Circuit breaker service life prediction system based on machine learning

Abstract

The invention relates to the technical field of circuit breaker service life prediction systems, in particular to a circuit breaker service life prediction system based on machine learning, which comprises a circuit breaker body analysis module, an environment analysis module, a control module, a performance judgment module and a communication module, the circuit breaker body analysis module is used for analyzing and obtaining related information of the circuit breaker body; the environment analysis module is used for analyzing and obtaining related information of the environment; the control module obtains the real-time health index of the circuit breaker according to the related information of the circuit breaker body and the related information of the environment; the performance judgment module obtains information about whether the performance of the circuit breaker is good or poor according to the real-time health index of the circuit breaker and transmits the information about whether the performance of the circuit breaker is good or poor to the communication module; and the communication module transmits the good or poor performance information of the circuit breaker to a user side. The operation state of the circuit breaker is comprehensively monitored, an accurate data basis is provided for subsequent health assessment, and the practicability is high.

Description

Technical Field

[0001] The present invention relates to the technical field of circuit breaker life prediction systems, and in particular to a circuit breaker life prediction system based on machine learning. Background Art

[0002] Circuit breakers are "smart insurance" in the power system. They can not only switch circuits, but also automatically cut off power for protection. They are the core guarantee equipment for electrical safety.

[0003] The application document with publication number CN116413589A discloses a method and device for predicting the electrical life of a circuit breaker. The method includes: step S1: simulating through an arc-contact coupling ablation model of simulation software to obtain a correlation data set between contact parameters and contact ablation thickness; step S2: performing an electrical life experimental test on the circuit breaker to obtain a correlation data set between the ablation thickness of the contact and the remaining life of the circuit breaker; step S3: performing machine learning training based on the correlation data set between the contact parameters and the ablation thickness of the contact obtained in step S1, and the correlation data set between the ablation thickness and the remaining life of the circuit breaker obtained in step S2, to obtain an electrical life prediction model between contact parameters and the remaining life of the circuit breaker; step S4: inputting the contact parameters of the circuit breaker to be tested into the trained electrical life prediction model to obtain the remaining life of the circuit breaker to be tested; the device executes a computer program to implement the above method.

[0004] Existing technologies lack a complete closed loop of status assessment, performance judgment, and user feedback, resulting in poor overall practicality. Summary of the Invention

[0005] The purpose of the present invention is to address the above-mentioned shortcomings and propose a circuit breaker life prediction system based on machine learning.

[0006] The present invention adopts the following technical solutions: A machine learning-based circuit breaker life prediction system comprises a circuit breaker body analysis module, an environmental analysis module, a control module, a performance judgment module, and a communication module. The circuit breaker body analysis module is used to analyze and obtain relevant information about the circuit breaker body and transmit it to the control module. The environmental analysis module is used to analyze and obtain relevant information about the environment and transmit it to the control module. The control module obtains a real-time health index of the circuit breaker based on the relevant information about the circuit breaker body and the relevant information about the environment and transmits the real-time health index to the performance judgment module. The performance judgment module obtains information on whether the circuit breaker is performing well or poorly based on the real-time health index and transmits the information to the communication module. The communication module transmits the information on whether the circuit breaker is performing well or poorly to a user.

[0007] Optionally, the circuit breaker body analysis module is used to analyze and obtain the rated current, rated operating life, maximum allowable wear, rated temperature rise, calculation time interval, real-time operating current, real-time cumulative operating time, real-time contact wear and real-time internal temperature rise, and transmit them to the control module; the environmental analysis module is used to analyze and obtain the real-time ambient humidity, real-time dust concentration, rated allowable humidity and rated allowable dust concentration, and transmit them to the control module; the control module obtains the real-time environmental stress influencing factor based on the real-time ambient humidity, real-time dust concentration, rated allowable humidity and rated allowable dust concentration, obtains the real-time basic health index and the previous moment basic health index based on the rated current, rated operating life, maximum allowable wear, rated temperature rise, real-time operating current, real-time cumulative operating time, real-time contact wear and real-time internal temperature rise, and obtains the circuit breaker real-time health index based on the real-time basic health index, the previous moment basic health index, calculation time interval and real-time environmental stress influencing factor.

[0008] Optionally, the circuit breaker body analysis module includes an information setting submodule and a real-time detection submodule; the information setting submodule is used to set the calculation time interval, rated current, rated operating life, maximum allowable wear and rated temperature rise, and transmit them to the control module; the real-time detection submodule is used to detect and obtain the real-time operating current, real-time cumulative operating time, real-time contact wear and real-time internal temperature rise, and transmit them to the control module.

