Method and device for risk early warning of continuous overcharge review of charging facilities based on state transition probability, and medium

Abstract

The application discloses a charging facility continuous over-limit review risk early warning method and device based on state transition probability, medium, and relates to the technical field of charging facility measurement and verification, which comprises the following steps: obtaining the historical measurement records of the charging facility, calculating the measurement error of each record, comparing the error limit value, and forming a binary state sequence arranged in time sequence; setting the size of the sliding window N, and intercepting the latest N records in the binary state sequence as the observation window; counting the state transition events of adjacent records in the sliding window, calculating the transition probability of still being over-limit after over-limit, and calculating the event transition probability P risk from over-limit state to over-limit state twice risk ; comparing the P risk with the preset threshold value to output an early warning signal. The application converts the review condition into a probability problem to realize the calculation of the continuous over-limit review risk index, so that the application changes from post-discovery to pre-warning, and outputs quantitative early warning in the risk accumulation stage without waiting for the complete formation of the continuous over-limit chain.

Description

Technical Field

[0001] This invention relates to the technical field of metering verification of charging facilities, and in particular to a method and device for risk warning of continuous out-of-tolerance verification of charging facilities based on state transition probability. Background Technology

[0002] With the rapid growth of electric vehicle ownership, the accuracy of charging facility metering is subject to increasingly stringent regulations. Even minor deviations in the metering results of charging facilities can accumulate over time, leading to substantial errors in electricity bill settlements and harming the legitimate rights and interests of consumers and operating companies. Therefore, metrology regulatory agencies have gradually established an online remote verification system for charging facilities. Represented by JJG (Anhui) 138-2026, "Online Remote Verification Procedure for Electric Vehicle Charging Facilities," this procedure requires provincial online remote verification platforms to remotely calculate the error for each charging order from each charging facility and compare it with the prescribed error limits. Specifically, JJG stipulates that if more than two consecutive records of exceeding the error limit are found, on-site verification is required.

[0003] Currently, provincial-level online remote verification platforms generally implement the above procedures in the following manner: The platform receives metering data for each charging facility, calculates the metering error for each order, and determines whether it exceeds the prescribed error limits (±1% for first-level charging piles and ±2% for second-level charging piles). When the same charging facility has three or more consecutive records of exceeding the tolerance, the system automatically marks the facility and triggers a review work order, notifying regulatory personnel or operation and maintenance units to conduct on-site verification.

[0004] However, the above mechanism is essentially a post-event discovery mechanism. Its technical defects are: (1) Delayed discovery. The platform only triggers an alarm after the continuous out-of-tolerance chain has been fully formed. This means that before the charging facility is triggered for review, it may have been operating in an out-of-tolerance state for a long time, and a large number of charging orders generated during this period have measurement deviations. Taking a public charging pile with a daily charging volume of several hundred kilowatt-hours as an example, for each day the discovery is delayed, the cumulative power error can reach several kilowatt-hours or even higher. (2) Lack of trend prediction capability. The existing platform only counts and compares the out-of-tolerance results and does not analyze the changing trend of the out-of-tolerance state. When the metering performance of the charging facility is in the transition stage from "occasional out-of-tolerance" to "continuous out-of-tolerance", the platform cannot identify this risk accumulation process. Occasional out-of-tolerance and continuous out-of-tolerance are completely different in terms of risk meaning. The former may be caused by short-term environmental disturbances, while the latter often means that the metering device has undergone irreversible precision degradation. The existing technology cannot distinguish between the two, nor can it issue an early warning before the continuous out-of-tolerance is fully formed. Summary of the Invention

[0005] The purpose of this section is to outline some aspects of embodiments of the present invention and to briefly describe some preferred embodiments. Simplifications or omissions may be made in this section, as well as in the abstract and title of this application, to avoid obscuring the purpose of these documents; however, such simplifications or omissions should not be construed as limiting the scope of the invention.

[0006] To address the technical problems of delayed detection and lack of trend prediction capabilities in the metering error detection of charging facilities in existing technologies, this invention provides a risk early warning method for continuous out-of-tolerance verification of charging facilities based on state transition probability, comprising the following steps: Obtain historical metering records of charging facilities and calculate metering errors for each record; The measurement error of each measurement record is compared with the error limit. If it exceeds the limit, it is judged as an out-of-tolerance state and recorded as 1; otherwise, it is judged as a normal state and recorded as 0, forming a binary state sequence arranged in time sequence. Set the sliding window size N, and select the latest N records from the binary state sequence obtained in step two as the observation window. The observation window is updated in real time as new measurement records are added. Statistically analyze the state transition events of adjacent records within the sliding window, and calculate the transition probability P of remaining out of tolerance after exceeding the tolerance. 11 :P 11 = X UU / (X UN + X UU ); Among them, X UU X represents the number of consecutive deviations. UN This indicates the number of times the error immediately returned to normal. Assuming the latest record within the observation window has been determined to be out of tolerance, and given that three consecutive out-of-tolerance states are expected to occur, the event transition probability from two consecutive out-of-tolerance states to the current out-of-tolerance state is calculated as: P risk =P 11 ×P 11 = P 11 2 ; According to the P risk Compared with a preset threshold, when P risk When the value is greater than or equal to the preset threshold, an early warning signal is output.

