Method and system for evaluating usability of power conversion by rider

Abstract

The invention discloses a method and a system for evaluating the power conversion usability of a rider, relates to the field of power conversion usability evaluation, and solves the problems that the evolution of power conversion behaviors of the rider is difficult to dynamically adapt and cross-time dynamic behavior analysis and adaptability evaluation are difficult to realize due to lack of customization capability for a personalized evaluation system of the rider. The technical problem of insufficient insight in the space-time relevance of the behavior characteristics of the rider is solved. The method comprises the following steps: acquiring reference data of a rider in target time; at the end of the current credit management time period; determining an active time period and an inactive time period of the rider in the next credit management time period based on the battery replacement time; determining a credit factor I of the rider in an active period and a credit factor II of the rider in an inactive period based on the reference data, and sending a credit report of the current rider to a manager based on the credit factor I and the credit factor II; setting a battery replacement priority for the rider based on the credit factor I and the credit factor II; according to the invention, the accuracy of power conversion telecommunication evaluation can be improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of battery replacement credit evaluation, in particular to a rider battery replacement credit evaluation method and system. BACKGROUND

[0002] The rider battery replacement credit evaluation system is to solve the problems that may occur in the process of replacing the battery by the rider of an electric bicycle, an electric motorcycle and the like, such as battery abuse, improper charging, theft and the like, so as to ensure the safe use of the battery and the service quality of the rider. With the rapid development of instant delivery industries such as take-out and express delivery, electric bicycles and electric motorcycles have become the main delivery tools for riders. As a key component of these tools, the use and maintenance of the battery are directly related to the work efficiency and delivery safety of the rider.

[0003] At present, most of the rider battery replacement credit evaluation methods and systems lack the customization capability of the personalized evaluation system for the rider in the battery replacement credit evaluation, which leads to the difficulty in dynamically adapting to the evolution of the rider's battery replacement behavior, so that the evaluation system cannot be changed from "post-punishment" to "pre-prevention", and the maintenance cost is increased; at the same time, most of the rider battery replacement credit evaluation methods and systems are difficult to realize dynamic behavior analysis and adaptive evaluation across time periods, which leads to insufficient insight into the spatio-temporal correlation of the rider's behavior characteristics, and further causes inaccurate data analysis.

[0004] Therefore, the present application discloses a rider battery replacement credit evaluation method and system to solve the above technical problems. SUMMARY

[0005] The present application aims to at least solve one of the technical problems existing in the prior art; for this purpose, the present application proposes a rider battery replacement credit evaluation method and system to solve the technical problems that in the existing battery replacement credit evaluation, the customization capability of the personalized evaluation system for the rider is lacking, which leads to the difficulty in dynamically adapting to the evolution of the rider's battery replacement behavior; and it is difficult to realize dynamic behavior analysis and adaptive evaluation across time periods, which leads to insufficient insight into the spatio-temporal correlation of the rider's behavior characteristics.

[0006] To achieve the above-mentioned purpose, the first aspect of the present application provides a rider battery replacement credit evaluation method, comprising: obtaining reference data of the rider within a target time; wherein the target time is obtained according to the rider's historical battery replacement time; the reference data includes the battery replacement time, the battery replacement frequency and the number of abnormal power consumption times; At the end of the credit management period in which the current time is located, the active and inactive periods of the rider in the next credit management period are determined based on the battery swapping time; the credit factor one of the rider in the active period and the credit factor two of the inactive period are determined based on reference data, and the current credit report of the rider is issued to the manager based on the credit factor one and the credit factor two; wherein, the credit management period is determined based on the historical credit factor one and credit factor two. The rider's battery swapping priority is set based on credit factor one and credit factor two.

[0007] Preferably, obtaining the rider's reference data within the target time period includes: The battery swapping time of each battery swapping operation by the rider is obtained through the battery swapping cabinet. The number of battery swaps by the rider within the target time is counted based on the battery swapping time. The number of battery swaps is divided by the number of days in the target time to obtain the battery swapping frequency of the rider on each day within the target time. The time interval between two consecutive battery swaps is determined by the rider's battery swap time within the target time. Battery swaps with time intervals less than a short-time threshold are marked as ultra-short-time battery swaps, and battery swaps with time intervals greater than a long-time threshold are marked as ultra-long-time battery swaps. The number of ultra-short-time and ultra-long-time battery swaps is added together to obtain the number of abnormal power consumptions. The short-time and long-time thresholds are both determined by the battery swap time intervals of other riders.

[0008] Preferably, the target time is obtained based on the rider's historical battery swapping time, including: Extract the rider's battery swap times from historical records, and then retrieve the time intervals between the n most recent battery swaps, arranged chronologically from front to back. The time variation ratio between two adjacent battery swaps is determined based on the calculation formula (1). Obtain several change ratios The average value TZ; where n is obtained manually and the value of n is not less than 20; When the average value TZ is greater than 0, the duration MC of the target time is determined based on the calculation formula (2); when the average value TZ is not greater than 0, the duration MC of the target time is determined based on the calculation formula (3); when the duration MC of the target time is greater than the maximum value of the standard duration range, the maximum value is used to update the duration MC of the target time; when the duration MC of the target time is less than the minimum value of the standard duration range, the minimum value is used to update the duration MC of the target time; the range of the target time is determined according to the duration MC of the target time and the end time of the credit management period in which the current time is located; wherein, the standard duration range is updated regularly according to the number of battery swaps of the rider group, for example, the median and upper and lower limits are recalculated every quarter through cluster analysis, and the upper and lower limits of the standard duration range are inversely proportional to the number of battery swaps of the rider group every quarter; The calculation formula (1) is as follows: ; The calculation formula (2) is as follows: ; The calculation formula (3) is as follows: ; In the formula, and The amplitude adjustment coefficient is determined by the number of abnormal power consumptions by the current rider, in order to optimize the prediction accuracy of the target time (MC). It is directly proportional to the number of abnormal power consumptions by the current rider. It is inversely proportional to the number of abnormal power consumptions by the current rider, and and The values ​​of are all (0,2]; ZC is the median of the standard duration range.

