An embedded electric energy metering box structure parameter optimization method and system

Abstract

The application discloses an embedded electric energy metering box structure parameter optimization method and system, and belongs to the technical field of parameter optimization. The method obtains harmonic current characteristics and mechanical vibration characteristics generated by a stacking interface in the working process of an embedded electric energy metering box with communication interruption. A failure process digital reproduction and evaluation model is constructed, the harmonic current characteristics and the mechanical vibration characteristics are taken as model inputs, and a contact resistance growth rate is output. Based on a genetic algorithm, the failure process digital reproduction and evaluation model is optimized with the minimum contact resistance growth rate as the target. The embedded electric energy metering box structure parameters corresponding to the minimum contact resistance growth rate are taken as the embedded electric energy metering box structure parameter optimization result. Through the application, the structure and material parameter combination that can minimize the interface contact resistance growth rate can be systematically found, so that the vicious cycle of the failure process is inhibited at the source, and the long-term physical connection stability of the data communication link is ensured.

Description

Technical Field

[0001] This invention belongs to the field of parameter optimization technology, specifically relating to a method and system for optimizing the structural parameters of an embedded power metering box. Background Technology

[0002] An electricity metering box is a collection of metering instruments and auxiliary equipment necessary for measuring electrical energy, including electricity meters, voltage and current transformers and their secondary circuits, electricity metering panels, cabinets, and boxes. After the electricity metering boxes have been in operation for a period of time, the power grid management department may receive reports of abnormal electricity metering from some users' electricity metering boxes. The reports do not show significant deviations in electricity readings, but rather irregular, short-term data communication interruptions recorded in the system's background logs for specific electricity metering boxes. Initially, these records were attributed to network fluctuations or occasional software failures. On-site technicians inspected the problematic electricity metering boxes individually, replacing one or two units deemed "potentially defective," but the problem was not eradicated. Instead, over time, the number of electricity metering boxes with data communication interruption records slowly increased, and the frequency of interruptions also showed an upward trend.

[0003] Further investigation revealed that these power metering boxes exhibiting communication anomalies were not randomly distributed, but rather concentrated in specific cabinets, typically located in the upper-middle part of a stacked array. The users corresponding to these cabinets shared a common characteristic: a large number of inverters and switching power supplies. When these nonlinear loads are operating, they inject significant amounts of harmonic current into the power supply network, particularly high-order harmonic components such as the third, fifth, and seventh harmonics.

[0004] Harmonic currents are superimposed on the 50 Hz fundamental current and transmitted through a vertical power supply busbar formed by the stacked interfaces of the entire row of energy metering boxes. The stacked interfaces were originally designed to carry power frequency currents, and while the internal current-carrying contacts have sufficient design margins, they were not specifically designed for the characteristics of high-frequency currents. Therefore, when high-frequency harmonic currents flow through contacts with limited cross-sectional area, the skin effect becomes significant. The current tends to concentrate on the surface of the contacts, resulting in a reduction in the actual effective conductive area and a sharp increase in local current density, thus generating Joule heating at these tiny contact points that far exceeds expectations.

[0005] The aforementioned heat generation is not uniformly distributed across the entire interface, but rather highly concentrated on a few copper contacts responsible for carrying the main current. The signal contacts adjacent to them, used for data communication, and the polymer base serving as structural support and insulation, do not directly carry large currents and therefore have much lower temperatures. This creates a temperature gradient field at a microscopic scale. Due to the difference in thermal expansion coefficients between copper and the insulating polymer material, the thermal expansion of the current-carrying contacts is significantly greater than that of the surrounding base. This inconsistent expansion disrupts the uniformly distributed internal preload balance achieved at room temperature through precision mold design and assembly. This generates minute, continuous mechanical stress within the interface, increasing the pressure of the current-carrying contacts on their sockets, and potentially causing imperceptible shifts in the fit between the signal contacts and the base.

[0006] More importantly, other areas within the building where the electricity metering box causing the communication interruption is located may experience the start-up and shutdown of large equipment, or the high-speed rotation and feed of cutting tools during precision machine processing, all of which generate continuous, low-amplitude mechanical vibrations within the building structure. These vibrations are transmitted through the base of the cabinet to the vertically stacked array of electricity metering boxes. For an interface in an ideal state, this level of micro-vibration is usually insufficient to pose a threat. However, for an interface already under "tension" due to internal thermal stress, the situation is entirely different.

[0007] Continuous fretting, like a tireless hand, repeatedly applies minute shear and normal perturbations to the contact surface of contacts that have been excessively compressed due to thermal stress; this phenomenon is called fretting. Under the repeated action of fretting, the metal plating (usually tin or silver) on the contact surface is gradually worn away and oxidized, forming a thin oxide film that is either insulating or semiconductor—a process known as fretting corrosion. This oxide film directly leads to a slow increase in contact resistance.

[0008] This creates a vicious cycle: harmonic currents cause localized heating, which in turn leads to uneven thermal stress. This thermal stress, combined with environmental micro-vibrations, triggers fretting corrosion. Fretting corrosion increases contact resistance, and this increased resistance generates even more heat under the same current (according to P=I²R), further exacerbating localized temperature rise and thermal stress. This positive feedback process slowly erodes the electrical performance of the interface.

[0009] Ultimately, this cumulative effect began to manifest. On one hand, the continuous overheating of the current-carrying contacts caused the performance of the surrounding insulation materials to slowly deteriorate; on the other hand, the microstructural deformation caused by thermal stress and fretting began to affect the smaller data communication contacts, which required higher contact precision. The contact pressure of these signal contacts could become unstable, sometimes too high, and sometimes experiencing momentary poor contact due to slight deformation. This perfectly explains the intermittent, short-term communication interruptions recorded in the background logs. It was not a complete failure of a single component, but a systemic process of gradually decreasing connection reliability caused by the interplay of multiple physical phenomena involving electricity, heat, and mechanics. Field technicians could not solve the problem by simply replacing a single power meter box, because the root cause lay in the interface behavior of the entire stacked system under specific operating conditions, rather than a quality defect in a single device. Summary of the Invention

[0010] This invention provides a method and system for optimizing the structural parameters of an embedded power metering box. The aim is to systematically find a combination of structural and material parameters that minimizes the rate of increase in interface contact resistance, thereby suppressing the vicious cycle of failure at the source and ensuring the long-term physical connection stability of the data communication link.

[0011] The first objective of this invention is to provide a method for optimizing the structural parameters of an embedded energy metering box, comprising:

[0012] In an embedded power metering box that experiences communication interruption, the characteristics of harmonic current and mechanical vibration generated by the stacked interface during operation are obtained.

[0013] A digital reproduction and evaluation model of the failure process is constructed, with harmonic current characteristics and mechanical vibration characteristics as model inputs, and the contact resistance growth rate is output.

[0014] Based on a genetic algorithm, the model for digital reproduction and evaluation of the failure process is optimized with the goal of minimizing the contact resistance growth rate.

[0015] The structural parameters of the embedded energy metering box corresponding to the minimum contact resistance growth rate are taken as the optimization result of the embedded energy metering box structural parameters.

[0016] Furthermore, the characteristics of harmonic currents and mechanical vibrations generated by the stacking interface during operation in the embedded power metering box where communication is interrupted include:

[0017] In an embedded power metering box where communication is interrupted, a high-bandwidth current probe and a power quality analyzer are deployed to continuously collect the current waveform of the power supply busbar during the typical working cycle of the equipment. The harmonic current spectrum of the current waveform is obtained through Fourier transform, and the amplitude and phase information of the main harmonic components are recorded as harmonic current characteristics.

[0018] Accelerometers are installed on the columns, beams, and inside the embedded power metering box to collect vibration signals transmitted from the floor during operation; the vibration signals are subjected to spectrum analysis to identify the frequency range and amplitude of the energy concentration vibration, which serve as mechanical vibration characteristics.

[0019] The harmonic current characteristics are converted into a time-varying heat source generation rate applied to the finite element geometry of the stacked interface current-carrying contacts in the embedded energy metering box.

[0020] The mechanical vibration characteristics are converted into mechanical excitation functions applied to the support points of the stacked interface structure in the embedded power metering box.