[0009] Optionally, the real-time detection submodule includes a current detection unit, a time detection unit, a wear detection unit and a temperature detection unit; the current detection unit is used to detect and obtain the real-time operating current, and transmit it to the control module; the time detection unit is used to detect and obtain the real-time accumulated running time, and transmit it to the control module; the wear detection unit is used to detect and obtain the real-time contact wear amount, and transmit it to the control module; the temperature detection unit is used to detect and obtain the real-time internal temperature rise, and transmit it to the control module.

[0010] Optionally, the temperature detection unit includes an internal temperature detector, an external temperature detector and a calculator; the internal temperature detector is used to detect and obtain the real-time internal temperature of the circuit breaker, and transmit it to the calculator; the external temperature detector is used to detect and obtain the real-time ambient temperature, and transmit it to the calculator; the calculator obtains the real-time internal temperature rise based on the real-time internal temperature of the circuit breaker and the real-time ambient temperature, and transmits it to the control module.

[0011] Optionally, the environmental analysis module includes a humidity detection submodule, a dust detection submodule and a rated value setting submodule; the humidity detection submodule is used to detect and obtain real-time environmental humidity, and transmit it to the control module; the dust detection submodule is used to detect and obtain real-time dust concentration, and transmit it to the control module; the rated value setting submodule is used to set the rated allowable humidity and rated allowable dust concentration, and transmit them to the control module.

[0012] Optionally, when the control module calculates the real-time health index of the circuit breaker, the following formula is satisfied: ; in, Real-time health index for circuit breakers, It is a real-time basic health index. is the basic health index at the previous moment, To calculate the time interval, is the real-time environmental stress influencing factor.

[0013] The beneficial effects achieved by the present invention are: 1. By setting up a circuit breaker body analysis module, the present invention can collect and analyze the key operating parameters of the circuit breaker (including operating current, cumulative operating time, contact wear, temperature rise, etc.) in real time, realize comprehensive monitoring of the circuit breaker operating status, and provide an accurate data basis for subsequent health assessment, which is highly practical. 2. By setting up an environmental analysis module, real-time monitoring of external environmental stresses such as ambient humidity and dust concentration is achieved, which makes up for the defect of existing technologies that ignore the impact of environmental factors on circuit breaker life, and improves the accuracy of life prediction and the adaptability of the model; 3. By setting up a control module, integrating circuit breaker data and environmental data, and outputting the real-time health index of the circuit breaker based on the health index calculation model, this index can dynamically reflect the operating status of the equipment and has stronger real-time and intelligent features compared to the traditional fixed threshold judgment method; 4. A performance judgment module is set up to evaluate the current performance status of the circuit breaker according to the health index, realize intelligent identification of information such as "good status / poor status", and provide a basis for fault warning and maintenance decision-making; 5. A communication module is set up to transmit the analysis results to the user terminal in real time, realizing a closed loop of the entire process from information collection, calculation, evaluation to alarm push, improving the timeliness and response efficiency of operation and maintenance, and is suitable for remote monitoring and intelligent management scenarios.

[0014] To further understand the features and technical contents of the present invention, please refer to the following detailed description and drawings of the present invention. However, the drawings provided are only for reference and illustration and are not intended to limit the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 It is a schematic diagram of the overall structure of the present invention; Figure 2 Schematic diagram of the structure of the temperature detection unit in the present invention; Figure 3 Schematic diagram of the structure of the real-time detection submodule in the present invention; Figure 4 This is an effect diagram of the real-time environmental stress influencing factors of the present invention; Figure 5 This is a schematic diagram of the overall structure of the second embodiment of the present invention; Figure 6 This is an effect diagram of the real-time remaining life of a circuit breaker according to the second embodiment of the present invention. DETAILED DESCRIPTION

[0016] The following is an explanation of the embodiments of the present invention through specific embodiments. Those skilled in the art can understand the advantages and effects of the present invention from the contents disclosed in this specification. The present invention can be implemented or applied through other different specific embodiments, and the details in this specification can also be modified and changed based on different viewpoints and applications without departing from the spirit of the present invention. In addition, the drawings of the present invention are only for simple schematic illustrations and are not depicted according to actual dimensions. It is stated in advance. The following embodiments will further explain the relevant technical contents of the present invention in detail, but the disclosed contents are not intended to limit the scope of protection of the present invention.

[0017] Example 1: This example provides a circuit breaker life prediction system based on machine learning, combined with Figures 1 to 4 shown.

[0018] A machine learning-based circuit breaker life prediction system comprises a circuit breaker body analysis module, an environmental analysis module, a control module, a performance judgment module, and a communication module. The circuit breaker body analysis module is used to analyze and obtain relevant information about the circuit breaker body and transmit it to the control module. The environmental analysis module is used to analyze and obtain relevant information about the environment and transmit it to the control module. The control module obtains a real-time health index of the circuit breaker based on the relevant information about the circuit breaker body and the relevant information about the environment and transmits the real-time health index to the performance judgment module. The performance judgment module obtains information on whether the circuit breaker is performing well or poorly based on the real-time health index and transmits the information to the communication module. The communication module transmits the information on whether the circuit breaker is performing well or poorly to a user.