[0009] Preferably, the determination of a rider's active and inactive periods within the next credit management period based on battery swapping time includes: The day is divided into several analysis periods with a fixed duration as one unit. The number of battery swaps for the current rider in each analysis period in history is extracted. The analysis periods with more than one battery swap are marked as active periods, and the analysis periods with less than or equal to one battery swap are marked as inactive periods. The fixed duration is set manually based on experience and can be 30 minutes, 1 hour, or 2 hours. The number of swaps threshold is obtained based on the average number of battery swaps of the rider group, and the number of swaps threshold is proportional to the average number of battery swaps.

[0010] Preferably, determining the rider's credit factor one during active periods and credit factor two during inactive periods based on reference data includes: Extract the battery swapping time, battery swapping frequency, number of abnormal power consumptions, and corresponding credit factor one and credit factor two for each rider within the historical target time period from the historical target data. The historical target data includes the battery swapping time, battery swapping frequency, number of abnormal power consumptions, and credit factor one and credit factor two set by experts based on the battery swapping time, battery swapping frequency, and number of abnormal power consumptions within the target time period for several riders. The system integrates rider battery swapping time, battery swapping frequency, and abnormal power consumption times within a target time period into several sets of training and testing data. The training data is used to train the artificial intelligence model, and the testing data is used to test the trained model. The model is then adjusted based on the testing results. The final result is a credit assessment model with the rider's battery swapping time, battery swapping frequency, and abnormal power consumption times within the target time period as input, and the rider's credit factor one during active periods and credit factor two during inactive periods as output. The artificial intelligence model includes a BP neural network model and an RBF neural network model. Input the rider's battery swapping time, battery swapping frequency, and number of abnormal power consumptions within the target time into the credit assessment model to obtain the rider's credit factor one during active periods and credit factor two during inactive periods.

[0011] Preferably, the step of issuing the current rider's credit report to the manager based on credit factor one and credit factor two includes: When a rider's credit factor 1 is lower than the first percentile of the rider group's credit factor 1, and the rider's credit factor 2 is lower than the second percentile of the rider group's credit factor 2, a report on the rider's low credit is sent to the manager, a text message is sent to the rider reminding them to pay attention to their contract compliance, and the manager is reminded to pay close attention to the rider. When a rider's credit factor 1 is lower than the value of the rider group's credit factor 1 at the first percentile, or when a rider's credit factor 2 is lower than the value of the rider group's credit factor 2 at the second percentile, a report on the rider's low credit is sent to the manager, and a text message is sent to the rider reminding him to pay attention to his contract compliance. When a rider's credit factor 1 is not lower than the value of the rider group's credit factor 1 at the first percentile, and the rider's credit factor 2 is not lower than the value of the rider group's credit factor 2 at the second percentile, a report is sent to the manager that the rider's current credit is good; where the first and second percentiles are both determined through experience.

[0012] Preferably, the credit management period is determined based on historical credit factor one and credit factor two, including: Determine whether the total number of times the current rider has appeared in the history of credit factor one and credit factor two exceeds the corresponding factor number threshold; the factor number threshold is set manually based on experience, for example, the factor number threshold is 30, 50, or 100. Yes, characteristic credit factor one and characteristic credit factor two are obtained from historical credit factor one and credit factor two. Based on the characteristic credit factor one and characteristic credit factor two, the duration of the current rider in the next credit management period is determined by calculation formula (4). If the duration of the next credit management period is greater than the maximum value of the duration range of the standard management period, the duration of the next credit management period is set to the maximum value of the duration range of the standard management period. If the duration of the next credit management period is less than the minimum value of the duration range of the standard management period, the duration of the next credit management period is set to the minimum value of the duration range of the standard management period. Starting from the end time of the credit management period where the current time is located, the next credit management period is planned based on the duration of the next credit management period. The duration range of the standard management period is obtained based on the average daily electricity exchange of the rider group, and the maximum and minimum values ​​of the duration range of the standard management period are inversely proportional to the average daily electricity exchange. No, use the median of the standard management period duration range as the duration of the next credit management period for the current rider. Starting from the end time of the current credit management period, plan the next credit management period based on the duration of the next credit management period. The calculation formula (4) is as follows: ; In the formula, TY1 is characteristic credit factor one, TY2 is characteristic credit factor two; PY1 is the average credit factor one of the rider group, PY2 is the average credit factor two of the rider group; and the median of the duration range of the BC standard management period.

[0013] Preferably, obtaining characteristic credit factor one and characteristic credit factor two from historical credit factor one and credit factor two includes: Credit factor 1 from the past is integrated into credit group 1, and credit factor 2 from the past is integrated into credit group 2. Each credit group is extracted sequentially, and the variance of the credit group is obtained. It is then determined whether the variance is less than the corresponding variance threshold. If yes, the credit factors of the credit group are retained. If no, the credit factor with the largest absolute value of the difference from the average value of the credit factors in the credit group is removed, and the variance is re-evaluated until the variance of the credit group is less than the corresponding variance threshold. In this case, the remaining credit factors of the credit group are retained. The variance threshold is determined empirically. Obtain the maximum, minimum, and average values ​​of the data retained in the extracted credit groups. Calculate the feature values ​​of the credit groups by weighting the maximum, minimum, and average values. Label the feature values ​​of credit group one as feature credit factor one, and label the feature values ​​of credit group two as feature credit factor two.

[0014] Preferably, the step of setting battery swapping priorities for riders based on credit factor one and credit factor two includes: Extract credit factor one and credit factor two for each rider, and determine whether the extracted rider's credit factor one is lower than the value of the rider group's credit factor one at the first percentile; if yes, set the current rider's battery swapping priority to low during the active period in the next credit management period, and prioritize providing the current rider with old batteries during the active period; if no, set the current rider's battery swapping priority to high during the active period in the next credit management period, and prioritize providing the current rider with new batteries during the active period. Determine whether the extracted rider's credit factor 2 is lower than the value of the rider group's credit factor 2 at the second percentile; if yes, set the current rider's battery swap priority to low during the inactive period of the next credit management period, and prioritize providing the current rider with an old battery during the inactive period; if no, set the current rider's battery swap priority to high during the inactive period of the next credit management period, and prioritize providing the current rider with a new battery during the inactive period.