[0021] Furthermore, a digital reproduction and evaluation model of the failure process is constructed, using harmonic current characteristics and mechanical vibration characteristics as model inputs, and outputting the contact resistance growth rate, including:

[0022] A digital reproduction and evaluation model of the failure process is constructed; the model includes a transient electrothermal analysis module, a thermodynamic analysis module, a forced vibration analysis module, and a failure evaluation module; among which,

[0023] Input the heat source generation rate into the transient electrothermal analysis module to calculate the temperature distribution inside the stacked interface of the embedded power metering box over time under the action of harmonic current, and obtain a fine temperature field.

[0024] Using a fine temperature field as a load, and based on a thermal analysis module, the stress distribution state inside the stacked interface of the embedded power metering box caused by uneven heating is calculated.

[0025] Using the mechanical excitation function as the load, and based on the forced vibration analysis module, the small relative sliding displacement generated between the contact surfaces of the stacked interfaces in the embedded power metering box under dual loads is calculated.

[0026] Inputting a small relative sliding displacement into the failure assessment module outputs the contact resistance growth rate.

[0027] Furthermore, the heat source generation rate is input into the transient electrothermal analysis module to calculate the temperature distribution inside the stacked interface of the embedded energy metering box over time under the action of harmonic current, obtaining a fine temperature field, including:

[0028] Temperature data from a first temperature sensor and a second temperature sensor, which are stacked on an embedded energy metering box, are collected. The first temperature sensor is located at the current-carrying contact of the stacked interface in the embedded energy metering box, and the distance between the second temperature sensor and the first temperature sensor is set to a preset distance.

[0029] Based on the temperature data from the first temperature sensor, the rate of temperature rise is determined, and based on the temperature data from the first temperature sensor and the temperature data from the second temperature sensor, the instantaneous temperature difference between the first temperature sensor and the second temperature sensor is determined.

[0030] The rate of temperature rise is compared with a preset first rate threshold, and the instantaneous temperature difference is compared with a preset first temperature difference threshold to determine whether there is a temperature rise event caused by high total current.

[0031] If it is determined that there is a temperature rise event caused by high total current, a dynamic temperature reference is established based on the temperature state after the temperature rise event occurs, and the second temperature rise rate and the second instantaneous temperature difference corresponding to the subsequent temperature change are calculated based on the dynamic temperature reference.

[0032] The second temperature rise rate is compared with a preset second rate threshold, and the second instantaneous temperature difference is compared with a preset second temperature difference threshold to determine whether there are superimposed high-order harmonic events. Combined with the judgment result of whether there is a temperature rise event caused by high total current, the type of event causing the temperature rise is determined.

[0033] Based on event type and harmonic current characteristics, the temperature distribution of the stacked interfaces in the embedded power metering box is determined, and a fine temperature field is obtained.

[0034] Furthermore, if a temperature rise event caused by a high total current is determined to exist, a dynamic temperature reference is established based on the temperature state after the occurrence of the temperature rise event. Based on the dynamic temperature reference, the second temperature rise rate and the second instantaneous temperature difference corresponding to subsequent temperature changes are calculated, including:

[0035] Based on the initial temperature change of the temperature rise event, the background heat component corresponding to the temperature rise event caused by high total current is determined.

[0036] Based on the temperature data from the first and second temperature sensors, the background thermal component is removed to obtain the residual temperature component, wherein the background thermal component is used as a dynamic temperature reference.

[0037] The second temperature rise rate and the second instantaneous temperature difference are calculated based on the residual temperature component.

[0038] Furthermore, based on the initial temperature change of the temperature rise event, the background thermal components corresponding to the temperature rise event caused by the high total current are determined, including:

[0039] In the initial stage of a temperature rise event, an initial temperature sequence consisting of multiple time-ordered temperature data is obtained.

[0040] For each temperature data point in the initial temperature sequence, a local trend of change is determined based on its neighboring data in the sequence;

[0041] Identify candidate segments with persistent changing trends from the initial temperature sequence;

[0042] For each candidate segment, a fluctuation index is determined to characterize the degree of fluctuation within the candidate segment.

[0043] The fluctuation index is compared with the preset fluctuation threshold. If the fluctuation index is greater than the fluctuation threshold, the temperature data in the candidate segment is determined to be inconsistent with the preset persistence criterion.

[0044] The process continues until all temperature data that does not meet the preset persistence criteria are filtered out and treated as instantaneous fluctuation data.

[0045] Based on the remaining temperature data in the initial temperature sequence excluding instantaneous fluctuation data, a corrected temperature sequence is constructed.

[0046] Based on the corrected temperature sequence, the background thermal components corresponding to the temperature rise event caused by high total current are determined.

[0047] Furthermore, for candidate segments, fluctuation indicators are determined to characterize the degree of fluctuation within the candidate segments, including:

[0048] Determine the overall trend of temperature data within the candidate segment;

[0049] Based on the overall trend of change, determine the trend magnitude value that characterizes the overall trend of change;

[0050] Based on the overall trend and the temperature data within the candidate segments, determine the deviation value that characterizes the trend magnitude of the temperature data from the overall trend.

[0051] The volatility indicator is determined based on the deviation and amplitude values.

[0052] Furthermore, determine the overall trend of temperature data within the candidate segment, including:

[0053] The preset trend function is defined by parameters;

[0054] The parameter values ​​are determined with the goal of minimizing the aggregated difference between the calculated value of the trend function and the temperature data within the candidate segment.

[0055] Based on the parameters whose values ​​have been determined, the trend function is defined as the overall trend of change.

[0056] Furthermore, the small relative sliding displacement is input into the failure assessment module, which outputs the contact resistance growth rate, including:

[0057] The contact surface state of the stacked interfaces in the embedded power metering box is iteratively calculated over multiple calculation cycles, wherein within each calculation cycle:

[0058] Based on the temperature distribution and the contact surface state updated in the previous calculation cycle, the first impact of thermal oxidation on the contact surface state is determined.

[0059] Based on the first influence and mechanical vibration characteristics, the second influence of fretting wear on the contact surface state is determined, and the contact surface state is updated based on the second influence for calculation in the next calculation cycle.

[0060] After completing a preset number of calculation cycles, the contact resistance growth rate is determined based on the final updated contact surface state and internal stress state.

[0061] The second objective of this invention is to provide an embedded power metering box structural parameter optimization system, comprising:

[0062] The data acquisition module is used to acquire the harmonic current characteristics and mechanical vibration characteristics generated by the stacking interface in the embedded power metering box during operation when communication is interrupted.

[0063] The model building module is used to build a digital reproduction and evaluation model of the failure process. It takes harmonic current characteristics and mechanical vibration characteristics as model inputs and outputs the contact resistance growth rate.

[0064] The model optimization module is used to optimize the digital reproduction and evaluation model of the failure process based on the genetic algorithm, with the goal of minimizing the contact resistance growth rate.

[0065] The parameter optimization module is used to take the embedded energy metering box structural parameters corresponding to the minimum contact resistance growth rate as the optimization result of the embedded energy metering box structural parameters.

[0066] Compared with the prior art, the advantages of the present invention are:

[0067] The embedded energy metering box structural parameter optimization method of the present invention obtains the harmonic current characteristics and mechanical vibration characteristics generated by the stacking interface during operation in an embedded energy metering box with communication interruption; constructs a digital reproduction and evaluation model of the failure process, using the harmonic current characteristics and mechanical vibration characteristics as model inputs and outputting the contact resistance growth rate; optimizes the digital reproduction and evaluation model of the failure process based on a genetic algorithm with the goal of minimizing the contact resistance growth rate; and takes the embedded energy metering box structural parameters corresponding to the minimum contact resistance growth rate as the optimization result of the embedded energy metering box structural parameters. Through this invention, it is possible to systematically discover the combination of structural and material parameters that minimizes the interface contact resistance growth rate, thereby suppressing the vicious cycle of the failure process at its source and ensuring the long-term physical connection stability of the data communication link. Attached Figure Description

[0068] Figure 1A flowchart illustrating a method for optimizing the structural parameters of an embedded energy metering box provided by the present invention;

[0069] Figure 2 This is a schematic diagram of the structure of an embedded power metering box structural parameter optimization system provided by the present invention. Detailed Implementation

[0070] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0071] like Figure 1 As shown, the present invention provides a method for optimizing the structural parameters of an embedded power metering box, including:

[0072] S110: Harmonic current characteristics and mechanical vibration characteristics generated by the stacking interface during operation in an embedded power metering box that has experienced a communication interruption.

