Electric power material quality assurance management method and system

Abstract

The invention relates to the technical field of electric power, in particular to an electric power material quality assurance management method and system, and the method comprises the following steps: collecting real-time operation data, including voltage and current, of various types of electric power equipment, carrying out the real-time monitoring and recording, and generating a real-time performance monitoring record; and based on the real-time performance monitoring record, analyzing performance fluctuation by using a sliding window, and generating an equipment performance stability score. According to the invention, the operation data of the power equipment, such as voltage and current, are collected in real time, and continuous monitoring and recording are implemented, so that the accuracy and instantaneity of the data are improved. The performance fluctuations are analyzed and evaluated using a sliding window, and a device performance stability score is generated, allowing immediate identification of potential faults and performance degradation, thereby improving the efficiency of preventive maintenance and fault response. Meanwhile, by dynamically adjusting the quality guarantee period of the equipment, the management method based on the actual operation condition of the equipment brings higher flexibility and adaptability to equipment management.

Description

Technical Field

[0001] The present invention relates to the field of power technology, and particularly to a method and system for quality assurance management of power materials. Background Art

[0002] The power material quality assurance management method is a method in the power field that focuses on ensuring the quality and reliability of power equipment and components. It aims to ensure that all materials used in the power system, such as transformers, cables, protection equipment, etc., meet quality standards and performance requirements through systematic management measures and technical means. However, the existing technology generally adopts a fixed-cycle maintenance plan, which may lead to over-maintenance or under-maintenance, resulting in waste of resources or an increased risk of equipment failure. For example, overly frequent inspections and maintenance not only increase operating costs but also may affect production efficiency due to unnecessary downtime of equipment. Therefore, improvements are needed. Summary of the Invention

[0003] The purpose of the present invention is to solve the drawbacks existing in the prior art and to propose a power material quality assurance management method and system.

[0004] To achieve the above purpose, the present invention adopts the following technical solutions. The power material quality assurance management method includes the following steps: Collect real-time operation data of various power equipment, including voltage and current, conduct real-time monitoring and recording, and generate a real-time performance monitoring record; based on the real-time performance monitoring record, use a sliding window to analyze performance fluctuations and generate a device performance stability score.

[0005] Based on the device performance stability score, apply a multi-parameter weight analysis method to adjust the influence weight of each performance parameter and generate a weight-adjusted performance score; based on the weight-adjusted performance score, compare past data with threshold settings to verify whether the current operating state of each device is stable and generate a device stability state result.

[0006] According to the device stability state result, dynamically adjust the quality assurance cycle and generate an adjusted quality assurance cycle record; conduct a prediction model analysis on the device to predict the performance stability of the device in the future time period and generate a performance prediction result.

[0007] According to the device prediction result and real-time monitoring data, trigger an emergency quality assurance process for the device that triggers the danger threshold, feedback the quality assurance requirements to the supplier, update the maintenance or replacement plan, and generate an emergency quality assurance trigger record.

[0008] Preferably, the step of obtaining the real-time performance monitoring record is as follows: Real-time collect voltage and current data of power equipment through voltage and current sensors installed on the power equipment, mark each data point with a time stamp, and form a real-time data record of the power equipment; Based on the real-time data record of the power equipment, format and clean the data, calculate the average current and voltage levels of the equipment, and obtain the real-time performance monitoring record.

[0009] Preferably, the steps for obtaining the equipment performance stability score are as follows: Based on the real-time performance monitoring record, read and sort the data at each time point, divide it into time windows of a fixed size in chronological order, and generate partitioned time window records; According to the partitioned time window records, calculate the performance fluctuation value within each time window. The calculation formula is: ; Wherein, is the performance fluctuation value of the th time window, is the performance value at the th time point within the th window, is the performance mean value of the th window, is the number of time points within each window; According to the performance fluctuation value, calculate the equipment performance stability score. The calculation formula is: ; Wherein, is the equipment performance stability score, is the performance fluctuation value of the th time window, is the number of all time windows.

[0010] Preferably, the steps for obtaining the weight-adjusted performance score are as follows: Based on the equipment performance stability score, screen the parameters associated with the equipment performance, including power efficiency, temperature stability, vibration frequency, or start-up time, and generate a performance parameter list; Based on the performance parameter list, calculate the weight-adjusted performance score. The calculation formula is: ;

[0011] Wherein, is the weight-adjusted performance score, is the weight of the th performance parameter, is the equipment performance stability score of the th parameter, is the coefficient for adjusting the initial influence size, is the coefficient for adjusting the growth rate, and is the attenuation coefficient, and sn is the number of performance parameters.

[0012] Preferably, the step of obtaining the result of the device stability state is as follows: Adjust the performance score based on the weight, and at the same time collect the performance data of each device and the preset performance stability threshold to obtain a comparison list of historical data and thresholds; Based on the comparison list of historical data and thresholds, compare the current weight-adjusted performance score of each device with the historical data of the same period, analyze the trend of score changes, and obtain the analysis result of the performance change of each device; According to the analysis result of the performance change of each device, judge whether the current operating state is stable by referring to the preset performance stability threshold. If the score is lower than the preset performance stability threshold, it is determined to be unstable, otherwise it is stable, and a device stability state result is generated.

[0013] Preferably, the step of obtaining the adjusted warranty period record is as follows: Collect the result of the device stability state, classify the state of each device, and mark the devices with stable operation and fluctuating operation to obtain a device state classification record; Based on the device state classification record, calculate the adjusted warranty period. The calculation formula is: ; Wherein, is the adjusted warranty period, is the benchmark warranty period, and are adjustment coefficients, is the device performance stability score, is the device failure rate adjustment coefficient, is the expected failure rate of the device; According to the adjusted warranty period, update the warranty record of the device to obtain the adjusted warranty period record.