[0015] A second aspect of the present invention provides a rider battery swapping cost evaluation system, comprising: a data collection module, a credit evaluation module connected to the data collection module, and a battery swapping management module connected to the credit evaluation module. The data collection module is used to acquire reference data of the rider within a target time period; wherein, the target time is obtained based on the rider's historical battery swapping time; the reference data includes battery swapping time, battery swapping frequency, and number of abnormal power consumption times. The credit evaluation module is used to determine the rider's active and inactive periods in the next credit management period based on the battery swapping time at the end of the current credit management period; to determine the rider's credit factor one during the active period and credit factor two during the inactive period based on reference data; and to issue a credit report of the current rider to the manager based on credit factor one and credit factor two; wherein, the credit management period is determined based on historical credit factor one and credit factor two. The battery swapping management module is used to set battery swapping priorities for riders based on credit factor one and credit factor two.

[0016] Compared with the prior art, the beneficial effects of the present invention are: 1. This invention addresses the technical problems in existing battery swapping credit evaluation systems. These problems include: acquiring reference data of riders within a target time period; determining the rider's active and inactive periods in the next credit management period based on battery swapping time at the end of the current credit management period; determining the rider's credit factor one during active periods and credit factor two during inactive periods based on the reference data; issuing a credit report to the administrator based on credit factors one and two; and setting battery swapping priorities for riders based on credit factors one and two. This invention improves the accuracy and flexibility of rider battery swapping evaluation.

[0017] 2. This invention, by constructing an intelligent adjustment mechanism for dynamic credit periods, achieves a more precise, personalized, and efficient credit assessment system. It ensures the operational autonomy of high-credit riders while strengthening supervision of low-credit riders, forming an adaptive credit management ecosystem. Its innovation lies in transforming static credit assessment rules into a dynamic response mechanism. Through mathematical modeling, credit behavior characteristics are converted into quantifiable period adjustment parameters, ensuring that credit management maintains both the seriousness of the system and the flexibility to handle complex scenarios. This technological breakthrough provides a replicable solution for building an intelligent, data-driven credit management system, possessing significant practical value and promotional significance. This method utilizes the nonlinear characteristics of exponential functions to achieve gradient processing in credit assessment, allowing riders with different credit levels to receive management cycles matching their creditworthiness. This differentiated management strategy not only enhances the scientific rigor of credit assessment but also effectively reduces system operating costs through the dynamic adjustment mechanism, while simultaneously improving the response speed to abnormal credit behavior, ultimately forming a new paradigm of credit management that balances efficiency and accuracy. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a schematic diagram of the operation steps of the present invention; Figure 2 A schematic diagram illustrating the operation steps for setting battery swapping priorities in this invention; Figure 3 This is a schematic diagram of the system modules of the present invention. Detailed Implementation

[0020] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. 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.

[0021] Please see Figure 1 The first aspect of this invention provides a method for evaluating riders' mobile phone usage fees, including: Obtain reference data for the rider within the target time; the target time is obtained based on the rider's historical battery swapping time; the reference data includes battery swapping time, battery swapping frequency, and number of abnormal power consumptions; At the end of the credit management period in which the current time is located, the active and inactive periods of the rider in the next credit management period are determined based on the battery swapping time; the credit factor one of the rider in the active period and the credit factor two of the inactive period are determined based on reference data, and the current credit report of the rider is issued to the manager based on the credit factor one and the credit factor two; wherein, the credit management period is determined based on the historical credit factor one and credit factor two. The rider's battery swapping priority is set based on credit factor one and credit factor two.

[0022] This application obtains reference data on riders within the target time period, including: The battery swapping time of each battery swapping operation by the rider is obtained through the battery swapping cabinet. The number of battery swaps by the rider within the target time is counted based on the battery swapping time. The number of battery swaps is divided by the number of days in the target time to obtain the rider's battery swapping frequency on each day within the target time. The time interval between two consecutive battery swaps is determined by the rider's battery swap time within the target time. Battery swaps with time intervals less than the short-time threshold are marked as ultra-short-time battery swaps, and battery swaps with time intervals greater than the long-time threshold are marked as ultra-long-time battery swaps. The number of ultra-short-time battery swaps and ultra-long-time battery swaps is added together to obtain the number of abnormal power consumptions. The short-time threshold and the long-time threshold are both determined by the time intervals of battery swaps by other riders.

[0023] It should be noted that both the short-term and long-term thresholds are determined by the time intervals during which other riders swap batteries. Specifically, the short-term threshold is obtained by multiplying the average time intervals during which other riders swap batteries using the current model by a ratio one, and the long-term threshold is obtained by multiplying the average by a ratio two. Ratios one and two are set manually. The value range of ratio one is (0, 0.3], and the value of ratio two is greater than or equal to 2. For example: the average time interval for other riders to swap batteries using model A1 is 12 hours. The value for ratio one is 0.1, and the value for ratio two is 2. Therefore, the short-term threshold is: 12 × 0.1 = 1.2 hours, and the long-term threshold is: 12 × 2 = 24 hours.

[0024] It should be noted that short-term battery swapping may involve order fraud or fake swapping to obtain platform subsidies or points; long-term battery swapping may involve riders not swapping batteries at designated stations but charging them themselves, which poses certain safety risks.

[0025] The target time in this application is obtained based on the rider's historical battery swapping time, including: Extract the rider's battery swap times from historical records, and then retrieve the time intervals between the n most recent battery swaps, arranged chronologically from front to back. The time variation ratio between two adjacent battery swaps is determined based on the calculation formula (1). Obtain several change ratios The average value TZ; where n is obtained manually and the value of n is not less than 20; When the average value TZ is greater than 0, the duration MC of the target time is determined based on the calculation formula (2); when the average value TZ is not greater than 0, the duration MC of the target time is determined based on the calculation formula (3); when the duration MC of the target time is greater than the maximum value of the standard duration range, the maximum value is used to update the duration MC of the target time; when the duration MC of the target time is less than the minimum value of the standard duration range, the minimum value is used to update the duration MC of the target time; the range of the target time is determined according to the duration MC of the target time and the end time of the credit management period in which the current time is located; among which, the standard duration range is updated regularly according to the number of battery swaps of the rider group, for example, the median and upper and lower limits are recalculated by cluster analysis every quarter, and the median and upper and lower limits of the standard duration range are inversely proportional to the number of battery swaps of the rider group every quarter; The calculation formula (1) is as follows: ; The calculation formula (2) is as follows: ; The calculation formula (3) is as follows: ; In the formula, and The amplitude adjustment coefficient is determined by the number of abnormal power consumptions by the current rider, in order to optimize the prediction accuracy of the target time (MC). It is directly proportional to the number of abnormal power consumptions by the current rider. It is inversely proportional to the number of abnormal power consumptions by the current rider, and and The values ​​of are all (0,2]; ZC is the median of the standard duration range.

