Measurement switch whole life cycle health state early warning method based on digital twinning

Abstract

This invention relates to the field of condition early warning technology, specifically disclosing a method for early warning of the health status of a measurement switch throughout its entire lifecycle based on digital twins. The method includes the following steps: S1: Constructing a model that converts the actual operating time of the target switch to the equivalent operating time under a preset standard operating condition based on its historical load data; S2: Constructing a digital twin for switches of the same type in the power grid, obtaining performance degradation data sequences under different typical operating conditions through simulation, and generating a corresponding baseline health status degradation curve library accordingly; S3: Acquiring the load data of the target switch in real time, calculating its current equivalent operating time through the conversion model, and matching it with the baseline curve that best matches its real-time operating condition; S4: Comparing the real-time performance parameters of the target switch with the expected values ​​of the baseline curve at the current equivalent operating time point, and judging whether its health status is abnormal based on the deviation.

Description

Technical Field

[0001] This invention relates to the field of status early warning technology, specifically to a method for early warning of the health status of a measurement switch throughout its entire lifecycle based on digital twins. Background Technology

[0002] In power systems and related fields, measuring switches are key equipment, and their operational health status is crucial to the stable operation of the entire system.

[0003] In existing technologies, monitoring and early warning of the operational health status of measuring switches often fails to distinguish between normal data fluctuations and potential abnormal data, leading to frequent misjudgments. On one hand, the complex electromagnetic environment in which the measuring switches operate, their frequent long-term operation, and the alternating influence of different operating conditions result in a large amount of noise interference in the collected data. This interference intertwines with normal data fluctuations, making accurate differentiation difficult. On the other hand, the lack of effective models and algorithms to deeply mine data features makes it impossible to accurately identify early signs of anomalies hidden within large amounts of normal data. This situation not only wastes significant human and material resources on unnecessary maintenance due to misjudgments but may also lead to serious consequences such as equipment damage and power outages due to the failure to detect genuine potential faults in a timely manner, greatly impacting the reliability and economy of the power system. Summary of the Invention

[0004] The purpose of this invention is to provide a method for early warning of the health status of a measurement switch throughout its entire lifecycle based on digital twins, thereby solving the following technical problems.

[0005] The objective of this invention can be achieved through the following technical solutions: A digital twin-based method for early warning of the health status of measurement switches throughout their entire lifecycle includes the following steps: S1: Identify the target measurement switch in the power grid, obtain the historical load data of the target measurement switch, and preset a fixed standard operating condition. Construct an equivalent operating time conversion model based on the historical load data. The input of the equivalent operating time conversion model is the load data, and the output is the equivalent operating time of the target measurement switch under the fixed standard operating condition. S2: Construct digital twins of each measuring switch in the power grid; set several typical operating conditions, and for any typical operating condition, simulate each digital twin under the typical operating condition to obtain a degradation data sequence; based on all degradation data sequences, generate a library of baseline health state degradation curves for each typical operating condition. S3: Obtain the real-time load data of the target measurement switch, and obtain the current equivalent running time based on the real-time load data and the equivalent running time conversion model; match the typical working condition that is consistent with the real-time load data, record it as the current working condition, and obtain the baseline health status degradation curve library of the current working condition. S4: Based on the baseline health status degradation curve library under the current operating conditions and the current equivalent running time, determine the baseline expected value set of the target measurement switch; obtain the real-time performance parameters of the target measurement switch, obtain the deviation between the real-time performance parameters and the baseline expected value set, and determine whether there is an abnormality in the health status of the target measurement switch based on the deviation.

[0006] As a further aspect of the present invention: the historical load data includes several load parameters of a measuring switch of the same type as the target measuring switch in the past daily data, the load parameters including grid load current data, the number of times historical disconnection and opening operations were performed, the current amplitude when historical disconnection and opening operations were performed, and environmental data; the environmental data includes temperature data and humidity data.

[0007] As a further aspect of the present invention: the setting of the fixed standard operating conditions includes an electrical load of 80% of the rated current of the target measuring switch, an ambient temperature of 25°C, and the number of times the disconnect or open operation is performed as the daily typical operation count; the daily typical operation count is the average number of disconnect and open operations performed daily based on historical statistical data.