[0014] Preferably, the step of obtaining the performance prediction result is as follows: Collect the historical performance data of the device to be predicted, including voltage, current and temperature parameters, to obtain a historical performance data set; Based on the historical performance data set, apply the ARIMA model for trend analysis, set the moving average term number and autoregressive term number of the ARIMA model to obtain the model configuration; According to the model configuration, run the ARIMA model for performance prediction, analyze the trend of device performance stability in the future time period, and generate a performance prediction result.

[0015] Preferably, the step of obtaining the emergency warranty trigger record is as follows: Based on the performance prediction results, identify the devices that exceed the danger threshold to obtain a list of devices triggering emergency quality assurance. Based on the list of devices triggering emergency quality assurance, prepare a quality assurance requirement document, including requirements for device maintenance or replacement, record and send it to the supplier, and obtain the supplier feedback. According to the supplier feedback, update the device maintenance or replacement plan and generate an emergency quality assurance trigger record.

[0016] The present invention provides a management system, including: A data monitoring module that collects real-time operation data of power devices, including voltage and current, and generates real-time monitoring records. A performance analysis module that, based on the real-time monitoring records, applies a sliding window to analyze the performance fluctuations of power devices, obtains the performance stability indicators of the devices, and gets the device stability scores. A weight adjustment module that, based on the device stability scores, adjusts the weights of various parameters affecting the device performance to obtain a weight-adjusted performance score. A status verification module that, using the weight-adjusted performance score, compares the historical performance data of the device with a preset threshold to determine whether the current operating status meets the expectations and generates a device stability status result. A quality assurance management module that, according to the device stability status result, adjusts and records the quality assurance cycle, and at the same time predicts the performance stability of the device in the future time period to generate performance prediction results. Based on the performance prediction results and real-time monitoring records, it triggers an emergency quality assurance process for devices at risk, shares the requirements with the supplier, updates the maintenance or replacement plan, and obtains an emergency quality assurance trigger record.

[0017] Compared with the prior art, the advantages and positive effects of the present invention are as follows: The present invention improves the accuracy and timeliness of data by collecting real-time operation data of power devices, such as voltage and current, and implementing continuous monitoring and recording. Using a sliding window analysis to evaluate performance fluctuations and generate device performance stability scores allows for immediate identification of potential failures and performance degradation, thus improving the efficiency of preventive maintenance and fault response. At the same time, by dynamically adjusting the quality assurance cycle of the device, the management method based on the actual operating conditions of the device brings higher flexibility and adaptability to device management. This method extends the maintenance cycle of stable devices and shortens the cycle of fluctuating devices based on the device stability status result, saving maintenance costs and reducing unnecessary interventions. The application of prediction models such as ARIMA, combined with real-time monitoring data, further optimizes the trigger mechanism of the emergency quality assurance process and improves the reliability of device operation and the supply chain response speed. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 It is a step schematic diagram of the present invention. Specific Embodiments

[0019] In order to make the objectives, technical solutions and advantages of the present invention clearer and more understandable, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0020] Please refer to Figure 1 , the present invention provides a technical solution, a power material quality assurance management method, including the following steps: Collect real-time operation data of various power equipment, including voltage and current, conduct real-time monitoring and recording, and generate real-time performance monitoring records; based on the real-time performance monitoring records, use a sliding window to analyze performance fluctuations and generate equipment performance stability scores.

[0021] Based on the equipment performance stability scores, apply a multi-parameter weight analysis method to adjust the influence weights of each performance parameter and generate weighted-adjusted performance scores; based on the weighted-adjusted performance scores, compare past data with threshold settings to verify whether the current operating state of each device is stable and generate equipment stability state results.

[0022] According to the equipment stability state results, dynamically adjust the quality assurance cycle and generate adjusted quality assurance cycle records; conduct predictive model analysis on the equipment to predict the performance stability of the equipment in the future time period and generate performance prediction results.

[0023] According to the equipment prediction results and real-time monitoring data, trigger an emergency quality assurance process for equipment that triggers a danger threshold, feedback the quality assurance requirements to the supplier, update the maintenance or replacement plan, and generate an emergency quality assurance trigger record.

[0024] The steps for obtaining real-time performance monitoring records are as follows: Real-time collect voltage and current data of power equipment through voltage and current sensors installed on the power equipment, mark each data point with a timestamp, and form real-time data records of the power equipment; Based on the real-time data records of the power equipment, perform data formatting and cleaning, calculate the average current and voltage levels of the equipment, and obtain real-time performance monitoring records.

[0025] Specifically, select multiple installation locations in the power equipment and place voltage and current sensors. Determine the number and layout of these sensor installation points by combining the technical documents provided by the equipment manufacturer and past operation and maintenance experience. Set a sampling frequency of several times per second and assign a unique identification mark to distinguish the data sources of different sensors. Bind the voltage and current values collected each time with an accurate timestamp and record the corresponding readings at that moment. In actual operation, the rated voltage range of the equipment from 0V to 24V and the current range from 0A to 5A can be used as the initial safety range first. This range is given by the manufacturer according to the load capacity and insulation level, and is continuously adjusted to a more suitable level in combination with the fault library information accumulated by the operation and maintenance department over the years. For example, if the equipment has been in the voltage range of 10V to 20V for a long time without any abnormalities, this range can be established as a relatively stable section. Subsequently, when the operation and maintenance personnel find that the new equipment data reaches outside this stable section, they can conduct in-depth analysis. If the sensor readings instantaneously deviate from this safety range, an alarm needs to be issued in combination with the alarm threshold setting of the local monitoring terminal. This threshold is usually calculated by technical personnel based on parameters such as the overload characteristics, voltage withstand range, and ambient temperature of the equipment during the preliminary investigation. When detecting over-limit data, the historical samples in the current acquisition batch of the sensor can be continuously compared to screen for persistent abnormalities or instantaneous interference signals. Arrange the values obtained from this series of samples in chronological order, use a simple mapping table to correspond different sensor identifiers to their respective locations, and form a structured data set. Then, after all the samplings are completed, a unified comparison and verification are carried out. This verification process will eliminate obviously unreliable outliers through cross-reference of the hardware calibration value comparison and the environmental noise test results. If disconnection or sensor abnormalities occur during the acquisition process, mark this item of data and include it in the subsequent comprehensive analysis. Through the above operations, the on-site acquisition content of the equipment, which is sorted by timestamp and covers the voltage and current values at different sampling locations, is finally obtained, forming a real-time data record of the power equipment.

