Primary and secondary integrated set of column circuit breaker fault prediction method and system

Abstract

The application provides a secondary fusion complete set of on-column circuit breaker fault prediction method and system, belonging to the technical field of fault prediction. The method integrates the multi-dimensional historical operation data of the on-column circuit breaker to extract performance degradation features therefrom and determine the damage cumulative action quantity of the circuit breaker, thereby realizing quantitative evaluation and degradation trend determination of the comprehensive degradation degree of the circuit breaker under the coupling action of multiple factors; further, the long short-term memory network is used to perform time series fitting on the performance degradation feature sequence, to generate a continuous performance degradation trajectory that dynamically changes with the operation condition and solve the degradation dynamic change rate of each time node, so as to determine whether it is in the performance accelerated degradation stage and mark the abnormal degradation trajectory; finally, the abnormal offset data of the abnormal degradation trajectory is used to complete the fault prediction of the on-column circuit breaker, realize long-term and continuous tracking and early prediction of the performance degradation trend of the circuit breaker, and thereby improve the reliability and fault prevention and control capability of the power distribution network operation.

Description

Technical Field

[0001] This application relates to the field of fault prediction technology, specifically to a fault prediction method and system for integrated primary and secondary pole-mounted circuit breakers. Background Technology

[0002] Currently, the power distribution network is a crucial link in the power system that directly faces users. Integrated primary and secondary pole-mounted circuit breakers, as key equipment integrating protection, measurement, and control functions, are permanently installed on outdoor poles, bearing the heavy responsibility of line interruption and fault isolation. A fault in one of these circuit breakers can not only cause widespread power outages but also affect fault location and rapid power restoration. Existing methods for assessing the condition of such equipment rely on periodic inspections or post-fault analysis, lacking continuous tracking of the equipment's condition over a future period. Furthermore, due to the large number and dispersed distribution of circuit breakers, manual inspections cannot cover every change in operating conditions. By the time an anomaly is detected, it is often already close to a fault, impacting the stability of the power system.

[0003] The above information in the background section of the invention is only used to enhance the understanding of the background of this application, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0004] In view of this, this application provides a method and system for predicting faults in a primary and secondary integrated pole-mounted circuit breaker, which can predict circuit breaker faults in advance.

[0005] In a first aspect, embodiments of this application provide a method for predicting faults in a primary and secondary integrated pole-mounted circuit breaker. This method includes: acquiring multi-source historical operating data of the pole-mounted circuit breaker, wherein the multi-source historical operating data includes load current, arcing time, circuit voltage difference, circuit current, opening and closing coil current waveforms, ambient temperature, ambient humidity, cumulative temperature rise amplitude, and operating current of the energy storage motor; determining contact wear, circuit resistance increase, opening and closing time deviation, insulation dielectric loss factor, and energy storage attenuation based on the multi-source historical operating data, and constructing a cumulative damage effect based on the opening and closing time deviation, the insulation dielectric loss factor, and the energy storage attenuation; if the cumulative damage effect exceeds a preset effect threshold, then determining the fault... The pole-mounted circuit breaker exhibits a performance degradation trend. A performance degradation sequence is determined based on the contact wear, increased circuit resistance, opening / closing time deviation, and insulation dielectric loss factor. Based on this performance degradation sequence, a long short-term memory network (LSTM) is used to determine the continuous performance degradation trajectory of the pole-mounted circuit breaker. The dynamic rate of change of degradation at each time point is determined based on this continuous performance degradation trajectory. If the dynamic rate of change of degradation exceeds a preset rate of change threshold, it is determined that the pole-mounted circuit breaker exhibits an accelerated performance degradation trend, and the trajectory corresponding to the current time point is marked as an abnormal degradation trajectory. Abnormal offset trajectory data is determined based on the abnormal degradation trajectory, and the fault prediction result of the pole-mounted circuit breaker is determined based on the abnormal offset trajectory data.

[0006] Secondly, embodiments of this application provide a primary and secondary integrated pole-mounted circuit breaker fault prediction system. This system includes: an acquisition module, a first determination module, a second determination module, a third determination module, a marking module, and a fourth determination module. The acquisition module is used to acquire multi-source historical operating data of the pole-mounted circuit breaker, including load current, arcing time, circuit voltage difference, circuit current, opening and closing coil current waveforms, ambient temperature, ambient humidity, cumulative temperature rise amplitude, and the operating current of the energy storage motor. The first determination module is used to determine contact wear, circuit resistance increase, opening and closing time deviation, insulation dielectric loss factor, and energy storage attenuation based on the multi-source historical operating data, and to construct a cumulative damage effect based on the opening and closing time deviation, the insulation dielectric loss factor, and the energy storage attenuation. The second determination module is used to determine that the pole-mounted circuit breaker exhibits a performance degradation trend if the cumulative damage effect exceeds a preset effect threshold, and to determine the cumulative damage effect based on the... The system comprises four modules: a first module for determining a performance degradation sequence based on contact wear, circuit resistance increase, opening and closing time deviation, and insulation dielectric loss factor; a second module for determining a continuous performance degradation trajectory of the pole-mounted circuit breaker using a long short-term memory network based on the performance degradation sequence, and determining the dynamic rate of degradation change at each time point based on the continuous performance degradation trajectory; a third module for determining a continuous performance degradation trajectory of the pole-mounted circuit breaker using a long short-term memory network based on the performance degradation sequence, and determining the dynamic rate of degradation change at each time point based on the dynamic rate of degradation change; a fourth module for determining abnormal offset trajectory data based on the abnormal degradation trajectory, and determining the fault prediction result of the pole-mounted circuit breaker based on the abnormal offset trajectory data.

[0007] This application provides a method and system for predicting faults in pole-mounted circuit breakers that integrate primary and secondary circuit breaker systems. By integrating multi-dimensional historical operating data of the pole-mounted circuit breaker, performance degradation characteristics such as contact wear, circuit resistance increase, opening and closing time deviation, insulation dielectric loss factor, and energy storage attenuation are extracted. The degradation indicators of opening / closing, insulation, and energy storage dimensions are integrated into a cumulative damage effect, enabling a quantitative assessment of the overall degradation degree and determination of degradation trends of the circuit breaker under the coupled effects of multiple factors. Furthermore, a long short-term memory network is used to perform time-series fitting on the performance degradation characteristic sequence, generating a continuous performance degradation trajectory that dynamically changes with operating conditions. The dynamic rate of change of degradation at each time node is calculated to determine whether the circuit breaker is in an accelerated performance degradation stage and to mark abnormal degradation trajectories. Finally, fault prediction of the pole-mounted circuit breaker is completed based on the abnormal offset data of the abnormal degradation trajectory. This approach addresses the shortcomings of traditional operation and maintenance models that rely on manual periodic inspections and post-event analysis. It is adaptable to the dynamic evolution of circuit breaker degradation rates under complex outdoor environments and variable operating conditions, enabling long-term continuous tracking and early prediction of circuit breaker performance degradation trends. This significantly improves the initiative and accuracy of pole-mounted circuit breaker operation and maintenance, thereby enhancing the reliability and fault prevention capabilities of the distribution network. Attached Figure Description

[0008] To more clearly illustrate the technical solutions in the embodiments of this application or the conventional technology, the drawings used in the description of the embodiments or the conventional technology will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0009] Figure 1 This is a flowchart illustrating a secondary-secondary integrated pole-mounted circuit breaker fault prediction method provided in an exemplary embodiment of this application.

[0010] Figure 2 This is a flowchart illustrating a method for predicting faults in a primary and secondary integrated pole-mounted circuit breaker, provided in another exemplary embodiment of this application.

[0011] Figure 3 This is a flowchart illustrating a method for predicting faults in a primary and secondary integrated pole-mounted circuit breaker, as provided in another exemplary embodiment of this application.

[0012] Figure 4 This is a flowchart illustrating a method for predicting faults in a primary and secondary integrated pole-mounted circuit breaker, as provided in another exemplary embodiment of this application.

[0013] Figure 5 This is a flowchart illustrating a method for predicting faults in a primary and secondary integrated pole-mounted circuit breaker, as provided in another exemplary embodiment of this application.

[0014] Figure 6 This is a flowchart illustrating a method for predicting faults in a primary and secondary integrated pole-mounted circuit breaker, as provided in another exemplary embodiment of this application.

[0015] Figure 7 This is a flowchart illustrating a method for predicting faults in a primary and secondary integrated pole-mounted circuit breaker, as provided in another exemplary embodiment of this application.

[0016] Figure 8 This is a flowchart illustrating a method for predicting faults in a primary and secondary integrated pole-mounted circuit breaker, as provided in another exemplary embodiment of this application. Detailed Implementation

[0017] To enable those skilled in the art to better understand the technical solution, the present application will be described in detail below with reference to the embodiments. The description in this section is only exemplary and explanatory, and should not be used to limit the scope of protection of the present application in any way.

[0018] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, they are provided so that this application will be more comprehensive and complete, and will fully convey the concept of the exemplary embodiments to those skilled in the art. The described features, structures, or characteristics may be combined in any suitable manner in one or more embodiments. In the following description, numerous specific details are provided to give a full understanding of embodiments of this application.

[0019] The terms “a,” “one,” and “the” are used to indicate the existence of one or more elements / components / etc.; the terms “including” and “having” are used to indicate an open-ended inclusion and that other elements / components / etc. may exist in addition to those listed. The terms “first” and “second” are used only as markers and are not a limitation on the number of objects.

