Fault prediction method and system for measurement switch

Abstract

The invention relates to the field of fault prediction, in particular to a fault prediction method and system for a measurement switch, and the method comprises the steps: obtaining a driving current sequence and a power supply voltage sequence of an operation mechanism, and the temperature of an environment where the measurement switch is located; calculating an adaptive path constraint window width of a dog walking distance algorithm; based on the adaptive path constraint window width, calculating a dog walking distance between the driving current sequence and a stored reference healthy waveform sequence; and giving a fault alarm in response to the fact that the dog walking distance is greater than a preset fault threshold value. According to the invention, the anti-interference capability and accuracy of fault prediction under complex working conditions are improved.

Description

Technical Field

[0001] This invention relates to the field of fault prediction, and in particular to a fault prediction method and system for a measuring switch. Background Technology

[0002] With the development of smart grids, the reliability of low-voltage distribution network equipment is receiving increasing attention. Intelligent measurement switches, as core devices, integrate high-precision current sensors, measurement units, and the switch body, enabling real-time acquisition of parameters such as voltage, current, and temperature. In actual operation, the mechanical performance of the operating mechanism (such as a motor or electromagnet) is crucial in determining the reliability of the switch's operation. Traditional monitoring methods often rely on statistics of opening and closing times, which are subject to lag.

[0003] Chinese patent application CN119805145A discloses a method for evaluating the surge current withstand capability of silicon carbide diodes. This method involves collecting reverse IV characteristic data of the device at different temperatures during the wafer testing phase. Based on the positive correlation between the leakage current's rate of change with temperature and the device's peak surge current, the technique introduces a Frechet distance algorithm. In the evaluation process, a judgment threshold is set according to the rated surge current, and the sum of Frechet distances is calculated. By comparing the judgment threshold with the sum of Frechet distances, an accurate assessment of the device's surge current withstand capability can be made.

[0004] Waveform matching algorithms based on "dog-walking distance" are an effective early warning method that can identify mechanical faults by measuring the morphological differences of the driving current waveform. In existing technologies, when calculating waveform similarity using the "dog-walking distance" algorithm, a fixed-size time curvature constraint window is typically used to limit the matching path. However, in real-world scenarios, the operating speed of the measuring switch is significantly affected by the supply voltage. When the mains voltage is low, the output power of the operating mechanism decreases, causing the operating process to physically slow down, and the current waveform is stretched on the time axis. In this case, if a fixed constraint window is used, the algorithm may misjudge this normal slowdown caused by voltage fluctuations as a morphological distortion exceeding the constraint range, thus triggering a false alarm of mechanical jamming. Summary of the Invention

[0005] To address the problem of false alarms that easily occur during the detection of measuring switches, this invention provides a fault prediction method and system for measuring switches.

[0006] In a first aspect, the present invention provides a method for predicting the faults of a measuring switch, employing the following technical solution: A method for predicting faults in a measuring switch, comprising the following steps: The system acquires the drive current sequence, supply voltage sequence, and ambient temperature of the measuring switch; calculates the adaptive path constraint window width of the "walking distance" algorithm, whereby the adaptive path constraint window width is positively correlated with a preset base window width and an adaptive time warp allowance factor; the adaptive time warp allowance factor is positively correlated with the voltage drive performance deficiency and the difference between the current temperature and the reference temperature; the voltage drive performance deficiency is positively correlated with the difference between the rated voltage and the average value of the supply voltage sequence, and the standard deviation of the supply voltage sequence; based on the adaptive path constraint window width, the system calculates the "walking distance" between the drive current sequence and the stored baseline healthy waveform sequence; and issues a fault alarm in response to the "walking distance" exceeding a preset fault threshold.

[0007] By comprehensively considering the stability of the power supply voltage and the difference between the ambient temperature and the reference temperature, the path constraint window width in the dog-walking distance algorithm is dynamically calculated and adjusted. This enables the algorithm to distinguish between normal sluggishness caused by electrical or environmental factors and waveform distortion caused by mechanical failures. As a result, the accuracy of fault warning is significantly improved and the false alarm rate is reduced under complex working conditions.

[0008] Preferably, the expression for the adaptive path constraint window width is:

[0009] In the formula, This represents the width of the adaptive path constraint window based on the supply voltage, expressed in units of sampling points. This indicates the preset base window width under standard reference conditions. This represents the adaptive time warping allowance factor. This is the minimum safety margin. This represents the floor function.

