Charging device monitoring operation and maintenance management system and method based on multi-source data

Abstract

The application discloses a charging equipment monitoring operation and maintenance management system and method based on multi-source data, relates to the technical field of equipment monitoring, and comprises a target user determination module, a user classification module, a power consumption stage integration module, a fixed user screening module, a distinguishable feature coefficient analysis module, a charging feature recognition model construction module and an intelligent response adjustment module. The application analyzes the user portrait features of other users without fixed charging habits in different power consumption stages, constructs a recognition model, and maximally evaluates the use of charging equipment by users, so that the charging equipment responds to a reasonable charging time based on the judgment of user data and charging data, reduces the charging pressure in the peak power consumption stage, and improves the accuracy and intelligence of peak-shifting charging adjustment.

Description

Technical Field

[0001] This invention relates to the field of charging monitoring technology, specifically to a charging equipment monitoring and maintenance management system and method based on multi-source data. Background Technology

[0002] As electric bicycles become increasingly prevalent in urban transportation, their charging demand is having a more significant impact on the power grid. The rapid increase in the number of electric bicycles has led to a continuous rise in charging load. Under traditional charging methods, a large number of electric bicycles are charged during the evening hours after get off work, which overlaps with peak residential electricity consumption times, putting enormous pressure on the power grid.

[0003] Traditional electric vehicle charging methods rely heavily on user-managed operations, lacking effective intelligent management and control mechanisms. Users often connect chargers haphazardly upon arriving home, without considering the grid's load. This disorderly charging behavior makes it difficult for the grid to predict and balance electricity demand, further exacerbating power shortages during peak hours.

[0004] To mitigate the impact of electric vehicle charging on the power grid, the industry has experimented with some off-peak charging initiatives. Some regions have introduced time-of-use pricing policies to encourage users to charge during off-peak hours. However, this approach relies heavily on user self-discipline and lacks effective technical means for enforcement and precise guidance. Many users, unaware of the time-of-use pricing policy or constrained by their travel schedules, still choose to charge during peak hours, significantly diminishing the policy's effectiveness. Furthermore, while some existing so-called "smart charging" systems can provide simple timed charging, they lack data analysis based on grid load data, resulting in significant deficiencies in the accuracy and intelligence of off-peak charging. Summary of the Invention

[0005] The purpose of this invention is to provide a charging equipment monitoring and maintenance management system and method based on multi-source data, so as to solve the problems raised in the prior art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a method for monitoring and maintaining charging equipment based on multi-source data, the method comprising:

[0007] Step S100: Extract all logged-in users who respond to the charging device and have charging redundancy time, and mark them as target users. Charging redundancy time refers to the time interval between when a user starts charging the charging object and when the user uses the charging object, which is greater than the time required for the charging object to complete charging, and the difference is greater than the time threshold. Charging redundancy time is added as an indicator for user screening because users without charging redundancy time cannot achieve device start-stop adjustment in the time dimension during charging events. Based on the charging data of the target users, the target users are classified into non-fixed users and fixed users.

[0008] Step S200: Using one day as a monitoring unit, extract the electricity load data of all users recorded by the charging equipment within several monitoring units, and integrate the different electricity consumption stages within the monitoring units; determine the charging time characteristics of fixed users based on different electricity consumption stages; and screen out key fixed users.

[0009] Step S300: Extract user data of non-fixed users based on electricity consumption stage, analyze the distinguishable feature coefficients of non-fixed users based on electricity consumption stage; set the threshold of distinguishable feature coefficients, and construct a charging feature recognition model for non-fixed users based on the numerical relationship between the distinguishable feature coefficients and the threshold of distinguishable feature coefficients.

[0010] Step S400: When a new charging device response event is added, determine whether the user of the responding device is a key fixed user. If so, make intelligent adjustments based on the real-time charging status of the user; otherwise, make feature recognition and response adjustments based on the charging feature recognition model.

[0011] Furthermore, based on the target users' charging data, the target users are classified into non-fixed users and fixed users, including the following specific processes:

[0012] The charging data records the charging time, extracts the charging time t0 of all target users responding to the charging object and the power consumption time t1 of using the charging object, and constructs a reference time axis with the charging time t0 and power consumption time t1 of any target user.

[0013] Mark the charging data of the same target user under all unit monitoring durations on the reference time axis, and set the allowable error thresholds t0±a1 and t1±a1 on the reference time axis, where a1 is the set allowable time error amount;

[0014] When the charging time of each charging event recorded by the same target user satisfies t0∈[t0-a1,t0+a1] and the power consumption time satisfies t1∈[t1-a1,t1+a1], the target user is marked as a fixed user; otherwise, it is marked as a non-fixed user.

[0015] Furthermore, step S200 includes the following specific steps:

[0016] Within each monitoring unit duration, periods with electricity load data greater than or equal to the peak load threshold are marked as peak periods, periods with load data less than the off-peak load threshold are marked as off-peak periods, and the remaining periods are marked as normal periods; and a horizontal axis for period differentiation is generated with the monitoring unit duration as the axis length;

[0017] Find the horizontal axis that distinguishes the duration of several monitoring units within the monitoring period, and align all the horizontal axes of the time periods with the start and end points of the time.