[0026] Based on the real-time data records of power equipment obtained previously, select all valid data points and perform unified formatting operations. Connect the records of each sensor in sequence according to the time stamp. Check the rated current range of the equipment from 0A to 5A and the rated voltage range from 0V to 24V, and eliminate extreme values outside this range. The elimination threshold here is given by the equipment manufacturer during the test phase and corrected by the operation and maintenance personnel based on historical operation data. Therefore, if some extreme values only appear for a very short time and the sensor fault factor cannot be excluded, they will be uniformly marked as abnormal readings and temporarily retained for further verification later. After completing the above formatting steps, use basic arithmetic methods to compare the change range of each sampling point within adjacent two time stamps one by one. For the situation where there are large offsets for consecutive multiple time stamps, it is necessary to verify again whether the sensor status within this interval is normal. If it is determined that the sensor has not failed, it means that such offsets belong to the actual operation fluctuations of the equipment, and mark this fluctuation segment independently in a change interval table. Then, call the sorted normal interval data and calculate the arithmetic means of its voltage and current respectively. If the monitoring results of some equipment show that there are multiple cases exceeding the preliminary set threshold within the specified time period, it is necessary to manually confirm whether the set threshold is reasonable and make appropriate corrections. Here, the threshold is mostly determined by the equipment type and power level, and can be obtained by collecting the operation records of the same type of equipment in the past period of time and conducting statistics. If it is found through statistics that the average voltage of some models is more likely to be in the range of 10V to 12V, the corresponding threshold can be adjusted from 24V to 15V or 16V to more realistically reflect the actual situation. Finally, output the normal data interval after the above cleaning and screening and calculate the average levels of current and voltage to obtain the real-time performance monitoring record.

[0027] The steps to obtain the equipment performance stability score are as follows: Based on the real-time performance monitoring record, read and sort the data at each time point, divide them into time windows of fixed size in chronological order, and generate partitioned time window records; According to the partitioned time window records, calculate the performance fluctuation value within each time window. The calculation formula is: ; Among them, is the performance fluctuation value of the th time window, is the performance value of the th window at the th time point, is the performance mean value of the th window, is the number of time points within each window; According to the performance fluctuation value, calculate the equipment performance stability score. The calculation formula is: ; Among them, is the stability score of the device performance, is the performance fluctuation value of the th time window, and

[0028] is the total number of all time windows.

[0029] Specifically, based on the obtained real-time performance monitoring records, the data at each time point is read and sorted. First, the recording time axis is divided according to the internal sampling frequency of the device and the time points are numbered. Then, combined with the voltage range of 0V to 24V and the current range of 0A to 5A accumulated in the historical monitoring of the device as the basic reference standard, the voltage value and current value of each data point are compared with their respective allowable intervals. For example, if the voltage of some data points exceeds 24V or is lower than 0V, it is marked as a temporary abnormal reading, and it is verified whether the sensor reading is inaccurate or there is a short-term mutation in the subsequent steps. If it is judged as a faulty reading, the data point is excluded. If it is an instantaneous interference, it is retained and compared again in the subsequent data cleaning process. After the above basic comparison, the different time points are connected in sequence to form a continuous time series, and a time window of a fixed size is set for partitioning according to the needs of the actual equipment maintenance department. For example, in some operation and maintenance scenarios, the data monitored every 30 seconds is used as an independent window. This division size is generally determined according to the operating speed of the device and the law of rapid fault evolution, and the specific value is obtained through the actual fault occurrence duration recorded by the operation and maintenance personnel during on-site evaluation and multiple inspection experiments. If it is found through multiple rounds of tests that the device may have a significant voltage fluctuation within 15 seconds, the time window can be set at about 15 seconds. After the time window is divided, the data in each window is renumbered according to the time stamp order and a separate partition record is formed, which stores the voltage data sequence and current data sequence in the window respectively for subsequent detailed comparison and calculation. Finally, after all partitions are numbered and data is sorted, the time window record of the partition is obtained.

[0029] The advantage of the formula is that it quantifies the difference between the performance values and the mean value of each time point within each time window, so as to intuitively present the fluctuation degree within the current window; The acquisition step of is to sum up the performance values of the aforementioned 5 sampling points and then divide by 5. This operation process is based on the actual sampling situation of the device in this window. If some sampling points are marked as abnormal readings, they are not included in the summation, so as to obtain the performance mean value of this window; Calculation process: First, calculate the performance mean value , for example, record the performance values of 5 valid sampling points within this window as 41.0, 40.7, 39.5, 42.2, 40.0 respectively (the unit can be defined according to different monitoring items), then: ; Calculate the value of ; Divide by , and the obtained average value is , Take the square root of this average value, ; This result indicates that the fluctuation value within this time window is approximately 0.9215. If the operation and maintenance department stipulates in the previously set reference range that when exceeds 1.5, it is determined that the fluctuation is too large, then this value has not been reached in this window.

[0030] Formula: In the formula, first assign numerical values to all parameters of the formula and write out the operation process. During the current monitoring period, it is divided into time windows in total. The results obtained previously are 0.9215, 1.4000, and 2.0000 respectively. Specifically, they are obtained by performing the same fluctuation value operation and collation on the data collected within each window. Subsequently, substitute these three fluctuation values into the formula and calculate step by step. First, calculate respectively. For , there is: ; For , there is: ; For , there is: ; Then perform the summation, ; Finally, divide by , ; The result shows that the equipment performance stability score is approximately 0.4234. If the monitoring personnel had previously set the performance stability score threshold at 0.4000, a score lower than 0.4000 would be judged as an unstable state, while a score higher than or equal to 0.4000 would be tentatively considered a stable state. At this time, the score is slightly higher than 0.4000, indicating a relatively average level of stability in the overall evaluation. If more windows are further refined or a longer time period is observed in the future, the final stable state of the equipment can be determined by comparing multiple score results.