[0019] Specifically, the performance judgment module refers to the following principles when making judgments: when the real-time health index of the circuit breaker is greater than or equal to the selection threshold of the real-time health index of the circuit breaker, it indicates that the circuit breaker performance is good; when the real-time health index of the circuit breaker is less than the selection threshold of the real-time health index of the circuit breaker, it indicates that the circuit breaker performance is poor; the selection threshold of the real-time health index of the circuit breaker is set by those skilled in the art; in this embodiment, the selection threshold of the real-time health index of the circuit breaker is set to 0.7; the fixed value of "0.7" is obtained by collecting historical monitoring data of the device (the real-time health index of the circuit breaker, the fault type, and whether a fault occurs in each measurement), and then dividing all the data according to the health status. For example, the real-time health index of the circuit breaker is divided into several segments (such as 0.9-1, 0.8-0.9, 0.7-0.8, ..., 0-0.1), and then calculating the failure rate in each interval (failure times / total times). By drawing a curve (horizontal axis: real-time health index of the circuit breaker, from 1 to 0; vertical axis: failure rate%), it is found that an inflection point appears at the value of 0.7.

[0020] The real-time health index of the circuit breaker can be further divided. For example, when the range of the real-time health index of the circuit breaker is less than 0.5, it means that the circuit breaker is seriously degraded and there are obvious safety hazards. It should be repaired or replaced as a priority and is a high-risk device; when the range of the real-time health index of the circuit breaker is greater than or equal to 0.5 to less than 0.7, it means that the circuit breaker is moderately aged and there are some risks. It needs early warning and it is recommended to arrange maintenance or replacement in the medium term; when the range of the real-time health index of the circuit breaker is greater than or equal to 0.7 to less than 0.9, it means that the circuit breaker is slightly aged and the equipment performance begins to decline, but it can still operate normally. It is recommended to include it in planned maintenance; when the range of the real-time health index of the circuit breaker is greater than or equal to 0.9, it means that the circuit breaker is in good health, the circuit breaker is in a state close to that of a new device, all parameters are within the safety range, and the environmental impact is acceptable.

[0021] Among them, a machine learning prediction model is integrated inside the control module, which is used to intelligently evaluate the current state of the circuit breaker based on relevant information of the circuit breaker body and relevant information of the environment. The machine learning prediction model receives relevant information of the circuit breaker body and relevant information of the environment as input feature variables, and outputs a real-time health index of the circuit breaker based on the above input features. The machine learning prediction model can be a decision tree model, a random forest model, a support vector machine model, a neural network model or other algorithm model with nonlinear mapping and fitting capabilities to achieve high-precision intelligent prediction of the circuit breaker state; in this embodiment, the machine learning prediction model is a random forest model, which is composed of multiple decision trees and can perform nonlinear feature fitting and classification prediction on the health state of the circuit breaker, thereby achieving highly robust and high-precision intelligent judgment of the operating state of the circuit breaker.

[0022] Optionally, the circuit breaker body analysis module is used to analyze and obtain the rated current, rated operating life, maximum allowable wear, rated temperature rise, calculation time interval, real-time operating current, real-time cumulative operating time, real-time contact wear and real-time internal temperature rise, and transmit them to the control module; the environmental analysis module is used to analyze and obtain the real-time ambient humidity, real-time dust concentration, rated allowable humidity and rated allowable dust concentration, and transmit them to the control module; the control module obtains the real-time environmental stress influencing factor based on the real-time ambient humidity, real-time dust concentration, rated allowable humidity and rated allowable dust concentration, obtains the real-time basic health index and the previous moment basic health index based on the rated current, rated operating life, maximum allowable wear, rated temperature rise, real-time operating current, real-time cumulative operating time, real-time contact wear and real-time internal temperature rise, and obtains the circuit breaker real-time health index based on the real-time basic health index, the previous moment basic health index, calculation time interval and real-time environmental stress influencing factor.

[0023] Optionally, the circuit breaker body analysis module includes an information setting submodule and a real-time detection submodule; the information setting submodule is used to set the calculation time interval, rated current, rated operating life, maximum allowable wear and rated temperature rise, and transmit them to the control module; the real-time detection submodule is used to detect and obtain the real-time operating current, real-time cumulative operating time, real-time contact wear and real-time internal temperature rise, and transmit them to the control module.

[0024] Optionally, the real-time detection submodule includes a current detection unit, a time detection unit, a wear detection unit and a temperature detection unit; the current detection unit is used to detect and obtain the real-time operating current, and transmit it to the control module; the time detection unit is used to detect and obtain the real-time accumulated running time, and transmit it to the control module; the wear detection unit is used to detect and obtain the real-time contact wear amount, and transmit it to the control module; the temperature detection unit is used to detect and obtain the real-time internal temperature rise, and transmit it to the control module.