[0026] It is worth noting that this invention constructs an intelligent collection mechanism based on historical data and real-time behavioral characteristics by dynamically and personally collecting data on riders' battery swapping behavior. This can effectively improve the accuracy of rider credit evaluation and the efficiency of system operation. At the same time, the monitoring time period set based on individual differences of riders can effectively reduce redundant data collection and storage, reduce system storage pressure and energy consumption by shortening unnecessary monitoring cycles, and improve computing efficiency while ensuring the accuracy of data analysis. This invention extracts adjacent intervals from riders' historical battery swapping time series and calculates their change ratios. Combined with dynamic analysis of the arithmetic mean TZ, it comprehensively reflects the stability and regularity of riders' battery swapping behavior, avoiding bias caused by a single indicator, thus providing a scientific basis for setting personalized monitoring times. Secondly, it designs differentiated MC calculation formulas for positive and negative differences in TZ values. By introducing an amplitude adjustment coefficient and dynamically adjusting the value of the amplitude adjustment coefficient in conjunction with the number of abnormal power consumption by riders, it can not only accurately capture the fluctuation characteristics of rider behavior but also effectively balance the stability and sensitivity of prediction results, significantly improving the prediction accuracy of the target time MC. Thirdly, by constraining the MC value within a standard duration range and dynamically determining the monitoring range in conjunction with the end time of the credit management period, it ensures the rationality and comparability of the evaluation results and avoids calculation errors caused by extreme values. At the same time, through periodic updates of the standard duration range, such as recalculating the median and upper and lower limits every quarter, it ensures that the system adapts to the dynamic needs of changes in group behavior.

[0027] It should be noted that the amplitude adjustment coefficient and It is used to adjust several change ratios The influence of the average value TZ on the target time duration MC; when other conditions remain unchanged, or The longer the target time, the greater the impact on MC. or The shorter the target time, the less impact MC will have.

[0028] It should be noted that several change ratios The average value TZ is the percentage change. The arithmetic mean.

[0029] This application determines a rider's active and inactive periods within the next credit management period based on battery swapping time, including: The day is divided into several analysis periods with a fixed duration as one unit. The number of battery swaps for the current rider in each analysis period in history is extracted. The analysis periods with more than one battery swap are marked as active periods, and the analysis periods with less than or equal to one battery swap are marked as inactive periods. The fixed duration is set manually based on experience and can be 30 minutes, 1 hour, or 2 hours. The number of swaps threshold is obtained based on the average number of battery swaps of the rider group, and the number of swaps threshold is proportional to the average number of battery swaps.

[0030] It should be noted that the average number of battery swaps per rider is the total number of battery swaps per rider divided by the number of riders.

[0031] This application determines a rider's credit factor one during active periods and credit factor two during inactive periods based on reference data, including: Extract the battery swapping time, battery swapping frequency, number of abnormal power consumptions, and corresponding credit factor one and credit factor two for each rider within the historical target time period from the historical target data. The historical target data includes the battery swapping time, battery swapping frequency, number of abnormal power consumptions, and credit factor one and credit factor two set by experts based on the battery swapping time, battery swapping frequency, and number of abnormal power consumptions within the target time period for several riders. The system integrates rider battery swapping time, battery swapping frequency, and abnormal power consumption times within a target time period into several sets of training and testing data. The training data is used to train the artificial intelligence model, and the testing data is used to test the trained model. The model is then adjusted based on the testing results. The final result is a credit assessment model with the rider's battery swapping time, battery swapping frequency, and abnormal power consumption times within the target time period as input, and the rider's credit factor one during active periods and credit factor two during inactive periods as output. The artificial intelligence model includes a BP neural network model and an RBF neural network model. Input the rider's battery swapping time, battery swapping frequency, and number of abnormal power consumptions within the target time into the credit assessment model to obtain the rider's credit factor one during active periods and credit factor two during inactive periods.

[0032] It should be noted that Credit Factor 1 is a value used to evaluate a rider's creditworthiness during active periods, while Credit Factor 2 is a value used to evaluate a rider's creditworthiness during inactive periods. Both Credit Factor 1 and Credit Factor 2 are obtained by analyzing the rider's battery swapping time, battery swapping frequency, and number of abnormal power consumptions within the target time period. The closer the recorded number of battery swapping times is to the average number of battery swaps for the rider group within the current target time period, the higher the values ​​of Credit Factor 1 and Credit Factor 2. Similarly, the closer the battery swapping frequency is to the average battery swapping frequency for the rider group within the current target time period, the higher the values ​​of Credit Factor 1 and Credit Factor 2. The values ​​of Credit Factor 1 and Credit Factor 2 are inversely proportional to the number of abnormal power consumptions.

[0033] It should be noted that when analyzing credit factor one and credit factor two in the credit assessment model, a total credit factor is first obtained by analyzing the rider's battery replacement time, battery replacement frequency, and number of abnormal power consumption times within the target time. Then, the total credit factor is multiplied by volatility coefficient one to obtain credit factor one, and the total credit factor is multiplied by volatility coefficient two to obtain credit factor two. Among them, volatility coefficient one is obtained based on the number of abnormal power consumption times of the rider group during historical active periods, and volatility coefficient two is obtained based on the number of abnormal power consumption times of the rider group during historical inactive periods. The values ​​of volatility coefficient one and volatility coefficient two are both (0,1], with volatility coefficient one being greater than volatility coefficient two.