[0031] The steps to obtain the weight-adjusted performance score are as follows: Based on the equipment performance stability score, screen the parameters associated with the equipment performance, including power efficiency, temperature stability, vibration frequency, or startup time, to generate a list of performance parameters; Based on the list of performance parameters, calculate the weight-adjusted performance score. The calculation formula is: ; where, is the weight-adjusted performance score, is the weight of the th performance parameter, is the equipment performance stability score of the th parameter, is the coefficient to adjust the initial impact size, is the coefficient to adjust the growth rate, and are the attenuation coefficients, and sn is the number of performance parameters.

[0032] Specifically, based on the equipment performance stability score obtained previously, parameters in multiple dimensions such as power efficiency, temperature stability, vibration frequency and startup time that can reflect the operation status of the equipment are selected as references. Combined with the real-time operation data of the equipment and the performance parameter ranges accumulated in past maintenance records, the current values ​​of these parameters are compared with their corresponding threshold ranges. For example, the power efficiency is usually provided by the equipment manufacturer in a reasonable range (70% to 95%), and the temperature stability needs to be combined with the superposition of the ambient temperature and the internal heat of the equipment to determine the appropriate range (for example, 0°C to 90°C). The vibration frequency can refer to the data previously monitored by the vibration sensor and compared with the existing mechanical vibration safety standards (for example, 0Hz to 50Hz). The startup time can be compared in the records of multiple starts and stops through actual power-on tests, and the time consumed for multiple starts and stops can be compared with the equipment indicators. The normal startup time (for example, 3 to 6 seconds) is checked. If it is extremely short or far beyond the normal startup range, it is marked as abnormal data and handed over to the operation and maintenance personnel for manual identification. When these parameters are included in the evaluation process one by one, they need to be combined with factors such as equipment model and environmental requirements. If the operation and maintenance personnel need to monitor the temperature stability more strictly in a low-temperature environment, a more detailed threshold (for example, -10°C to 90°C) will be added to the temperature dimension. In the specific implementation process, it is necessary to check the original records of various parameters and parse and annotate them. The data points corresponding to each parameter are sorted in order in the time series and compared with the threshold. If a parameter is found to exceed the threshold multiple times, it is marked as having potential risks. After completing the above inspection and recording work, a list of performance parameters including power efficiency, temperature stability, vibration frequency and startup time is obtained.

[0033] The formula is useful in that it combines the performance stability scores with the corresponding weights, initial impact size, growth rate, and attenuation coefficient, thereby reflecting the different degrees of impact of each parameter on the operating status of the equipment in a comprehensive score; The steps for obtaining parameters are as follows: first, according to the design objectives of the equipment and the statistical results of the measured data, combined with the influence of each parameter on the failure rate in the summary of previous maintenance, the weight under the current working condition is defined in numerical form. If the temperature stability has a greater impact on a certain type of equipment, a higher value is assigned. , with a value range of 0.1 to 0.4. The operation and maintenance personnel record and summarize the contribution of each parameter to the failure probability when it deviates during the equipment operation cycle, and select the specific value after multiple rounds of comparison; The steps for obtaining parameters are as follows: use the previously obtained equipment performance stability score to break it down into corresponding parameter dimensions. For example, the temperature stability score obtained by collecting temperature data and calculating before is 0.82, which is calculated by comparing whether the equipment temperature parameter remains fluctuating smoothly within the range of 0°C to 90°C; The steps for obtaining parameters are as follows: according to the nominal values of the equipment before it is put into operation and the initial influence intensity obtained from the actual tests of similar equipment, the operation and maintenance personnel record the influence degree of each parameter on the initial failure rate during 300 hours of continuous operation monitoring, form a comparison table of the initial influence magnitude, and then obtain the value through weighted average; The steps for obtaining parameters are as follows: it is necessary to evaluate the growth rate of the influence of the performance parameter on the overall operation of the equipment when it gradually increases within its normal range. The operation and maintenance personnel extract multiple operation curves with different durations and different failure rates from the previous monitoring data, statistically calculate the change rates of each index in segments and summarize them into a growth rate table, and then perform linear fitting based on the growth rate table to give the quantification range (such as 1.1 to 1.3), and finally select it in combination with the specific equipment model; and The steps for obtaining parameters are as follows: for the possible attenuation effects of each parameter, usually run the same model equipment in the laboratory environment, record the attenuation degree of each parameter on the performance score under extreme environment or extreme load, collect and compare information such as output power, maximum temperature, vibration peak value, etc. during the operation of the equipment, and finally quantify these attenuation laws into coefficients and value range (for example is between 0.5 and 1.5, is between 0.2 and 0.8); Calculation process: Statistical quantity of performance parameters , for example, currently there are 4 performance parameters: power efficiency, temperature stability, vibration frequency, and startup time, and their weights , , , are obtained, and these values come from multiple rounds of failure statistics and expert evaluations; Obtain the , , , and values of each parameter. For example, the stability score of power efficiency, the temperature stability score of temperature stability, the vibration frequency score , startup time score , determined in combination with the foregoing method , , , etc. for specific values; Calculate item by item and divide by , and then multiply by the corresponding , and then add up all the results to obtain the final value; For example, if the calculation result is 0.1497, and if the operation and maintenance department defined in the previous monitoring that when is lower than 0.5, the equipment needs to be further checked, then when the comprehensive score is 0.49 or 0.48, it indicates a certain risk. When it exceeds 0.8, it is relatively stable. If it is between 0.5 and 0.8, it means that multiple parameters are within the normal range but some dimensions are slightly fluctuating. The comprehensive score within different threshold ranges helps with subsequent equipment management and maintenance determination.

