumulation sequence and arcing time sequence as inputs to a neural network.

Images