[0025] Optionally, the temperature detection unit includes an internal temperature detector, an external temperature detector, and a calculator; the internal temperature detector is used to detect and determine the real-time internal temperature of the circuit breaker and transmit it to the calculator; the external temperature detector is used to detect and determine the real-time ambient temperature and transmit it to the calculator; the calculator determines the real-time internal temperature rise based on the real-time internal temperature of the circuit breaker and the real-time ambient temperature and transmits it to the control module. High temperatures can cause insulating materials (such as plastic or rubber) inside the circuit breaker to gradually harden, become brittle, or decompose, reducing their insulation performance.

[0026] Optionally, the environmental analysis module includes a humidity detection submodule, a dust detection submodule and a rated value setting submodule; the humidity detection submodule is used to detect and obtain real-time environmental humidity, and transmit it to the control module; the dust detection submodule is used to detect and obtain real-time dust concentration, and transmit it to the control module; the rated value setting submodule is used to set the rated allowable humidity and rated allowable dust concentration, and transmit them to the control module.

[0027] Optionally, when the control module calculates the real-time health index of the circuit breaker, the following formula is satisfied: ; in, Real-time health index for circuit breakers, It is a real-time basic health index. is the basic health index at the previous moment, To calculate the time interval, is the real-time environmental stress impact factor. The 0.1 in the formula is set to avoid negative values ​​in the health index and is based on the minimum acceptable state of the equipment.

[0028] Optionally, the control module satisfies the following formula during calculation: ; ; in, For real-time operating current, is the rated current, To accumulate running time in real time, is the rated operating life, is the real-time contact wear amount, is the maximum allowable wear, is the real-time internal temperature rise, is the rated temperature rise; is the real-time ambient humidity, is the rated allowable humidity, For real-time dust concentration, is the rated allowable dust concentration.

[0029] To calculate the real-time health index of the circuit breaker, you can refer to the following program code: import numpy as np # √ Environmental stress factor S_env(t) def calc_S_env(H_t, H_rated, D_t, D_rated): """ Calculate the environmental stress factor S_env(t) Parameters: - H_t: real-time ambient humidity - H_rated: rated humidity - D_t: real-time dust concentration - D_rated: rated dust concentration Returns: - S_env(t): environmental stress factor """ return 0.5 * (H_t / H_rated)**2 + 0.5 * (D_t / D_rated)**2 # √ Base health index HI_base(t) def calc_HI_base(I_oper, I_rated, t_oper, t_rated, V_wear, V_max, T_temp, T_rated): """ Calculate the base health index HI_base(t) Parameters: - I_oper: real-time operating current - I_rated: rated current - t_oper: accumulated operating time - t_rated: rated service life - V_wear: real-time contact wear - V_max: maximum allowable wear - T_temp: real-time internal temperature rise - T_rated: rated temperature rise Returns: - HI_base(t): base health index """ exponent = -0.8 * ( 0.35 * (I_oper / I_rated)**2 + 0.25 * (t_oper / t_rated)**2 + 0.2 * (V_wear / V_max)**2 + 0.2 * (T_temp / T_rated)**2 ) return np.exp(exponent) # √ Real-time health index HI(t) def calc_HI(HI_base_t, HI_base_t_prev, dt, S_env_t): """ Calculate the real-time health index HI(t) Parameters: - HI_base_t: current base health index - HI_base_t_prev: previous base health index - dt: time interval between current and previous step - S_env_t: environmental stress factor at time t Returns: - HI(t): real-time health index """ derivative = (HI_base_t - HI_base_t_prev) / dt HI_t = max(0.1, HI_base_t * (1 + 0.3 * derivative + 0.4 * S_env_t)) return HI_t Specifically, the circuit breaker real-time health index can be understood as an assessment of the comprehensive real-time health status of the circuit breaker, which is used to guide operation and maintenance decisions. It refers to a dimensionless indicator that quantitatively evaluates the current operating status of the circuit breaker at a certain time t by integrating multiple dimensional factors such as electrical parameters, mechanical wear, temperature rise, and environmental stress.