[0010] By introducing a basic window width, an adaptive time warping allowance factor, and a minimum safety margin, the expansion of the constraint window is ensured to be linearly related to the severity of the operating conditions. This guarantees a basic tolerance space under standard operating conditions and allows for precise expansion of the search range under severe operating conditions based on the calculated allowance factor. In this way, the effects of time axis stretching are eliminated while avoiding matching failures caused by excessively large windows, thus ensuring the rigor of waveform comparison.

[0011] Preferably, the expression for the adaptive time-warping allowable factor is:

[0012] In the formula, This represents the adaptive time warping allowance factor. Indicates the degree of voltage drive performance deficiency. This indicates the ambient temperature of the current measuring switch. Indicates the reference temperature. Indicates the maximum allowable operating temperature of the equipment. This indicates the peak value of the starting current in the current action waveform. This represents the peak value of the starting current of the reference waveform.

[0013] A multidimensional adaptive factor model was constructed, which includes voltage driving efficiency, temperature rise effect and starting current decay. It can accurately quantify the degree of driving force reduction from a physical perspective, so that the adaptive factor not only depends on voltage fluctuations, but also reflects the physical fact of high temperature slowdown, thus providing a basis for adjusting the window width that is more in line with the actual physical characteristics.

[0014] Preferably, the expression for the voltage drive performance deficiency is:

[0015] In the formula, Indicates the degree of voltage drive performance deficiency. This indicates the rated operating voltage of the measuring switch. This represents the arithmetic mean of the voltage sequence during the operation. represents the standard deviation of the voltage sequence during the operation, and norm represents the normalization function.

[0016] By combining the arithmetic mean and standard deviation of voltage during the operation process, the abstract power quality problem is transformed into a specific numerical indicator. This indicator can accurately assess the fluctuation of driving torque caused by high internal resistance of the power supply or load impact, providing key data support for judging whether the sluggish operation is caused by electrical factors.

[0017] Preferably, the method for obtaining the reference health waveform sequence is as follows: under the power supply conditions of reference temperature and rated voltage, the driving current data under a standard operation is collected by the current sensor built into the measuring switch, and the reference health waveform sequence is obtained after noise reduction processing.

[0018] By establishing a standard waveform under ideal conditions, the bias of the baseline data itself is eliminated, ensuring that the subsequently calculated dog-walking distance can truly reflect the difference between the current equipment state and the optimal mechanical state, thus improving the reliability of comparative analysis.

[0019] Preferably, the reference temperature is the standard ambient temperature condition for performance testing, and the value is 23℃.

[0020] Preferably, the method for calculating the "dog-walking distance" between the driving current sequence and the stored reference health waveform sequence includes: constructing an Euclidean distance matrix between the real-time acquired driving current waveform sequence and the reference health waveform sequence, and calculating the distance near the diagonal of the distance matrix. Dynamic programming path search is performed within the area to calculate the dog walking distance. i represents the time position of the i-th sampling point in the baseline health waveform sequence, and j represents the time position of the j-th sampling point in the real-time acquired drive current waveform sequence.

[0021] By calculating the dog-walking distance within a restricted area, waveform time-axis stretching caused by voltage or temperature is effectively filtered out, thus forcing the calculation results to mainly reflect the waveform morphology differences, achieving effective decoupling between electrical interference and mechanical faults.

[0022] Preferably, the width of the basic window is 50 sampling points.

[0023] It provides an empirically optimized value for the base window width, balancing the computational efficiency of the algorithm with tolerance for minor time jitter, providing a stable starting point for adaptive adjustment, and preventing missed or false alarms caused by improper base window settings.

[0024] Preferably, the method for obtaining the drive current sequence, power supply voltage sequence, and ambient temperature of the measuring switch is as follows: the drive current sequence is acquired using a built-in current sensor, the power supply voltage sequence is acquired using a built-in voltage sensor, and the ambient temperature of the measuring switch is acquired using a built-in NTC thermistor.

[0025] Data acquisition is achieved by fully utilizing the built-in current sensor, voltage sensor, and NTC thermistor integrated into the measurement switch body, eliminating the need for external testing equipment. This allows fault prediction functionality to be implemented directly at the device end, reducing implementation costs and enabling real-time, in-situ monitoring of the device's operating environment and internal status.

[0026] Secondly, the present invention provides a fault prediction system for a measuring switch, which adopts the following technical solution: A fault prediction system for a measuring switch includes a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, a fault prediction method for a measuring switch as described above is implemented.

[0027] The above-mentioned fault prediction method for a measuring switch is generated into a computer program and stored in a memory so that it can be loaded and executed by a processor. Thus, a system can be built based on the memory and the processor for convenient use.