[0008] As a further aspect of the present invention: the construction process of the equivalent runtime conversion model includes: Based on the historical load data, the basic loss rate of the target measurement switch under fixed standard operating conditions is determined through digital twin simulation experiments. The basic loss rate is the change in loss of the target measurement switch per unit time, and the change in loss is the amount of contact wear or the degree of insulation aging. For any load parameter in the historical load data, the basic loss rate of the load parameter is determined by the control variable method and denoted as the single-parameter loss factor. Correction coefficients are set for each single-parameter loss factor, and a regression equation is constructed based on each single-parameter loss factor and its correction coefficient. Based on the historical load data, the values ​​of each correction coefficient are continuously adjusted using the least squares method to minimize the error between the basic loss rate after coupling each single-parameter loss factor and its correction coefficient and the historical load data. The final values ​​of each correction coefficient are recorded and denoted as the coupling correction coefficients. Each coupling correction coefficient is associated with and coupled with its corresponding single-parameter loss factor to obtain the comprehensive loss factor. The equivalent running time is then calculated. K i Let t be the comprehensive loss factor for day i in the historical load data. real-i Let n be the actual running time under the working conditions on day i, n be the total number of days in the historical load data, and i be the index.

[0009] As a further aspect of the present invention: the process of obtaining the degraded data sequence includes: A virtual operating environment matching typical operating conditions is constructed, and a physical field simulation model is invoked. This physical field simulation model includes electrical, mechanical, and thermodynamic physical fields to simulate the physical mechanism of performance degradation of the measurement switch under typical operating conditions. Load parameter combinations for typical operating conditions are input into the digital twin of the measurement switch to complete the simulation of the entire lifecycle of the measurement switch. During the simulation, the time scale followed by the internal process is converted into equivalent operating time, and the performance parameters of the digital twin of the measurement switch are collected synchronously to obtain a single-twin degradation data sequence showing the change of performance parameters with equivalent operating time. The performance parameters include several performance indices, including contact resistance, partial discharge, and mechanical vibration signal. The single-twin degradation data sequence of each measurement switch's digital twin is obtained, and the average value of the single-twin degradation data sequence of each digital twin under typical operating conditions is obtained to obtain the average single-twin degradation data sequence, which is denoted as the degradation data sequence under this typical operating condition.

[0010] As a further aspect of the present invention: the process of generating a library of baseline health state degradation curves for typical operating conditions includes: For any typical operating condition, obtain the degradation data sequence of the typical operating condition, split the degradation data sequence into several datasets of performance indicators changing with equivalent running time, and match a mathematical fitting model for each dataset. Input each dataset into the matching mathematical fitting model for fitting to obtain the fitting curve of each dataset. The fitting curves of each dataset constitute the baseline health state degradation curve library of the typical operating condition.

[0011] As a further aspect of the present invention: the process of matching typical operating conditions consistent with the real-time load data includes: Obtain all load parameters from the real-time load data to obtain a real-time load parameter set; for any typical working condition, obtain all load parameters of the typical working condition to obtain a working condition load parameter set; obtain the similarity between the real-time load parameter set and each working condition load parameter set, select the working condition load parameter set with the highest similarity to the real-time load parameter set, denoted as the best matching set, and denote the typical working condition corresponding to the best matching set as the current working condition.

[0012] As a further aspect of the present invention: the process of determining the reference expected value set of the target measurement switch includes: For any performance index fitting curve in the baseline health state degradation curve library for the current operating condition, substitute the current equivalent running time into the fitting curve, and obtain the ordinate of the fitting curve when the horizontal axis is the current equivalent running time. This ordinate is denoted as the baseline expected value of the performance index. Thus, the baseline expected values ​​of each performance index are obtained, resulting in a set of baseline expected values.