[0034] Specifically, the steps for testing the trained AI model using validation data and adjusting the AI ​​model based on the validation results are as follows: The battery swapping time, battery swapping frequency, and abnormal power consumption counts from the test data are input into the trained AI model to obtain corresponding credit factors one and two. These credit factors one and two are then compared with their counterparts in the test data. If the difference between the corresponding credit factor one and its counterpart is within a certain threshold, and the difference between the corresponding credit factor two and its counterpart is also within a certain threshold (thresholds are empirically determined), no parameter adjustment is needed, and the next set of test data is tested. If the difference is not within the threshold, the corresponding parameters are adjusted until the difference between the corresponding credit factor one and its counterpart is within the threshold, and the difference between the corresponding credit factor two and its counterpart is also within the threshold. Then, the next set of test data is tested. When the number of test data where credit factors one and its counterpart are within the threshold accounts for 90% or more of the total test data, a credit assessment model is obtained, with the input being the rider's battery swapping time, battery swapping frequency, and abnormal power consumption counts within the target time, and the output being the rider's credit factor one during active periods and credit factor two during inactive periods.

[0035] This application issues a credit report to the manager based on credit factor one and credit factor two, including: When a rider's credit factor 1 is lower than the first percentile of the rider group's credit factor 1, and the rider's credit factor 2 is lower than the second percentile of the rider group's credit factor 2, a report on the rider's low credit is sent to the manager, a text message is sent to the rider reminding them to pay attention to their contract compliance, and the manager is reminded to pay close attention to the rider. When a rider's credit factor 1 is lower than the value of the rider group's credit factor 1 at the first percentile, or when a rider's credit factor 2 is lower than the value of the rider group's credit factor 2 at the second percentile, a report on the rider's low credit is sent to the manager, and a text message is sent to the rider reminding him to pay attention to his contract compliance. When a rider's credit factor 1 is not lower than the value of the rider group's credit factor 1 at the first percentile, and the rider's credit factor 2 is not lower than the value of the rider group's credit factor 2 at the second percentile, a report is sent to the manager that the rider's current credit is good; where the first and second percentiles are both determined through experience.

[0036] It should be noted that the values ​​of rider group credit factor one at the first percentile and rider group credit factor two at the second percentile can be specifically explained as follows: If the credit factor of each rider in the rider group is sorted from largest to smallest, and the set percentile is 60%, then the data at the 60th position of the rider group's credit factor is the percentile of the rider group's credit factor. If there is no data at the 60th position of the rider group's credit factor, then the data closest to the 60th position is taken as the percentile of the rider group's credit factor. If the credit factor 2 of each rider in the rider group is sorted from largest to smallest, and the set percentile 2 is 65%, then the data at the 65th position of the rider group's credit factor 2 is the percentile of the rider group's credit factor 2. If there is no data at the 65th position of the rider group's credit factor 2, then the data closest to the 65th position is taken as the percentile of the rider group's credit factor 2.

[0037] The credit management period in this application is determined based on historical credit factor one and credit factor two, including: Determine whether the total number of times the current rider has appeared in the history of credit factor one and credit factor two exceeds the corresponding factor number threshold; the factor number threshold is set manually based on experience, for example, the factor number threshold is 30, 50, or 100. Yes, characteristic credit factor one and characteristic credit factor two are obtained from historical credit factor one and credit factor two. Based on the characteristic credit factor one and characteristic credit factor two, the duration of the current rider in the next credit management period is determined by calculation formula (4). If the duration of the next credit management period is greater than the maximum value of the duration range of the standard management period, the duration of the next credit management period is set to the maximum value of the duration range of the standard management period. If the duration of the next credit management period is less than the minimum value of the duration range of the standard management period, the duration of the next credit management period is set to the minimum value of the duration range of the standard management period. Starting from the end time of the credit management period where the current time is located, the next credit management period is planned based on the duration of the next credit management period. The duration range of the standard management period is obtained based on the average daily electricity exchange of the rider group, and the maximum and minimum values ​​of the duration range of the standard management period are inversely proportional to the average daily electricity exchange. No, use the median of the standard management period duration range as the duration of the next credit management period for the current rider. Starting from the end time of the current credit management period, plan the next credit management period based on the duration of the next credit management period. The calculation formula (4) is as follows: ; In the formula, TY1 is characteristic credit factor one, TY2 is characteristic credit factor two; PY1 is the average credit factor one of the rider group, PY2 is the average credit factor two of the rider group; and the median of the duration range of the BC standard management period.

[0038] It is worth noting that this technical solution, by constructing an intelligent adjustment mechanism for dynamic credit periods, achieves precision, personalization, and efficiency in the credit assessment system. This not only ensures the operational autonomy of high-credit riders but also strengthens the supervision of low-credit riders, forming an adaptive credit management ecosystem. Its innovation lies in transforming static rules of credit assessment into a dynamic response mechanism. Through mathematical modeling, credit behavior characteristics are converted into quantifiable time-period adjustment parameters, ensuring that credit management maintains the seriousness of the system while possessing the flexibility to cope with complex scenarios. This technological breakthrough provides a replicable solution for building an intelligent, data-driven credit management system, with significant practical value and promotional significance. This method achieves gradient processing of credit assessment through the nonlinear characteristics of exponential functions, allowing riders with different credit levels to receive management cycles that match their credit levels. This differentiated management strategy not only improves the scientific nature of credit assessment but also effectively reduces system operating costs through the dynamic adjustment mechanism, while enhancing the response speed to abnormal credit behavior, ultimately forming a new paradigm of credit management that balances efficiency and precision.

[0039] It is worth noting that this invention achieves an intelligent upgrade of the rider credit assessment system by constructing a credit management period mechanism based on dynamic adjustment of historical credit factors. This method introduces a quantitative analysis model of credit factor one and credit factor two, and combines manually set factor number thresholds with group characteristic parameters to form a multi-dimensional assessment system, effectively solving the inherent defects of traditional static management periods in terms of credit assessment accuracy and resource allocation efficiency. The core value of this technical solution lies in establishing an exponential relationship model between characteristic credit factors and group benchmark values, and using the calculation formula (4) to achieve dynamic and elastic adjustment of credit management periods. This nonlinear mapping mechanism based on exponential functions can not only accurately capture the fluctuation characteristics of rider credit behavior, but also construct an adaptive credit assessment framework through intelligent matching with the standard management period duration range. When the rider's historical credit data exceeds the preset threshold, the system extracts characteristic credit factors and performs exponential calculations by combining them with the group average value. The generated dynamic period length breaks through the rigid constraints of traditional fixed periods and ensures system stability through boundary control of maximum and minimum values. This dual-constraint mechanism effectively balances the flexibility and reliability of credit assessment. For riders who did not reach the threshold, the group median was used as the baseline time period, which maintained the fairness of the evaluation system and avoided evaluation bias due to insufficient data.