[0020] Currently, the power distribution network is a crucial link in the power system that directly faces users. Integrated primary and secondary pole-mounted circuit breakers, as key equipment integrating protection, measurement, and control functions, are permanently installed on outdoor poles, bearing the heavy responsibility of line interruption and fault isolation. A fault in one of these circuit breakers can not only cause widespread power outages but also affect fault location and rapid power restoration. Existing methods for assessing the condition of such equipment rely on periodic inspections or post-fault analysis, lacking continuous tracking of the equipment's condition over a future period. Furthermore, due to the large number and dispersed distribution of circuit breakers, manual inspections cannot cover every change in operating conditions. By the time an anomaly is detected, it is often already close to a fault, impacting the stability of the power system.

[0021] Specifically, because pole-mounted circuit breakers are exposed to the outdoors for extended periods, ambient temperature and humidity continuously affect the aging rate of their insulation and metal components. The magnitude of the load current determines the degree of heating and wear of the contacts and current-carrying parts, while the number of opening and closing operations directly consumes the lifespan of the mechanism. These factors do not act independently but rather cumulatively on the same circuit breaker, jointly pushing the critical components of the circuit breaker from a good state to an abnormal state. Therefore, under the long-term cumulative effects, the performance degradation of pole-mounted circuit breakers slowly evolves along a trajectory that changes continuously over time.

[0022] However, this degradation trajectory is constantly rewritten with new operating conditions. For example, the same circuit breaker may experience minimal contact wear during the first six months of operation in a low-load, dry environment; however, during the rainy and high-load summer season, insulation dampness and contact erosion will accelerate significantly, resulting in a completely different degradation rate. If we only use data collected at a single moment for evaluation, or apply fixed judgment rules to all circuit breakers, we cannot reflect this degradation process that changes with operating conditions. Consequently, it is difficult to identify which circuit breaker is heading towards failure weeks or months in advance, impacting the health management and maintenance strategies for circuit breakers.

[0023] Therefore, how to continuously utilize long-term operating data under different environments and working conditions to determine the trajectory of performance degradation of key components over time, and thus make fault predictions about the future state of the circuit breaker, has become a technical problem that needs to be solved to improve the initiative of operation and maintenance and the reliability of power supply of integrated primary and secondary pole-mounted circuit breakers.

[0024] This application provides a method for predicting faults in a primary and secondary integrated pole-mounted circuit breaker, such as... Figure 1 The method for fault prediction of a primary and secondary integrated pole-mounted circuit breaker is shown. This method may include the following steps: Step S110: Obtain multi-source historical operating data of the pole-mounted circuit breaker. The multi-source historical operating data includes load current, arcing time, circuit voltage difference, circuit current, opening and closing coil current waveform, ambient temperature, ambient humidity, cumulative temperature rise amplitude, and operating current of the energy storage motor. Step S120: Determine the contact wear, circuit resistance increase, opening and closing time deviation, insulation dielectric loss factor, and energy storage attenuation based on multi-source historical operating data, and construct and determine the cumulative damage effect based on the opening and closing time deviation, insulation dielectric loss factor, and energy storage attenuation. Step S130: If the cumulative damage exceeds the preset threshold, it is determined that the pole-mounted circuit breaker has a performance degradation trend, and the performance degradation sequence is determined based on the contact wear, the increase in circuit resistance, the deviation of opening and closing time, and the insulation dielectric loss factor. Step S140: Based on the performance degradation sequence, a long short-term memory network is used to determine the continuous performance degradation trajectory of the pole-mounted circuit breaker, and the degradation dynamic change rate corresponding to each time point is determined according to the continuous performance degradation trajectory. Step S150: If the degradation dynamic change rate exceeds the preset change rate threshold, it is determined that the pole-mounted circuit breaker has an accelerated performance degradation trend, and the trajectory corresponding to the current time point is marked as an abnormal degradation trajectory. Step S160: Determine the abnormal offset trajectory data based on the abnormal degradation trajectory, and determine the fault prediction result of the pole-mounted circuit breaker based on the abnormal offset trajectory data.

[0025] According to the primary and secondary integrated pole-mounted circuit breaker fault prediction method provided in this application, this method can integrate multi-dimensional historical operating data of the pole-mounted circuit breaker to extract performance degradation characteristics such as contact wear, circuit resistance increase, opening and closing time deviation, insulation dielectric loss factor, and energy storage attenuation. It integrates the degradation indicators of opening and closing, insulation, and energy storage dimensions into a cumulative damage effect, thereby achieving a quantitative assessment of the comprehensive degradation degree of the circuit breaker under the coupled effect of multiple factors and determining the degradation trend. Furthermore, it uses a long short-term memory network to perform time-series fitting on the performance degradation feature sequence, generating a continuous performance degradation trajectory that dynamically changes with operating conditions and solving the degradation dynamic change rate at each time node to determine whether it is in the accelerated performance degradation stage and mark the abnormal degradation trajectory. Finally, it completes the fault prediction of the pole-mounted circuit breaker based on the abnormal offset data of the abnormal degradation trajectory. This approach addresses the shortcomings of traditional operation and maintenance models that rely on manual periodic inspections and post-event analysis. It is adaptable to the dynamic evolution of circuit breaker degradation rates under complex outdoor environments and variable operating conditions, enabling long-term continuous tracking and early prediction of circuit breaker performance degradation trends. This, in turn, realizes fault prediction and health management strategies, thereby improving the reliability and fault prevention capabilities of the distribution network.

[0026] The following is a detailed description of each step in the fault prediction method for primary and secondary integrated pole-mounted circuit breakers provided in this application: It should be noted that the integrated primary and secondary pole-mounted circuit breaker in this application is a core integrated switchgear for 10kV overhead power distribution lines. It is a primary switch body responsible for high-voltage power switching and fault isolation, deeply integrated with a secondary measurement and control unit possessing measurement, protection, control, and communication functions. This complete set is then mounted on outdoor poles after factory assembly. Its primary section, centered on a vacuum interrupter, operating mechanism, and conductive circuits, bears the physical functions of normal load switching and fault current interruption. The secondary section integrates instrument transformers, protection control chips, operating circuits, and communication modules into a single, built-in design, replacing traditional external feeder terminals. This reduces potential fault points such as loose outdoor wiring and moisture aging at the hardware level, and integrates the combined capabilities of switch operation, relay protection, status monitoring, and remote interaction at the functional level. Furthermore, after factory calibration and commissioning, it significantly simplifies on-site construction and commissioning processes. This equipment is widely deployed at the segment, connection and branch nodes of power distribution lines. It is the core execution unit for rapid fault isolation, load transfer and line self-healing control in power distribution networks. At the same time, its built-in full-scale operation data acquisition capability also provides data support for equipment status assessment, degradation trend analysis and fault prediction.

[0027] In one embodiment of this application, in step S110, multi-source historical operating data of the pole-mounted circuit breaker is obtained. The multi-source historical operating data includes load current, arcing time, circuit voltage difference, circuit current, opening and closing coil current waveform, ambient temperature, ambient humidity, cumulative temperature rise amplitude, and operating current of the energy storage motor.

[0028] Specifically, the load current, circuit current, and circuit voltage difference can be periodically collected by the current transformer and voltage transformer of the main circuit of the pole-mounted circuit breaker; the ambient temperature and humidity can be periodically collected by the temperature and humidity sensor installed on the circuit breaker body; the temperature rise can be calculated and accumulated in real time by the circuit breaker measurement and control unit based on the difference between the temperature of the current-carrying component and the ambient temperature; when the opening and closing operation command or the energy storage motor start signal is detected, the high-speed acquisition mode is triggered to synchronously record the arcing time, the opening and closing coil current waveform, and the operating current of the energy storage motor.

[0029] Among them, the load current is the power frequency load current flowing through the main circuit of the pole-mounted circuit breaker under normal operating conditions, measured in amperes. It is an electrical parameter reflecting the load level of the equipment and used to calculate contact electrical wear. The arcing time is the duration of the arc generated at the instant the moving and stationary contacts separate during the circuit breaker's opening operation, measured in milliseconds. It is a key transient parameter for quantifying the degree of contact electrical erosion and evaluating the performance of the arc-extinguishing chamber. The circuit voltage difference is the voltage difference between the incoming and outgoing lines when the circuit breaker is closed, measured in millivolts. Combined with the circuit current, it can be used to calculate the circuit resistance and is a parameter reflecting contact performance and conductor connection status. The circuit current is the current flowing through the conductive circuit when the main circuit of the circuit breaker is closed, measured in amperes. It is used in conjunction with the circuit voltage difference to calculate the circuit resistance and can also verify the consistency of the load current data collection. The opening and closing coil current waveform is a continuous waveform showing the change of current in the operating mechanism's opening and closing coil over time during the circuit breaker's opening and closing operations. The unit is amperes, and it reflects the coil's health status, the degree of mechanism jamming, and the core's travel, among other mechanical characteristics. Ambient temperature is the outdoor air temperature at the circuit breaker's installation location, measured in degrees Celsius. It is a key environmental parameter affecting insulation aging, metal component oxidation, and circuit temperature rise. Ambient humidity is the relative humidity of the outdoor environment at the circuit breaker's installation location, measured as a percentage. It is a key environmental parameter affecting the rate of moisture absorption by insulation components, corrosion of metal components, and aging of sealing rings. The cumulative temperature rise amplitude is the cumulative integral of the difference between the temperature of the current-carrying components in the circuit breaker's main circuit and the ambient temperature over a historical period, measured in Kelvin per hour. It is a cumulative parameter reflecting the long-term cumulative heating effect and assessing the degree of thermal aging of insulation and contacts. The operating current of the energy storage motor is the operating current in the circuit breaker's spring operating mechanism when the energy storage motor performs spring energy storage operations, measured in amperes. It reflects the motor's health status, the lubrication status of the transmission mechanism, and the spring load level.