[0013] The beneficial effects of this invention are: This invention achieves accurate and early warning of the health status of measurement switches throughout their entire lifecycle by constructing an equivalent operating time conversion model and a multi-condition benchmark health status degradation curve library. By uniformly converting real and varied operating conditions to equivalent operating time under standard operating conditions, it establishes comparable aging ages for devices with different operating histories. Furthermore, by utilizing a benchmark curve library generated by digital twin cluster simulation and covering multiple typical operating conditions, it provides dynamically adaptable health benchmarks for individual devices. This transforms the early warning logic from traditional absolute threshold judgment to relative aging rate assessment, thereby enabling the keen identification of systematic and trend deviations in performance degradation and significantly reducing misjudgments and missed alarms caused by noise interference and changes in operating conditions. Attached Figure Description

[0014] The invention will now be further described with reference to the accompanying drawings.

[0015] Figure 1 This is a schematic diagram illustrating the steps of the method for early warning of the health status of a measurement switch throughout its entire lifecycle based on digital twins, as proposed in this invention. Detailed Implementation

[0016] 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 embodiments of the present invention, and not all embodiments. 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.

[0017] Please see Figure 1 As shown, this invention is a method for early warning of the health status of a measurement switch throughout its entire lifecycle based on digital twins, comprising the following steps: S1: Determine the target measurement switch in the power grid, obtain the historical load data of the target measurement switch, and preset a fixed standard operating condition. Construct an equivalent operating time conversion model based on the historical load data. The input of the equivalent operating time conversion model is the load data, and the output is the equivalent operating time of the target measurement switch under the fixed standard operating condition. In a preferred embodiment of the present invention, the historical load data includes several load parameters of a measuring switch of the same type as the target measuring switch on a past day. The load parameters include grid load current data, the number of times historical disconnection and opening operations were performed, the current amplitude during historical disconnection and opening operations, and environmental data. The environmental data includes temperature data and humidity data. In a preferred embodiment of the present invention, the setting of the fixed standard operating conditions includes an electrical load of 80% of the rated current of the target measuring switch, an ambient temperature of 25°C, and the number of times the disconnect or open operation is performed as the daily typical operation number; the daily typical operation number is the average number of disconnect and open operations performed daily based on historical statistical data. In a preferred embodiment of the present invention, the process of constructing the equivalent runtime conversion model includes: Based on the historical load data, the basic loss rate of the target measurement switch under fixed standard operating conditions is determined through digital twin simulation experiments. The basic loss rate is the change in loss of the target measurement switch per unit time, and the change in loss is the amount of contact wear or the degree of insulation aging. For any load parameter in the historical load data, the basic loss rate of the load parameter is determined by the control variable method and denoted as the single-parameter loss factor. Correction coefficients are set for each single-parameter loss factor, and a regression equation is constructed based on each single-parameter loss factor and its correction coefficient. Based on the historical load data, the values ​​of each correction coefficient are continuously adjusted using the least squares method to minimize the error between the basic loss rate after coupling each single-parameter loss factor and its correction coefficient and the historical load data. The final values ​​of each correction coefficient are recorded and denoted as the coupling correction coefficients. Each coupling correction coefficient is associated with and coupled with its corresponding single-parameter loss factor to obtain the comprehensive loss factor. The equivalent running time is then calculated. K i Let t be the comprehensive loss factor for day i in the historical load data. real-i Let n be the actual running time under the working conditions on day i, n be the total number of days in the historical load data, and i be the index. Specifically, based on the historical load data, the basic loss rate per unit time under the fixed standard operating condition is determined through digital twin simulation or accelerated aging experiments. For each load parameter in the historical load data, the loss characteristics of each load parameter acting alone are analyzed using the controlled variable method, and the single-parameter loss factor of each load parameter relative to the standard operating condition is quantified. Through multivariate regression analysis of historical data, the coupling effect when multiple load parameters act simultaneously is corrected to obtain a comprehensive multi-parameter coupling loss factor, i.e., a comprehensive loss factor. A conversion formula is established based on the loss equivalence principle, whereby the equivalent running time is equal to the integral of the multi-parameter coupling loss factor over time during the actual running time. For discretized historical load data, the conversion model calculates the cumulative