[0028] The present invention has the following technical effects: By constructing a multidimensional adaptive factor model that includes voltage driving performance, temperature drift, and starting current, and dynamically adjusting the constraint window width of waveform matching, this method effectively decouples the normal time-varying electrical environment from the deformation caused by mechanical faults, and improves the anti-interference ability and accuracy of fault prediction under complex working conditions. Attached Figure Description

[0029] Figure 1 This is a flowchart of a fault prediction method for a measuring switch according to the present invention.

[0030] Figure 2 This is a diagram illustrating the matching of existing technologies.

[0031] Figure 3 This is a schematic diagram of the matching process of the present invention.

[0032] Figure 4 This is a diagram illustrating the quantitative comparison of fault determination results. Detailed Implementation

[0033] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0034] This invention discloses a fault prediction method for a measuring switch, referring to... Figure 1 This includes the following steps: S1: Collect driving current data, power supply voltage data, and temperature data of the measuring switch.

[0035] Utilizing the high-precision current and voltage sensors built into the intelligent measuring switch, high-frequency sampling (e.g., 10kHz) is triggered the instant an action command is received, acquiring the drive current sequence and power supply voltage sequence of the operating mechanism control circuit. Simultaneously, the ambient temperature of the measuring switch is obtained using an NTC sensor built into the measuring switch to monitor terminal temperature.

[0036] S2: Analyze the power supply voltage data and temperature characteristics, and calculate the adaptive time bending constraint window width for each action waveform matching.

[0037] S21: Analyze the characteristics of the power supply voltage data at the moment of operation to obtain the voltage drive performance deficiency.

[0038] The driving force of the operating mechanism comes directly from the voltage. If the average voltage during operation is lower than the rated value, it will lead to insufficient output torque and slower operation. In specific analysis, the greater the drop in the average voltage during operation relative to the rated voltage, the more scarce the driving energy, the more reasonable the sluggish operation, and the greater the corresponding voltage drive efficiency deficiency. Furthermore, voltage stability is also crucial. More severe voltage fluctuations during operation indicate a higher internal resistance of the power supply or a greater possibility of load surges, which will further deteriorate the smoothness of the motor torque, resulting in a longer operation time than expected. Therefore, the more severe the voltage fluctuations, the greater the voltage drive efficiency deficiency should be. Based on the above principles, the voltage drive efficiency deficiency is calculated as follows:

[0039] In the formula, Indicates the degree of voltage drive performance deficiency. This indicates the rated operating voltage of the measuring switch. This represents the arithmetic mean of the voltage sequence during the operation. This represents the standard deviation of the voltage sequence during the operation, and norm represents the normalization function, specifically the linear normalization algorithm.

[0040] In the formula The larger the value, the greater the voltage drop relative to the rated voltage, the weaker the driving force, the slower the action, and the greater the corresponding lack of voltage driving efficiency; in the formula The larger the value, the more severe the voltage fluctuation during the operation, the worse the power supply quality, the greater the damage to the smoothness of the operation, the longer the operation time is than expected, and the greater the corresponding lack of voltage drive efficiency.

[0041] S22: Analyze the temperature characteristics of the operating mechanism coil and calculate the adaptive time bending allowable factor in combination with the voltage drive efficiency deficiency.

[0042] By calculating the voltage drive efficiency deficiency, the potential for slower operation can be reflected from the voltage perspective. In real-world physical scenarios, there exists a special case of high temperature and high resistance: even with a normal supply voltage, if the measuring switch is in a high-temperature environment or the operating coil heats up due to frequent operation, the copper resistance of the operating coil will increase significantly. According to Ohm's law, with a constant voltage, increased resistance leads to a decrease in drive current, which in turn reduces the Ampere force, ultimately resulting in slower operation.

[0043] The difference between high-temperature, high-resistance detection methods and those considering only voltage lies in the fact that voltage-only methods assume normal speed when voltage is normal, ignoring the weakening effect of temperature on resistance and driving force. Therefore, this step introduces temperature characteristics to optimize the voltage drive efficiency deficiency, resulting in an adaptive time bending allowable factor. The higher the current temperature relative to the reference temperature, the greater the increase in coil resistance, the more severe the drive force attenuation, and the more reasonable the slowdown in action. Therefore, it is necessary to amplify and correct the voltage drive efficiency deficiency, resulting in a larger adaptive time bending allowable factor. It should be noted that the reference temperature refers to the standard ambient temperature conditions specified for performance testing and measurement of the intelligent measurement switch, preset to 23℃. Simultaneously, to verify whether the driving force has truly attenuated, the change in the peak starting current can be analyzed. If the current peak starting current is significantly lower than the reference value, it indicates that the driving force is indeed insufficient, rather than mechanically jammed, and the resulting adaptive time bending allowable factor should also be larger. Based on the above principles, the adaptive time bending allowable factor is calculated as follows:

[0044] In the formula, This represents the adaptive time warping allowance factor. Indicates the degree of voltage drive performance deficiency. This indicates the ambient temperature of the current measuring switch. This indicates the reference temperature, which is the standard ambient temperature condition specified when performing performance testing and measurement on the intelligent measuring switch. The default setting is 23℃. Indicates the maximum allowable operating temperature of the equipment. This indicates the peak value of the starting current in the current action waveform. The peak value of the starting current, representing the reference waveform, can be obtained from the performance test data before the equipment leaves the factory.

[0045] In the formula The larger the value, the higher the current temperature, the more pronounced the coil thermal resistance effect, and the greater the allowable deviation in operating time at high temperatures, resulting in a larger adaptive time bending allowable factor. In the formula... The larger the value, the more severe the attenuation of the current peak starting current relative to the reference value, and the greater the possibility of a decrease in driving force. This further indicates that the slowdown in action is more likely due to electrical factors rather than mechanical failure. Therefore, the larger the adaptive time warping allowance factor, the greater the tolerance given to the algorithm.

[0046] S23: Based on the adaptive time curvature allowable factor, the adaptive path constraint window width based on the supply voltage is calculated.

[0047] A larger adaptive time warping allowance factor indicates more severe voltage and temperature conditions, increasing the likelihood of slower switching action. Therefore, a larger constraint window is needed to allow the dog-walking algorithm to perform significant time warping to eliminate computational errors caused by time stretching. Conversely, a smaller allowance factor indicates better operating conditions, requiring faster switching action. In this case, the window should be tightened to maintain high sensitivity to mechanical jamming. Based on these principles, the adaptive path constraint window width is calculated using the following expression:

[0048] In the formula, This represents the width of the adaptive path constraint window based on the supply voltage, expressed in units of sampling points. This indicates the preset base window width under standard reference conditions. , This represents the adaptive time warping allowance factor. This is the minimum safety margin. , This represents the floor function.

[0049] S3: Calculate waveform morphology differences using the optimized dog-walking distance algorithm and provide fault warnings.

[0050] Through the above steps, we obtained the adaptive path constraint window width optimized for the current working conditions. This parameter is then used to directly optimize the calculation process of the dog-walking distance. The specific method is as follows: To obtain a baseline health waveform sequence, during the pre-shipment performance test of the intelligent measurement switch, under reference temperature (23℃) and rated voltage power supply conditions, the drive current data under a single standard operation is collected using the current sensor built into the measurement switch. After noise reduction, the data is stored in the memory of the measurement unit, constructing an Euclidean distance matrix between the real-time acquired drive current waveform sequence and the baseline health waveform sequence. When calculating the discrete dog-walking distance, an adaptive path constraint window width is introduced as a constraint condition, i.e., only the area near the diagonal of the distance matrix... The dog-walking distance is calculated by performing dynamic programming path search within the area. This refers to setting a fault warning threshold based on the purely mechanical morphological differences after eliminating the effects of voltage fluctuations and temperature drift. ,like If the signal is detected, it indicates a risk of mechanical jamming or wear in the transmission chain, requiring a timely maintenance alarm. It should be noted that i represents the time position of the i-th sampling point in the baseline health waveform sequence, and j represents the time position of the j-th sampling point in the real-time acquired drive current waveform sequence.

[0051] Combination Figure 2As shown in the figure, this illustrates the limitations of existing technologies using fixed-width constraint windows. Low voltage or high temperature slows motor movement, causing the waveform to stretch on the time axis, resulting in a significant shift in the optimal matching path and collisions with the window boundary. The path is forcibly truncated outside the window, leading the algorithm to calculate incorrect maximum distance values ​​and misinterpreting normal sluggishness as mechanical failure.

[0052] Combination Figure 3 As shown in the figure, this demonstrates the effectiveness of the adaptive window algorithm of this invention. The window width calculated by the algorithm in this scheme is dynamically and significantly widened. Matching paths that experience time-axis stretching are completely enveloped within the constraint window without touching the boundaries. This indicates that the algorithm tolerates non-faulty time lags, maintains the calculated distance at a low level, and correctly determines the device health.