[0030] For example, taking a 10kV integrated primary and secondary pole-mounted circuit breaker as an example, multi-source historical operating data of the equipment over the past 90 days was retrieved. The data acquisition configuration was as follows: steady-state quantities were collected every 15 minutes, transient quantities were triggered by events, the sampling frequency was 10 kHz, and the time synchronization accuracy was 1 millisecond. In the steady-state operation scenario, the average daytime load current of the circuit breaker was 320 amps, and the peak load current was 485 amps. The corresponding circuit current in the closed state was 320 amps, and the circuit voltage difference was 9.6 millivolts. The average daytime outdoor ambient temperature was 28 degrees Celsius, the lowest nighttime temperature was 16 degrees Celsius, the average daytime ambient humidity was 62%, and the peak humidity on rainy days was 94%. The cumulative temperature rise over the past 90 days, calculated by combining circuit heating and ambient temperature, was 1280 Kelvin per hour. In the opening and closing operation scenario, a line fault opening operation triggered transient data acquisition, and the arcing time of this opening was recorded as 8.2 milliseconds. The peak value of the current waveform of the trip coil was 5.8 amps, and the rise time of the waveform was 12 milliseconds, corresponding to the normal closing process of the iron core. When the energy storage motor performed the closing spring energy storage, the average operating current of the energy storage motor was 2.3 amps, the peak current was 4.1 amps, and the total energy storage time was 14 seconds.

[0031] In one embodiment of this application, step S120, which determines the contact wear, circuit resistance increase, opening and closing time deviation, insulation dielectric loss factor, and energy storage attenuation based on multi-source historical operating data, further includes the following steps: Figure 2 As shown, the specific content is as follows: Step S210: Determine the total historical accumulated arc energy based on the load current and arcing time, and determine the contact wear amount based on the total historical accumulated arc energy; Step S220: Determine the loop resistance value based on the loop voltage difference and loop current, and determine the loop resistance increment based on the loop resistance value; Step S230: Determine the opening and closing time deviation based on the current waveform of the opening and closing coils, and correct the opening and closing time deviation based on the contact wear and the cumulative amplitude of temperature rise; Step S240: Obtain the phase angle between the voltage across the insulating medium and the leakage current across the insulating medium, determine the insulating medium loss factor based on the phase angle, and correct the insulating medium loss factor based on ambient temperature, ambient humidity, cumulative temperature rise amplitude, and loop resistance increase. Step S250: Obtain the theoretical work value of the energy storage motor, determine the energy storage attenuation based on the theoretical work value and the operating current, and correct the energy storage attenuation based on the load current and the cumulative temperature rise.

[0032] Specifically, the interrupted load current and corresponding arcing time for each tripping operation can be extracted from multi-source historical operating data. First, the arc energy of a single tripping operation is calculated, and then the total historical accumulated arc energy is obtained by summing all historical tripping operations. Therefore, the single arc energy... The calculation formula is:

[0033] In the formula, The arc voltage constant is... This represents the load current at the moment of circuit breaker tripping. This refers to the duration of a single arc ignition.

[0034] Therefore, the total energy accumulated throughout history of electric arcs The sum of the arc energy from all opening operations within the statistical period:

[0035] Then, contact wear amount The calculation formula is:

[0036] In the formula, The contact wear conversion factor can be calibrated based on the ablation test of copper-tungsten contact materials commonly used in 10kV primary and secondary integrated pole-mounted circuit breakers.

[0037] Furthermore, the loop voltage difference and corresponding loop current under steady-state closing conditions can be extracted, and the DC resistance value of the current main circuit can be calculated based on Ohm's law. Then, a preset initial loop resistance reference value is retrieved. Calculate the increment of the current loop resistance relative to the reference value, i.e., the increase in loop resistance. The calculation formula is:

[0038] In the formula, For the loop voltage difference, For loop current, This is the initial circuit resistance reference value, which is the typical factory-set circuit resistance value of the pole-mounted circuit breaker.

[0039] Furthermore, feature point identification is performed on the current waveform of the opening and closing coils. The time difference between the start of energization of the positioning coil and the moment when the iron core reaches its position is the current opening and closing action time. Retrieve the preset opening and closing reference time Calculate the original opening and closing time deviation. The formula is as follows:

[0040] Subsequently, correction factors for contact wear and cumulative temperature rise were introduced to calibrate the original deviation. Contact wear reduces contact overtravel and slightly shortens the operating time; increased temperature rise increases coil resistance, reduces pull-in force, and slows down the operating time. Therefore, the corrected opening and closing time deviation... The calculation formula is:

[0041] In the formula, This is the contact wear correction factor. The cumulative temperature rise correction factor was obtained through environmental testing of the same type of circuit breaker. This refers to the amount of contact wear. This represents the cumulative temperature rise.

[0042] Furthermore, the power frequency voltage signal and the corresponding leakage current signal at both ends of the insulating medium are collected, and the phase difference between the voltage and the current is calculated. Insulating dielectric loss factor The tangent of this phase angle is calculated using the following formula:

[0043] The original dielectric loss factor is significantly affected by ambient temperature and humidity, temperature rise, and loop resistance, requiring multi-factor correction. High temperature and high humidity can lead to inflated dielectric loss measurements, and the additional heat generated by increased loop resistance accelerates insulation aging. The corrected insulation dielectric loss factor... The calculation formula is:

[0044] In the formula, , , , These are the correction factors corresponding to ambient temperature, ambient humidity, loop resistance, and cumulative temperature rise, respectively. For reference temperature, This is for reference humidity.

[0045] Among these, the correction factor for ambient temperature is used as a subtraction correction term in the correction formula because high outdoor temperatures accelerate the dipole polarization of epoxy resin. Under the same actual aging degree of insulation, the original dielectric loss factor measured at high temperatures is higher. Its magnitude needs to match the temperature aging test characteristics of epoxy resin. The correction factor for ambient humidity is also used as a subtraction correction term in the correction formula because high humidity forms a conductive water film on the insulation surface, increasing leakage current and raising the measured value of the original dielectric loss. Its magnitude needs to match the insulation's moisture polarization mechanism. The correction factor for increased loop resistance is used as an addition correction term in the correction formula because increased loop resistance represents contact degradation. Long-term current carrying by the equipment will continuously generate additional heat, accelerating irreversible aging of the insulation and causing an increase in the actual insulation loss. This factor needs to match the steady-state heat generation power increment corresponding to the increase in loop resistance, reflecting the actual insulation aging increment caused by long-term conductive heat generation. Regarding the correction factor for the cumulative temperature rise amplitude, since the cumulative temperature rise amplitude is the long-term integral total of the heat generated by the current carrier, continuous high-temperature accumulation will damage the epoxy resin molecular structure and permanently increase the dielectric loss, which is an irreversible aging increment. Therefore, it is used as an additional correction term in the correction formula. The unit of the coefficient matches the cumulative temperature rise amplitude and is obtained by fitting long-term accelerated thermal aging test data.

[0046] Furthermore, the preset rated theoretical work value of the energy storage motor is retrieved. This serves as the ideal energy benchmark for the energy storage system. Subsequently, the operating current waveform of the energy storage motor is integrated, and the actual input electrical energy is calculated by combining this with the motor's rated voltage. This is then converted into the actual stored energy value of the spring. The calculation formula is as follows:

[0047] In the formula, The rated voltage of the energy storage motor, This is the motor operating current. This refers to the duration of a single energy storage session. This represents the initial transmission efficiency of the energy storage system.

[0048] The energy storage attenuation is the relative difference between the theoretical work done and the actual effective energy storage, and its calculation formula is as follows:

[0049] In the formula, This refers to the energy storage attenuation. Doing work for theory For practical and effective energy storage.

[0050] The physical quantities generated in step S120 are explained in detail below: Contact wear is the thickness of the contact surface material lost due to the high-temperature erosion of the opening arc, measured in millimeters. It is a parameter reflecting the remaining electrical life of the arc-extinguishing chamber contacts; a larger value indicates more severe contact electrical degradation. Circuit resistance increment is the increase in the main circuit resistance of the circuit breaker relative to the factory initial reference value, measured in microohms. It reflects the cumulative degree of contact degradation and conductor loosening; a larger value indicates more significant degradation of circuit conductivity. Opening and closing time deviation is the difference between the current opening and closing action time and the factory reference time, measured in milliseconds. It reflects the degree of mechanical performance degradation of the spring operating mechanism; a positive value indicates delayed action, and a negative value indicates premature action. Insulation dielectric loss factor is a dimensionless parameter characterizing the degree of power loss of the insulation dielectric, i.e., the tangent of the dielectric loss angle. It reflects the degree of aging, moisture absorption, and degradation of the insulation material; a larger value indicates poorer insulation performance. Energy storage attenuation is the percentage reduction in the actual effective energy stored in a spring relative to its theoretical work value. It reflects the degree of overall performance degradation of the energy storage motor, transmission mechanism, and closing spring. The larger the value, the more significant the decrease in the efficiency of the energy storage system.