equivalent running time by summing the product of the multi-parameter coupling loss factor and the actual running time of the corresponding period. It should be noted that the core logic of the equivalent running time is that the total loss under real operating conditions equals the total loss under fixed standard operating conditions, while the total loss under real operating conditions... , treal Let t be the actual operating time under real-world conditions, L0 be the base loss rate, and K(t) be the comprehensive loss factor at time t under real-world conditions. The total loss under standard operating conditions is L... eq =t eq ×L0; Let L real =L eq Eliminating L0 yields the conversion formula. For discrete historical load data, the conversion formula is simplified to: ; S2: Obtain all measurement switches in the power grid and construct digital twins for each measurement switch; set several typical operating conditions, and for any typical operating condition, simulate each digital twin under the typical operating condition to obtain a degradation data sequence; based on the degradation data sequence, generate a library of baseline health state degradation curves for each typical operating condition. In a preferred embodiment of the present invention, the process of setting several typical working conditions includes performing cluster analysis on the historical load data, identifying several clusters, and combining a set of load parameters defined by the center point of each cluster as a typical working condition. Specifically, based on the historical operation records of the same type of measurement switch group corresponding to the digital twin sample cluster, cluster analysis is performed on the historical load data to identify the mainstream operation modes with different characteristics; according to the results of the cluster analysis, load parameter center points or feature intervals that can represent each mainstream operation mode are selected, and a set of load parameter conditions defined by each center point or feature interval is set as a typical operating condition. In a preferred embodiment of the present invention, the process of obtaining the degraded data sequence includes: A virtual operating environment matching typical operating conditions is constructed, and a physical field simulation model is invoked. This model includes electrical, mechanical, and thermodynamic physical fields to simulate the physical mechanism of performance degradation of the measurement switch under typical operating conditions. Load parameter combinations for typical operating conditions are input to the digital twin of the measurement switch to complete the simulation of the entire lifecycle of the measurement switch. During the simulation, the time scale followed by the internal processes is converted to equivalent operating time, and the performance parameters of the digital twin of the measurement switch are simultaneously collected to obtain a single-twin degradation data sequence showing the change of performance parameters with equivalent operating time. These performance parameters include several performance indices, such as contact resistance, partial discharge, and mechanical vibration signals. The single-twin degradation data sequences of each measurement switch's digital twin are obtained, and the average value of the single-twin degradation data sequences of each digital twin under typical operating conditions is obtained to obtain the average single-twin degradation data sequence, which is denoted as the degradation data sequence under that typical operating condition. Specifically, for each typical operating condition, a matching virtual operating environment is built, and a multiphysics coupled simulation model covering electrical, mechanical, and thermal aspects is invoked to simulate the physical mechanism of performance degradation of the measuring switch under that condition. An accelerated simulation strategy based on equivalent operating time is adopted to compress the real full life cycle timescale into the simulation timescale, and the simulation time step is dynamically adjusted according to the life cycle stage of the measuring switch. According to the full life cycle stages from the break-in period, stable operation period to the aging period, the digital twin is driven to run continuously under the load of the typical operating condition until the preset life end threshold is reached. During the simulation, the equivalent operating time is used as the time scale to collect a set of evolution data of key performance parameters in real time. The collected evolution data of the key performance parameters synchronized with the time scale is organized and output as the degradation data sequence. In a preferred embodiment of the present invention, the process of generating the baseline health state degradation curve library for typical operating conditions includes: For any typical working condition, obtain the degradation data sequence of the typical working condition, split the degradation data sequence into several datasets of performance indicators changing with equivalent running time, and match a mathematical fitting model for each dataset. Input each dataset into the matching mathematical fitting model for fitting to obtain the fitting curve of each dataset. The fitting curves of each dataset constitute the baseline health state degradation curve library of the typical working condition. Specifically, for each typical operating condition, the degradation data sequence is split according to the type of key performance parameters, forming contact resistance dataset, partial discharge dataset, and mechanical vibration signal dataset corresponding to the equivalent operating time. Based on the physical degradation law of each key performance parameter, a corresponding mathematical fitting model is matched for each dataset. Specifically, a linear or polynomial regression model is matched for the contact resistance dataset, an exponential growth model is matched for the partial