[0073] In high-density deployed embedded power metering boxes, to avoid data communication interruptions, a parameter optimization method needs to be established. This method uses the measured specific harmonic current spectrum and continuous environmental micro-vibration characteristics as initial input conditions. By reproducing the complete failure chain from harmonic heating to fretting corrosion, a set of structural and material parameter combinations that minimizes the growth rate of interface contact resistance can be systematically found. This suppresses the vicious cycle of failure at its source and ensures the long-term physical connection stability of the data communication link.

[0074] Specifically, the steps for obtaining harmonic current characteristics and mechanical vibration characteristics include:

[0075] S111: In an embedded power metering box where communication is interrupted, a high-bandwidth current probe and a power quality analyzer are deployed to continuously collect the current waveform of the power supply busbar during the typical working cycle of the equipment; through Fourier transform, the harmonic current spectrum of the current waveform is obtained, and the amplitude and phase information of the main harmonic components are recorded as harmonic current characteristics.

[0076] The amplitude and phase information of the main harmonic components includes the amplitude and phase information of the 3rd, 5th, and 7th main harmonic components.

[0077] S112: Accelerometers are installed in the columns, beams and inside the embedded power metering box to collect vibration signals transmitted from the floor during operation; the vibration signals are subjected to spectrum analysis to identify the frequency range and amplitude of the energy concentration vibration, which are used as mechanical vibration characteristics.

[0078] S113: Converts harmonic current characteristics into a time-varying heat source generation rate applied to the finite element geometry of the stacked interface current-carrying contacts in the embedded energy metering box.

[0079] S114: Convert the mechanical vibration characteristics into a mechanical excitation function applied to the support points of the stacked interface structure in the embedded power metering box.

[0080] S120: Construct a digital reproduction and evaluation model of the failure process, using harmonic current characteristics and mechanical vibration characteristics as model inputs, and outputting the contact resistance growth rate.

[0081] In the embodiments of the present invention, the purpose of constructing a digital reproduction and evaluation model of the failure process is to calculate the contact resistance growth rate through harmonic current characteristics and mechanical vibration characteristics. Therefore, the digital reproduction and evaluation model of the failure process can be regarded as a method or tool for calculating the contact resistance growth rate. Specifically, the digital reproduction and evaluation model of the failure process in this embodiment can be considered as a black box for calculating the contact resistance growth rate. By inputting harmonic current characteristics and mechanical vibration characteristics into the black box for calculating the contact resistance growth rate, the contact resistance growth rate can be obtained.

[0082] In this embodiment, the digital reproduction and evaluation model of the failure process includes:

[0083] The transient electrothermal analysis module is used to perform transient electrothermal analysis. Taking the converted heat source generation rate as input, it calculates the temperature distribution inside the interface (especially the surface of the current-carrying contact) over time under the action of harmonic current, considering the skin effect, to obtain a fine temperature field.

[0084] The thermal analysis module is used to perform thermal analysis. The temperature field calculated in the previous step is applied as a load to the interface structure, which includes contacts, bases, and other components made of different materials. Based on the differences in the thermal expansion coefficients of the different materials, the internal stress distribution caused by uneven heating is calculated, determining the preload changes and micro-deformations at key locations such as signal contacts.

[0085] The forced vibration analysis module is used to perform forced vibration analysis. On a structure already subjected to thermal stress, the "mechanical load" obtained in the first step is applied. The module analyzes and calculates the minute relative sliding displacement generated between the contact surfaces under this dual action.

[0086] The failure assessment module is used to perform failure assessments. It substitutes the calculated relative sliding displacement into a preset fretting wear calculation formula related to the contact plating material to estimate the increase in contact resistance due to fretting corrosion per unit time. This "contact resistance growth rate" is defined as a core indicator for evaluating interface performance; the smaller the value, the higher the long-term reliability of the interface.

[0087] The steps for obtaining the fine temperature field include the following:

[0088] Temperature data from a first temperature sensor and a second temperature sensor, which are stacked on an embedded energy metering box, are collected. The first temperature sensor is located at the current-carrying contact of the stacked interface in the embedded energy metering box, and the distance between the second temperature sensor and the first temperature sensor is set to a preset distance.

[0089] Based on the temperature data from the first temperature sensor, the rate of temperature rise is determined, and based on the temperature data from the first temperature sensor and the temperature data from the second temperature sensor, the instantaneous temperature difference between the first temperature sensor and the second temperature sensor is determined.

[0090] The rate of temperature rise is compared with a preset first rate threshold, and the instantaneous temperature difference is compared with a preset first temperature difference threshold to determine whether there is a temperature rise event caused by high total current.

[0091] If a temperature rise event caused by high total current is determined, a dynamic temperature reference is established based on the temperature state after the temperature rise event occurs. Based on the dynamic temperature reference, the second temperature rise rate and the second instantaneous temperature difference corresponding to subsequent temperature changes are calculated. In this step, the background thermal component corresponding to the temperature rise event caused by high total current is determined based on the initial temperature change of the temperature rise event. Based on the temperature data from the first and second temperature sensors, the background thermal component is removed to obtain the residual temperature component, which is used as the dynamic temperature reference. Based on the residual temperature component, the second temperature rise rate and the second instantaneous temperature difference are calculated.

[0092] The second temperature rise rate is compared with a preset second rate threshold, and the second instantaneous temperature difference is compared with a preset second temperature difference threshold to determine whether there are superimposed high-order harmonic events. Combined with the judgment result of whether there is a temperature rise event caused by high total current, the type of event causing the temperature rise is determined.

[0093] Based on event type and harmonic current characteristics, the temperature distribution of the stacked interfaces in the embedded power metering box is determined, and a fine temperature field is obtained.

[0094] Furthermore, the steps for determining the background heat component corresponding to a temperature rise event caused by high total current based on the initial temperature change of the temperature rise event include: acquiring an initial temperature sequence consisting of multiple time-ordered temperature data during the initial stage of the temperature rise event; determining a local trend of change for each temperature data in the initial temperature sequence based on its adjacent data in the sequence; identifying candidate segments with persistent trends of change from the initial temperature sequence; determining the overall trend of temperature data within the candidate segments; determining the trend amplitude value characterizing the overall trend of change based on the overall trend of change; determining the deviation value characterizing the trend amplitude value of temperature data deviating from the overall trend of change based on the overall trend of change and the temperature data within the candidate segments; determining a fluctuation index based on the deviation value and the amplitude value; comparing the fluctuation index with a preset fluctuation threshold; if the fluctuation index is greater than the fluctuation threshold, determining that the temperature data within the candidate segments does not meet the preset persistence criterion; continuing to judge until all temperature data that does not meet the preset persistence criterion are filtered out as instantaneous fluctuation data; constructing a corrected temperature sequence based on the remaining temperature data in the initial temperature sequence excluding instantaneous fluctuation data; and determining the background heat component corresponding to the temperature rise event caused by high total current based on the corrected temperature sequence.

[0095] Specifically, determining the overall trend of temperature data within the candidate segment includes: pre-setting a trend function defined by parameters; determining the parameter values ​​with the goal of minimizing the aggregation difference between the calculated value of the trend function and the temperature data within the candidate segment; and determining the trend function as the overall trend based on the parameters whose values ​​have been determined.

[0096] S130: Based on a genetic algorithm, the model for digital reproduction and evaluation of the failure process is optimized with the goal of minimizing the contact resistance growth rate.

[0097] This step is an automated optimization loop. The optimization of the failure process digital reproduction and evaluation model, aimed at minimizing the contact resistance growth rate, is essentially an optimization of the design parameters of the power metering box stacking interface. Specifically,

[0098] Define a batch of interface design parameters to be optimized as variables. The interface design parameters can be set as the initial contact positive pressure of the current-carrying contact, the plating material options of the signal contact (such as tin, silver, gold), the plating thickness, the material elastic modulus of the insulating base, etc.

[0099] The first set of parameter combinations is randomly generated and input into the digital reproduction and evaluation model of the failure process to calculate the contact resistance growth rate corresponding to the set of parameters.