[0018] The intersection of peak hours on the horizontal axis is defined as the peak electricity consumption phase, the intersection of off-peak hours is defined as the off-peak electricity consumption phase, and the remaining time periods are defined as the normal electricity consumption phase.

[0019] Extract charging data from the historical records of fixed users and match the charging data with the characteristics of the electricity consumption stage at the charging time. When the charging data involves two types of electricity consumption stages and includes the feature that the electricity consumption stage changes from the peak electricity consumption stage to the off-peak electricity consumption stage over time, output the corresponding fixed user as the key fixed user. When the charging data involves a single type of electricity consumption stage, output the corresponding fixed user as the priority fixed user.

[0020] Furthermore, step S300 includes the following specific steps:

[0021] Step S310: Mark the electricity consumption cycle that progresses from the peak electricity consumption stage to other electricity consumption stages over time as the fluctuating electricity consumption stage; extract the total number of non-fixed users D1 recorded in the fluctuating electricity consumption stage within each monitoring unit time period, including the reference time axis; use the formula: G=(1 / m)×∑(D1 / D0) to calculate the distinguishable feature coefficient G corresponding to the fluctuating electricity consumption stage, where D0 represents the total number of target users included in the fluctuating electricity consumption stage, and m represents the total number of records within the monitoring unit time period; similarly calculate the distinguishable feature coefficient G1 corresponding to the peak electricity consumption stage within the same monitoring period;

[0022] Step S320: Set the distinguishable characteristic coefficient threshold G0. If G > G0 and the difference between G and G1 is greater than the difference threshold, mark the start time point t of the peak electricity consumption phase. 起 A charging feature recognition model is constructed for t. 实 ≥t 起 ;t 实 This indicates the moment when the user turns on device charging.

[0023] The above conditions indicate that the charging characteristics of non-fixed users are relatively concentrated during this period of fluctuating electricity consumption. They all show that the charging equipment starts charging during the peak electricity consumption period and only starts charging during the low electricity consumption period. Therefore, using the time relationship as an identification model can quickly and efficiently adjust the response.

[0024] If the condition G>G0 and G-G1 is not met, extract the charging equipment monitoring events recorded by each non-fixed user during the fluctuating power consumption phase and the peak power consumption phase; the charging equipment monitoring events store the initial charging remaining power of the power consumption equipment and the log data of the user equipment; calculate the power profile index U for each non-fixed user's different charging equipment monitoring events recorded during the fluctuating power consumption phase and the peak power consumption phase, U=(W2-W1) / W1, where W1 represents the remaining power in the fluctuating power consumption phase and W2 represents the remaining power in the peak power consumption phase; combine the monitoring events of the same non-fixed user in the two power consumption phases to calculate all power profile indices, calculate the dispersion coefficient of each non-fixed user with respect to the power profile index, iterate through the dispersion coefficients of all non-fixed users to calculate the mean of the dispersion coefficients, and set the dispersion coefficient threshold;

[0025] When the mean of the coefficient of variation is less than the threshold of the coefficient of variation, the remaining power of the device after initial charging is marked as a valid feature; otherwise, it is not marked. The smaller the coefficient of variation, the smaller the degree of dispersion of the value reflected by the power profile index. This shows that when non-fixed users execute charging events during fluctuating and peak power consumption phases, there is a certain correlation between the remaining power of the user's device and the power consumption phase experienced during the final completion of the charging event. Therefore, the remaining power can be used as a basis for judging whether to perform time-adjustable operations for non-fixed users or new users when responding to charging device events.

[0026] Extract log data from the user's device before the response to a charging device monitoring event, based on two types of electricity consumption phases for the same non-fixed user. The log data includes running applications. The log data before the response to a charging device monitoring event during fluctuating electricity consumption phases is marked as the first log, and the log data before the response to a charging device monitoring event during peak electricity consumption phases is marked as the second log.

[0027] The common log is obtained by taking the intersection of the first log and the second log for each non-fixed user record. The running application with the common log removed from the first log is regarded as the state maintenance application, and the running application with the common log removed from the second log is regarded as the fluctuation application.

[0028] Iterate through each non-fixed user to obtain the corresponding state maintenance application and fluctuation application. Take the intersection of the state maintenance applications of all non-fixed users to obtain the first target application, and take the intersection of the fluctuation applications of all non-fixed users to obtain the second target application; construct the charging feature recognition model Y = f1 + f2.

[0029] f1 represents the characteristic value when the real-time remaining power of a user belongs to the power range of the effective characteristic quantity corresponding to the two types of power consumption stages. The power range is composed of the maximum and minimum values ​​of the remaining power recorded in the historical analysis of the two types of power consumption stages; f1 = {1, 0}, f1 = 1 indicates the power range belonging to the fluctuating power consumption stage, and f1 = 0 indicates the power range of the unmarked effective characteristic quantity or the peak power consumption stage.