discharge dataset, and a piecewise function or spline fitting model is matched for the mechanical vibration signal dataset. Each dataset is input into its matched mathematical fitting model, and the parameters are solved by the least squares method to obtain the continuous fitting equation of each key performance parameter changing with the equivalent operating time. Based on each fitting equation, a baseline health state degradation curve library corresponding to the typical operating condition is generated. S3: Obtain the real-time load data of the target measurement switch, and obtain the current equivalent operating time of the target measurement switch based on the real-time load data and the equivalent operating time conversion model; match the typical working condition that is consistent with the real-time load data, record it as the current working condition, and obtain the reference health status degradation curve library corresponding to the current working condition. In a preferred embodiment of the present invention, the process of the equivalent running time of the target measurement switch includes: The real-time load data is input into the equivalent running time conversion model. The equivalent running time conversion model generates the current comprehensive loss factor based on the real-time load data, which is denoted as the real-time comprehensive loss factor. Based on the real-time comprehensive loss factor, the model outputs the current equivalent running time. Specifically, the load data of the target measurement switch is acquired in real time, and the load data contains the same parameters as the historical load data; the real-time load data is input into the equivalent running time conversion model, which outputs the real-time comprehensive loss factor at the current moment based on a predefined loss factor mapping relationship; according to the loss equivalence principle followed by the conversion model, the integral of the real-time comprehensive loss factor over time in the current period is added to the existing historical accumulated value of the equivalent running time of the target measurement switch, thereby updating its current equivalent running time; In a preferred embodiment of the present invention, the process of matching typical operating conditions consistent with the real-time load data includes: Obtain all load parameters from the real-time load data to obtain a real-time load parameter set; for any typical working condition, obtain all load parameters of the typical working condition to obtain a working condition load parameter set; obtain the similarity between the real-time load parameter set and each working condition load parameter set, select the working condition load parameter set with the highest similarity to the real-time load parameter set, denoted as the best matching set, and denote the typical working condition corresponding to the best matching set as the current working condition; S4: Based on the baseline health status degradation curve library of the current operating condition and the current equivalent running time, determine the baseline expected value set of the target measurement switch; obtain the real-time performance parameters of the target measurement switch, obtain the deviation between the real-time performance parameters and the baseline expected value set, and determine whether there is an abnormality in the health status of the target measurement switch based on the deviation. In a preferred embodiment of the present invention, the process of determining the reference expected value set of the target measurement switch includes: For any performance index fitting curve in the baseline health state degradation curve library for the current operating condition, substitute the current equivalent running time into the fitting curve, and obtain the ordinate of the fitting curve when the horizontal axis is the current equivalent running time. This ordinate is denoted as the baseline expected value of the performance index. Thus, the baseline expected values ​​of each performance index are obtained, resulting in a set of baseline expected values. In a preferred embodiment of the present invention, the process of obtaining the deviation between real-time performance parameters and the set of baseline expected values ​​includes: The real-time performance parameters include several real-time performance indices. For any real-time performance index, the benchmark expected value of the corresponding performance index is obtained from the benchmark expected value set, thus obtaining the individual deviation between the real-time performance index and the benchmark expected value. , where R real For real-time performance metrics, R ref The baseline expected value is used as the benchmark. The deviation of each individual real-time performance indicator from its corresponding baseline expected value is obtained, and the average value of all individual deviations is obtained, which is recorded as the deviation between the real-time performance parameter and the set of baseline expected values. In a preferred embodiment of the present invention, the process of determining whether the health status of the target measurement switch is abnormal based on the deviation includes: A deviation threshold is set. If the deviation is greater than or equal to the deviation threshold, it is directly determined that the health status of the target measurement switch is abnormal. If the deviation is less than the deviation threshold, the deviation of the target measurement switch is continuously collected within a preset time window to obtain a deviation sequence. The deviation sequence is fitted using the least squares method to generate a proportional regression line, and the slope of the regression line is obtained. If the slope is greater than a preset slope threshold, the health status of the target measurement switch is abnormal. If the slope is less than or equal to the preset slope threshold, the deviation of the target measurement switch is monitored.