[0051] For example, taking a 10kV integrated primary and secondary pole-mounted circuit breaker as an example, the equipment has a rated current of 630A and has been in operation for 3 years. Multi-source historical operating data from the past 90 days is extracted for calculation. A total of 12 tripping operations were performed in the past 90 days, with an average interrupted load current of 350 Amps, an average arcing time of 8 milliseconds, and an arc voltage of 200V. The average arc energy per operation = 200 × 350 × 0.008 = 560 Joules, and the total historical cumulative arc energy = 12 × 560 = 6720 Joules. The calculated contact wear amount = The voltage difference in the circuit under steady-state closing is 11.96 millivolts, corresponding to a circuit current of 230 amperes. The current circuit resistance is calculated as 0.01196 / 230 ≈ 52 microohms. The initial circuit resistance reference value is 45 microohms, therefore the circuit resistance increase is 52 - 45 = 7 microohms. Next, the current tripping action time is extracted from the tripping coil current waveform as 49 milliseconds, the reference tripping time is 45 milliseconds, and the original time deviation is 49 - 45 = 4 milliseconds. Combining the contact wear of 0.054 mm and the cumulative temperature rise of 1200 Kelvin per hour, after calibration with a correction factor, the final tripping and closing time deviation is 3.2 milliseconds. The phase angle between the insulation dielectric voltage and the leakage current corresponds to an original dielectric loss factor of 0.008. The current ambient temperature is 30 degrees Celsius and the relative humidity is 70%. The reference room temperature conditions are 20 degrees Celsius and 50% relative humidity. Combining the circuit resistance increase of 7 microohms and the cumulative temperature rise of 1200 Kelvin per hour, after multi-factor correction, the insulation dielectric loss factor correction value is 0.0062, more accurately reflecting the degree of insulation aging. The energy storage motor has a rated voltage of 24 volts, a single energy storage duration of 12 seconds, an average operating current of 2 amps, and an initial transmission efficiency of 0.2; the calculated actual energy storage value is 24 × 2 × 12 × 0.2 = 115.2 joules. The theoretical work value is 120 joules, and the initial energy storage attenuation is (120 - 115.2) / 120 × 100% = 4%; after correction based on the load current and cumulative temperature rise, the final energy storage attenuation is 3.5%, reflecting the degree of performance degradation of the energy storage system.

[0052] In one embodiment of this application, step S120, which involves determining the cumulative damage effect based on the opening and closing time deviation, the insulation dielectric loss factor, and the energy storage attenuation, further includes the following steps: Figure 3 As shown, the specific content is as follows: Step S310: Obtain the weights corresponding to the opening and closing time deviation, the insulation dielectric loss factor, and the energy storage attenuation amount, respectively. Step S320: Based on the weights corresponding to the opening and closing time deviation, the insulation dielectric loss factor, and the energy storage attenuation, perform a weighted calculation on the opening and closing time deviation, the insulation dielectric loss factor, and the energy storage attenuation, and use the result of the weighted calculation as the daily damage intensity. Step S330: Within a preset historical time period, perform cumulative calculation on the daily damage intensity and use the result of the cumulative calculation as the cumulative damage amount.

[0053] Specifically, three pre-calibrated weighting coefficients can be retrieved from the equipment parameter configuration module: the weight corresponding to the opening and closing time deviation, the weight corresponding to the insulation dielectric loss factor, and the weight corresponding to the energy storage attenuation. The weight for the opening and closing time deviation can be set based on the failure mode impact analysis of the integrated primary and secondary pole-mounted circuit breaker and historical fault statistics of the distribution network. Faults in the opening and closing mechanism account for the highest proportion of total equipment faults, directly determining whether the circuit breaker can reliably perform opening and closing operations, and thus having the greatest impact on equipment operational reliability. The weight for the insulation dielectric loss factor can be set based on the fact that insulation degradation is the second largest category of fault causes for pole-mounted circuit breakers. Temperature and humidity fluctuations and long-term temperature rises in outdoor environments continuously accelerate insulation aging, which can lead to insulation breakdown and flashover faults in severe cases. The weight for the energy storage attenuation can be set based on the fact that energy storage system performance degradation mainly affects the reliability of closing operations, and its probability of occurrence and degree of impact are relatively lower than faults in the opening and closing mechanism itself and insulation faults.

[0054] Furthermore, the daily opening and closing time deviation, insulation dielectric loss factor, and energy storage attenuation are normalized. A value of 0 is set as the lower limit for each parameter, and the corresponding failure threshold value as the upper limit. This converts parameters with different dimensions into relative damage values ​​within the range of 0 to 1, eliminating the incomparability caused by dimensional differences. Subsequently, a weighted summation operation is performed according to the corresponding weight coefficients of the three parameters. The operation logic is as follows: multiply the normalized opening and closing time deviation by its corresponding weight, multiply the normalized insulation dielectric loss factor by its corresponding weight, and multiply the normalized energy storage attenuation by its corresponding weight. The products of these three factors are then added together, and the final value is the daily damage intensity. The daily damage intensity ranges from 0 to 1; a higher value indicates more severe overall degradation damage to the equipment on that day.

[0055] Furthermore, using a preset historical time period as the statistical boundary, the daily damage intensity values ​​corresponding to all natural days within that time period are extracted. These values ​​are then cumulatively added together in chronological order to obtain the total damage cumulative effect. This cumulative damage cumulative effect reflects the total cumulative degradation damage level of the equipment across three dimensions—mechanical operation, insulation performance, and energy storage system—within the statistical period. It can be directly used for subsequent comparison with a preset cumulative threshold to determine whether key components have experienced performance degradation. The preset historical time period can be set to the past 90 days. The performance degradation of pole-mounted circuit breakers is a slow time-varying process; a 90-day statistical period can fully cover seasonal temperature and humidity fluctuations, load peak and valley changes, and the cumulative impact of multiple operations, thus fully reflecting the cumulative damage trend of the equipment.

[0056] For example, a simplified calculation is given using a 10kV integrated primary and secondary pole-mounted circuit breaker. The preset weights for the opening and closing time deviation are 0.4, the insulation dielectric loss factor is 0.35, and the energy storage attenuation is 0.25. The failure thresholds for these three parameters are 20 milliseconds, 0.02, and 25%, respectively. The preset historical time period is the past 90 days. Taking a typical operating day, the opening and closing time deviation is 8 milliseconds, the insulation dielectric loss factor is 0.007, and the energy storage attenuation is 6%. First, the three parameters are normalized by comparing them to their corresponding failure thresholds, resulting in relative damage values ​​of 0.4, 0.35, and 0.24. Then, a weighted summation is performed according to the corresponding weights, calculating the daily damage intensity for that day to be 0.3425. Following the same calculation logic, the daily damage intensity for each of the past 90 days is calculated sequentially. The values ​​for the 90 days are then summed up, ultimately yielding a cumulative damage effect of approximately 27.9 for the equipment.

[0057] In one embodiment of this application, in step S130, if the cumulative damage exceeds a preset threshold, it is determined that the pole-mounted circuit breaker exhibits a performance degradation trend, and a performance degradation sequence is determined based on contact wear, increased circuit resistance, opening and closing time deviation, and insulation dielectric loss factor. The step also includes the following steps: Figure 4 As shown, the specific content is as follows: Step S410: Based on the contact wear amount, the increase in circuit resistance, the opening and closing time deviation, and the insulation dielectric loss factor, a time stamp alignment algorithm is used to generate a time-aligned performance degradation sequence. Step S420: Use random forest to clean up anomalies in the performance degradation sequence to remove noise interference.

[0058] Specifically, the cumulative damage effect value can be compared with a preset effect threshold. If the cumulative damage effect value is significantly greater than the preset effect threshold, the pole-mounted circuit breaker is determined to have a performance degradation trend. If the cumulative damage effect value is less than or equal to the preset effect threshold, the pole-mounted circuit breaker is determined to be in a healthy operating phase and has not shown a clear degradation trend. Furthermore, to avoid misjudgments caused by single data fluctuations, the trend of cumulative damage effect value changes over three consecutive statistical periods can be reviewed simultaneously. If the value shows a continuous increase and exceeds the limit in the current period, the judgment conclusion is further confirmed. The preset effect threshold can be obtained based on historical operating data and fault case statistics of a large number of pole-mounted circuit breakers of the same model with integrated primary and secondary circuits. The sample covers outdoor operating equipment in different regions, with different years of operation and different load levels.

[0059] Furthermore, a timestamp alignment algorithm is employed to construct a unified time-series coordinate axis based on a preset 1-day time step, and differentiated resampling processing is performed for parameters with different sampling frequencies. For parameters with sampling frequencies higher than the daily level, the arithmetic mean of all sampled values ​​for that day is taken as the value for that time node. For parameters with sampling frequencies lower than the daily level and time intervals, linear interpolation is used to fill in the missing daily values. After resampling all data, the above four parameters at the same time node are concatenated into a multi-dimensional state vector in a fixed-dimensional order, and then all state vectors are arranged sequentially according to time order to generate a time-series fully aligned performance degradation sequence.

[0060] Furthermore, the time-aligned performance degradation sequence is input into a pre-trained random forest anomaly detection model. This model consists of multiple independent isolated decision trees, each of which performs path depth calculation for each data point in the sequence. The average path depth of each data point across all decision trees is converted into an anomaly score in the range of 0 to 1. Data points with anomaly scores greater than a preset anomaly threshold are marked as noise anomalies. These anomalies are typically caused by transient sensor interference, communication transmission jumps, or transient operating condition fluctuations, and do not represent the true performance degradation trend of the equipment. After identifying all anomalies, the system replaces the anomaly point values ​​with linear interpolation of adjacent normal points, completing the smoothing repair of the sequence and finally outputting a high-performance quality degradation sequence after removing noise interference.