[0030] f2 represents the number of applications running on the user device that are the same as the first target application before the real-time user record charging device monitoring event response.

[0031] Furthermore, step S400 includes the following steps:

[0032] When the users of the real-time response device are key fixed users and are in a period of fluctuating electricity consumption, the estimated charging time is obtained, and the actual start time of charging is adjusted based on the real-time charging time, with the principle of maximizing coverage of the off-peak or normal period.

[0033] When the user of the real-time response device is a priority fixed user and it is during the peak power consumption period, the device charging will be turned on in real time.

[0034] When the user of the real-time response device is a non-fixed user, and the charging feature recognition model is t 实 ≥t 起 At that time, obtain the response time of the charging device, in t 实 ≥t 起 The actual start time of charging is adjusted according to the principle of maximizing coverage of off-peak or normal periods; if the principle is not met, the device charging is started in real time.

[0035] When the user of the real-time response device is a non-fixed user, and the charging feature recognition model is Y = f1 + f2, the remaining charging power of the non-fixed user and the log data of the user device are extracted and used to calculate the output charging feature recognition value. The charging feature recognition value of each device that is charging in real time is calculated, and the average value is calculated. When the real-time charging feature recognition value is greater than the average value, the actual charging start time is adjusted according to the principle of maximizing coverage of the off-peak or normal stage; otherwise, the device charging is started in real time.

[0036] The charging equipment monitoring and maintenance management system based on multi-source data includes a target user identification module, a user classification module, an electricity consumption stage integration module, a fixed user screening module, a distinguishable feature coefficient analysis module, a charging feature recognition model construction module, and an intelligent response adjustment module.

[0037] The target user identification module is used to extract all logged-in users who respond to the charging device and have redundant charging time, and mark them as target users.

[0038] The user classification module is used to classify target users into non-fixed users and fixed users based on their charging data.

[0039] The electricity consumption phase integration module is used to integrate different electricity consumption phases within a monitoring unit's duration.

[0040] The fixed user screening module is used to determine the charging time characteristics of fixed users based on different electricity consumption stages and to screen out key fixed users.

[0041] The distinguishable characteristic coefficient analysis module is used to analyze the distinguishable characteristic coefficients of non-fixed users based on electricity consumption stages;

[0042] The charging feature recognition model building module is used to build a charging feature recognition model for non-fixed users;

[0043] The intelligent response adjustment module is used to adjust the charging time intelligently when a new charging device responds to an event.

[0044] Furthermore, the distinguishable characteristic coefficient analysis module includes a unit for determining fluctuating electricity consumption stages and a unit for calculating distinguishable characteristic coefficients;

[0045] The fluctuating electricity consumption phase determination unit is used to mark the electricity consumption cycle that progresses from the peak electricity consumption phase to other electricity consumption phases over time as a fluctuating electricity consumption phase;

[0046] The distinguishable characteristic coefficient calculation unit is used to extract the total number of non-fixed users recorded in the fluctuating electricity consumption phase within each monitoring unit time period, including the reference time axis, and calculate the distinguishable characteristic coefficient corresponding to the fluctuating electricity consumption phase.

[0047] Furthermore, the charging feature recognition model construction module includes a first-class recognition model construction unit and a second-class recognition model construction unit;

[0048] A type of identification model building unit is used to build a charging feature identification model based on the starting time point when the distinguishable feature coefficient is greater than the distinguishable feature coefficient threshold and the difference between the distinguishable feature coefficients of the corresponding fluctuating electricity consumption stage and the peak electricity consumption node is greater than the difference threshold.

[0049] The Class II identification model construction unit is used to calculate the power profile indicators of each non-fixed user during the fluctuating power consumption phase and peak power consumption phase, recording different charging equipment monitoring events, when the construction conditions of the Class I identification model are not met. It calculates the dispersion coefficient of each non-fixed user with respect to the power profile indicators, iterates through the dispersion coefficients of all non-fixed users to calculate the mean of the dispersion coefficients, sets the dispersion coefficient threshold, and marks the effective feature quantities based on the relationship between the dispersion coefficients and the threshold; thus constructing the Class II identification model.

[0050] Furthermore, the intelligent response adjustment module includes a real-time user determination unit, a data input unit, and a maximum coverage principle response unit;

[0051] The real-time user determination unit is used to determine the real-time user type;

[0052] The data input unit is used to input real-time acquired user data into the recognition model;

[0053] The maximum coverage principle response unit is used to adjust the actual start time of charging in response to the charging device based on the principle of maximizing coverage during off-peak or normal periods.

[0054] Compared with the prior art, the beneficial effects of the present invention are:

[0055] 1. This invention identifies user groups within a large dataset who can identify patterns in user charging data to implement peak charging adjustments over time by screening target users, thus saving the system the overall computational power required for analyzing numerous and complex user data.