[0018] The foregoing has provided a detailed description of one embodiment of the present invention, but this description is merely a preferred embodiment and should not be construed as limiting the scope of the invention. All equivalent variations and modifications made within the scope of the present invention should still fall within the scope of the present invention.

Claims

1. A method for early warning of the health status of a measurement switch throughout its entire lifecycle based on digital twins, characterized in that, Includes the following steps: S1: Identify the target measurement switch in the power grid, obtain the historical load data of the target measurement switch, and preset a fixed standard operating condition. Construct an equivalent operating time conversion model based on the historical load data. The input of the equivalent operating time conversion model is the load data, and the output is the equivalent operating time of the target measurement switch under the fixed standard operating condition. S2: Construct digital twins of each measuring switch in the power grid; set several typical operating conditions, and for any typical operating condition, simulate each digital twin under the typical operating condition to obtain a degradation data sequence; based on all degradation data sequences, generate a library of baseline health state degradation curves for each typical operating condition. S3: Obtain the real-time load data of the target measurement switch, and obtain the current equivalent running time based on the real-time load data and the equivalent running time conversion model; match the typical working condition that is consistent with the real-time load data, record it as the current working condition, and obtain the baseline health status degradation curve library of the current working condition. S4: Based on the current operating condition's baseline health status degradation curve library and the current equivalent operating time, determine the baseline expected value set for the target measurement switch; Obtain the real-time performance parameters of the target measurement switch, obtain the deviation between the real-time performance parameters and the set of baseline expected values, and determine whether there is any abnormality in the health status of the target measurement switch based on the deviation.

2. The method for early warning of the full life cycle health status of a measurement switch based on digital twin as described in claim 1, characterized in that, The historical load data includes several load parameters of the same type of measuring switch as the target measuring switch on a past day. The load parameters include grid load current data, the number of times historical disconnect and open operations were performed, the current amplitude during historical disconnect and open operations, and environmental data. The environmental data includes temperature data and humidity data.

3. The method for early warning of the full life cycle health status of a measurement switch based on digital twin as described in claim 1, characterized in that, The fixed standard operating conditions are set as follows: the electrical load is 80% of the rated current of the target measuring switch, the ambient temperature is 25°C, and the number of times the disconnect or open operation is performed is the daily typical operation count; the daily typical operation count is the average number of disconnect and open operations performed daily based on historical statistical data.

4. The method for early warning of the full life cycle health status of a measurement switch based on digital twin as described in claim 1, characterized in that, The construction process of the equivalent runtime conversion model includes: Based on the historical load data, the basic loss rate of the target measurement switch under fixed standard operating conditions is determined through digital twin simulation experiments. The basic loss rate is the change in loss of the target measurement switch per unit time, and the change in loss is the amount of contact wear or the degree of insulation aging. For any load parameter in the historical load data, the basic loss rate of the load parameter is determined by the control variable method and denoted as the single-parameter loss factor. Correction coefficients are set for each single-parameter loss factor, and a regression equation is constructed based on each single-parameter loss factor and its correction coefficient. Based on the historical load data, the values ​​of each correction coefficient are continuously adjusted using the least squares method to minimize the error between the basic loss rate after coupling each single-parameter loss factor and its correction coefficient and the historical load data. The final values ​​of each correction coefficient are recorded and denoted as the coupling correction coefficients. Each coupling correction coefficient is associated with and coupled with its corresponding single-parameter loss factor to obtain the comprehensive loss factor. The equivalent running time is then calculated. K i Let t be the comprehensive loss factor for day i in the historical load data. real-i Let n be the actual running time under the working conditions on day i, n be the total number of days in the historical load data, and i be the index.