[0034] The steps to obtain the equipment stability status result are as follows: Adjust the performance score based on the weight, and at the same time collect the performance data of each device and the preset performance stability threshold to obtain a comparison list of historical data and thresholds; Based on the comparison list of historical data and thresholds, compare the current weight-adjusted performance score of each device with the historical data of the same period, analyze the trend of score changes, and obtain the performance change analysis result of each device; According to the performance change analysis result of each device, judge whether the current operating state is stable by comparing with the preset performance stability threshold. If the score is lower than the preset performance stability threshold, it is determined to be unstable, otherwise it is stable, and the equipment stability status result is generated.

[0035] Specifically, based on the performance scores adjusted by the weights obtained previously, it is necessary to obtain the operation index statistics of each device in different time periods and collect the preset performance stability thresholds corresponding to these devices. First, determine the allowable range of key indicators with reference to the device technical specifications. For example, the temperature range is from 0°C to 90°C, the pressure range is from 0 MPa to 2 MPa, the current range is from 0 A to 5 A, and the voltage range is from 0 V to 24 V. During the device operation and maintenance process, further combine the results of the failure rate investigation to continuously correct the specific upper and lower limits. Match the historical data of these indicators item by item with the corresponding timestamps and compare them with the threshold ranges. If some data fall on the interval edge or exceed this interval multiple times, it may indicate related problems. It is possible to further verify these data to check whether there are abnormal readings from the sensors. If it is confirmed that the sensors are normal after manual or automatic verification, the records that exceed or are lower than the threshold can be marked as potential risks. At the same time, during this process, it is also necessary to collect the operation history of the same batch of devices in the same or similar environments to check whether there are significant differences between devices. For example, whether the temperature increase trends of the same model devices are consistent under the same load conditions. If individual devices deviate significantly from most devices, they can be marked in the summary table and key attention can be paid in subsequent links. Through these operations, finally, integrate the numerical changes of each device in the past multiple operation cycles and the comparison with the thresholds to form a list of historical data and threshold comparisons for this device.

[0036] Based on the previously obtained list of historical data and threshold comparisons, it is necessary to cross-compare the current weight-adjusted performance scores of each device with the historical data in the same period. First, arrange the scores in the same time period or adjacent time periods to check whether there are continuous increases or decreases in parameters such as power efficiency, temperature stability, current load, and vibration frequency. Attach the corresponding timestamps and notes on the operating environment to each comparison result, such as whether the ambient temperature is 3°C to 5°C higher than usual or other factors that may cause score fluctuations. Subsequently, conduct a retrospective analysis of the score change trend. If the score shows a gradual decrease or a significant increase in fluctuations at consecutive measurement points, it is necessary to verify whether there are problems such as lagging device maintenance or aging of some key components. Such problems can often be confirmed by referring to the data records of previous operation weeks and cooperating with the regular maintenance files. If the score changes relatively smoothly during the continuous monitoring period, it can be temporarily treated as normal. For the convenience of subsequent statistics, the multiple score results of each device in the key sections can be summarized. For example, the scores under high-temperature conditions are in the range of 0.6 to 0.7 or the scores under high-load conditions are in the range of 0.7 to 0.8. Combine the scores in these different scenarios with their corresponding environmental and operating conditions. In this way, the change law of the score with the operating conditions and historical status can be clearly presented. Finally, according to the above comparison and analysis operations, the performance change analysis results of each device can be obtained.

[0037] According to the performance change analysis results of each device obtained previously, it is necessary to judge the current weight-adjusted performance score against the preset performance stability threshold. The specific method is to list this score and its threshold for each device in the evaluation form. If it is found that the score is lower than the threshold range provided by the device design or operation and maintenance department, it will be marked as an unstable state during evaluation. For example, if the threshold of a certain unit is specified to be between 0.50 and 1.00, and its recent scores are repeatedly lower than 0.50, subsequent handling will be immediately triggered. If the score remains at 0.60 or higher, it will be classified as a stable state. To avoid missed or misjudgments, a critical range (such as 0.45 to 0.50) can be set around the threshold. For devices within the critical range, the monitoring frequency will be increased during subsequent operation periods or short-term inspections will be arranged. If it is still found that the score is lower than the threshold multiple times during the inspection process, it will be clearly marked as unstable. This mark will be associated with the device number and timestamp and saved. Through this set of screening logics, the current stability identifier of each device can be viewed in the dispatching system, and subsequent investigations or maintenance can be carried out for unstable devices. Through the above steps, the device stability status results are generated.

[0038] The steps for obtaining the adjusted warranty period record are as follows: Collect the device stability status results, classify the status of each device, mark the devices with stable operation and fluctuating operation, and obtain the device status classification record; Based on the device status classification record, calculate the adjusted warranty period. The calculation formula is: ; Wherein, is the adjusted warranty period, is the benchmark warranty period, and are adjustment coefficients, is the device performance stability score, is the device failure rate adjustment coefficient, is the device expected failure rate; According to the adjusted warranty period, update the device warranty record to obtain the adjusted warranty period record.

[0039] Specifically, collect the stability status results of each device and classify them. First, combine all the obtained stability status identifiers with the corresponding historical maintenance list of the device, and extract the operation records of each device in the previous stage, including whether the performance stability score has been lower than a specific threshold (such as in the range of 0.50 to 0.60) multiple times, whether the temperature has exceeded 90°C multiple times, whether the current has exceeded 5A multiple times, etc. Through these classification marks, identify the devices that conform to the two types of stable operation and fluctuating operation. When grouping devices of the same type, it is necessary to refer to information such as device model and power level to ensure that devices of the same model or the same power level are clustered together for easy comparison. If some of these devices indicate in the fault record that they have triggered overload alarms multiple times or have had an abnormally long start-up time (such as exceeding 7 seconds of the predetermined standard), special markings can be made in this record, and more intensive inspections or detections can be carried out when necessary. If the threshold established by the operation and maintenance department requires further investigation when the current fluctuates at 5A and the voltage fluctuates at 24V simultaneously, such situations can be listed separately, and the same batch of devices can be compared through the centralized query function. If different devices show significant differences or frequent fluctuations under the same load conditions, record these devices as in the fluctuating operation state and add them to the corresponding group. After completing the classification of stable and fluctuating, update the latest status of each device in the form of a timestamp and a reason annotation, and classify and summarize to obtain the device status classification record.