[0056] 2. This invention divides target users according to their charging habits, and prioritizes the power consumption adjustment of users with fixed charging habits during power consumption stages that meet the adjustable charging time, so as to quickly achieve optimized adjustment of power load sharing based on simple analysis algorithms;

[0057] This invention analyzes user profile characteristics of users without fixed charging habits at different electricity consumption stages, constructs an identification model, and evaluates users' usage of charging equipment to the greatest extent. Based on the judgment of user data and charging data, it enables the charging equipment to respond with reasonable charging time, reduces charging pressure during peak electricity consumption stages, and improves the accuracy and intelligence of off-peak charging regulation. Attached Figure Description

[0058] Figure 1 This is a schematic diagram of the charging equipment monitoring and maintenance management method based on multi-source data according to the present invention. Detailed Implementation

[0059] 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.

[0060] Example: Figure 1 As shown, this invention provides a method for monitoring and maintaining charging equipment based on multi-source data. The method includes:

[0061] Step S100: Extract all logged-in users who respond to the charging device and have charging redundancy time, and mark them as target users. Charging redundancy time refers to the time interval between when a user starts charging the charging object and when the user uses the charging object, which is greater than the time required for the charging object to complete charging, and the difference is greater than the time threshold. Charging redundancy time is added as an indicator for user screening because users without charging redundancy time cannot achieve device start-stop adjustment in the time dimension during charging events. Based on the charging data of the target users, the target users are classified into non-fixed users and fixed users.

[0062] Step S200: Using one day as a monitoring unit, extract the electricity load data of all users recorded by the charging equipment within several monitoring units, and integrate the different electricity consumption stages within the monitoring units; determine the charging time characteristics of fixed users based on different electricity consumption stages; and screen out key fixed users.

[0063] Step S300: Extract user data of non-fixed users based on electricity consumption stage, analyze the distinguishable feature coefficients of non-fixed users based on electricity consumption stage; set the threshold of distinguishable feature coefficients, and construct a charging feature recognition model for non-fixed users based on the numerical relationship between the distinguishable feature coefficients and the threshold of distinguishable feature coefficients.

[0064] Step S400: When a new charging device response event is added, determine whether the user of the responding device is a key fixed user. If so, make intelligent adjustments based on the real-time charging status of the user; otherwise, make feature recognition and response adjustments based on the charging feature recognition model.

[0065] Based on the target users' charging data, target users are classified into non-fixed users and fixed users, including the following specific processes:

[0066] The charging data records the charging time, extracts the charging time t0 of all target users responding to the charging object and the power consumption time t1 of using the charging object, and constructs a reference time axis with the charging time t0 and power consumption time t1 of any target user.

[0067] Mark the charging data of the same target user under all unit monitoring durations on the reference time axis, and set the allowable error thresholds t0±a1 and t1±a1 on the reference time axis, where a1 is the set allowable time error amount;

[0068] When the charging time of each charging event recorded by the same target user satisfies t0∈[t0-a1,t0+a1] and the power consumption time satisfies t1∈[t1-a1,t1+a1], the target user is marked as a fixed user; otherwise, it is marked as a non-fixed user.

[0069] Step S200 includes the following specific steps:

[0070] Within each monitoring unit duration, periods with electricity load data greater than or equal to the peak load threshold are marked as peak periods, periods with load data less than the off-peak load threshold are marked as off-peak periods, and the remaining periods are marked as normal periods; and a horizontal axis for period differentiation is generated with the monitoring unit duration as the axis length;

[0071] Find the horizontal axis that distinguishes the duration of several monitoring units within the monitoring period, and align all the horizontal axes of the time periods with the start and end points of the time.

[0072] The intersection of peak hours on the horizontal axis is defined as the peak electricity consumption phase, the intersection of off-peak hours is defined as the off-peak electricity consumption phase, and the remaining time periods are defined as the normal electricity consumption phase.

[0073] Extract charging data from the historical records of fixed users and match the charging data with the characteristics of the electricity consumption stage at the charging time. When the charging data involves two types of electricity consumption stages and includes the feature that the electricity consumption stage changes from the peak electricity consumption stage to the off-peak electricity consumption stage over time, output the corresponding fixed user as the key fixed user. When the charging data involves a single type of electricity consumption stage, output the corresponding fixed user as the priority fixed user.

[0074] Step S300 includes the following specific steps:

[0075] Step S310: Mark the electricity consumption cycle that progresses from the peak electricity consumption stage to other electricity consumption stages over time as the fluctuating electricity consumption stage; extract the total number of non-fixed users D1 recorded in the fluctuating electricity consumption stage within each monitoring unit time period, including the reference time axis; use the formula: G=(1 / m)×∑(D1 / D0) to calculate the distinguishable feature coefficient G corresponding to the fluctuating electricity consumption stage, where D0 represents the total number of target users included in the fluctuating electricity consumption stage, and m represents the total number of records within the monitoring unit time period; similarly calculate the distinguishable feature coefficient G1 corresponding to the peak electricity consumption stage within the same monitoring period;

[0076] Step S320: Set the distinguishable characteristic coefficient threshold G0. If G > G0 and the difference between G and G1 is greater than the difference threshold, mark the start time point t of the peak electricity consumption phase. 起 A charging feature recognition model is constructed for t. 实 ≥t 起 ;t 实 This indicates the moment when the user turns on device charging.