[0061] For example, taking a 10kV integrated primary and secondary pole-mounted circuit breaker as an example, the calculated cumulative damage effect over a 90-day period is 27.9, exceeding the preset threshold of 25, indicating a performance degradation trend in the pole-mounted circuit breaker. Four degradation parameters from the past 90 days are extracted: contact wear and circuit resistance increase are calculated daily at midnight; opening and closing time deviation data are only generated on opening and closing operation days, with a valid value appearing on average every 3 to 5 days; insulation dielectric loss factor is collected hourly, with 24 sets of samples per day. A timestamp alignment algorithm is used to construct a timeline with a unified time step of one day. The final result is a time-series alignment performance degradation sequence containing 90 time nodes, each node corresponding to 4-dimensional parameters. For example, the state vector for day 30 is [0.032 mm, 5 μ ohms, 5 ms, 0.006], and the state vector for day 60 is [0.048 mm, 8 μ ohms, 7 ms, 0.0072].

[0062] In one embodiment of this application, step S140, which involves determining the continuous performance degradation trajectory of the pole-mounted circuit breaker using a long short-term memory network based on the performance degradation sequence, further includes the following steps: Figure 5 As shown, the specific content is as follows: Step S510: Determine the performance degradation feature vector based on the performance degradation sequence; Step S520: Based on the performance degradation feature vector, a long short-term memory network is used to generate a discrete degradation point set; Step S530: Based on the discrete degenerate point set, a cubic spline interpolation algorithm is used to generate a continuous performance degradation trajectory.

[0063] Specifically, a sliding window of preset length is used to perform frame-by-frame sliding truncation on the performance degradation sequence in chronological order. For each continuous subsequence segment, the max-min normalization process is performed on all parameters within the subsequence to uniformly map parameters of different dimensions to the interval between 0 and 1. Then, the parameters are concatenated in chronological and dimensional order to form a one-dimensional performance degradation feature vector, and the performance degradation feature vectors corresponding to all time steps are generated in sequence.

[0064] Furthermore, the performance degradation feature vectors are input into a pre-trained Long Short-Term Memory (LSTM) network model in chronological order. After entering the network model, the feature data undergoes a complete gating process: the input gate filters valid input information, the forget gate filters redundant historical information, the cell state updates the memory of long-term degradation patterns, and the output gate extracts the degradation features at the current moment. This process accurately captures the long-term temporal dependencies and nonlinear evolution patterns in the performance degradation sequence. The model ultimately outputs a one-dimensional comprehensive degradation state value for each time step. All output degradation state values ​​are matched one-to-one with their corresponding time nodes and arranged chronologically to form a discrete degradation point set. This discrete point set completely covers all time nodes of the preset historical time period to reconstruct the discrete evolution state of the device degradation process.

[0065] Furthermore, all time points and corresponding comprehensive degradation state values ​​in the discrete degradation point set are used as interpolation control points, and a cubic spline interpolation algorithm is invoked to perform trajectory smoothing. The interpolation algorithm constructs a cubic polynomial function between every two adjacent discrete control points. By constraining the function values, first derivative values, and second derivative values ​​of adjacent piecewise functions at the connection points to be continuous and equal, it ensures a smooth transition of the entire trajectory at all control points, without any broken line inflection points or abrupt slope changes. Finally, a continuous performance degradation trajectory is generated that covers the entire statistical time period and allows querying of the corresponding degradation state value at any time point.

[0066] For example, taking a 10kV integrated primary and secondary pole-mounted circuit breaker as an example, the equipment has been determined to have a performance degradation trend. The corresponding 90-day performance degradation sequence has been anomaly cleaned. Each time point includes four parameters: contact wear, circuit resistance increase, opening and closing time deviation, and insulation dielectric loss factor. The preset sliding window length is 30 days, the number of hidden layer units in the long short-term memory network is 64, and the cubic spline adopts natural spline boundary conditions. With a window length of 30 days, sliding truncation starts from day 1 to day 30, generating a subsequence every day. After performing max-min normalization on the four parameters in each subsequence, they are spliced ​​into a 120-dimensional performance degradation feature vector, finally generating 61 sets of performance degradation feature vectors corresponding to days 30 to 90. Sixty-one feature vectors are sequentially input into a pre-trained Long Short-Term Memory (LSTM) network, which outputs the comprehensive degradation state values ​​for corresponding time nodes in sequence. For example, day 30 corresponds to 0.32, day 60 to 0.54, and day 90 to 0.73. The degradation values ​​for all 61 time nodes are arranged chronologically, forming a discrete degradation point set covering days 30 to 90. Using these 61 discrete points as control points, cubic spline interpolation is performed using natural spline boundary conditions to generate smooth cubic curve segments between every two adjacent day nodes. All segmented curves are connected end-to-end and have continuous derivatives, ultimately resulting in a smooth and continuous performance degradation trajectory covering the 61-day time range.

[0067] In one embodiment of this application, step S140, which determines the dynamic rate of degradation at each time point based on the continuous performance degradation trajectory, further includes the following steps: Figure 6 As shown, the specific content is as follows: Step S610: Generate a time point sequence based on the continuous performance degradation trajectory, and determine the node drop value corresponding to each time node based on the time point sequence; Step S620: If the node drop value exceeds the preset drop threshold, the current node drop value is smoothed to determine the smoothed node drop value, and the smoothed node drop value is used as the degradation rate component corresponding to each time node. Step S630: In the continuous performance degradation trajectory, calculate the change in the trajectory tangent direction at adjacent time nodes to generate the degradation acceleration component corresponding to each time node; Step S640: Based on the degradation velocity component and the degradation acceleration component, a random forest regression algorithm is used for nonlinear fusion to generate the degradation dynamic change rate corresponding to each time point.

[0068] Specifically, the complete time range and corresponding state values ​​of the continuous performance degradation trajectory can be read. Following a preset 1-day sampling step, sampling is performed at equal intervals from the start to the end of the trajectory, extracting a set of time nodes with uniform time intervals arranged in chronological order. All nodes together constitute a time point sequence, and each time node corresponds to a specific comprehensive degradation state value. Subsequently, a first-order backward difference operation is performed on the state values ​​corresponding to the time point sequence, that is, the comprehensive degradation state value of the next time node is subtracted from the comprehensive degradation state value of the previous time node. The difference obtained is the node drop value corresponding to the next node. The node drop value directly reflects the change in the comprehensive degradation state of the integrated primary and secondary pole-mounted circuit breakers within two adjacent days. Positive values ​​indicate a continuously deepening degradation degree, and larger values ​​indicate a more significant daily degradation amplitude.

[0069] Furthermore, the node drop value corresponding to each time node is compared with a preset drop threshold one by one. If the node drop value is less than or equal to the preset drop threshold, it indicates that the drop falls within the normal range of gradual degradation fluctuations and there is no obvious noise interference. The node drop value is directly used as the degradation rate component of the corresponding time node. If the node drop value exceeds the preset drop threshold, it indicates that the drop likely includes abnormal jumps caused by instantaneous operating condition fluctuations and measurement interference, and cannot represent the true degradation rate. Then, a preset 3-point moving average window is called to perform an arithmetic mean operation on the node drop values ​​of the current node and its adjacent nodes before and after it to calculate a smoothed node drop value. The smoothed node drop value is used as the degradation rate component of the current time node. The degradation rate component characterizes the rate of degradation of the primary and secondary integrated pole-mounted circuit breaker at the corresponding time node, corresponding to the first-order change characteristics of the degradation process. The preset drop threshold can be set based on statistical analysis of historical data from a large number of devices of the same model during their healthy operation phase.

[0070] Furthermore, based on the continuously differentiable continuous performance degradation trajectory, the slope of the tangent line at each time node is calculated through differentiation. This slope represents the instantaneous degradation velocity of the corresponding node. Subsequently, the system performs a first-order backward difference operation on the instantaneous degradation velocities of all nodes. That is, the instantaneous tangent slope of the next time node is subtracted from the instantaneous tangent slope of the previous time node. The difference is the change in the tangent direction of the trajectory corresponding to the next node, which is the degradation acceleration component of the corresponding time node. The degradation acceleration component characterizes the rate of change of the degradation velocity itself, corresponding to the second-order change characteristics of the degradation process. A positive value indicates that the degradation velocity continues to accelerate, and the equipment enters the accelerated degradation stage; a negative value indicates that the degradation velocity gradually slows down, and the degradation trend of the equipment tends to stabilize.

[0071] Furthermore, the degradation velocity component and degradation acceleration component corresponding to each time point are combined into a two-dimensional feature vector, which is then input into the pre-trained random forest regression model. Each decision tree within the model independently calculates a predicted rate of change based on the input two-dimensional features. The arithmetic mean of the outputs of all decision trees is then calculated, and the final average value is the dynamic rate of change of degradation corresponding to that time point. The calculation results of all time points are arranged sequentially to form a complete sequence of dynamic rate of change. The random forest regression algorithm can automatically fit the nonlinear coupling effect of degradation velocity and degradation acceleration on the degradation trend. Compared with linear weighted summation, it better reflects the true physical laws of equipment degradation. The output dynamic rate of change of degradation comprehensively reflects the degradation magnitude and acceleration trend, and can be directly used in the subsequent abnormal offset judgment process.