[0030] In the formula The fixed values ​​mentioned in the following are explained below. "1" can be understood as the base term, indicating that when there is no degradation, the circuit breaker health index is directly equal to the baseline real-time basic health index. "0.3" can be understood as the rate of change coefficient, which measures the contribution of the health index change rate to the health impact. A greater rate of change indicates that the device may be experiencing accelerated degradation. A value of 0.3 indicates that the "health change rate" contributes 30% to the total health impact. "0.4" can be understood as the environmental stress term weight, which measures the contribution of environmental factors (such as temperature, humidity, and dust) to health. 0.4 represents a 40% contribution to the health impact of environmental factors, representing an additional load stress correction term. To verify the rationality of the two fixed values ​​of 0.3 and 0.4, this embodiment uses a validation method based on sensitivity analysis. By comparing the contribution of influencing factors to the circuit breaker health index with typical device parameters, the parameter settings are verified to be consistent with the device degradation mechanism and actual engineering application requirements. Sensitivity analysis is a common model validation method. Its basic concept is to adjust the value of a single input factor while keeping other variables unchanged and observe the magnitude of the change in the output result. By comparing the output change degree caused by the adjustment of each factor, the relative importance of each factor can be judged. The following are the steps for verification: (1) Set a set of typical circuit breaker operating parameters and select is 0.8, Take 0.10, Take 0.40. Substitute these parameters into the circuit breaker real-time health index calculation formula of this embodiment to obtain the results. On this basis, and Perform individual disturbance adjustments, for example, first The baseline value is increased by 10%, that is, adjusted from 0.10 to 0.11, while maintaining Calculate the change of the real-time health index of the circuit breaker after adjustment and observe the cause. Increase the fluctuation amplitude of the circuit breaker real-time health index caused by the increase, and then the recovery To the baseline value of 0.10, The baseline value was increased by 10%, that is, adjusted from 0.40 to 0.44, while maintaining No change, calculate the real-time health index of the circuit breaker again, and observe The sensitivity analysis results under the above typical parameters show that the change range of the real-time health index of the circuit breaker caused by the increase is The change in the real-time health index of the circuit breaker caused by a 10% increase is approximately The change caused by a 10% increase is about 5 times that of the previous one. This result directly reflects that The impact on the real-time health index of the circuit breaker is more significant than that of , It has a higher contribution ratio in the health index assessment. Based on this analysis, when designing the health assessment model, this embodiment reasonably sets the weight of the environmental stress item to be greater than the change rate coefficient to reflect the actual impact difference between the two in the health decay process. This weight distribution not only conforms to the impact ranking relationship revealed by the sensitivity analysis, but also takes into account the engineering operability of the model and the flexibility of parameter adjustment. It is suitable for application in equipment online monitoring and status assessment scenarios. After obtaining the above results, it is necessary to verify the rationality of the setting of specific numerical values. For example, the change rate coefficient range is set between 0.2 and 0.4, and the environmental stress item weight is set between 0.3 and 0.5. Through the above interval settings, multiple groups of parameter combinations are formed (for example: 0.2 / 0.3, 0.2 / 0.35, 0.25 / 0.4, 0.3 / 0.4, 0.35 / 0.45, 0.4 / 0.5 The evaluation test is to substitute different parameter combinations (for example, 0.2 / 0.3, 0.2 / 0.35, 0.25 / 0.4, 0.3 / 0.4, 0.35 / 0.45, 0.4 / 0.5, etc.) for a batch of circuit breaker devices with known degradation conditions (or maintenance records) to see whether the degraded devices can be accurately found, whether healthy devices will be falsely reported, and whether the health index stratification is clear. It was found that when the rate of change coefficient is lower than 0.25 (such as 0.2) and the weight of the environmental stress item is higher than 0.45 (such as 0.45 or 0.5), the model is not sensitive enough to the short-term performance changes of the equipment, which is manifested in a decrease in the ability to identify equipment in the accelerated degradation stage and an increase in the missed detection rate. When the rate of change coefficient is higher than 0.35 (such as 0.4) or the weight of the environmental stress item is higher than 0.45 (such as 0.5), the model over-responds to the evaluation results of some healthy devices, especially when the environmental stress fluctuation amplitude is high. When the degree is small, false alarms are likely to occur and the false detection rate increases significantly. When the rate of change coefficient is set to 0.3 and the weight of the environmental stress item is set to 0.4, the model performs best in all evaluation indicators. For example, the health index distribution remains well differentiated between different levels, the output results are stable, and it is not overly sensitive to small parameter disturbances. Therefore, the weight set of 0.3 / 0.4 is located in the middle of the optimal range of recognition effect in multiple rounds of tests. It avoids the lack of sensitivity caused by too low a rate of change coefficient and prevents excessive response caused by too high a weight of the environmental stress item. It is a configuration that achieves the optimal balance between "recognition sensitivity" and "evaluation stability".