[0077] The above conditions indicate that the charging characteristics of non-fixed users are relatively concentrated during this period of fluctuating electricity consumption. They all show that the charging equipment starts charging during the peak electricity consumption period and only starts charging during the low electricity consumption period. Therefore, using the time relationship as an identification model can quickly and efficiently adjust the response.

[0078] If the condition G>G0 and G-G1 is not met, extract the charging equipment monitoring events recorded by each non-fixed user during the fluctuating power consumption phase and the peak power consumption phase; the charging equipment monitoring events store the initial charging remaining power of the power consumption equipment and the log data of the user equipment; calculate the power profile index U for each non-fixed user's different charging equipment monitoring events recorded during the fluctuating power consumption phase and the peak power consumption phase, U=(W2-W1) / W1, where W1 represents the remaining power in the fluctuating power consumption phase and W2 represents the remaining power in the peak power consumption phase; combine the monitoring events of the same non-fixed user in the two power consumption phases to calculate all power profile indices, calculate the dispersion coefficient of each non-fixed user with respect to the power profile index, iterate through the dispersion coefficients of all non-fixed users to calculate the mean of the dispersion coefficients, and set the dispersion coefficient threshold;

[0079] When the mean of the coefficient of variation is less than the threshold of the coefficient of variation, the remaining power of the device after initial charging is marked as a valid feature; otherwise, it is not marked. The smaller the coefficient of variation, the smaller the degree of dispersion of the value reflected by the power profile index. This shows that when non-fixed users execute charging events during fluctuating and peak power consumption phases, there is a certain correlation between the remaining power of the user's device and the power consumption phase experienced during the final completion of the charging event. Therefore, the remaining power can be used as a basis for judging whether to perform time-adjustable operations for non-fixed users or new users when responding to charging device events.

[0080] Extract log data from the user's device before the response to a charging device monitoring event, based on two types of electricity consumption phases for the same non-fixed user. The log data includes running applications. The log data before the response to a charging device monitoring event during fluctuating electricity consumption phases is marked as the first log, and the log data before the response to a charging device monitoring event during peak electricity consumption phases is marked as the second log.

[0081] The common log is obtained by taking the intersection of the first log and the second log for each non-fixed user record. The running application with the common log removed from the first log is regarded as the state maintenance application, and the running application with the common log removed from the second log is regarded as the fluctuation application.

[0082] Iterate through each non-fixed user to obtain the corresponding state maintenance application and fluctuation application. Take the intersection of the state maintenance applications of all non-fixed users to obtain the first target application, and take the intersection of the fluctuation applications of all non-fixed users to obtain the second target application; construct the charging feature recognition model Y = f1 + f2.

[0083] f1 represents the characteristic value when the real-time remaining power of a user belongs to the power range of the effective characteristic quantity corresponding to the two types of power consumption stages. The power range is composed of the maximum and minimum values ​​of the remaining power recorded in the historical analysis of the two types of power consumption stages; f1 = {1, 0}, f1 = 1 indicates the power range belonging to the fluctuating power consumption stage, and f1 = 0 indicates the power range of the unmarked effective characteristic quantity or the peak power consumption stage.

[0084] f2 represents the number of applications running on the user device that are the same as the first target application before the real-time user record charging device monitoring event response.

[0085] As shown in the example: the application that the user runs before charging is a food delivery application, and the power consumption stage after the user responds to charging is a fluctuating power consumption stage, that is, a transition from peak power consumption stage to off-peak power consumption stage or a transition from peak power consumption stage to normal power consumption stage.

[0086] Furthermore, if the food delivery app is still retained after taking the intersection of all non-fixed user applications, then the food delivery app can be identified as a user profile feature that may record cross-stage electricity consumption by non-fixed users. The more user data that match this user profile feature, the greater the likelihood that the interval between the charging time and the start-up time of the electrical equipment is much longer than the time required to complete charging. This means there will be more room for adjusting the charging time, and the greater the room for adjustment based on the needs of such users' electrical equipment.

[0087] Step S400 includes the following steps:

[0088] When the users of the real-time response device are key fixed users and are in a period of fluctuating electricity consumption, the estimated charging time is obtained, and the actual start time of charging is adjusted based on the real-time charging time, with the principle of maximizing coverage of the off-peak or normal period.

[0089] As shown in the example: if the user responds to the charging device at 19:00, this is during a period of fluctuating power consumption. This period of fluctuating power consumption is based on the user's determination, that is, the user's usual charging period covers both peak and off-peak power consumption periods; and the end point of the user's reference time axis is 9:00; with an error of 1 hour.

[0090] The period from 22:00 to 7:00 the next day is considered a low-peak electricity consumption period. During this time, the estimated charging time is 6 hours. Therefore, the maximum coverage principle means that users who respond to charging at 19:00 can postpone the charging time to the earliest 22:00 and the latest 8:00 the next day, rather than charging in the period from 19:00 to 22:00, which would occupy the peak electricity load.