[0072] For example, taking a 10kV integrated primary and secondary pole-mounted circuit breaker as an example, this equipment has generated a continuous performance degradation trajectory covering the past 90 days, with the comprehensive degradation state value gradually increasing from 0.2 on day 1. The preset time point sequence sampling step size is 1 day, the preset drop threshold is 0.02, the smoothing process uses a 3-point moving average window, and the random forest regression model contains 100 decision trees. 90 time points are extracted from the continuous performance degradation trajectory with a step size of 1 day to form a time point sequence. The node drop value of each node is calculated through differential operations. For example, the comprehensive degradation state value on day 30 is 0.32, the comprehensive degradation state value on day 31 is 0.332, and the corresponding node drop value on day 31 is 0.012. The comprehensive degradation state value on day 59 is 0.52, and on day 60, due to instantaneous measurement interference, the comprehensive degradation state value jumps to 0.55, and the corresponding node drop value on day 60 is 0.03. Each node drop value was compared with a preset drop threshold. The node drop value on day 31, 0.012, was less than the preset drop threshold, so 0.012 was directly used as the degradation rate component for day 31. The node drop value on day 60, 0.03, exceeded the preset drop threshold, triggering smoothing processing. The arithmetic mean of the node drop values ​​on days 59 (0.015), 60 (0.03), and 61 (0.018) was calculated, resulting in a smoothed node drop value of approximately 0.021. This value was used as the degradation rate component for day 60, effectively eliminating noise interference from instantaneous jumps.

[0073] Next, the slope of the tangent line of the continuous performance degradation trajectory at each time point is calculated. For example, the slope of the tangent line on day 50 is 0.014 per day, and the slope on day 51 is 0.017 per day. The difference between the two is 0.003, which is the degradation acceleration component of day 51, 0.003. This value is positive, indicating that the degradation rate of the equipment at the corresponding stage is gradually accelerating, and it has entered the accelerated degradation stage from the uniform degradation stage. The degradation rate component and degradation acceleration component of each node are combined into a two-dimensional feature and input into the random forest regression model. For example, the degradation rate component on day 51 is 0.017, and the degradation acceleration component is 0.003. After nonlinear fusion calculation using 100 decision trees, the degradation dynamic change rate for that day is output as 0.28. The smoothed degradation rate component on day 60 is 0.021, and the degradation acceleration component is 0.006. After fusion, the degradation dynamic change rate is output as 0.42.

[0074] In one embodiment of this application, in step S150, if the degradation dynamic change rate exceeds a preset change rate threshold, it is determined that the pole-mounted circuit breaker has an accelerated performance degradation trend, and the trajectory corresponding to the current time point is marked as an abnormal degradation trajectory.

[0075] Specifically, the degradation dynamic change rate at the current time point is compared with a preset change rate threshold. If the degradation dynamic change rate is significantly greater than the preset change rate threshold, the integrated primary and secondary pole-mounted circuit breaker is determined to have an accelerated performance degradation trend. Optionally, to avoid misjudgments caused by single-point data fluctuations or instantaneous operating condition interference, the degradation dynamic change rate values ​​of the previous three consecutive time points can be simultaneously traced back. If the values ​​show a continuous upward trend and the current point officially exceeds the limit, the accelerated performance degradation trend is also confirmed. The preset change rate threshold k is statistically calibrated based on a large dataset of full lifecycle degradation data of the same model of integrated primary and secondary pole-mounted circuit breakers. The sample must cover outdoor operating equipment with different operating years, different load levels, and different environmental conditions.

[0076] After confirming the accelerated performance degradation trend, the first time point of consecutive limit violations is taken as the anomaly starting point, and the entire segment of continuous performance degradation trajectory after this starting point is marked as an anomalous degradation trajectory. In addition, attribute tags are added to this segment of anomalous degradation trajectory, including the anomaly start time, the current peak dynamic rate of degradation change, and the cumulative duration of limit violations. The marked anomalous degradation trajectory is then archived and stored independently.

[0077] For example, taking a 10kV integrated primary and secondary pole-mounted circuit breaker installed on a heavy-load line in a suburban area as an example, this integrated primary and secondary pole-mounted circuit breaker has a rated current of 630A and has been in operation for 3 years, operating under high load conditions for a long period in summer. The preset threshold value for the rate of change is 0.35, which is lowered to 0.3 under heavy-load conditions. During the initial stable operation phase, the degradation dynamic rate of change was stable in the range of 0.20 to 0.28. The degradation dynamic rate of change was 0.31 on day 62, 0.32 on day 63, 0.34 on day 64, and reached 0.38 on day 65. Among them, the value of the reduced preset threshold of 0.33 was first exceeded on day 64, and the value further increased on day 65. The value exceeded the limit for two consecutive time points and showed a continuous upward trend, which finally confirmed that this integrated primary and secondary pole-mounted circuit breaker had an accelerated performance degradation trend. After completing the trend determination, day 64 is taken as the starting point of the anomaly. The continuous performance degradation trajectory segments from day 64 onwards are marked as abnormal degradation trajectories. The attribute information such as the anomaly start time and the current peak degradation dynamic change rate of 0.38 are simultaneously labeled and stored independently in the abnormal trajectory database.

[0078] In one embodiment of this application, step S160, determining the abnormal offset trajectory data based on the abnormal degradation trajectory, further includes the following steps: Figure 7 As shown, the specific content is as follows: Step S710: Obtain the normal degradation trajectory and determine the offset amplitude time series based on the normal degradation trajectory and the abnormal degradation trajectory; Step S720: Determine the offset duration and offset amplitude peak value based on the offset amplitude sequence, and determine the offset fluctuation frequency distribution based on the offset amplitude time series using Fast Fourier Transform; Step S730: Construct abnormal offset trajectory data based on offset duration, peak offset amplitude, and offset fluctuation frequency distribution.

[0079] Specifically, a normal degradation trajectory matching the current equipment model and operating conditions can be obtained to confirm that the time range and sampling step size of the normal degradation trajectory completely correspond to the abnormal degradation trajectory. Then, using the timestamp as an alignment reference, the abnormal degradation trajectory and the normal degradation trajectory are mapped onto the same time-series coordinate axis. For each time node, a difference operation is performed: the comprehensive degradation state value of the abnormal degradation trajectory at that node is subtracted from the baseline degradation state value of the corresponding node of the normal degradation trajectory, yielding the offset magnitude value for that time node. All offset magnitude values ​​for all time nodes are arranged in chronological order to form an offset magnitude time series.

[0080] Furthermore, temporal feature extraction is performed on the offset amplitude time series: Traversing backwards from the beginning of the sequence, the time node where the offset amplitude first exceeds the offset judgment threshold is located, denoted as the offset start time. Traversing backwards from the end of the sequence, the time node where the offset amplitude last exceeds the offset judgment threshold is located, denoted as the offset end time. The time span between these two times is the offset duration. Simultaneously, all values ​​in the offset amplitude time series are traversed, and the maximum value is found; this maximum value is the offset amplitude peak. After completing the temporal feature extraction, the system calls the Fast Fourier Transform (FFT) algorithm to perform frequency domain transformation on the offset amplitude time series: first, the sequence is mean-removed, then a Hanning window function is applied to suppress spectral leakage, followed by an FFT operation to obtain the spectral distribution of the offset amplitude sequence. The energy proportion of each frequency component is statistically analyzed, and the dominant and secondary dominant frequencies with the highest energy proportions are identified, ultimately forming the offset fluctuation frequency distribution. Finally, the output offset duration, offset amplitude peak, and offset fluctuation frequency distribution are structurally concatenated according to a preset dimensional order to construct the abnormal offset trajectory data.

[0081] For example, taking a 10kV integrated primary and secondary pole-mounted circuit breaker as an example, this equipment has been determined to have an accelerated performance degradation trend, with the corresponding abnormal degradation trajectory covering a time range of 27 days from day 64 to day 90. The preset offset judgment threshold is 0.05, and the Fast Fourier Transform uses a 256-point Hanning window configuration. The normal degradation trajectory is constructed from historical data of 50 healthy devices of the same model. Normal degradation trajectories of the same model are retrieved from the benchmark library and aligned point-by-point with the abnormal degradation trajectory using timestamps as a reference, with the offset magnitude calculated daily. For example, on day 64, the abnormal trajectory degradation state value is 0.38, the normal trajectory benchmark value is 0.32, and the offset magnitude is 0.06; on day 75, the abnormal trajectory degradation state value is 0.52, the normal trajectory benchmark value is 0.40, and the offset magnitude is 0.12; on day 90, the abnormal trajectory degradation state value is 0.73, the normal trajectory benchmark value is 0.55, and the offset magnitude is 0.18. Finally, a time series of offset magnitudes containing 27 time nodes is generated. By traversing the time series of offset amplitudes, day 64 was identified as the start of the offset (the offset amplitude of 0.06 first exceeded the threshold of 0.05), and day 90 as the end of the offset, resulting in a calculated offset duration of 27 days. The maximum offset amplitude of 0.18 was found to occur on day 90. A Fast Fourier Transform was then performed on the offset amplitude time series, and a Hanning window was applied to obtain the spectral distribution. Analysis revealed that the dominant frequency with the highest energy proportion was approximately 0.14 weeks per day, corresponding to a fluctuation cycle of about 7 days, while the secondary dominant frequency was approximately 0.07 weeks per day, corresponding to a fluctuation cycle of about 14 days, thus obtaining the offset fluctuation frequency distribution. The three features—offset duration of 27 days, peak offset amplitude of 0.18, and offset fluctuation frequency distribution (dominant frequency 0.14 weeks per day, secondary dominant frequency 0.07 weeks per day)—were structurally concatenated in dimensional order, and metadata such as device number and anomaly start time were added to construct a complete abnormal offset trajectory data.