5. The method for early warning of the full life cycle health status of a measurement switch based on digital twin as described in claim 1, characterized in that, The process of obtaining the degraded data sequence includes: A virtual operating environment matching typical operating conditions is constructed, and a physical field simulation model is invoked. This physical field simulation model includes electrical, mechanical, and thermodynamic physical fields to simulate the physical mechanism of performance degradation of the measurement switch under typical operating conditions. Load parameter combinations for typical operating conditions are input into the digital twin of the measurement switch to complete the simulation of the entire lifecycle of the measurement switch. During the simulation, the time scale followed by the internal process is converted into equivalent operating time, and the performance parameters of the digital twin of the measurement switch are collected synchronously to obtain a single-twin degradation data sequence showing the change of performance parameters with equivalent operating time. The performance parameters include several performance indices, including contact resistance, partial discharge, and mechanical vibration signal. The single-twin degradation data sequence of each measurement switch's digital twin is obtained, and the average value of the single-twin degradation data sequence of each digital twin under typical operating conditions is obtained to obtain the average single-twin degradation data sequence, which is denoted as the degradation data sequence under this typical operating condition.

6. The method for early warning of the full life cycle health status of a measurement switch based on digital twin as described in claim 1, characterized in that, The process of generating the baseline health status degradation curve library for typical operating conditions includes: For any typical operating condition, obtain the degradation data sequence of the typical operating condition, split the degradation data sequence into several datasets of performance indicators changing with equivalent running time, and match a mathematical fitting model for each dataset. Input each dataset into the matching mathematical fitting model for fitting to obtain the fitting curve of each dataset. The fitting curves of each dataset constitute the baseline health state degradation curve library of the typical operating condition.

7. The method for early warning of the full life cycle health status of a measurement switch based on digital twin as described in claim 1, characterized in that, The process of matching typical operating conditions consistent with the real-time load data includes: Obtain all load parameters from the real-time load data to obtain a real-time load parameter set; for any typical working condition, obtain all load parameters of the typical working condition to obtain a working condition load parameter set; obtain the similarity between the real-time load parameter set and each working condition load parameter set, select the working condition load parameter set with the highest similarity to the real-time load parameter set, denoted as the best matching set, and denote the typical working condition corresponding to the best matching set as the current working condition.

8. The method for early warning of the full life cycle health status of a measurement switch based on digital twin as described in claim 1, characterized in that, The process of determining the reference expected value set of the target measurement switch includes: For any performance index fitting curve in the baseline health state degradation curve library for the current operating condition, substitute the current equivalent running time into the fitting curve, and obtain the ordinate of the fitting curve when the horizontal axis is the current equivalent running time. This ordinate is denoted as the baseline expected value of the performance index. Thus, the baseline expected values ​​of each performance index are obtained, resulting in a set of baseline expected values.

9. The method for early warning of the full life cycle health status of a measurement switch based on digital twin as described in claim 1, characterized in that, The process of obtaining the deviation between real-time performance parameters and the set of baseline expected values ​​includes: The real-time performance parameters include several real-time performance indices. For any real-time performance index, the benchmark expected value of the corresponding performance index is obtained from the benchmark expected value set, thus obtaining the individual deviation between the real-time performance index and the benchmark expected value. , where R real For real-time performance metrics, R ref The baseline expected value is used as the benchmark. The deviation of each individual real-time performance indicator from its corresponding baseline expected value is obtained, and the average value of all individual deviations is obtained, which is denoted as the deviation between the real-time performance parameter and the set of baseline expected values.

10. The method for early warning of the full life cycle health status of a measurement switch based on digital twin as described in claim 1, characterized in that, The process of determining whether the health status of the target measurement switch is abnormal based on the deviation includes: A deviation threshold is set. If the deviation is greater than or equal to the deviation threshold, it is directly determined that the health status of the target measurement switch is abnormal. If the deviation is less than the deviation threshold, the deviation of the target measurement switch is continuously collected within a preset time window to obtain a deviation sequence. The deviation sequence is fitted using the least squares method to generate a proportional regression line, and the slope of the regression line is obtained. If the slope is greater than a preset slope threshold, the health status of the target measurement switch is abnormal. If the slope is less than or equal to the preset slope threshold, the deviation of the target measurement switch is monitored.

Images