[0040] The advantage of the formula lies in integrating the performance stability score, failure rate factor, and benchmark warranty period of the device during operation, enabling the warranty period to be closer to the real-time usage status of the device; The steps for obtaining the parameters are as follows: Before the device is put into use, the manufacturer and the operation and maintenance party jointly give them based on the factory test of the device and the test results under standard working conditions, and combine the historical life statistics of the same type of device to form a basic duration (such as 12 months or 24 months); and The steps for obtaining the parameters are as follows: The device monitoring personnel conduct centralized tests on the operation characteristics of the device under different load environments, different temperature ranges, and different vibration intensities, and set the adjustment range ( between 0.1 and 0.5) and the exponential decay rate ( between 0.2 and 0.8) by collecting a series of operation and maintenance data and combining the actual loss status of the device under high load conditions. After multiple rounds of comparison tests, record the optimal values as the corresponding and ; The obtaining steps are as follows: extract the final score value of the device (such as 0.75 or 0.80) from the device performance stability score obtained previously. This score is obtained through the calculation of the data fluctuation degree within the time window and weight adjustment in the early stage; The obtaining steps of the parameter are as follows: observe the failure frequency of the device under different working conditions such as normal load and extreme load during continuous operation monitoring, compare with the average failure rate of similar devices in the industrial field and conduct comprehensive summary in a weighted manner. Therefore It may be in the range of 0.1 to 0.3, and the specific value is obtained through multiple evaluations and statistics; The obtaining steps of the parameter are as follows: calculate the historical failure rate of the main vulnerable components during the operation of the device, extract the damage situation of the relevant components from the fault location data after the formal operation and calculate the time-average failure rate, compare this value with the existing industrial data, and then make corrections to obtain the expected failure rate of the device (such as 0.02 or 0.03); Calculation process: Determine the values of each parameter, such as months, , , , , , First calculate , Calculate , Sum , Multiply by , that is (unit: number of months); This result indicates that the calculated adjusted warranty period is about 14.55 months. If the monitoring personnel predefine that it needs to be rechecked when it exceeds 15 months, then the boundary situation of this value and 15 months can be compared. If it is found in subsequent operation observation that the device performance gradually improves (causing to increase or to decrease), then it may be further extended during the recalculation. If there is a significant increase, then it will be correspondingly shortened; ​According to the adjusted warranty period obtained previously, for each device, continue to track the operation data in the new cycle and summarize important indicators such as current and voltage. Then, compare the information collected in real-time with the previously determined failure rate reference range (for example, the number of failures per quarter shall not exceed 2 times). If it is found that the number of device failures significantly exceeds this reference range within a certain statistical cycle, record the device in the operation and maintenance system and check whether its specific failure rate has increased (such as rising from 0.03 to 0.05). If there is an increase, recalculate the warranty period and immediately notify the relevant department to intervene in the maintenance. After completing this series of query and comparison operations, write the adjusted new warranty period into the device file and form an update identifier. For those devices that have not fluctuated significantly during this detection cycle, retain the previous cycle information and append the current date stamp to obtain the adjusted warranty period record.

[0041] The steps to obtain the performance prediction results are as follows: Collect the historical performance data of the devices that need to be predicted, including voltage, current, and temperature parameters, to obtain the historical performance data set; Based on the historical performance data set, apply the ARIMA model for trend analysis, set the moving average term number and autoregressive term number of the ARIMA model, and obtain the model configuration; According to the model configuration, run the ARIMA model for performance prediction, analyze the device performance stability trend in the future time period, and generate the performance prediction results.

[0042] Specifically, when collecting the historical performance data of the devices that need to be predicted, it is necessary to first confirm the collection range and collection frequency, and then find the voltage, current, and temperature values of the device in the past several cycles (such as consecutive 180 days) from the previous operation and maintenance records, and associate each record with the time stamp. If the operation and maintenance department sets a safety range of 0V to 24V for voltage, a rated range of 0A to 5A for current, and a working boundary of 0°C to 90°C for temperature, it is necessary to compare whether each record exceeds the corresponding range. If it is found that some voltage points are higher than 24V or the temperature data exceeds 90°C, mark them as potential anomalies and confirm these marks later. For example, call the dedicated fault record information to check whether a fault alarm was triggered at that time. If it is confirmed that there is no sensor error, retain the abnormal reading for analyzing its impact on subsequent predictions. After completing the comparison of the overall data and forming a continuous time series, it is necessary to additionally identify the sub-intervals of the device under different working conditions (such as high load or high temperature environment) and record their numerical distributions. This sub-interval information can provide a basis for the differential analysis of the prediction model. For example, if the voltage fluctuation characteristics of the device in the high temperature interval are significantly different from those in the normal temperature interval, the interval can be modeled separately or processed by adding feature quantities when establishing the model later. After these operations are completed, the historical performance data set can be obtained.

[0043] Based on the historical performance dataset obtained previously, it is necessary to apply the ARIMA model for trend analysis. First, select the time series segments of the device in the standard load environment from the dataset and ensure that the sampling frequency is sufficiently uniform, such as once per minute or once every ten seconds. Subsequently, arrange these time series data in chronological order and construct training samples. Then, it is necessary to set the number of moving average terms and the number of autoregressive terms of the ARIMA model. The specific orders can be selected based on the observations of the ACF (Autocorrelation Function) and PACF (Partial Autocorrelation Function) graphs by the device operation and maintenance department in previous experiments. For example, it may be found through observation that the number of autoregressive terms is 2 and the number of moving average terms is 1, which can better fit the fluctuation trend of the device. At the same time, seasonal or periodic factors need to be considered and additional processing is required if they exist. For the situation where the temperature reading rises rapidly and falls slowly, the temperature can be regarded as a key feature and the voltage and current can be used as parallel features and input into the time series model together during modeling. In addition, if there are many device types or a large sample span, grouped modeling needs to be carried out, and each group of devices is matched with a similar model configuration according to its power level and thermal balance characteristics. After finally completing these settings, the configuration items of the ARIMA model are obtained.