[0091] When the user of the real-time response device is a priority fixed user and it is during the peak power consumption period, the device charging will be turned on in real time.

[0092] When the user of the real-time response device is a non-fixed user, and the charging feature recognition model is t 实 ≥t 起 At that time, obtain the response time of the charging device, in t 实 ≥t 起 The actual start time of charging is adjusted according to the principle of maximizing coverage of off-peak or normal periods; if the principle is not met, the device charging is started in real time.

[0093] When the user of the real-time response device is a non-fixed user, and the charging feature recognition model is Y = f1 + f2, the remaining charging power of the non-fixed user and the log data of the user device are extracted and used to calculate the output charging feature recognition value. The charging feature recognition value of each device that is charging in real time is calculated, and the average value is calculated. When the real-time charging feature recognition value is greater than the average value, the actual charging start time is adjusted according to the principle of maximizing coverage of the off-peak or normal stage; otherwise, the device charging is started in real time.

[0094] The charging equipment monitoring and maintenance management system based on multi-source data includes a target user identification module, a user classification module, an electricity consumption stage integration module, a fixed user screening module, a distinguishable feature coefficient analysis module, a charging feature recognition model construction module, and an intelligent response adjustment module.

[0095] The target user identification module is used to extract all logged-in users who respond to the charging device and have redundant charging time, and mark them as target users.

[0096] The user classification module is used to classify target users into non-fixed users and fixed users based on their charging data.

[0097] The electricity consumption phase integration module is used to integrate different electricity consumption phases within a monitoring unit's duration.

[0098] The fixed user screening module is used to determine the charging time characteristics of fixed users based on different electricity consumption stages and to screen out key fixed users.

[0099] The distinguishable characteristic coefficient analysis module is used to analyze the distinguishable characteristic coefficients of non-fixed users based on electricity consumption stages;

[0100] The charging feature recognition model building module is used to build a charging feature recognition model for non-fixed users;

[0101] The intelligent response adjustment module is used to adjust the charging time intelligently when a new charging device responds to an event.

[0102] The distinguishable characteristic coefficient analysis module includes a unit for determining fluctuating power consumption phases and a unit for calculating distinguishable characteristic coefficients;

[0103] The fluctuating electricity consumption phase determination unit is used to mark the electricity consumption cycle that progresses from the peak electricity consumption phase to other electricity consumption phases over time as a fluctuating electricity consumption phase;

[0104] The distinguishable characteristic coefficient calculation unit is used to extract the total number of non-fixed users recorded in the fluctuating electricity consumption phase within each monitoring unit time period, including the reference time axis, and calculate the distinguishable characteristic coefficient corresponding to the fluctuating electricity consumption phase.

[0105] The charging feature recognition model construction module includes a first-class recognition model construction unit and a second-class recognition model construction unit;

[0106] A type of identification model building unit is used to build a charging feature identification model based on the starting time point when the distinguishable feature coefficient is greater than the distinguishable feature coefficient threshold and the difference between the distinguishable feature coefficients of the corresponding fluctuating electricity consumption stage and the peak electricity consumption node is greater than the difference threshold.

[0107] The Class II identification model construction unit is used to calculate the power profile indicators of each non-fixed user during the fluctuating power consumption phase and peak power consumption phase, recording different charging equipment monitoring events, when the construction conditions of the Class I identification model are not met. It calculates the dispersion coefficient of each non-fixed user with respect to the power profile indicators, iterates through the dispersion coefficients of all non-fixed users to calculate the mean of the dispersion coefficients, sets the dispersion coefficient threshold, and marks the effective feature quantities based on the relationship between the dispersion coefficients and the threshold; thus constructing the Class II identification model.

[0108] The intelligent response adjustment module includes a real-time user determination unit, a data input unit, and a maximum coverage principle response unit;

[0109] The real-time user determination unit is used to determine the real-time user type;

[0110] The data input unit is used to input real-time acquired user data into the recognition model;

[0111] The maximum coverage principle response unit is used to adjust the actual start time of charging in response to the charging device based on the principle of maximizing coverage during off-peak or normal periods.