[0100] Specifically, the optimization steps described above can be implemented on a computing platform that integrates an optimization algorithm library and multiphysics simulation software.

[0101] The optimization program can be specifically a script written in Python, implemented using an open-source genetic algorithm library (such as the DEAP library).

[0102] The specific steps include:

[0103] Define the parameter variable to be optimized and its value range, i.e., "gene coding":

[0104] 1. Coating material: A discrete variable whose value set is ['tin', 'silver', 'gold'], represented by integers 0, 1, and 2 in the program.

[0105] 2. Base material: A discrete variable whose value set is ['Standard polycarbonate PC', 'Glass fiber reinforced PBT'], represented by integers 0 and 1 in the program.

[0106] 3. Initial contact normal force: a continuous variable, with a value range of 5.0 Newtons to 10.0 Newtons.

[0107] When optimization begins, the genetic algorithm starts generating the first "individual" (i.e., the first set of parameter combinations) in the first generation population. Random number generation operation:

[0108] Randomly select a number from [0, 1, 2] for the "coating material". For example, selecting 1 represents "silver".

[0109] Randomly select a number in [0, 1] for "Base Material". For example, selecting 0 represents "Standard Polycarbonate PC".

[0110] Generate a random floating-point number for the "Initial Contact Positive Pressure" within the range [5.0, 10.0], for example, 7.8.

[0111] At this point, the first set of parameters `{'Coating Material': 'Silver', 'Base Material': 'PC', 'Initial Contact Positive Pressure': 7.8}` has been generated.

[0112] Input these parameters into a pre-written multiphysics simulation software (such as ANSYS Workbench or COMSOL Multiphysics). The software automatically modifies the corresponding material properties and boundary conditions in the simulation model, sequentially executing the electrothermal analysis, thermodynamic analysis, forced vibration analysis, and failure simulation workflow defined in the model for digital reproduction and evaluation of the failure process. Upon completion, a text result file is output, containing the calculated "contact resistance growth rate" value, for example, 0.045 (representing a monthly growth of 4.5%). A Python script reads this result file and records 0.045 as the fitness score (i.e., performance rating) for the first "individual".

[0113] This growth rate is used as the criterion for evaluating the merits of this set of parameters, and a new generation of parameter combinations is generated according to established rules (such as selection, crossover, and mutation). This process is repeated, with each iteration aiming to find a parameter combination that results in a lower "contact resistance growth rate".

[0114] S140: The embedded energy metering box structural parameters corresponding to the minimum contact resistance growth rate are taken as the optimization result of the embedded energy metering box structural parameters.

[0115] The loop terminates after satisfying a preset convergence condition (such as the growth rate no longer decreasing significantly or reaching the upper limit of the number of iterations). The last set of parameters output is the optimal parameter combination that can most effectively suppress fretting corrosion and ensure connection stability under this specific field condition.

[0116] In an embodiment of the present invention, the contact surface state of the stacked interface in the embedded power metering box is iteratively calculated in multiple calculation cycles. In each calculation cycle: based on the temperature distribution and the contact surface state updated in the previous calculation cycle, a first influence of thermal oxidation on the contact surface state is determined; based on the first influence and mechanical vibration characteristics, a second influence of fretting wear on the contact surface state is determined, and the contact surface state is updated based on the second influence for calculation in the next calculation cycle; after completing a preset number of calculation cycles, the contact resistance growth rate is determined based on the finally updated contact surface state and internal stress state.

[0117] The working principle of the method of the present invention is to transform a complex connector interface progressive failure problem that occurs in the physical world and is caused by the combined effect of multiple physical phenomena into an engineering calculation problem that can be accurately solved in a computer and has a clear optimization objective.

[0118] Its core logic is to establish a digital causal chain from the "specific on-site environment" to the "interface design parameters." First, it accurately captures the two key environmental factors that cause the problem through on-site measurements: the harmonic current spectrum generated by specific electrical equipment, and the continuous micro-vibration characteristics generated by the operation of buildings and equipment. This is equivalent to creating a precise profile of the "environment" in which the problem occurs.

[0119] This method constructs a digital model in a computer that closely resembles the real interface. Instead of analyzing electrical, thermal, or mechanical problems in isolation, it connects them in series to simulate real physical processes: It uses measured harmonic current as a heat source input to calculate the precise temperature distribution within the interface caused by high-frequency effects; then, it uses this temperature distribution as a load to calculate the internal mechanical stress caused by the inconsistent thermal expansion and contraction of different materials; finally, it applies measured environmental vibrations to this model already subjected to thermal stress to calculate the minute slippage occurring between the contacts.

[0120] This series of calculations ultimately points to a core quantitative metric: the projected growth rate of contact resistance. This metric directly reflects the speed at which interface performance deteriorates.

[0121] This invention employs an automated optimization program with the sole objective of minimizing the rate of contact resistance growth. It iteratively adjusts various design parameters of the interface (such as materials, plating, and structural dimensions), performing the aforementioned full-chain simulation for each parameter combination. Through hundreds or thousands of iterative calculations, the program automatically filters and converges to an optimal set of parameters. The design corresponding to this set of parameters exhibits the strongest resistance and the longest stable operating life under the specific harsh environment initially measured.

[0122] The method of this invention establishes a direct and quantitative path from "specific field environment" to "optimal design parameters." It does not rely on general design principles or engineers' empirical guesses, but rather transforms the complex failure problem involving the interaction of multiple physical phenomena into a computational problem with a clear optimization objective. By fully reproducing the entire chain from cause (harmonics, fretting) to result (increased resistance), this method can accurately find the optimal parameter settings under specific harsh environments, providing a rigorous and efficient approach to solving such systemic reliability problems caused by complex factors.

[0123] The following is a specific embodiment provided by the present invention:

[0124] In the absence of the solution proposed in this invention (conventional processing method):

[0125] 1. Problem Discovery: The maintenance personnel discovered frequent instantaneous offline communication records in multiple power metering boxes (such as No. 305, No. 307, and No. 308) in cabinet C during the background monitoring system.

[0126] 2. On-site handling: The technician arrived at the site and inspected the physical connections, finding no obvious looseness. Based on his experience, he judged that the problem might be with the interface of metering box 307, so he retrieved a brand new metering box of the same model from the spare parts warehouse and replaced 307.

[0127] 3. Outcome: Within a few days of the replacement, communication with unit 307 returned to normal. However, a week later, the frequency of communication interruptions with units 305 and 308 increased, and even the newly replaced unit 307 experienced intermittent offline records again. The maintenance team was trapped in a reactive, "whack-a-mole" approach to repairs, unable to eradicate the problem, and resulting in wasted spare parts and long-term unreliability of data collection.

[0128] 4. Reasons for failure: Traditional methods attribute problems to occasional failures of individual products, ignoring the unique environment of the entire cabinet. The replacement spare parts have the exact same design parameters as the original products, and they are equally incapable of resisting the high-order harmonics and continuous micro-vibrations generated by the 3D printer cluster in the field, so the problem will inevitably recur.

[0129] When this solution is adopted (optimized processing flow):

[0130] Step 1: On-site feature acquisition and load conversion

[0131] Engineers brought a power quality analyzer and multiple miniature accelerometers to the site. Current probes were connected to the main power busbar of rack C, and accelerometers were attached to the vertical columns and middle shelves of the rack. During the normal operation of the power meter in rack C, current waveforms and vibration signals were continuously recorded for 8 hours.

[0132] The purpose of this analysis was not simply to determine whether there was "poor power quality" or "vibration," but to precisely quantify these disturbances. The analysis revealed significant 5th and 7th harmonics in the current, with amplitudes reaching 15% and 10% of the fundamental frequency, respectively; the vibration signal also exhibited a distinct energy peak near 120 Hz. These specific data points identified the root cause of the interface failure and formed the foundation for all subsequent analyses.

[0133] Step 2: Digital Reproduction and Evaluation of the Failure Process

[0134] Display the 3D model of the interface of this type of metering box.

[0135] The measured harmonic current data was used as a heat source input for electrothermal analysis. The calculation results show that under the action of harmonic current, the local temperature of the current-carrying contact is 25 degrees Celsius higher than that when only power frequency current is present.