[0044] According to the model configuration obtained previously, read the historical time series data of the device point by point during the training process and correspond the three parameters of voltage, current, and temperature with the corresponding timestamps. During the training phase, it is necessary to fit the ARIMA parameters batch by batch with the data of past periods and compare the difference between the prediction and the actual value after each batch. If the difference exceeds the allowable range specified in advance by the operation and maintenance department (for example, the prediction deviation is greater than ±5% or the temperature prediction deviation from the actual value exceeds 3°C), then the model order or other fine-tuning items are corrected after the end of this iteration and the next batch of training continues. After multiple rounds of iteration, if the prediction deviation can be maintained within the allowable range in most periods, it indicates that the model is basically applicable. Then, when entering the actual prediction phase, the recently obtained voltage and current information can be input into the model in the same order, and the model automatically recursively predicts the operating trends of temperature and other key parameters in the future time period. If it is monitored in the results that the device voltage or temperature has a continuous tendency to develop towards the boundary limit, a prompt can be added in the operation and maintenance system. Through this prediction operation, comprehensively analyze the performance stability trend of the device in the future time period and output the corresponding data sequence, and finally generate the performance prediction result.

[0045] The steps to obtain the emergency quality assurance trigger record are as follows: Based on the performance prediction results, identify the devices that exceed the danger threshold to obtain the list of devices triggering emergency quality assurance; Based on the list of devices triggering emergency quality assurance, prepare the quality assurance requirement document, including the requirements for device maintenance or replacement, record and send it to the supplier, and obtain the supplier feedback; Based on supplier feedback, update the equipment maintenance or replacement plan and generate emergency warranty trigger records.

[0046] Specifically, based on the performance prediction results obtained above, it is necessary to compare the operating data of all equipment in the future period and identify equipment with dangerous conditions. The monitoring data should be compared with the dangerous thresholds set in advance by the operation and maintenance department one by one. For example, for equipment with a temperature range of 0°C to 90°C, 90°C can be used as the dangerous threshold, and for equipment with a current in the range of 0A to 5A, 5A can be used as the corresponding dangerous threshold. If the current exceeds 5A or the temperature exceeds 90°C for many times during monitoring or prediction, the equipment is judged to be at high risk. Then, the alarm records in the operation and maintenance experience database in recent months are used to check whether an alarm has occurred. If there are multiple situations where edge alarms are triggered, these devices are listed as those that may fail or have a failure rate that increases too quickly. Then, based on the specific device model, power level, operating environment and other details, it is confirmed whether they will be added to the danger list. For example, if the current of the same model of equipment generally increases to around 4.7A after the load increases by 10% but does not break the 5A threshold, it will not be included for the time being. If individual devices continue to fluctuate sharply to 5.2A in a short period of time, it will be immediately determined to be out of limit. All devices that meet or exceed the danger threshold are summarized and fixedly marked in subsequent steps to form an ordered list, and a list of devices that trigger emergency warranty is obtained.

[0047] Based on the list of equipment that triggers emergency warranty obtained above, it is necessary to prepare a warranty demand document and search for the main risk reasons of each equipment. For example, the number of failures, startup time, temperature abnormality points, etc. in the past 30 days may be queried to list them as actual needs for maintenance or replacement. If the startup time of a certain equipment is continuously higher than the predetermined standard set by the operation and maintenance department (for example, 7 seconds) and the temperature exceeds 90°C several times, it will be included in the replacement plan. If other equipment only has a short-term current overload, it will be included in the maintenance plan. These needs are uniformly recorded and accompanied by detailed information such as equipment identification number, model, and operating time. After completing the list, it is sent to the corresponding manufacturing supplier and wait for its supply feedback on replacement parts or maintenance materials. If it is found in the supplier's feedback that there are no standing spare parts for certain models and they need to be ordered in advance or the fault point needs to be further confirmed, continue to communicate with the supplier in the next stage. If the supplier is sure that it can deliver the goods in time or provide a spare model, it can be confirmed in this link. Finally, a new round of maintenance and replacement schedule is formed based on all feedback information, and feedback from the supplier is obtained.

[0048] According to the supplier feedback obtained previously, it is necessary to update the maintenance or replacement plan for the corresponding equipment. For example, the equipment with faulty parts that can be delivered within a week will be scheduled first and included in the maintenance list for next week. At the same time, the equipment that needs to borrow other models of accessories temporarily will be sorted. Combining the failure rate registered in the operation and maintenance system with high-risk indicators such as temperature and current, if the high-risk indicators are prominent, it will be ranked higher for priority handling. If the supplier reports a shortage of accessories of the same model and it will take at least two weeks to deliver the goods, then record this message and evaluate whether the current equipment can operate at a low load within two weeks to avoid secondary failures. Then, unify and organize all the equipment maintenance and replacement plans and combine them with the equipment numbers and the estimated maintenance periods. During this process, if it is confirmed that some equipment needs to be replaced, it is necessary to explain the matching of the replacement time point with the rhythm of the factory production line. After completing the above operations, an emergency quality assurance trigger record can be generated.

[0049] The above are only the preferred embodiments of the present invention, and do not limit the present invention in other forms. Any person skilled in the art may use the technical content disclosed above to make changes or modifications into equivalent embodiments with equivalent changes and apply them to other fields. However, as long as it does not depart from the technical solution content of the present invention, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention still fall within the protection scope of the technical solution of the present invention.