[0040] This method establishes an inverse correlation between credit management periods and the average daily battery swapping volume of riders, achieving deep coupling between credit assessment and actual operational indicators, ensuring that the period length reflects real operational needs. In practice, the system continuously monitors rider credit behavior. The dynamic period length generated by the index calculation model significantly improves the management efficiency of high-credit riders by extending their credit assessment cycle, reducing data processing frequency and thus reducing computational resource consumption and system load. Simultaneously, it implements more refined monitoring cycle compression for low-credit riders, improving the sensitivity of abnormal behavior identification by shortening the assessment interval. This differentiated management strategy effectively optimizes the overall data analysis performance.

[0041] It should be noted that the average daily swap volume is calculated by dividing the total swap volume of the rider group over a certain number of historical days by the number of days in the statistics, and then dividing by the number of riders in the group. If the number of riders fluctuates, the number of riders in the group is taken as the maximum number of riders over a certain number of historical days.

[0042] This application obtains Feature Credit Factor 1 and Feature Credit Factor 2 from historical Credit Factor 1 and Credit Factor 2, including: Credit factor 1 from the past is integrated into credit group 1, and credit factor 2 from the past is integrated into credit group 2. Each credit group is extracted sequentially, and the variance of the credit group is obtained. It is then determined whether the variance is less than the corresponding variance threshold. If yes, the credit factors of the credit group are retained. If no, the credit factor with the largest absolute value of the difference from the average value of the credit factors in the credit group is removed, and the variance is re-evaluated until the variance of the credit group is less than the corresponding variance threshold. In this case, the remaining credit factors of the credit group are retained. The variance threshold is determined empirically. Obtain the maximum, minimum, and average values ​​of the data retained in the extracted credit groups. Calculate the feature values ​​of the credit groups by weighting the maximum, minimum, and average values. Label the feature values ​​of credit group one as feature credit factor one, and label the feature values ​​of credit group two as feature credit factor two.

[0043] It should be noted that in the process of removing the credit factor with the largest absolute value of the difference from the average value of the credit factors in the credit group and re-performing the variance assessment, the average value of the credit factors is the average value of the credit factors retained in the current variance assessment step.

[0044] It should be noted that, among the credit factors in the credit group with the largest absolute difference from the average credit factor, if there are two credit factors with the largest absolute difference from the average credit factor, the largest credit factor will be removed first.

[0045] It should be noted that if, after removing 90% of the credit factors, the variance of the remaining credit factors is still greater than the temperature variance threshold, then the average value of the credit factors in the credit group will be used as the feature value.

[0046] It should be noted that the weights of the feature values ​​of the credit group obtained by weighting the maximum, minimum and average values ​​are set manually based on experience; for example, the weight of the maximum value can be 0.3, the weight of the minimum value can be 0.2, and the weight of the average value can be 0.5.

[0047] Please see Figure 2 This application sets battery swapping priorities for riders based on credit factor one and credit factor two, including: Extract credit factor one and credit factor two for each rider, and determine whether the extracted rider's credit factor one is lower than the value of the rider group's credit factor one at the first percentile. If yes, set the current rider's battery swap priority to low during the active period in the next credit management period, and prioritize providing the current rider with old batteries during the active period. If no, set the current rider's battery swap priority to high during the active period in the next credit management period, and prioritize providing the current rider with new batteries during the active period. Determine whether the extracted rider's credit factor 2 is lower than the value of the rider group's credit factor 2 at the second percentile; if yes, set the current rider's battery swap priority to low during the inactive period of the next credit management period, and prioritize providing the current rider with old batteries during the inactive period; if no, set the current rider's battery swap priority to high during the inactive period of the next credit management period, and prioritize providing the current rider with new batteries during the inactive period.

[0048] It is worth noting that the dynamic credit rating control mechanism in this invention constructs a multi-dimensional battery resource allocation system by accurately quantifying riders' credit performance and combining it with differentiated service strategies for different time periods. This not only significantly improves the efficiency of battery lifecycle management but also demonstrates systemic value in operational optimization, resource fairness, and user behavior guidance. This mechanism uses a dual-dimensional evaluation model of credit factor one and credit factor two to transform riders' credit levels into quantifiable numerical indicators. It combines group benchmark values ​​with percentiles one and two to achieve accurate identification, avoiding the one-sidedness of single-dimensional evaluation and achieving refined allocation of service resources through time-based differentiated strategies. When a rider's credit factor is below a preset threshold, the system automatically classifies them into a low-priority sequence, prioritizing the provision of older batteries during active periods to reduce resource waste, and further strengthening the constraint mechanism during inactive periods. This ensures basic service needs while creating effective incentives. Meanwhile, high-credit riders receive priority access to new batteries, creating positive feedback and reinforcing the economic rewards of their compliant behavior. This two-way adjustment mechanism effectively balances the fairness and efficiency of resource allocation through data-driven intelligent decision-making. It not only ensures the service experience of high-quality riders, but also guides riders to improve their credit behavior through differentiated strategies, thus forming a virtuous cycle.

[0049] From a battery management perspective, this solution significantly improves battery asset turnover efficiency and reduces equipment wear and replacement costs by extending the lifespan of old batteries and optimizing the deployment schedule of new batteries. Simultaneously, it avoids resource misallocation by dynamically adjusting strategies to adapt to the usage characteristics of different riders. Furthermore, this mechanism implicitly includes a behavioral guidance function, encouraging riders to proactively maintain good credit records through a strong correlation between credit rating and service benefits, thereby improving overall operational order.

[0050] From a system perspective, this design achieves closed-loop management of credit evaluation, resource allocation, and behavioral incentives, providing a replicable intelligent solution for battery management in smart food delivery scenarios. Its value lies not only in short-term operational optimization but also in building a sustainable credit ecosystem, providing a data foundation and implementation paradigm for larger-scale distributed resource scheduling in the future.