[0112] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for monitoring, operation, and maintenance management of charging equipment based on multi-source data, characterized in that: The method includes: Step S100: Extract all logged-in users who respond to the charging device and have a charging redundancy time and mark them as target users. The charging redundancy time refers to the time interval between when a user starts charging the charging object and when the user uses the charging object, which is longer than the time required for the charging object to be fully charged and the difference is greater than a time threshold. Based on the charging data of the target users, classify the target users into non-fixed users and fixed users. The process of classifying target users into non-fixed users and fixed users based on their charging data includes the following specific steps: The charging data records the charging time, extracts the charging time t0 of all target users responding to the charging object and the power consumption time t1 of using the charging object, and constructs a reference time axis with the charging time t0 and power consumption time t1 of any target user. Mark the charging data of the same target user under all unit monitoring durations on the reference time axis, and set the allowable error thresholds t0±a1 and t1±a1 on the reference time axis, where a1 is the set allowable time error amount; When the charging time of each charging event recorded by the same target user satisfies t0∈[t0-a1,t0+a1] and the power consumption time satisfies t1∈[t1-a1,t1+a1], the target user is marked as a fixed user; otherwise, it is marked as a non-fixed user. Step S200: Using one day as a monitoring unit, extract the electricity load data of all users recorded by the charging equipment within several monitoring units, and integrate the different electricity consumption stages within the monitoring units; determine the charging time characteristics of fixed users based on different electricity consumption stages; and screen out key fixed users. Step S200 includes the following specific steps: Within each monitoring unit duration, periods with electricity load data greater than or equal to the peak load threshold are marked as peak periods, periods with load data less than the off-peak load threshold are marked as off-peak periods, and the remaining periods are marked as normal periods; and a horizontal axis for period differentiation is generated with the monitoring unit duration as the axis length; Find the horizontal axis that distinguishes the duration of several monitoring units within the monitoring period, and align all the horizontal axes of the time periods with the start and end points of the time. The intersection of peak hours on the horizontal axis is defined as the peak electricity consumption phase, the intersection of off-peak hours is defined as the off-peak electricity consumption phase, and the remaining time periods are defined as the normal electricity consumption phase. Extract charging data from the historical records of fixed users and match the characteristics of the electricity consumption stage at the charging time. When the charging data involves two types of electricity consumption stages and includes the feature that the electricity consumption stage changes from the peak electricity consumption stage to the off-peak electricity consumption stage over time, output the corresponding fixed user as the key fixed user. When the charging data involves a single type of electricity consumption stage, output the corresponding fixed user as the priority fixed user. Step S300: Extract user data of non-fixed users based on electricity consumption stage, analyze the distinguishable feature coefficients of non-fixed users based on electricity consumption stage; set the threshold of distinguishable feature coefficients, and construct a charging feature recognition model for non-fixed users based on the numerical relationship between the distinguishable feature coefficients and the threshold of distinguishable feature coefficients. Step S300 includes the following specific steps: Step S310: Mark the electricity consumption cycle that progresses from the peak electricity consumption stage to other electricity consumption stages over time as the fluctuating electricity consumption stage; extract the total number of non-fixed users D1 recorded in the fluctuating electricity consumption stage within each monitoring unit time period, including the reference time axis; use the formula: G=(1 / m)×∑(D1 / D0) to calculate the distinguishable feature coefficient G corresponding to the fluctuating electricity consumption stage, where D0 represents the total number of target users included in the fluctuating electricity consumption stage, and m represents the total number of records in the monitoring unit time period within the monitoring cycle; similarly calculate the distinguishable feature coefficient G1 corresponding to the peak electricity consumption stage within the same monitoring cycle; Step S320: Set the distinguishable characteristic coefficient threshold G0. If G > G0 and the difference between G and G1 is greater than the difference threshold, mark the start time point t of the peak electricity consumption phase. 起 A charging feature recognition model is constructed for t. 实 ≥t 起 ;t 实 This indicates the moment when the user turns on device charging. If the condition G>G0 and G-G1 is not met, extract the charging equipment monitoring events recorded by each non-fixed user during the fluctuating power consumption phase and the peak power consumption phase; the charging equipment monitoring events store the initial charging remaining power of the power consumption equipment and the log data of the user equipment; calculate the power profile index U for each non-fixed user's different charging equipment monitoring events recorded during the fluctuating power consumption phase and the peak power consumption phase, U=(W2-W1) / W1, where W1 represents the remaining power in the fluctuating power consumption phase and W2 represents the remaining power in the peak power consumption phase; combine the monitoring events of the same non-fixed user in the two power consumption phases to calculate all power profile indices, calculate the dispersion coefficient of each non-fixed user with respect to the power profile index, iterate through the dispersion coefficients of all non-fixed users to calculate the mean of the dispersion coefficients, and set the dispersion coefficient threshold; When the mean of the coefficients of variation is less than the threshold of the coefficients of variation, the remaining charge of the electrical equipment after initial charging is marked as a valid feature; otherwise, it is not marked. Extract log data from the user's device before the response to a charging device monitoring event, based on two types of electricity consumption phases, for the same non-fixed user. The log data includes running an application that marks the log data before the response to the charging device monitoring event during fluctuating electricity consumption phases as the first log and the log data before the response to the charging device monitoring event during peak electricity consumption phases as the second log. The common log is obtained by taking the intersection of the first log and the second log for each non-fixed user record. The running application with the common log removed from the first log is regarded as the state maintenance application, and the running application with the common log removed from the second log is regarded as the fluctuation application. Iterate through each non-fixed user to obtain the corresponding state maintenance application and fluctuation application. Take the intersection of the state maintenance applications of all non-fixed users to obtain the first target application, and take the intersection of the fluctuation applications of all non-fixed users to obtain the second target application; construct the charging feature recognition model Y=f1+f2. f1 represents the characteristic value when the real-time remaining power of a user belongs to the power range corresponding to the effective characteristic quantity of the two types of power consumption stages. The power range is composed of the maximum and minimum values ​​of the remaining power recorded in the historical analysis of the two types of power consumption stages; f1={1,0}, f1=1 indicates the power range belonging to the fluctuating power consumption stage, and f1=0 indicates the power range that is not marked with effective characteristic quantity or belongs to the peak power consumption stage. f2 represents the number of applications running on the user device that are the same as the first target application before the real-time user record charging device monitoring event response; Step S400: When a new charging device response event is added, determine whether the user of the responding device is a key fixed user. If so, make intelligent adjustments based on the real-time charging status of the user; otherwise, make feature recognition and response adjustments based on the charging feature recognition model. Step S400 includes the following steps: When the users of the real-time response device are key fixed users and are in a period of fluctuating electricity consumption, the estimated charging time is obtained, and the actual start time of charging is adjusted based on the real-time charging time, with the principle of maximizing coverage of the off-peak or normal period. When the user of the real-time response device is a priority fixed user and it is during the peak power consumption period, the device charging will be turned on in real time. When the user of the real-time response device is a non-fixed user, and the charging feature recognition model is t 实 ≥t 起 At that time, obtain the response time of the charging device, in t 实 ≥t 起 The actual start time of charging is adjusted according to the principle of maximizing coverage of off-peak or normal periods; if the principle is not met, the device charging is started in real time. When the user of the real-time response device is a non-fixed user, and the charging feature recognition model is Y=f1+f2, the remaining charging power of the non-fixed user and the log data of the user device are extracted and used to calculate the output charging feature recognition value. The charging feature recognition value of each device that is charging in real time is calculated, and the average value is calculated. When the real-time charging feature recognition value is greater than the average value, the actual charging start time is adjusted according to the principle of maximizing coverage of the off-peak or normal stage; otherwise, the device charging is started in real time.