[0031] In the formula The fixed values ​​involved in the calculation are explained as follows. Among them, "-0.8" can be understood as the attenuation rate, which is used to adjust the factor of the body degradation speed. This form is closer to the aging mechanism of the circuit breaker. In the actual calculation process, if the value is set to -1 or -2, the attenuation is too fast, which will lead to It drops to a very low level at the beginning. If the value is set to -0.3 or -0.5, the decay is too slow to reflect the health status of the circuit breaker. Therefore, the corresponding "-0.8" is an average decay rate adjusted after regression analysis of the actual circuit breaker aging speed data. It can reflect the actual decay trend of the equipment during its life cycle. The "0.35" in the figure can be understood as the current weight coefficient. A large current load accelerates wear and is the most direct factor in degradation. The "0.2" in the value can be understood as the weight coefficient of the running time. Time is an important dimension of degradation, but degradation is not determined solely by time (it needs to work together with the load). The "0.25" in the figure can be understood as the wear weight coefficient. Wear directly reflects the state of the mechanical structure and affects the life of the contact. The impact is significant. The "0.2" in the figure can be understood as the temperature weight coefficient. Temperature rise is an accelerated aging factor, but it is usually environmental and the changes are relatively controllable. In summary, current load and wear are active degradation sources and are direct factors. Time and temperature are passive or auxiliary degradation sources and are cumulative effects. In actual operation and maintenance, it is found that circuit overload has a greater impact than wear, so the corresponding current weight coefficient has the largest value. The above combination data (0.35 / 0.2 / 0.25 / 0.2) is also obtained through sensitivity analysis. For example, multiple rounds of testing with different combinations (such as 0.2 / 0.3 / 0.3 / 0.2, 0.25 / 0.25 / 0.25 / 0.25, 0.35 / 0.15 / 0.3 / 0.2, etc.) are conducted to see which group works best in accurately identifying degraded equipment and distinguishing healthy equipment. In the formula The fixed values ​​involved in the following explanations are given respectively: The "0.5" and The “0.5” in the formula can be understood as the degradation intensity of humidity and dust is roughly equivalent, which can avoid the situation where the numerical value is biased during the calculation process. and The square of the calculation is mainly controlled within a reasonable nonlinear amplification range to reflect the nonlinear acceleration characteristics of the degradation process and avoid underestimation of the degradation rate under linear estimation.

[0032] The real-time basic health index reflects the real-time basic operating status of the device and is used for detailed analysis and diagnosis.

[0033] The previous moment's basic health index is calculated in the same way as the real-time basic health index.

[0034] The real-time environmental stress impact factor reflects the real-time impact of the environment and the degree of influence of the external environment on stress. It is more in line with actual operating conditions and can truly reflect the impact of external conditions on equipment life. For example, excessive humidity can easily cause equipment insulation to become damp and age. In addition, dust can reduce insulation performance and accelerate wear.

[0035] The unit of the calculation time interval is seconds, which is set to 60 seconds in this embodiment.

[0036] The units of real-time operating current and rated current are both amperes. Excessive real-time operating current will accelerate equipment aging. The rated current is the equipment safety operation benchmark set by the manufacturer.

[0037] The units of real-time accumulated running time and rated operating life are both seconds. Real-time accumulated running time can reflect the set running time. Long-term operation of the equipment will cause accumulated aging. The rated operating life is the expected safe operating life specified by the equipment manufacturer.

[0038] The units of real-time contact wear and maximum allowable wear are both millimeters. The maximum allowable wear is the maximum wear limit specified in the manufacturer's safety specifications. The principle of obtaining real-time contact wear is as follows: the contact travel change is collected by a displacement sensor installed on the circuit breaker's closing and opening drive mechanism, and compared and calculated in combination with the initial position, thereby realizing real-time non-contact detection during equipment operation.

[0039] The units of real-time internal temperature rise and rated temperature rise are both in degrees Celsius. The rated temperature rise is the maximum temperature rise allowed for safety as specified by the manufacturer. The real-time internal temperature rise is the increase in the internal temperature of the circuit breaker relative to the ambient temperature when the circuit breaker is operating. It is used to reflect the heat generation and heat dissipation capacity of the equipment body due to factors such as current carrying, friction, and contact resistance.

[0040] The real-time ambient humidity and rated allowable humidity are both expressed in percentage.

[0041] The units of real-time dust concentration and rated allowable dust concentration are both milligrams per cubic meter.

[0042] The above units are only examples, and those skilled in the art may set different units according to actual needs when implementing this solution.

[0043] This embodiment solves the problem of low practicality of traditional life prediction systems. By providing a circuit breaker body analysis module, the present invention can collect and analyze the key operating parameters of the circuit breaker (including operating current, cumulative operating time, contact wear, temperature rise, etc.) in real time, realize comprehensive monitoring of the circuit breaker operating status, and provide an accurate data basis for subsequent health assessment, thus having strong practicality.

[0044] Example 2: This example includes all the contents of Example 1, and provides a circuit breaker life prediction system based on machine learning, combined with Figure 5 and Figure 6 shown.