[0051] It should be noted that the criteria for distinguishing between new and old batteries are as follows: batteries that have been on the market for no more than 3 months are new batteries, and batteries that have been on the market for more than 3 months are old batteries.

[0052] It should be noted that the specific plan for prioritizing the provision of old batteries to current riders is as follows: when a rider applies for a battery swap at the battery swapping station, the rider will obtain a fully charged battery from the current battery swapping station and then provide the rider with the battery that was first put on the market among the fully charged batteries. The specific plan for prioritizing the provision of new batteries to current riders is as follows: when a rider applies for a battery swap at a battery swapping station, the rider will obtain a fully charged battery from the current battery swapping station, and then provide the rider with the battery that was put on the market most recently among the fully charged batteries.

[0053] Please see Figure 3 A second aspect of the present invention provides a rider battery swapping cost evaluation system, comprising: a data collection module, a credit evaluation module connected to the data collection module, and a battery swapping management module connected to the credit evaluation module. The data collection module is used to acquire reference data of the rider within a target time period; wherein, the target time is obtained based on the rider's historical battery swapping time; the reference data includes battery swapping time, battery swapping frequency, and number of abnormal power consumption times. The credit evaluation module is used to determine the rider's active and inactive periods in the next credit management period based on the battery swapping time at the end of the current credit management period; to determine the rider's credit factor one during the active period and credit factor two during the inactive period based on reference data; and to issue a credit report of the current rider to the manager based on credit factor one and credit factor two; wherein, the credit management period is determined based on historical credit factor one and credit factor two. The battery swapping management module is used to set battery swapping priorities for riders based on credit factor one and credit factor two.

[0054] Some of the data in the above formula are calculated by removing dimensions and taking their numerical values. The formula is the closest to the real situation obtained by software simulation of a large amount of collected data. The preset parameters and preset thresholds in the formula are set by those skilled in the art according to the actual situation or obtained through simulation of a large amount of data.

[0055] Working principle of the invention: This invention first obtains reference data of the rider within the target time period; then, at the end of the credit management period in which the current time is located, it determines the rider's next credit management period based on historical credit factor one and credit factor two, determines the rider's active and inactive periods within the next credit management period based on the battery swapping time, determines the rider's credit factor one during the active period and credit factor two during the inactive period based on the reference data, and issues the current rider's credit report to the manager based on credit factor one and credit factor two; finally, it sets the rider's battery swapping priority based on credit factor one and credit factor two.

[0056] The above embodiments are only used to illustrate the technical methods of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical methods of the present invention without departing from the spirit and scope of the technical methods of the present invention.

Claims

1. A method for evaluating riders' telecommunications usage, characterized in that, include: Obtain reference data for the rider within the target time; the target time is obtained based on the rider's historical battery swapping time; the reference data includes battery swapping time, battery swapping frequency, and number of abnormal power consumptions; At the end of the credit management period in which the current time is located, the active and inactive periods of the rider in the next credit management period are determined based on the battery swapping time; the credit factor one of the rider in the active period and the credit factor two of the inactive period are determined based on reference data, and the current credit report of the rider is issued to the manager based on the credit factor one and the credit factor two; wherein, the credit management period is determined based on the historical credit factor one and credit factor two. The rider's battery swapping priority is set based on credit factor one and credit factor two.

2. The method for evaluating rider's mobile phone usage fees according to claim 1, characterized in that, The acquisition of the rider's reference data within the target time period includes: The battery swapping time of each battery swapping operation by the rider is obtained through the battery swapping cabinet. The number of battery swaps by the rider within the target time is counted based on the battery swapping time. The number of battery swaps is divided by the number of days in the target time to obtain the battery swapping frequency of the rider on each day within the target time. The time interval between two consecutive battery swaps is determined by the rider's battery swap time within the target time. Battery swap behaviors corresponding to time intervals less than a short-time threshold are marked as ultra-short-time battery swap behaviors, and battery swap behaviors corresponding to time intervals greater than a long-time threshold are marked as ultra-long-time battery swap behaviors. The number of ultra-short-time battery swap behaviors and ultra-long-time battery swap behaviors are added together to obtain the number of abnormal power consumption behaviors.

3. The method for evaluating riders' mobile phone usage fees according to claim 1, characterized in that, The target time is obtained based on the rider's historical battery swapping time, including: Extract the rider's battery swap times from historical records, and then retrieve the time intervals between the n most recent battery swaps, arranged chronologically from front to back. The time variation ratio between two adjacent battery swaps is determined based on the calculation formula (1). Obtain several change ratios The average value of TZ; When the average value TZ is greater than 0, the duration MC of the target time is determined based on formula (2); when the average value TZ is not greater than 0, the duration MC of the target time is determined based on formula (3); when the duration MC of the target time is greater than the maximum value of the standard duration range, the maximum value is used to update the duration MC of the target time; when the duration MC of the target time is less than the minimum value of the standard duration range, the minimum value is used to update the duration MC of the target time; the range of the target time is determined according to the duration MC of the target time and the end time of the credit management period in which the current time is located. The calculation formula (1) is as follows: ； The calculation formula (2) is as follows: ； The calculation formula (3) is as follows: ； In the formula, and The amplitude adjustment coefficient is determined by the number of abnormal power consumptions by the current rider, and and The values ​​of are all (0,2]; ZC is the median of the standard duration range.

4. The method for evaluating rider's mobile phone usage fees according to claim 1, characterized in that, The method of determining a rider's active and inactive periods in the next credit management period based on battery swapping time includes: The day is divided into several analysis periods with a fixed duration as a unit of time. The number of battery swaps for the current rider in each analysis period in history is extracted. The analysis periods with more than one battery swap are marked as active periods, and the analysis periods with fewer than one battery swap are marked as inactive periods.