2. A charging equipment monitoring and maintenance management system based on multi-source data, using the charging equipment monitoring and maintenance management method based on multi-source data as described in claim 1, characterized in that: The system includes a target user identification module, a user classification module, an electricity consumption phase integration module, a fixed user screening module, a distinguishable feature coefficient analysis module, a charging feature recognition model construction module, and an intelligent response adjustment module. The target user determination module is used to extract all logged-in users who respond to the charging device and have charging redundancy time, and mark them as target users. The user classification module is used to classify target users into non-fixed users and fixed users based on the target users' charging data. The electricity consumption phase integration module is used to integrate different electricity consumption phases within a monitoring unit time period; The fixed user screening module is used to determine the charging time characteristics of fixed users based on different electricity consumption stages; And select key, regular users; The distinguishable characteristic coefficient analysis module is used to analyze the distinguishable characteristic coefficients of non-fixed users based on electricity consumption stages; The charging feature recognition model construction module is used to construct a charging feature recognition model for non-fixed users; The intelligent response adjustment module is used to adjust the intelligent response charging time when a new charging device responds to an event.

3. The charging equipment monitoring and maintenance management system based on multi-source data according to claim 2, characterized in that: The distinguishable characteristic coefficient analysis module includes a fluctuating power consumption stage determination unit and a distinguishable characteristic coefficient calculation unit; The fluctuating electricity consumption phase determination unit is used to mark the electricity consumption cycle in which the electricity consumption phase progresses from the peak electricity consumption phase to other electricity consumption phases over time as a fluctuating electricity consumption phase; The distinguishable feature coefficient calculation unit is used to extract the total number of non-fixed users recorded in the fluctuating electricity consumption phase within each monitoring unit time period, including the reference time axis, and calculate the distinguishable feature coefficient corresponding to the fluctuating electricity consumption phase.

4. The charging equipment monitoring and maintenance management system based on multi-source data according to claim 3, characterized in that: The charging feature recognition model construction module includes a first-class recognition model construction unit and a second-class recognition model construction unit; The first type of identification model construction unit is used to construct a charging feature identification model based on the starting time point when the distinguishable feature coefficient is greater than the distinguishable feature coefficient threshold and the difference between the distinguishable feature coefficients of the corresponding fluctuating electricity consumption stage and the peak electricity consumption node is greater than the difference threshold. The second-class identification model construction unit is used to calculate the power profile index of each non-fixed user in the fluctuating power consumption stage and peak power consumption stage when the construction conditions of the first-class identification model are not met. It calculates the discrete coefficient of each non-fixed user with respect to the power profile index, iterates through the discrete coefficients of all non-fixed users to calculate the mean of the discrete coefficients, sets the discrete coefficient threshold, and marks the effective feature quantity based on the relationship between the discrete coefficient and the threshold. Thus, a binary recognition model is constructed.

5. The charging equipment monitoring and maintenance management system based on multi-source data according to claim 4, characterized in that: The intelligent response adjustment module includes a real-time user determination unit, a data input unit, and a maximum coverage principle response unit. The real-time user determination unit is used to determine the real-time user type; The data input unit is used to input real-time acquired user data into the recognition model; The maximum coverage principle response unit is used to adjust the actual start time of charging in response to the charging device based on the principle of maximizing coverage during off-peak or normal periods.

Images