[0082] In one embodiment of this application, step S160, determining the fault prediction result of the pole-mounted circuit breaker based on the abnormal offset trajectory data, further includes the following steps: Figure 8 As shown, the specific content is as follows: Step S810: Based on the abnormal offset trajectory data, a long short-term memory network is used to determine the fault prediction value of the pole-mounted circuit breaker; Step S820: Determine the fault prediction result based on the fault prediction value. The fault prediction result includes red warning, yellow warning and blue warning.

[0083] Specifically, structured abnormal offset trajectory data can be read to extract offset duration, peak offset amplitude, and offset fluctuation frequency distribution. Simultaneously, the historical performance degradation sequence of the corresponding primary and secondary integrated pole-mounted circuit breakers is retrieved as time-series basis data. The abnormal features and historical time-series data are concatenated and combined to form a multi-dimensional extrapolation feature set conforming to the model input format. This extrapolation feature set is then input into a pre-trained Long Short-Term Memory (LSTM) network extrapolation model. The model relies on a gating mechanism to memorize long-term degradation patterns and, combined with the acceleration characteristics of the current abnormal offset, uses an autoregressive rolling prediction method to progressively extrapolate the comprehensive degradation state value for each day within a preset 90-day time window, outputting a complete future degradation state prediction sequence. All values ​​in the prediction sequence are traversed, and the maximum value of the comprehensive degradation state is extracted as the fault prediction value, while simultaneously recording the expected occurrence time corresponding to the predicted value.

[0084] Furthermore, the system reads the predicted fault values ​​and compares them with the preset three-level warning thresholds in descending order of priority. If the predicted fault value is greater than or equal to the red warning threshold, it is determined to be at the highest risk level, generating a red warning fault prediction result, along with a maintenance and repair recommendation for immediate power outage. If the predicted fault value is within the yellow warning threshold range, it is determined to be at a medium risk level, generating a yellow warning fault prediction result, along with a recommendation to include it in the recent maintenance plan and increase status monitoring. If the predicted fault value is within the blue warning threshold range, it is determined to be at a low risk level, generating a blue warning fault prediction result, along with a recommendation to increase monitoring priority and track degradation trends. The three-level warning thresholds need to be set to minimize both the false negative rate and the missed detection rate. For example, the blue warning threshold ensures that early signs of accelerated degradation are not missed, the false alarm rate for long-term equipment deterioration is controlled within 10%, and the false alarm probability for normal equipment is less than 12%; the yellow warning threshold accurately identifies moderate degradation, achieves 85% sensitivity in identifying functional faults within six months, and has no large number of invalid maintenance prompts; the red warning threshold provides 100% coverage of equipment nearing failure, has near-zero false alarm rate for high-risk equipment, and only a very small number of instantaneous noises cause false alarms. After completing the level determination, the warning level, fault prediction value, estimated risk time, and maintenance suggestions can be integrated into a complete fault prediction result and simultaneously pushed to the distribution network maintenance dispatch system and equipment ledger management system.

[0085] For example, taking a 10kV integrated primary and secondary pole-mounted circuit breaker installed on a heavy-load line in a suburban area as an example, the corresponding abnormal offset trajectory data includes an offset duration of 27 days, a peak offset amplitude of 0.18, and a dominant fluctuation frequency of 0.14 weeks per day. The preset prediction time window is the next 90 days, and the three warning thresholds are: red warning threshold of 0.85, yellow warning threshold range of 0.65-0.85, and blue warning threshold range of 0.45-0.65. The three features of the abnormal offset trajectory are concatenated and combined with the performance degradation sequence of the equipment over the past 90 days, and input into a pre-trained two-layer long short-term memory network extrapolation model. The model outputs a degradation state prediction sequence for the next 90 days through autoregressive rolling extrapolation. The sequence shows that the comprehensive degradation state value reaches a peak of 0.92 on day 72, and it is predicted that the equipment will approach the critical failure state on day 68. The peak value of 0.92 is extracted as the fault prediction value of the integrated primary and secondary pole-mounted circuit breaker. The fault prediction value of 0.92 was compared with the level 3 warning threshold one by one. 0.92 is greater than the red warning threshold of 0.85, meeting the criteria for the highest risk level. A red warning fault prediction result was generated, along with maintenance recommendations: This integrated primary and secondary pole-mounted circuit breaker has an extremely high risk of arc chamber failure or insulation breakdown within the next 70 days. It is recommended to schedule a power outage for maintenance within 15 days to replace the contact assembly and seals, and to verify the operating mechanism's energy storage system. Simultaneously, this result was pushed to the distribution network maintenance dispatch platform and included in the emergency maintenance plan. If the fault prediction value of another integrated primary and secondary pole-mounted circuit breaker of the same model is 0.72, which falls within the yellow warning range of 0.65 to 0.85, the system will generate a yellow warning fault prediction result. It is recommended to include this equipment in the next quarter's maintenance plan and shorten the status monitoring cycle from 7 days to 2 days to track changes in degradation trends.

[0086] This application also provides a primary and secondary integrated pole-mounted circuit breaker fault prediction system. This system may include an acquisition module, a first determination module, a second determination module, a third determination module, a marking module, and a fourth determination module. The acquisition module acquires multi-source historical operating data of the pole-mounted circuit breaker, including load current, arcing time, circuit voltage difference, circuit current, opening and closing coil current waveforms, ambient temperature, ambient humidity, cumulative temperature rise, and the operating current of the energy storage motor. The first determination module determines contact wear, circuit resistance increase, opening and closing time deviation, insulation dielectric loss factor, and energy storage attenuation based on the multi-source historical operating data, and constructs a cumulative damage effect based on the opening and closing time deviation, insulation dielectric loss factor, and energy storage attenuation. The second determination module determines that if the cumulative damage effect exceeds a preset effect threshold, the pole-mounted circuit breaker exhibits a performance degradation trend. The system employs a first-order determination module to determine the performance degradation sequence based on contact wear, circuit resistance increase, opening and closing time deviation, and insulation dielectric loss factor. The second-order determination module uses a long short-term memory network to determine the continuous performance degradation trajectory of the pole-mounted circuit breaker based on the performance degradation sequence, and determines the dynamic rate of degradation change at each time point based on the continuous performance degradation trajectory. The third-order determination module identifies the dynamic rate of degradation if it exceeds a preset threshold, indicating an accelerated performance degradation trend in the pole-mounted circuit breaker, and marks the trajectory at the current time point as an abnormal degradation trajectory. The fourth-order determination module determines abnormal offset trajectory data based on the abnormal degradation trajectory and determines the fault prediction result of the pole-mounted circuit breaker based on the abnormal offset trajectory data.

[0087] It should be noted that the embodiments of the primary and secondary integrated pole-mounted circuit breaker fault prediction system provided in this application can be used to execute the processing flow of the embodiments of the primary and secondary integrated pole-mounted circuit breaker fault prediction method in the above embodiments. Its functions will not be repeated here, but can be referred to the detailed description of the above method embodiments.

[0088] This application also provides an electronic device, which includes one or more processors and memory resources represented by a memory for storing instructions executable by the processor, such as application programs. The application programs stored in the memory may include one or more modules, each corresponding to a set of instructions. Furthermore, the processor is configured to execute instructions to perform the aforementioned method for predicting faults in a primary and secondary integrated pole-mounted circuit breaker.

[0089] The electronic device may also include a power supply component configured to perform power management of the electronic device, a wired or wireless network interface configured to connect the electronic device to a network, and an input / output (I / O) interface. The electronic device can be operated based on operating devices stored in memory, such as Windows Server™, Mac OS X™, Unix™, Linux™, FreeBSD™, or similar.

[0090] In one embodiment, a computer device, which may be a server, is also provided. The computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is connected to the system bus via the I / O interfaces. The processor of the computer device provides computing and control capabilities. The memory of the computer device includes non-volatile storage media and internal memory. The non-volatile storage media stores an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The database of the computer device stores data. The I / O interfaces of the computer device are used for exchanging information between the processor and external devices. The communication interface of the computer device is used for communicating with external terminals via a network connection. When the computer program is executed by the processor, it implements a method for predicting faults in a primary and secondary integrated pole-mounted circuit breaker.

[0091] In one embodiment, a computer device is provided, which may be a terminal. The computer device includes a processor, memory, input / output interface, communication interface, display unit, and input device. The processor, memory, and input / output interface are connected via a system bus, and the communication interface, display unit, and input device are also connected to the system bus via the input / output interface. The processor of the computer device provides computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The input / output interface of the computer device is used for exchanging information between the processor and external devices. The communication interface of the computer device is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, mobile cellular networks, NFC (Near Field Communication), or other technologies. When the computer program is executed by the processor, it implements a method for predicting faults in a primary and secondary integrated pole-mounted circuit breaker. The display unit of the computer device is used to form a visually visible image and may be a display screen, a projection device, or a virtual reality imaging device. The display screen can be an LCD screen or an e-ink screen. The input device of the computer device can be a touch layer covering the display screen, or buttons, trackballs, or touchpads set on the casing of the computer device, or external keyboards, touchpads, or mice, etc.

[0092] This application also provides a non-transitory computer-readable storage medium, which, when the instructions in the storage medium are executed by the processor of the electronic device, enables the electronic device to perform a method for a primary and secondary integrated pole-mounted circuit breaker fault prediction system.