5. The method for evaluating rider's mobile phone usage fees according to claim 1, characterized in that, The process of determining a rider's credit factor one during active periods and credit factor two during inactive periods based on reference data includes: Extract the battery swapping time, battery swapping frequency, number of abnormal power consumptions, and corresponding credit factor one and credit factor two for each rider within the historical target time period from the historical target data. The historical target data includes the battery swapping time, battery swapping frequency, number of abnormal power consumptions, and credit factor one and credit factor two set by experts based on the battery swapping time, battery swapping frequency, and number of abnormal power consumptions within the target time period for several riders. The system integrates rider battery swapping time, battery swapping frequency, and abnormal power consumption times within a target time period into several sets of training and testing data. The training data is used to train the artificial intelligence model, and the testing data is used to test the trained model. The model is then adjusted based on the testing results. The final result is a credit assessment model with the rider's battery swapping time, battery swapping frequency, and abnormal power consumption times within the target time period as input, and the rider's credit factor one during active periods and credit factor two during inactive periods as output. The artificial intelligence model includes a BP neural network model and an RBF neural network model. Input the rider's battery swapping time, battery swapping frequency, and number of abnormal power consumptions within the target time into the credit assessment model to obtain the rider's credit factor one during active periods and credit factor two during inactive periods.

6. The method for evaluating riders' mobile phone usage fees according to claim 1, characterized in that, The process of issuing a credit report to the manager based on credit factor one and credit factor two includes: When a rider's credit factor 1 is lower than the first percentile of the rider group's credit factor 1, and the rider's credit factor 2 is lower than the second percentile of the rider group's credit factor 2, a report on the rider's low credit is sent to the manager, a text message is sent to the rider reminding them to pay attention to their contract compliance, and the manager is reminded to pay close attention to the rider. When a rider's credit factor 1 is lower than the value of the rider group's credit factor 1 at the first percentile, or when a rider's credit factor 2 is lower than the value of the rider group's credit factor 2 at the second percentile, a report on the rider's low credit is sent to the manager, and a text message is sent to the rider reminding him to pay attention to his contract compliance. When a rider's credit factor 1 is not lower than the value of the rider group's credit factor 1 at the first percentile, and the rider's credit factor 2 is not lower than the value of the rider group's credit factor 2 at the second percentile, a report is sent to the manager that the rider's current credit is good.

7. The method for evaluating rider's mobile phone usage fees according to claim 1, characterized in that, The credit management period is determined based on historical credit factor one and credit factor two, including: Determine whether the total number of times the current rider has appeared in the history of credit factor one and credit factor two exceeds the corresponding factor number threshold; Yes, characteristic credit factor one and characteristic credit factor two are obtained from historical credit factor one and credit factor two. Based on the characteristic credit factor one and characteristic credit factor two, the duration of the current rider in the next credit management period is determined by calculation formula (4). If the duration of the next credit management period is greater than the maximum value of the duration range of the standard management period, the duration of the next credit management period is set to the maximum value of the duration range of the standard management period. If the duration of the next credit management period is less than the minimum value of the duration range of the standard management period, the duration of the next credit management period is set to the minimum value of the duration range of the standard management period. Starting from the end time of the credit management period where the current time is located, the next credit management period is planned based on the duration of the next credit management period. No, use the median of the standard management period duration range as the duration of the next credit management period for the current rider. Starting from the end time of the current credit management period, plan the next credit management period based on the duration of the next credit management period. The calculation formula (4) is as follows: ； In the formula, TY1 is characteristic credit factor one, TY2 is characteristic credit factor two; PY1 is the average credit factor one of the rider group, PY2 is the average credit factor two of the rider group; and the median of the duration range of the BC standard management period.

8. The method for evaluating rider's mobile phone usage fees according to claim 7, characterized in that, The step of obtaining characteristic credit factor one and characteristic credit factor two from historical credit factor one and credit factor two includes: Integrate historical credit factor one into credit group one, and integrate historical credit factor two into credit group two; extract each credit group sequentially, obtain the variance of the credit group, and determine whether the variance is less than the corresponding variance threshold; if yes, retain the credit factors of the credit group; if no, remove the credit factor in the credit group with the largest absolute value of the difference from the average value of the credit factors, and re-perform the variance judgment until the variance of the credit group is less than the corresponding variance threshold, then retain the remaining credit factors of the credit group. Obtain the maximum, minimum, and average values ​​of the data retained in the extracted credit groups. Calculate the feature values ​​of the credit groups by weighting the maximum, minimum, and average values. Label the feature values ​​of credit group one as feature credit factor one, and label the feature values ​​of credit group two as feature credit factor two.

9. The method for evaluating riders' mobile phone usage fees according to claim 6, characterized in that, The method of setting battery swapping priorities for riders based on credit factor one and credit factor two includes: Extract credit factor one and credit factor two for each rider, and determine whether the extracted rider's credit factor one is lower than the value of the rider group's credit factor one at the first percentile; if yes, set the current rider's battery swapping priority to low during the active period in the next credit management period, and prioritize providing the current rider with old batteries during the active period; if no, set the current rider's battery swapping priority to high during the active period in the next credit management period, and prioritize providing the current rider with new batteries during the active period. Determine whether the extracted rider's credit factor 2 is lower than the value of the rider group's credit factor 2 at the second percentile; if yes, set the current rider's battery swap priority to low during the inactive period of the next credit management period, and prioritize providing the current rider with an old battery during the inactive period; if no, set the current rider's battery swap priority to high during the inactive period of the next credit management period, and prioritize providing the current rider with a new battery during the inactive period.

10. A rider's mobile phone credit consumption evaluation system, used to run a rider's mobile phone credit consumption evaluation method according to any one of claims 1 to 9, characterized in that, include: The data collection module, the credit evaluation module connected to the data collection module, and the battery swapping management module connected to the credit evaluation module; The data collection module is used to acquire reference data of the rider within a target time period; wherein, the target time is obtained based on the rider's historical battery swapping time; the reference data includes battery swapping time, battery swapping frequency, and number of abnormal power consumption times. The credit evaluation module is used to determine the rider's active and inactive periods in the next credit management period based on the battery swapping time at the end of the current credit management period; to determine the rider's credit factor one during the active period and credit factor two during the inactive period based on reference data; and to issue a credit report of the current rider to the manager based on credit factor one and credit factor two; wherein, the credit management period is determined based on historical credit factor one and credit factor two. The battery swapping management module is used to set battery swapping priorities for riders based on credit factor one and credit factor two.

Images