[0136] The non-uniform temperature field was applied to the structural model for thermodynamic analysis. The results showed that the expansion of the copper contacts increased the contact pressure with the socket by 30% compared to the initial design value, while also causing a slight shift of 5 micrometers in the position of the adjacent signal contacts.

[0137] The measured 120 Hz vibration load was applied to this model, which was already under thermal stress. Analysis showed that under this "overstressed" state, the vibration caused 120 reciprocating slides per second with an amplitude of 2 micrometers between the contact surfaces of the signal contacts.

[0138] Substituting the sliding amplitude and contact pressure into a fretting corrosion model based on tin-plated copper, the calculated contact resistance of the interface is expected to increase by 8% per month. This figure clearly indicates that the interface performance is rapidly deteriorating.

[0139] This step makes the invisible physical processes transparent. It verifies the correctness of the failure chain of "harmonic heating - thermal stress - micro-vibration - fretting corrosion" and provides a quantitative degradation indicator of "8% monthly resistance increase", so that the severity of the problem is no longer a subjective feeling, but a specific value that can be compared and optimized.

[0140] Step 3: Iterative optimization of parameters for failure suppression

[0141] Start the optimization program and set the optimization objective as "minimize the monthly growth rate of contact resistance". Define adjustable design parameter variables:

[0142] Variable 1: Signal contact plating material (options: tin, silver, gold)

[0143] Variable 2: Insulating base material (Options: standard polycarbonate PC, glass fiber reinforced PBT)

[0144] Variable 3: Initial contact force of the current-carrying contact (range: 5 Newtons to 10 Newtons)

[0145] The program begins automatic iterative calculations. For example, it tests the "tin-PC-5 Newton" combination, achieving an 8% monthly growth rate; then it tests the "gold-PC-5 Newton" combination, where gold offers better corrosion resistance, resulting in a 3% monthly growth rate; next, it tests "tin-PBT-5 Newton," where PBT's coefficient of thermal expansion is closer to copper, reducing thermal stress and achieving a 2.5% monthly growth rate. After hundreds of combination tests, the program finds that the "silver-PBT-7.5 Newton" combination can reduce the monthly resistance growth rate to below 0.5%, an acceptable level of stability.

[0146] This process replaces expensive and time-consuming physical prototype trial and error. It systematically explores the entire design space, discovers the interactions between different parameters, and finds the global optimum. For example, it may reveal that moderately increasing the initial contact pressure (from 5 Newtons to 7.5 Newtons) in combination with changing the base material (PBT) is far more effective than simply using an expensive gold plating.

[0147] This method provides an optimized design recommendation for this shared manufacturing center environment: replace the insulating base material of the metering box interface with glass fiber reinforced PBT, use silver plating for the signal contacts, and set the initial contact positive pressure of the current-carrying contacts to 7.5 Newtons. After the metering boxes manufactured according to this new specification were used to replace all units in rack C, the communication interruption problem was completely resolved and did not recur in subsequent long-term observations.

[0148] Compared to existing technologies, this technical solution does not focus on improving a single technical point, but rather on a fundamental breakthrough in its problem-solving approach and methodology. Existing technologies, when dealing with similar connection reliability issues, often involve isolated, empirical enhancements of a particular aspect, such as using more corrosion-resistant materials, adding vibration-resistant structures, or installing filtering devices. This is a "general enhancement" approach that lacks specificity and may lead to unnecessary cost increases.

[0149] The technical solution of this invention establishes a quantitative causal relationship from a specific environment to a specific design: This solution proposes and implements for the first time a complete closed-loop path, from the precise capture of on-site environmental characteristics (specific harmonics and micro-vibrations), to the sequential coupling analysis of multiple physical processes (electrical, thermal, mechanical, and wear), and then to the automated optimization of design parameters. Its non-obvious aspect lies in recognizing that connection failure is not dominated by a single factor, but rather the result of the cascading amplification of multiple physical effects under specific conditions. Those skilled in the art typically handle electrical and mechanical problems separately, while this solution treats them as a continuous, interacting failure chain for overall modeling and analysis, thereby enabling the identification of the optimal strategy to suppress the development of this chain from its root cause.

[0150] The technical solution of this invention realizes a shift from "passive maintenance" to "active immunity": the traditional method is to passively replace components after a failure occurs, which is only a temporary solution. This solution, through "pre-diagnosis" of specific harsh environments, designs products with "immunity" to those environments. This forward-looking design optimization enables the product to withstand specific combined stresses before it is put into use, thereby fundamentally eliminating potential failures, significantly improving the long-term reliability and stability of the system, and greatly reducing the total life-cycle maintenance costs.

[0151] The technical solution of this invention provides an efficient and precise solution, avoiding over-design: by optimizing the core physical quantity of "contact resistance growth rate," this solution can accurately assess the actual effects of changes in various design parameters. It can discover the optimal balance between cost and benefit; for example, calculations may show that replacing the insulation base material with one that increases cost only slightly is far more effective than using an expensive precious metal plating. This data-driven, refined design avoids the experience-based, generalized "over-design" by engineers, ensuring both high performance and cost-effectiveness.

[0152] This paper proposes a technical approach to establish a foundation for subsequent analysis through on-site feature acquisition. The aim is to use the electrical and mechanical operating characteristics collected on-site as input for interface parameter optimization. The process begins by deploying high-bandwidth current probes and miniature accelerometers at the cabinet experiencing communication interruptions. During the typical operating cycle of the equipment, the current waveform of the power supply busbar and the vibration signal of the cabinet are continuously collected. Subsequently, the collected signals are processed to obtain the detailed spectrum of harmonic currents and the main vibration frequency range, and this information is converted into heat sources and mechanical excitations applied to the interface model. This process provides the basic data for subsequent failure process reproduction. However, in some application scenarios, this feature acquisition process may face unique challenges. For example, in a shared manufacturing space providing hardware incubation services for startups, the power supply system also uses such a high-density deployment of power metering boxes. This space is characterized by high tenant turnover and highly uncertain R&D projects. A technical team responsible for collecting on-site features, following standard operating procedures, conducted continuous 8-hour feature data collection on a cabinet experiencing frequent communication interruptions from 9:00 AM to 5:00 PM on Tuesday. During this period, the cabinet primarily powered the equipment of three tenants: one engaged in 3D printing of drone parts, one in small-scale CNC milling of consumer electronics casings, and another in routine circuit board soldering and debugging. The technical team obtained a stable set of electrical load data, mainly consisting of the 5th and 7th harmonics generated by the 3D printers and CNC equipment. To support cutting-edge research, the shared space allowed tenants to use high-power experimental equipment at night, provided they gave prior notice and independently accounted for electricity costs. One tenant, a team developing specialized power supplies, used a high-frequency pulse load tester to test the resilience of gallium nitride power devices only on Wednesdays and Fridays from 2:00 AM to 4:00 AM. When operating, this device injected extremely high-amplitude 19th and 21st harmonic currents into the grid for only a few minutes. These higher harmonic components caused a much more severe skin effect than the 5th and 7th harmonics, resulting in instantaneous and severe localized temperature rises at the current-carrying contacts of the metering box stack interface. Because the technical team's data collection was conducted during the day, the acquired "electrical imprints" completely missed the instantaneous, high-intensity harmonic impacts generated during the nighttime experiments. The "digital reproduction of the failure process" completed by the team based on daytime data, and the calculated contact resistance growth rate, could not explain the actual observed interface degradation rate. Ultimately, the parameter combination optimized based on this incomplete set of field data, while capable of handling routine 3D printing and CNC machining conditions, still lacked sufficient protection against the occasional, pulsed high-order harmonic impacts at night that truly accelerated failure. The physical connection stability issue of the interface was not eradicated after replacing it with an optimized product; only the failure reproduction cycle was slightly prolonged.

[0153] In a shared manufacturing environment where tenant composition and electricity consumption behavior are highly uncertain and critical degradation events are sporadic and hidden, how can we establish a field feature acquisition method? This method should be able to reliably identify and record transient strong harmonic events generated by atypical equipment during irregular periods, which can lead to accelerated deterioration of interface connection performance, under the premise of passive information acquisition and unpredictable equipment activation time. This will provide complete and accurate electrical load input for subsequent parameter optimization, and avoid the failure of optimization schemes due to the omission of key data.