[0053] Combination Figure 4 As shown in the figure, the relationship between the normalized dog walking distance and the alarm threshold of the two methods under different working conditions is intuitively compared. The difference lies in the high temperature interference group in the middle. The value of the existing technology is as high as 2.50, which is far beyond the threshold and produces false alarms. The value of the present invention is reduced to 0.25, which is far below the threshold and achieves correct identification. The data on the right shows that both maintain high sensitivity to real faults.

[0054] This invention also discloses a fault prediction system for a measuring switch, including a processor and a memory. The memory stores computer program instructions, which, when executed by the processor, implement a fault prediction method for a measuring switch according to the present invention.

[0055] The system also includes other components well known to those skilled in the art, such as communication buses and communication interfaces, the settings and functions of which are known in the art and will not be described in detail here.

[0056] The above are all preferred embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Therefore, all equivalent changes made in accordance with the structure, shape and principle of the present invention should be covered within the scope of protection of the present invention.

Claims

1. A method of failure prediction of a measuring switch, characterized in that, Including the following steps: The system acquires the drive current sequence and supply voltage sequence of the operating mechanism, as well as the ambient temperature of the measuring switch. It calculates the adaptive path constraint window width for the dog-walking distance algorithm. This adaptive path constraint window width is positively correlated with a preset base window width and an adaptive time curvature allowable factor. The adaptive time curvature allowable factor is positively correlated with the voltage drive performance deficiency and the difference between the current temperature and the reference temperature. The voltage drive performance deficiency is positively correlated with the difference between the rated voltage and the average value of the supply voltage sequence, and the standard deviation of the supply voltage sequence. Based on the adaptive path constraint window width, the dog-walking distance between the driving current sequence and the stored reference health waveform sequence is calculated; in response to the dog-walking distance being greater than a preset fault threshold, a fault alarm is issued.

2. The fault prediction method for a measuring switch according to claim 1, characterized in that, The expression for the width of the adaptive path constraint window is: In the formula, This represents the width of the adaptive path constraint window based on the supply voltage, expressed in units of sampling points. This indicates the preset base window width under standard reference conditions. This represents the adaptive time warping allowance factor. This is the minimum safety margin. This represents the floor function.

3. The fault prediction method for a measuring switch according to claim 1, characterized in that, The expression for the adaptive time warping allowance factor is: In the formula, This represents the adaptive time warping allowance factor. Indicates the degree of voltage drive performance deficiency. This indicates the ambient temperature of the current measuring switch. Indicates the reference temperature. Indicates the maximum allowable operating temperature of the equipment. This indicates the peak value of the starting current in the current action waveform. This represents the peak value of the starting current of the reference waveform.

4. The fault prediction method for a measuring switch according to claim 1, characterized in that, The expression for the voltage drive performance deficiency is: In the formula, Indicates the degree of voltage drive performance deficiency. This indicates the rated operating voltage of the measuring switch. This represents the arithmetic mean of the voltage sequence during the operation. represents the standard deviation of the voltage sequence during the operation, and norm represents the normalization function.

5. The fault prediction method for a measuring switch according to claim 1, characterized in that, The method for obtaining the reference health waveform sequence is as follows: under the power supply conditions of reference temperature and rated voltage, the driving current data under a standard operation is collected by the current sensor built into the measuring switch, and the reference health waveform sequence is obtained after noise reduction processing.

6. The fault prediction method for a measuring switch according to claim 5, characterized in that, The reference temperature is the standard ambient temperature condition for performance testing, and its value is 23℃.

7. The fault prediction method for a measuring switch according to claim 2, characterized in that, The method for calculating the "dog-walking distance" between the drive current sequence and the stored reference health waveform sequence includes: constructing an Euclidean distance matrix between the real-time acquired drive current waveform sequence and the reference health waveform sequence, and calculating the distance near the diagonal of the distance matrix. Dynamic programming path search is performed within the area to calculate the dog walking distance. i represents the time position of the i-th sampling point in the baseline health waveform sequence, and j represents the time position of the j-th sampling point in the real-time acquired drive current waveform sequence.

8. The fault prediction method for a measuring switch according to claim 2, characterized in that, The basic window width is 50 sampling points.

9. The fault prediction method for a measuring switch according to claim 1, characterized in that, The method for obtaining the drive current sequence, power supply voltage sequence, and ambient temperature of the measuring switch is as follows: the drive current sequence is acquired using a built-in current sensor, the power supply voltage sequence is acquired using a built-in voltage sensor, and the ambient temperature of the measuring switch is acquired using a built-in NTC thermistor.

10. A fault prediction system for a measuring switch, characterized in that, include: A processor and a memory, the memory storing computer program instructions that, when executed by the processor, implement a fault prediction method for a measuring switch according to any one of claims 1-9.

Images