[0093] This application may take the form of a computer program product implemented on one or more storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing program code. Computer-readable storage media include permanent and non-permanent, removable and non-removable media, and information storage can be implemented by any method or technology. Information may be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to: phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transfer medium that can be used to store information accessible by a computing device.

[0094] It should be noted that although the steps of the primary and secondary integrated pole-mounted circuit breaker fault prediction method of this application are described in a specific order in the accompanying drawings, this does not require or imply that these steps must be performed in that specific order, or that all the steps shown must be performed to achieve the desired result. Additional or alternative steps, such as omitting certain steps, combining multiple steps into one step, and / or decomposing one step into multiple steps, should all be considered part of this application.

[0095] It should be understood that this application is not limited to the detailed structure and arrangement of the modules in the primary and secondary integrated pole-mounted circuit breaker fault prediction system proposed in this specification. This application can have other embodiments and can be implemented and executed in various ways. The foregoing variations and modifications fall within the scope of this application. It should be understood that the invention and definition of this application extend to all alternative combinations of two or more individual features mentioned or apparent in the text and / or drawings. All these different combinations constitute multiple alternative aspects of this application. The embodiments described in this specification illustrate the best known mode for implementing this application and will enable those skilled in the art to utilize this application.

Claims

1. A method for predicting faults in a primary and secondary integrated pole-mounted circuit breaker, characterized in that, include: Acquire multi-source historical operating data of the pole-mounted circuit breaker, including load current, arcing time, circuit voltage difference, circuit current, opening and closing coil current waveform, ambient temperature, ambient humidity, cumulative temperature rise amplitude, and operating current of the energy storage motor. Based on the multi-source historical operating data, the contact wear, circuit resistance increase, opening and closing time deviation, insulation dielectric loss factor, and energy storage attenuation are determined, and the cumulative damage effect is constructed based on the opening and closing time deviation, the insulation dielectric loss factor, and the energy storage attenuation. If the cumulative damage exceeds a preset threshold, it is determined that the pole-mounted circuit breaker has a performance degradation trend, and the performance degradation sequence is determined based on the contact wear, the increase in circuit resistance, the opening and closing time deviation, and the insulation dielectric loss factor. Based on the performance degradation sequence, a long short-term memory network is used to determine the continuous performance degradation trajectory of the pole-mounted circuit breaker, and the degradation dynamic change rate corresponding to each time point is determined according to the continuous performance degradation trajectory. If the degradation dynamic change rate exceeds the preset change rate threshold, it is determined that the pole-mounted circuit breaker has an accelerated performance degradation trend, and the trajectory corresponding to the current time point is marked as an abnormal degradation trajectory. Based on the abnormal degradation trajectory, abnormal offset trajectory data is determined, and based on the abnormal offset trajectory data, the fault prediction result of the pole-mounted circuit breaker is determined.

2. The method for fault prediction of a primary and secondary integrated pole-mounted circuit breaker according to claim 1, characterized in that, The determination of contact wear, circuit resistance increase, opening and closing time deviation, insulation dielectric loss factor, and energy storage attenuation based on the multi-source historical operating data includes: The total historical accumulated arc energy is determined based on the load current and the arcing time, and the contact wear is determined based on the total historical accumulated arc energy. The loop resistance value is determined based on the loop voltage difference and the loop current, and the loop resistance increment is determined based on the loop resistance value. The opening and closing time deviation is determined based on the current waveform of the opening and closing coil, and the opening and closing time deviation is corrected based on the contact wear and the cumulative temperature rise. The phase angle between the voltage across the insulating medium and the leakage current across the insulating medium is obtained. The loss factor of the insulating medium is determined based on the phase angle. The loss factor of the insulating medium is then corrected based on the ambient temperature, the ambient humidity, the cumulative temperature rise, and the increase in the loop resistance. Obtain the theoretical work value of the energy storage motor, determine the energy storage attenuation amount based on the theoretical work value and the operating current, and correct the energy storage attenuation amount based on the load current and the cumulative temperature rise.

3. The method for predicting faults in a primary and secondary integrated pole-mounted circuit breaker according to claim 1, characterized in that, The determination of the cumulative damage effect based on the opening and closing time deviation, the insulation dielectric loss factor, and the energy storage attenuation includes: The weights corresponding to the opening and closing time deviation, the insulation dielectric loss factor, and the energy storage attenuation are obtained respectively. Based on the weights corresponding to the opening and closing time deviation, the insulation dielectric loss factor, and the energy storage attenuation, a weighted calculation is performed on the opening and closing time deviation, the insulation dielectric loss factor, and the energy storage attenuation, and the result of the weighted calculation is used as the daily damage intensity. Within a preset historical time period, the daily damage intensity is accumulated, and the result of the accumulation is used as the cumulative damage amount.

4. The method for predicting faults in a primary and secondary integrated pole-mounted circuit breaker according to claim 1, characterized in that, The determination of the performance degradation sequence based on the contact wear, the increase in circuit resistance, the opening and closing time deviation, and the insulation dielectric loss factor includes: Based on the contact wear amount, the increase in circuit resistance, the opening and closing time deviation, and the insulation dielectric loss factor, a time-aligned performance degradation sequence is generated using a timestamp alignment algorithm. Random forests were used to clean the performance degradation sequences to remove noise interference.

5. The method for fault prediction of a primary and secondary integrated pole-mounted circuit breaker according to claim 1, characterized in that, The step of determining the continuous performance degradation trajectory of the pole-mounted circuit breaker using a long short-term memory network based on the performance degradation sequence includes: Determine the performance degradation feature vector based on the performance degradation sequence; Based on the aforementioned performance degradation feature vector, a Long Short-Term Memory (LSTM) network is used to generate a discrete degradation point set. Based on the discrete set of degradation points, the continuous performance degradation trajectory is generated using a cubic spline interpolation algorithm.

6. The fault prediction method for a primary and secondary integrated pole-mounted circuit breaker according to claim 1, characterized in that, The step of determining the dynamic rate of change of degradation at each time point based on the continuous performance degradation trajectory includes: A time point sequence is generated based on the continuous performance degradation trajectory, and the node drop value corresponding to each time node is determined based on the time point sequence. If the node drop value exceeds the preset drop threshold, the current node drop value is smoothed to determine the smoothed node drop value, and the smoothed node drop value is used as the degradation rate component corresponding to each time node. In the continuous performance degradation trajectory, the change in the trajectory tangent direction at adjacent time nodes is calculated to generate the degradation acceleration component corresponding to each time node; Based on the degradation velocity component and the degradation acceleration component, a random forest regression algorithm is used for nonlinear fusion to generate the degradation dynamic change rate corresponding to each time point.

7. The method for predicting faults in a primary and secondary integrated pole-mounted circuit breaker according to claim 1, characterized in that, The step of determining the abnormal offset trajectory data based on the abnormal degradation trajectory includes: Obtain the normal degradation trajectory, and determine the offset amplitude time series based on the normal degradation trajectory and the abnormal degradation trajectory; The offset duration and offset amplitude peak are determined based on the offset amplitude sequence, and the offset fluctuation frequency distribution is determined by fast Fourier transform based on the offset amplitude time series. The abnormal offset trajectory data is constructed based on the offset duration, the peak offset amplitude, and the offset fluctuation frequency distribution.

8. The method for predicting faults in a primary and secondary integrated pole-mounted circuit breaker according to claim 1, characterized in that, The step of determining the fault prediction result of the pole-mounted circuit breaker based on the abnormal offset trajectory data includes: Based on the abnormal offset trajectory data, a long short-term memory network is used to determine the fault prediction value of the pole-mounted circuit breaker; The fault prediction result is determined based on the fault prediction value, and the fault prediction result includes red warning, yellow warning and blue warning.

9. A primary and secondary integrated pole-mounted circuit breaker fault prediction system, characterized in that, include: The acquisition module is used to acquire multi-source historical operating data of the pole-mounted circuit breaker. The multi-source historical operating data includes load current, arcing time, circuit voltage difference, circuit current, opening and closing coil current waveform, ambient temperature, ambient humidity, cumulative temperature rise amplitude, and operating current of the energy storage motor. The first determining module is used to determine the contact wear amount, circuit resistance increase, opening and closing time deviation, insulation dielectric loss factor and energy storage attenuation amount based on the multi-source historical operating data, and to construct and determine the cumulative damage effect amount based on the opening and closing time deviation, the insulation dielectric loss factor and the energy storage attenuation amount. The second determining module is used to determine that the pole-mounted circuit breaker has a performance degradation trend if the cumulative damage exceeds a preset threshold, and to determine the performance degradation sequence based on the contact wear, the increase in circuit resistance, the opening and closing time deviation, and the insulation dielectric loss factor. The third determining module is used to determine the continuous performance degradation trajectory of the pole-mounted circuit breaker based on the performance degradation sequence using a long short-term memory network, and to determine the degradation dynamic change rate corresponding to each time point according to the continuous performance degradation trajectory. The marking module is used to determine that the pole-mounted circuit breaker has an accelerated performance degradation trend if the degradation dynamic change rate exceeds a preset change rate threshold, and to mark the trajectory corresponding to the current time point as an abnormal degradation trajectory. The fourth determining module is used to determine abnormal offset trajectory data based on the abnormal degradation trajectory, and to determine the fault prediction result of the pole-mounted circuit breaker based on the abnormal offset trajectory data.

Images