[0154] The core idea of ​​this method is to shift the monitoring focus from directly capturing complex and unpredictable "electrical events" to continuously monitoring their inevitable and easily measurable "physical consequences"—abnormal temperature rises at interface contacts. By deploying miniature temperature sensors at key locations on each metering box interface, a low-cost, 24 / 7 "thermal sentinel" system is established. This system does not concern itself with current waveform details, but focuses solely on identifying and recording "thermal shock events" that far exceed normal operating ranges. Once such an event is captured, the system records its detailed timestamp and temperature change process. Maintenance personnel can use this "event report" to retrospectively query building access control records, tenant equipment usage reports, or time-of-use billing data, thereby accurately pinpointing the source tenant and equipment causing the shock. Finally, a targeted electrical characteristic reproduction and collection is performed on the identified source to obtain the crucial, previously overlooked electrical load data. This method cleverly bypasses the difficulty of directly capturing sporadic electrical events through a reverse engineering approach of "first grasping the consequences, then tracing the causes." It utilizes low-cost temperature monitoring to reliably identify and label critical degradation events, then uses data correlation to accurately pinpoint the source of the problem, ultimately guiding targeted resampling of missing critical payloads. This method transforms an open-ended, uncertain monitoring problem into an investigative problem with clear clues, achieving a low-cost, high-efficiency solution.

[0155] The specific implementation is as follows:

[0156] Hardware Deployment: Within the stacked interface of each embedded power metering box, a miniature temperature sensor (e.g., a surface-mount NTC thermistor) with high sensitivity and fast thermal response is fixed to its polymer-insulated base, adjacent to the root of the copper contacts carrying the main current, via dispensing or embedding. The signal line of this sensor is connected to the main control processor inside the metering box.

[0157] Autonomous learning within the normal temperature range: During the initial phase after installation and commissioning (e.g., 48 hours), the metering box enters "learning mode." During this period, the main control processor collects temperature data at a high frequency (e.g., 10 times per second), recording temperature changes caused by the start-up and shutdown of all routine equipment and load fluctuations. After the learning period, the system automatically generates a "normal operating temperature range" based on the massive amount of data collected. This range is not a fixed line, but rather a range encompassing the amplitude and rate of temperature fluctuations.

[0158] Identification and Recording of Thermal Shock Events: After the learning period, the system enters "monitoring mode." The main control processor continuously collects temperature data and compares it with the established "normal operating temperature range." A "thermal shock event" is determined when the system detects a temperature change that simultaneously meets both of the following conditions:

[0159] Condition 1: The rate of temperature rise significantly exceeds any normal rate recorded during the learning period.

[0160] Condition 2: The peak temperature exceeds the upper limit of the "normal operating temperature range" and reaches a preset absolute difference (e.g., 15 degrees Celsius higher).

[0161] This dual-condition judgment mechanism can effectively filter out normal temperature rises caused by slow changes in ambient temperature or the start-up of routine equipment, and only lock out those sudden and severe abnormal heat generation.

[0162] Once a "thermal shock event" is identified, the main control processor immediately packages the following information into a log and stores it in the internal non-volatile memory: the unique number of the event, the precise start time of the event (year / month / day / hour / minute / second), the duration of the event, and the complete temperature change curve data during the event.

[0163] Event backtracking and data correlation: Maintenance personnel regularly read the "thermal shock event" logs stored in each metering box. If multiple metering boxes in a cabinet record thermal shock events at the same time (e.g., 2:05 AM on Wednesday), maintenance personnel can use this precise timestamp to correlate and query other information systems. For example, querying the independent billing records for that period might reveal a large electricity bill incurred by the tenant's "Special Power Supply R&D Team"; or querying the building access control system might show entry and exit records for team members during that period. This method allows for pinpointing the source of the problem.

[0164] Targeted acquisition of critical loads: After identifying the source of the problem, the operations and maintenance team can communicate with the tenant, requesting them to notify in advance before conducting similar experiments next time, and arrange for technicians to bring a power quality analyzer to conduct a prepared and targeted on-site current waveform acquisition. The high-order harmonic data obtained in this way is the true critical load causing accelerated interface failure.

[0165] This technical solution effectively addresses the challenge of obtaining complete and accurate electrical load data for interface parameter optimization in highly uncertain shared manufacturing environments. Traditional methods, relying on periodic sampling, are prone to missing transient strong harmonic events generated by atypical equipment during irregular periods, which can decisively degrade interface performance. This results in subsequent optimization analyses based on incomplete data, and the optimization schemes are unable to eradicate the problem. This solution shifts its approach, monitoring the physical consequences—interface temperature—instead of directly measuring the electrical causes. It establishes a continuous "thermal shock" event capture mechanism, reliably identifying and marking the occurrence time of each hazardous event. This precise time stamp acts as an index, allowing maintenance personnel to trace back through other information systems and accurately pinpoint the event's culprit. This process ensures that hidden electrical load information that contributes most to interface failure is no longer overlooked. Through subsequent targeted data collection, engineers can obtain complete electrical load data that truly reflects the worst operating conditions. By using this data as input to optimize the interface parameters, the resulting design solution can resist all key threats in this specific environment, thereby fundamentally solving the problem of physical connection stability of the interface.

[0166] The above solution addresses the challenge of collecting critical electrical load data in environments with uncertain electricity consumption patterns. By deploying temperature sensors at each electricity metering box interface, abnormal temperature rises caused by transient strong harmonics—i.e., "thermal shock events"—are identified and recorded. Once maintenance personnel obtain the event logs with precise timestamps, they can compare the data with billing data or access control records at that time point to pinpoint the source device causing the event. This process provides clues for subsequent targeted load data collection.

[0167] However, in a science and technology park with a highly mixed range of business types, the process of tracing the event and correlating it with data may fail. To attract different types of businesses, the park has divided the same floor into a hardware R&D area and a small data hosting area. A single server rack contains both power meters serving hardware R&D tenants and power meters serving data hosting clients. One tenant in the hardware R&D area is a company developing plasma etching equipment, whose equipment injects high-amplitude, high-order harmonics into the power grid at certain stages of the process. The client in the data hosting area is a financial services company that has leased multiple racks to deploy high-density computing servers.

[0168] According to the service agreement with the data hosting customer, the park's operations and maintenance team is required to perform a 15-minute full-load stress test on the hosted servers every Thursday at 3:00 AM to verify the redundancy of the power supply and cooling system. This test will cause the server's power to reach its peak in a short period of time, generating a huge total current.

[0169] At 3:15 AM on a certain Thursday, a prototype of the plasma etching equipment developed by the research and development team experienced an unexpected 23rd and 25th harmonic shock during operation, lasting only two minutes, due to a malfunction in its internal radio frequency power module. This caused a sharp increase in the temperature of the current-carrying contact at the corresponding metering box interface, which was correctly recorded by the system as a "thermal shock event" with a log timestamp of 03:15.

[0170] Almost simultaneously, at 3:00 AM, the operations and maintenance team's automated system initiated a scheduled stress test on the servers throughout the entire data hosting area. At 3:15 AM, the data hosting customer's server cluster was operating at full load, with its total operating current far exceeding that of the plasma etching prototype. Consequently, the building's time-sharing billing system also recorded a clear event in the background where the data hosting customer's power usage reached its peak.

[0171] When the operations and maintenance (O&M) personnel reviewed the "thermal shock event" logs afterward, they discovered an alarm record with a timestamp of 03:15. Following procedure, they checked the billing system and found a clear record of high power consumption at that time, caused by server stress testing. Since stress testing is a known, planned operation that generates high heat, the O&M personnel directly attributed the "thermal shock event" to a normal result of stress testing. They assumed that the large total current caused the overall temperature rise of the interface, triggering the alarm, failing to recognize that the root cause was a more destructive higher-order harmonic generated by another tenant. Ultimately, the true source of the accelerated interface performance degradation was obscured by this high-power server stress test event, the investigation was wrongly closed, and targeted parameter optimization was impossible.