Claims

1. A method for quality assurance management of electric power materials, characterized in that: The following steps are involved: Collect real-time operating data of various power equipment, including voltage and current, conduct real-time monitoring and recording, and generate real-time performance monitoring records; Based on the real-time performance monitoring records, a sliding window is used to analyze performance fluctuations and generate a device performance stability score; Based on the equipment performance stability score, a multi-parameter weight analysis method is applied to adjust the influence weight of each performance parameter to generate a weight-adjusted performance score; Adjusting the performance score based on the weight, comparing past data and threshold settings, verifying whether the current operating state of each device is stable, and generating a device stability state result; According to the equipment stability status result, dynamically adjust the warranty period and generate an adjusted warranty period record; Perform predictive model analysis on the equipment, predict the performance stability of the equipment in the future time period, and generate performance prediction results; Based on the equipment prediction results and real-time monitoring data, the emergency warranty process is triggered for the equipment that triggers the danger threshold, the warranty requirements are fed back to the supplier, the maintenance or replacement plan is updated, and an emergency warranty trigger record is generated.

2. The power material quality assurance management method according to claim 1, characterized in that: The steps for obtaining the real-time performance monitoring record are: The voltage and current sensors installed on the power equipment collect the voltage and current data of the power equipment in real time, and timestamp each data point to form a real-time data record of the power equipment; Based on the real-time data record of the power equipment, the data is formatted and cleaned, the average current and voltage levels of the equipment are calculated, and the real-time performance monitoring record is obtained.

3. The power material quality assurance management method according to claim 1, characterized in that: The steps for obtaining the equipment performance stability score are as follows: Based on the real-time performance monitoring records, the data at each time point is read and sorted, divided into time windows of fixed size in chronological order, and a partitioned time window record is generated; According to the time window records of the partition, the performance fluctuation value in each time window is calculated, and the calculation formula is: ； in, For the The performance fluctuation value of a time window, For the In the window The performance value at a time point, For the The performance average of the windows is is the number of time points within each window; According to the performance fluctuation value, the equipment performance stability score is calculated using the following formula: ； in, Score the device performance stability, For the The performance fluctuation value of a time window, is the number of all time windows.

4. The power material quality assurance management method according to claim 1, characterized in that: The steps for obtaining the weight adjustment performance score are: Based on the device performance stability score, filter parameters associated with device performance, including power efficiency, temperature stability, vibration frequency or startup time, and generate a performance parameter list; Based on the performance parameter list, the weighted adjusted performance score is calculated using the following formula: ； in, Adjust the performance score for the weights, For the The weight of each performance parameter, For the The device performance stability score of each parameter, To adjust the coefficient for the initial effect size, To adjust the growth rate coefficient, and is the attenuation coefficient, and sn is the number of performance parameters.

5. The power material quality assurance management method according to claim 1, characterized in that: The steps for obtaining the device stability status result are: Adjust the performance score based on the weight, collect the performance data of each device and the preset performance stability threshold, and obtain a historical data and threshold comparison list; Based on the historical data and the threshold comparison list, the current weight-adjusted performance score of each device is compared with the historical data of the same period, and the score change trend is analyzed to obtain the performance change analysis result of each device; According to the performance change analysis results of each device, the preset performance stability threshold is compared to determine whether the current operating state is stable. If the score is lower than the preset performance stability threshold, it is determined to be unstable, otherwise it is stable, and the device stability state result is generated.

6. The power material quality assurance management method according to claim 1, characterized in that: The steps for obtaining the adjusted warranty period record are as follows: Collect the equipment stability status results, classify the status of each device, mark the devices with stable operation and those with fluctuating operation, and obtain a classification record of the equipment status; Based on the equipment status classification record, the adjusted warranty period is calculated using the following formula: ； in, For the adjusted warranty period, As the benchmark warranty period, and is the adjustment factor, Score the device performance stability, is the equipment failure rate adjustment coefficient, is the expected failure rate of the equipment; According to the adjusted warranty period, the warranty record of the equipment is updated to obtain the adjusted warranty period record.

7. The power material quality assurance management method according to claim 1, characterized in that: The steps for obtaining the performance prediction results are: Collect historical performance data of the equipment to be predicted, including voltage, current and temperature parameters, to obtain a historical performance data set; Based on the historical performance data set, an ARIMA model is applied to perform trend analysis, and the number of moving average items and autoregressive items of the ARIMA model are set to obtain a model configuration; According to the model configuration, the ARIMA model is run to perform performance prediction, analyze the equipment performance stability trend in the future time period, and generate performance prediction results.

8. The power material quality assurance management method according to claim 1, characterized in that: The steps for obtaining the emergency warranty trigger record are: Based on the performance prediction results, identify the equipment that exceeds the danger threshold and obtain a list of equipment that triggers emergency warranty; Based on the list of equipment that triggers emergency warranty, prepare warranty requirement documents, including equipment maintenance or replacement requirements, record and send to suppliers, and obtain feedback from suppliers; Based on the supplier feedback, update the equipment maintenance or replacement plan and generate an emergency warranty trigger record.

9. The management system of the power material quality assurance management method according to any one of claims 1 to 8, characterized in that: include: Data monitoring module, which collects power equipment operation data in real time, including voltage and current, and generates real-time monitoring records; The performance analysis module uses a sliding window to analyze the performance fluctuations of power equipment based on real-time monitoring records, obtains the performance stability index of the equipment, and obtains the equipment stability score; The weight adjustment module adjusts the weights of various parameters affecting the performance of the device based on the device stability score to obtain a weight-adjusted performance score; The status verification module uses weights to adjust the performance score, compares the historical performance data of the device with the preset threshold, determines whether the current operating status meets expectations, and generates the device stability status result; The warranty management module adjusts and records the warranty period based on the equipment stability status results, predicts the performance stability of the equipment in the future time period, generates performance prediction results, triggers emergency warranty processes for equipment at risk based on the performance prediction results and real-time monitoring records, shares requirements with suppliers, updates maintenance or replacement plans, and obtains emergency warranty trigger records.

Images