[0045] A circuit breaker life prediction system based on machine learning, the system also includes a life storage module; The life storage module is used to store the rated life of the circuit breaker and transmit it to the control module; The control module calculates the real-time remaining life of the circuit breaker according to the rated life of the circuit breaker and the real-time health index of the circuit breaker, and transmits the real-time remaining life of the circuit breaker to the communication module; The communication module transmits the real-time remaining life of the circuit breaker to the user end.

[0046] Optionally, when the control module calculates the real-time remaining life of the circuit breaker, the following formula is satisfied: ; in, The real-time remaining life of the circuit breaker, is the rated life of the circuit breaker.

[0047] When the control module calculates the real-time remaining life of the circuit breaker, refer to the following program code: import math # ------------------------------- # Function 1: Calculate the basic health index HI_base(t) # ------------------------------- def calculate_hi_base(I_oper, I_rated, t_oper, t_rated, V_wear, V_max, T_temp, T_rated, gamma=0.8, w1=0.3, w2=0.2, w3=0.3, w4=0.2, n=2, m=2, p=2, q=2): term1 = w1 * (I_oper / I_rated) ** n term2 = w2 * (t_oper / t_rated) ** m term3 = w3 * (V_wear / V_max) ** p term4 = w4 * (T_temp / T_rated) ** q HI_base = math.exp(-gamma * (term1 + term2 + term3 + term4)) return HI_base # ------------------------------- # Function 2: Calculate environmental stress factor S_env(t) # ------------------------------- def calculate_s_env(H, H_rated, D, D_rated): return 0.5 * (H / H_rated) ** 2 + 0.5 * (D / D_rated) ** 2 # ------------------------------- # Function 3: Calculate real-time health index HI(t) # ------------------------------- def calculate_hi(HI_base, HI_base_prev, delta_t, S_env, alpha=0.3, beta=0.4): trend = (HI_base - HI_base_prev) / delta_t return HI_base * (1 + alpha * trend + beta * S_env) # ------------------------------- # Function 4: Calculate the remaining life RUL(t) # ------------------------------- def calculate_rul(HI, L_rated, delta=1.2, kappa=2.0): return L_rated * (HI ** delta) * math.exp(-kappa * (1 - HI)) # ------------------------------- # Sample input data (can be replaced with real-time data) # ------------------------------- input_data = { 'I_oper': 80, 'I_rated': 100, 't_oper': 500000, 't_rated': 1000000, 'V_wear': 1.2, 'V_max': 3.0, 'T_temp': 60, 'T_rated': 80, 'H': 70, 'H_rated': 85, 'D': 0.2, 'D_rated': 0.3, 'HI_base_prev': 0.92, 'delta_t': 60, 'L_rated': 100000# Rated life (unit can be hours or number of operations) } # ------------------------------- # Main process: Calculate HI and RUL # ------------------------------- HI_base = calculate_hi_base( input_data['I_oper'], input_data['I_rated'], input_data['t_oper'], input_data['t_rated'], input_data['V_wear'], input_data['V_max'], input_data['T_temp'], input_data['T_rated'] ) S_env = calculate_s_env(input_data['H'], input_data['H_rated'],input_data['D'], input_data['D_rated']) HI = calculate_hi(HI_base, input_data['HI_base_prev'], input_data['delta_t'], S_env) RUL = calculate_rul(HI, input_data['L_rated']) # ------------------------------- # Output results # ------------------------------- print(f"Basic health index HI_base(t): {HI_base:.4f}") print(f"Environmental stress factor S_env(t): {S_env:.4f}") print(f"Real-time health index HI(t): {HI:.4f}") print(f"Real-time remaining life RUL(t): {RUL:.2f}") Specifically, the units of the real-time remaining life of the circuit breaker and the rated life of the circuit breaker are both hours. The rated life of the circuit breaker is the longest service life during which the circuit breaker can operate reliably under normal working conditions, according to the usage specifications and working environment set by the manufacturer.

[0048] In the formula The fixed values ​​involved in are explained as follows, for example, The control in the middle is the "nonlinear amplification" of the linear influence of the circuit breaker's real-time health index on its lifespan. "1.2" indicates that it is slightly higher than the linearity, but not as radical as the square, which is in line with the actual law. The selection of "-2" is a balance point. If this coefficient is set to -1, the decay will be very slow. If this coefficient is set to -3, the decay will be very fast. in The design logic is to flip the real-time health index of the circuit breaker. The higher the real-time health index of the circuit breaker, the faster it will decay. When it is equal to 1, it can be understood as complete health, and the exponential term becomes , the real-time remaining life of the circuit breaker is equal to the rated life of the circuit breaker, which is understood as no attenuation, assuming in " is replaced by " ", so the calculated result is contrary to the actual situation, so the formula is set to " " is set to "1".

[0049] The above units are only examples, and those skilled in the art may set different units according to actual needs when implementing this solution.