[0172] In mixed power consumption scenarios where hardware R&D and data hosting services coexist, potentially leading to overlapping of high-order harmonic events and high total current events, how can a method be established that can distinguish the nature of events based solely on information collected by the metering box itself? This method needs to accurately attribute a "thermal shock" to a more destructive high-order harmonic source rather than a misleading high total current source by analyzing the intrinsic characteristics of the temperature rise process, without relying on external billing or access control data. This ensures that subsequent failure investigations and parameter optimizations can pinpoint the true technical root cause.

[0173] The core of this method lies in shifting from monitoring the peak value of a single temperature point to analyzing the relative relationships and rates of temperature change among multiple points within a critical area of ​​the interface. By deploying a set of miniature temperature sensors at specific locations on the interface and continuously analyzing the differences in their readings and the rates of change of each sensor, the physical processes of heat generation can be identified, thereby distinguishing the fundamental nature of the event.

[0174] Multi-point temperature data acquisition layout: At least two temperature sensors are arranged on the insulating base of each metering box interface, along the path of heat conduction from the current-carrying contact to the base material. One sensor is close to the root of the current-carrying contact to capture the direct response of the heat source; the other sensor is kept at a fixed small distance from the first sensor to observe the conduction of heat in the insulating material.

[0175] Temperature rise process characteristic extraction: The processor inside the metering box collects the temperature readings of the two sensors in real time and performs two key calculations: First, calculate the temperature rise rate of the sensor adjacent to the heat source; second, calculate the instantaneous temperature difference between the two sensors.

[0176] Event Classification: A set of discrimination rules based on the above two characteristic quantities is established. When the temperature rise rate is extremely high and the temperature difference between the two points increases sharply, it is judged as a "local rapid heating event" dominated by the skin effect of higher harmonics. When the temperature rise rate is relatively slow and the temperature difference between the two points increases synchronously within a small range, it is judged as a "global slow heating event" dominated by high total current.

[0177] Categorized Logs and Alarms: The system records the judgment results along with the event timestamp and temperature curve. Alarm information is no longer a simple "temperature exceeds limit," but is clearly marked as "suspected harmonic impulse" or "high load overheating," providing maintenance personnel with a clear and targeted preliminary diagnostic conclusion.

[0178] The technical solution of this embodiment is implemented as follows:

[0179] Sensor Installation: On the polymer base of the metering box interface, select the root of a current-carrying contact as the center of the heat source. Attach or embed the first fast-response surface-mount thermistor at this location, denoted as T1. Along the direction away from the contact, approximately 2 to 3 millimeters from T1, install a second thermistor of the same type, denoted as T2. Connect the signal lines of T1 and T2 to the main control processor of the metering box.

[0180] Data Processing and Feature Calculation: The main control processor acquires the temperature values ​​of T1 and T2 at a frequency of 20 times per second. In each acquisition cycle, the following calculations are performed:

[0181] Calculate the rate of increase of T1: Rate(T1) = (T1_current - T1_previous) / Δt, where Δt is the sampling time interval (0.05 seconds). Here, T1_current is the current temperature value, and T1_previous is the temperature value Δt prior to the current time.

[0182] Calculate the temperature difference between two points: Diff(T) = T1_current - T2_current.

[0183] The discrimination logic is implemented as follows: The system internally presets two thresholds: a rate threshold V_rate (e.g., 4 degrees Celsius / second) and a temperature difference threshold V_diff (e.g., 6 degrees Celsius). When a temperature rise event is triggered (e.g., T1 exceeds the normal operating range), the system enters the discrimination procedure:

[0184] Harmonic event determination: If at any moment during the temperature rise process, Rate(T1)>V_rate and Diff(T)>V_diff are satisfied simultaneously, the system immediately marks the event as "harmonic impact type" and records the moment when the conditions are met and the corresponding characteristic value.

[0185] Total current event determination: If the temperatures of T1 and T2 continue to rise throughout the entire temperature rise process, but neither of the above two conditions is met, i.e. Rate(T1) ≤ V_rate and Diff(T) ≤ V_diff, then the system will mark the event as "high load heating type".

[0186] Information output: When maintenance personnel query the logs, they will see the following record: "Thursday 03:15, metering box 308, alarm: harmonic impulse type. Peak rate: 5.2 degrees Celsius / second, peak temperature difference: 8.1 degrees Celsius." This information clearly points to higher harmonics as the problem, rather than simply high current.

[0187] This technical solution effectively solves the problem of incorrect attribution of fault sources caused by the time overlap of high-order harmonic events and high total current events in mixed power consumption scenarios.

[0188] The effectiveness of this approach lies in its treatment of temperature not as a single, vague indicator, but as a process rich in information. By introducing multi-point differential measurement, this method successfully captures the unique "thermal characteristics" corresponding to different physical causes. The skin effect of higher harmonics inevitably leads to a high concentration of heat at the contact surface instantaneously before it slowly conducts to the surroundings. In our measurement system, this manifests as a sharp rise in temperature for T1, while the response of T2 is significantly delayed, resulting in a large rate of temperature increase and an instantaneous temperature difference. Conversely, the global heating caused by the high total current of the server cluster is generated throughout the conductor and relatively evenly transferred to the base. Therefore, the temperatures of T1 and T2 rise more consistently, with both their rates of increase and temperature difference remaining at a lower level.

[0189] By transforming these inherent physical differences into clear discrimination rules, the metering box itself possesses the ability to "see clearly." It can provide a preliminary conclusion as soon as an event occurs, such as "this is a harmonic problem" or "this is an overload problem." This conclusion does not rely on external billing systems or access control records, thus completely severing the path of erroneous correlation. When maintenance personnel face two suspicious event sources with overlapping times, this classification report based on physical phenomena provided by the device itself becomes decisive evidence for pinpointing the true source of the fault, ensuring that subsequent investigations and optimization efforts can directly address the core issue and fundamentally solve the problem.

[0190] like Figure 2 As shown, this invention proposes an embedded power metering box structural parameter optimization system 200, comprising:

[0191] Data acquisition module 210 is used to acquire the harmonic current characteristics and mechanical vibration characteristics generated by the stacking interface in the embedded power metering box during operation when communication is interrupted.

[0192] The model building module 220 is used to build a digital reproduction and evaluation model of the failure process. It takes harmonic current characteristics and mechanical vibration characteristics as model inputs and outputs the contact resistance growth rate.

[0193] The model optimization module 230 is used to optimize the digital reproduction and evaluation model of the failure process based on the genetic algorithm with the goal of minimizing the contact resistance growth rate.

[0194] The parameter optimization module 240 is used to take the embedded energy metering box structural parameters corresponding to the minimum contact resistance growth rate as the embedded energy metering box structural parameter optimization result.

[0195] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of the invention. In this specification, illustrative expressions of the terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0196] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this invention, "a plurality of" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0197] Any process or method description in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing custom logic functions or processes, and the scope of preferred embodiments of the invention includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as will be understood by those skilled in the art to which embodiments of the invention pertain.

[0198] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. An embedded electric energy metering box structure parameter optimization method, characterized in that, The method comprises the following steps: Obtaining the harmonic current characteristics and mechanical vibration characteristics generated by the stacking interface in the embedded power metering box during operation; Building a failure process digital replication and evaluation model, taking the harmonic current characteristics and the mechanical vibration characteristics as input of the model, and outputting the contact resistance growth rate; Based on the genetic algorithm, the failure process digital replication and evaluation model is optimized with the minimum contact resistance growth rate as the target; When the contact resistance growth rate is minimum, the corresponding structure parameters of the embedded power metering box are taken as the optimization results of the structure parameters of the embedded power metering box; The harmonic current characteristics are converted into a time-varying heat source generation rate applied to the finite element geometry of the current-carrying contact point of the stacking interface in the embedded power metering box; The mechanical vibration characteristics are converted into a mechanical excitation function applied to the structural support point of the stacking interface in the embedded power metering box; The failure process digital replication and evaluation model takes the harmonic current characteristics and the mechanical vibration characteristics as input of the model, and outputs the contact resistance growth rate, which comprises the following steps: Building the failure process digital replication and evaluation model; the failure process digital replication and evaluation model comprises a transient electro-thermal analysis module, a thermal analysis module, a forced vibration analysis module and a failure evaluation module; wherein, The heat source generation rate is input into the transient electro-thermal analysis module to calculate the change of the temperature distribution inside the stacking interface of the embedded power metering box with time under the action of the harmonic current, and a fine temperature field is obtained; The fine temperature field is taken as a load, and the stress distribution state inside the stacking interface of the embedded power metering box due to uneven heating is calculated based on the thermal analysis module; The mechanical excitation function is taken as a load, and the micro relative sliding displacement generated between the contact surfaces of the stacking interface of the embedded power metering box under the action of double loads is calculated based on the forced vibration analysis module; The micro relative sliding displacement is input into the failure evaluation module, and the contact resistance growth rate is output.

2. The method for optimizing the structure parameters of an embedded electric energy metering box according to claim 1, characterized in that, The harmonic current characteristics and mechanical vibration characteristics generated by the stacking interface in the embedded power metering box during operation, which comprises the following steps: In the embedded power metering box where communication interruption occurs, a high-bandwidth current probe and a power quality analyzer are deployed to continuously collect the current waveform of the power supply busbar within a typical working period of the device; the harmonic current spectrum of the current waveform is obtained by Fourier transform, and the amplitude and phase information of the main harmonic components are recorded as the harmonic current characteristics; Acceleration sensors are arranged in the columns, beams and box body of the embedded power metering box to collect vibration signals transmitted by the floor during operation; the vibration signals are subjected to frequency spectrum analysis to identify the vibration frequency range and amplitude with high energy, which are taken as the mechanical vibration characteristics.

3. The method for optimizing the structure parameters of an embedded electric energy metering box according to claim 1, characterized in that, The heat source generation rate is input into the transient electro-thermal analysis module to calculate the change of the temperature distribution inside the stacking interface of the embedded power metering box with time under the action of the harmonic current ​ collecting temperature data of a first temperature sensor and a second temperature sensor disposed at the stacked interface of the embedded electric energy metering box; wherein the first temperature sensor is disposed at a current-carrying contact of the stacked interface of the embedded electric energy metering box, and a distance between the first temperature sensor and the second temperature sensor is set as a preset distance; determining a temperature rising rate based on the temperature data of the first temperature sensor, and determining an instantaneous temperature difference between the first temperature sensor and the second temperature sensor based on the temperature data of the first temperature sensor and the temperature data of the second temperature sensor; comparing the temperature rising rate with a preset first rate threshold, and comparing the instantaneous temperature difference with a preset first temperature difference threshold, to determine whether a temperature rising event caused by high total current exists; if it is determined that the temperature rising event caused by high total current exists, establishing a dynamic temperature reference based on a temperature state after the temperature rising event occurs, and calculating a second temperature rising rate and a second instantaneous temperature difference corresponding to subsequent temperature changes based on the dynamic temperature reference; comparing the second temperature rising rate with a preset second rate threshold, and comparing the second instantaneous temperature difference with a preset second temperature difference threshold, to determine whether a superimposed high-order harmonic event exists, and determining an event type causing temperature rising in combination with a determination result of whether the temperature rising event caused by high total current exists; based on the event type and the harmonic current characteristic, determining a temperature distribution of the stacked interface of the embedded electric energy metering box, to obtain the fine temperature field.

4. The method for optimizing the structure parameters of an embedded electric energy metering box according to claim 3, characterized in that, if it is determined that the temperature rising event caused by high total current exists, establishing a dynamic temperature reference based on a temperature state after the temperature rising event occurs, and calculating a second temperature rising rate and a second instantaneous temperature difference corresponding to subsequent temperature changes based on the dynamic temperature reference, including: determining a background heat component corresponding to the temperature rising event caused by high total current based on an initial temperature change of the temperature rising event; removing the background heat component based on the temperature data of the first temperature sensor and the second temperature sensor, obtaining a residual temperature component, wherein the background heat component is used as the dynamic temperature reference; calculating the second temperature rising rate and the second instantaneous temperature difference based on the residual temperature component.

5. The method for optimizing the structure parameters of an embedded electric energy metering box according to claim 4, characterized in that, determining the background heat component corresponding to the temperature rising event caused by high total current based on an initial temperature change of the temperature rising event, including: in an initial stage of the temperature rising event, obtaining an initial temperature sequence composed of a plurality of temperature data sorted by time; for each temperature data in the initial temperature sequence, determining a local change trend based on adjacent data thereof in the sequence; identifying a candidate segment with a persistent change trend from the initial temperature sequence; for the candidate segment, determining a fluctuation index for representing a change fluctuation degree within the candidate segment; comparing the fluctuation index with a preset fluctuation threshold, and if the fluctuation index is greater than the fluctuation threshold, determining that the temperature data in the candidate segment does not meet a preset persistence criterion; continuously judging until all temperature data not meeting the preset continuity criterion is screened out as transient fluctuation data; constructing a modified temperature sequence based on the remaining temperature data in the initial temperature sequence except the transient fluctuation data; determining a background thermal component corresponding to the temperature rise event caused by the high total current based on the modified temperature sequence.

6. The method for optimizing the structure parameters of an embedded electric energy metering box according to claim 5, characterized in that, For determining, for the candidate segment, a fluctuation indicator for characterizing the degree of internal fluctuation of the candidate segment, comprising: determining the overall change trend of the temperature data in the candidate segment; determining a trend amplitude value characterizing the overall change trend based on the overall change trend; determining a deviation value characterizing the deviation of the temperature data in the candidate segment from the trend amplitude value based on the overall change trend and the temperature data in the candidate segment; determining the fluctuation indicator based on the deviation value and the trend amplitude value.

7. The method according to claim 6, wherein, determining the overall change trend of the temperature data in the candidate segment, comprising: presetting a trend function defined by parameters; minimizing the aggregate difference between the calculated value of the trend function and the temperature data in the candidate segment to determine the value of the parameters; determining the trend function as the overall change trend based on the parameters determined by the value. 8.The method of claim 1, wherein, inputting the micro relative sliding displacement into the failure assessment module to output the contact resistance growth rate, comprising: iteratively calculating the contact surface state of the stacked interface in the embedded electric energy metering box in multiple calculation periods, wherein in each calculation period: determining the first influence of thermal oxidation on the contact surface state based on the temperature distribution and the updated contact surface state of the previous calculation period; determining the second influence of fretting wear on the contact surface state based on the first influence and the mechanical vibration characteristics, and updating the contact surface state based on the second influence for calculation in the next calculation period; after completing a preset number of calculation periods, determining the contact resistance growth rate based on the finally updated contact surface state and internal stress state.

9. An embedded electric energy metering box structure parameter optimization system, characterized in that, comprising: a data acquisition module for acquiring harmonic current characteristics and mechanical vibration characteristics generated by the stacked interface in the embedded electric energy metering box during operation in the communication interruption; a model construction module for constructing a failure process digital reproduction and evaluation model, taking the harmonic current characteristics and the mechanical vibration characteristics as model inputs, and outputting a contact resistance growth rate; a model optimization module for optimizing the failure process digital reproduction and evaluation model based on a genetic algorithm with the minimum contact resistance growth rate as the target; a parameter optimization module for taking the embedded electric energy metering box structure parameters corresponding to the minimum contact resistance growth rate as the embedded electric energy metering box structure parameter optimization result; the data acquisition module is also used for: converting the harmonic current characteristics into a time-varying heat source generation rate applied to the finite element geometry of the current-carrying contact point of the stacked interface in the embedded electric energy metering box; converting the mechanical vibration characteristics into a mechanical excitation function applied to the structural support point of the stacked interface in the embedded electric energy metering box; The model construction module is further configured to: construct a failure process digital replication and evaluation model; the failure process digital replication and evaluation model comprises a transient electro-thermal analysis module, a thermal force analysis module, a forced vibration analysis module and a failure evaluation module; wherein, input the heat source generation rate into the transient electro-thermal analysis module, calculate the change of temperature distribution inside the stacking interface in the embedded electric energy metering box with time under the action of harmonic current, and obtain a fine temperature field; input the fine temperature field as a load, calculate the stress distribution state inside the stacking interface in the embedded electric energy metering box due to uneven heating based on the thermal force analysis module; input the mechanical excitation function as a load, calculate the micro relative sliding displacement generated between the contact surfaces of the stacking interface in the embedded electric energy metering box under the action of double loads based on the forced vibration analysis module; input the micro relative sliding displacement into the failure evaluation module, and output the contact resistance growth rate.

Images