[0050] This embodiment solves the problem of the relatively single life prediction system of the traditional system. The system introduces a life storage module, which can centrally manage and long-term preserve the rated life information of the circuit breaker, provide a reliable data basis for the remaining life calculation, and improve the accuracy and traceability of the system.

[0051] The contents disclosed above are only preferred feasible embodiments of the present invention and do not limit the scope of protection of the present invention. Therefore, all equivalent technical changes made using the contents of the present invention description and drawings are included in the scope of protection of the present invention. In addition, the elements therein can be updated as technology develops.

Claims

1. A circuit breaker life prediction system based on machine learning, characterized in that: The system includes a circuit breaker body analysis module, an environment analysis module, a control module, a performance judgment module and a communication module; The circuit breaker body analysis module is used to analyze and obtain relevant information of the circuit breaker body and transmit it to the control module; The environment analysis module is used to analyze and obtain relevant information of the environment and transmit it to the control module; The control module obtains the real-time health index of the circuit breaker based on the relevant information of the circuit breaker body and the relevant information of the environment, and transmits the real-time health index of the circuit breaker to the performance judgment module; The performance judgment module obtains information on whether the circuit breaker performance is good or poor based on the real-time health index of the circuit breaker, and transmits the information on whether the circuit breaker performance is good or poor to the communication module; The communication module transmits information about whether the circuit breaker performance is good or bad to the user end.

2. The circuit breaker life prediction system based on machine learning according to claim 1, characterized in that: The circuit breaker body analysis module is used to analyze and obtain the rated current, rated service life, maximum allowable wear, rated temperature rise, calculation time interval, real-time operating current, real-time accumulated operating time, real-time contact wear and real-time internal temperature rise, and transmit them to the control module; The environmental analysis module is used to analyze and obtain the real-time environmental humidity, real-time dust concentration, rated allowable humidity and rated allowable dust concentration, and transmit them to the control module; The control module calculates a real-time environmental stress impact factor based on the real-time ambient humidity, the real-time dust concentration, the rated allowable humidity, and the rated allowable dust concentration; calculates a real-time basic health index and a previous-moment basic health index based on the rated current, the rated operating life, the maximum allowable wear, the rated temperature rise, the real-time operating current, the real-time accumulated operating time, the real-time contact wear, and the real-time internal temperature rise; and calculates a real-time health index of the circuit breaker based on the real-time basic health index, the previous-moment basic health index, the calculation time interval, and the real-time environmental stress impact factor.

3. The machine learning-based circuit breaker life prediction system according to claim 2, characterized in that: The circuit breaker body analysis module includes an information setting submodule and a real-time detection submodule; The information setting submodule is used to set the calculation time interval, rated current, rated operating life, maximum allowable wear and tear, and rated temperature rise, and transmit them to the control module; The real-time detection submodule is used to detect and obtain the real-time operating current, real-time accumulated operating time, real-time contact wear and real-time internal temperature rise, and transmit them to the control module.

4. The machine learning-based circuit breaker life prediction system according to claim 3, characterized in that: The real-time detection submodule includes a current detection unit, a time detection unit, a wear detection unit and a temperature detection unit; The current detection unit is used to detect and obtain the real-time operating current and transmit it to the control module; The time detection unit is used to detect and obtain the real-time accumulated running time and transmit it to the control module; The wear detection unit is used to detect and obtain the real-time contact wear amount and transmit it to the control module; The temperature detection unit is used to detect and obtain real-time internal temperature rise and transmit it to the control module.

5. The circuit breaker life prediction system based on machine learning according to claim 4, characterized in that: The temperature detection unit includes an internal temperature detector, an external temperature detector and a calculator; The internal temperature detector is used to detect and obtain the real-time internal temperature of the circuit breaker and transmit it to the computer; The external temperature detector is used to detect and obtain the real-time ambient temperature and transmit it to the computer; The calculator calculates the real-time internal temperature rise according to the real-time internal temperature of the circuit breaker and the real-time ambient temperature, and transmits the result to the control module.

6. The circuit breaker life prediction system based on machine learning according to claim 2, characterized in that: The environmental analysis module includes a humidity detection submodule, a dust detection submodule and a rated value setting submodule; The humidity detection submodule is used to detect and obtain the real-time environmental humidity and transmit it to the control module; The dust detection submodule is used to detect and obtain real-time dust concentration and transmit it to the control module; The rated value setting submodule is used to set the rated allowable humidity and the rated allowable dust concentration, and transmit the results to the control module.

7. The circuit breaker life prediction system based on machine learning according to claim 2, characterized in that: When the control module calculates the real-time health index of the circuit breaker, the following formula is satisfied: ； in, Real-time health index for circuit breakers, It is a real-time basic health index. is the basic health index at the previous moment, To calculate the time interval, is the real-time environmental stress influencing factor.

Images