[0007] As a preferred embodiment of the charging facility continuous deviation verification risk warning method based on state transition probability described in this invention, the historical metering records of the charging facility are obtained through a cloud platform database.

[0008] As a preferred embodiment of the charging facility continuous deviation verification risk early warning method based on state transition probability described in this invention, the preset threshold includes a first threshold, a second threshold, and a third threshold, wherein the first threshold is less than the second threshold, and the second threshold is less than the third threshold. When P risk When the value is less than the first threshold, it is determined to be an occasional deviation, at which point the risk is low, and monitoring should continue; when the P value is less than the first threshold, it is considered an occasional deviation, at which point monitoring should continue; risk If P is greater than or equal to the first threshold and less than the second threshold, the warning level is determined to be Level 1 warning; when P risk If P is greater than or equal to the second threshold and less than the third threshold, the warning level is determined to be a Level II warning; when P risk If the value is greater than or equal to the third threshold, the warning level is determined to be a Level 3 warning, and it is recommended to review it immediately.

[0009] As a preferred embodiment of the charging facility continuous deviation verification risk early warning method based on state transition probability described in this invention, the method further includes: for each sliding window, calculating P... risk Compared with the preset threshold, record the highest warning level reached by the window, when P risk When any threshold is exceeded, an early warning signal is output, which includes the device identifier, the current risk value, the warning level, and recommended measures.

[0010] As a preferred embodiment of the charging facility continuous deviation verification risk early warning method based on state transition probability described in this invention, the early warning signal includes the device identifier, the current P... risk Values, warning levels, and recommended measures.

[0011] As a preferred embodiment of the charging facility continuous deviation review risk early warning method based on state transition probability described in this invention, the early warning suggestion of the first-level early warning is to pay attention to the risk; the early warning suggestion of the second-level early warning is to arrange inspection; and the early warning suggestion of the third-level early warning is to conduct immediate review.

[0012] As a preferred embodiment of the charging facility continuous deviation review risk warning method based on state transition probability described in this invention, the first threshold is 0.2, the second threshold is 0.3, and the third threshold is 0.4.

[0013] As a preferred embodiment of the charging facility continuous deviation verification risk early warning method based on state transition probability described in this invention, the method further includes: based on the P risk Establish a work order priority processing mechanism, which includes: if the current sliding window has triggered the out-of-tolerance state 3 times consecutively, and P risk If the value is greater than or equal to the preset threshold, the work unit is given the first processing priority; if the current sliding window has not triggered the out-of-tolerance state 3 times consecutively, but Prisk If the value is greater than or equal to the preset threshold, the processing unit will be given the second priority; if the current sliding window has triggered the out-of-tolerance state 3 times consecutively, and P risk If the error is less than the preset threshold, the work unit will be processed at the third priority; if the current sliding window has not triggered the out-of-tolerance state 3 times consecutively, but P risk If the value is less than the preset threshold, the processing unit will be placed in the fourth priority category.

[0014] This invention also discloses a risk warning device for continuous out-of-tolerance verification of charging facilities based on state transition probability, applied to the aforementioned risk warning method for continuous out-of-tolerance verification of charging facilities based on state transition probability. The device includes: The data acquisition module is used to acquire historical metering records of charging facilities and calculate metering errors for each record. The binary state sequence construction module is used to compare the measurement error of each measurement record with the error limit. If the error exceeds the limit, it is judged as an out-of-tolerance state and recorded as 1; otherwise, it is judged as a normal state and recorded as 0, forming a binary state sequence arranged in time sequence. The sliding window construction module is used to set a sliding window of size N, extract the latest N records in the binary state sequence as the observation window, and the observation window is updated in real time as new measurement records are added. The probability calculation module is used to count the state transition events of adjacent records within the sliding window, calculate the transition probability of remaining out of tolerance after exceeding the tolerance, and calculate the event transition probability from two consecutive out-of-tolerance states to the out-of-tolerance state if the latest record in the observation window has been determined to be out of tolerance and three consecutive out-of-tolerance states are to be formed in the future. And an output module, used to compare the event transition probability from two consecutive out-of-tolerance states with a preset threshold, when P risk When the value is greater than or equal to the preset threshold, an early warning signal is output.

[0015] The present invention also discloses a computer-readable storage medium storing a computer program thereon, wherein the computer program, when executed by a processor, implements the steps of the above-described method for risk warning of continuous out-of-tolerance review of charging facilities based on state transition probability.

[0016] The beneficial effects of this invention are: 1. The method of this invention transforms the procedure review conditions into a probability problem to calculate the risk index of continuous deviation review, thereby changing from post-discovery to pre-discovery warning. It does not require waiting for the continuous deviation chain to be fully formed and outputs quantitative warnings during the risk accumulation stage. Furthermore, the warning index is directly related to legal rules, and the warning results have clear legal meaning, which is convenient for regulatory acceptance.

[0017] 2. The method of this invention can effectively distinguish between occasional and persistent deviations. Because occasional deviations occur more frequently, P... 11 Low, P risk Keep low; only when the deviation is persistent, P risk The accuracy is significantly improved, and false alarms are greatly reduced. The method of this invention is simple to calculate and flexible to deploy, requiring only historical error judgment results as input, and can be deployed at the provincial platform or the edge of the charging station. Attached Figure Description

[0018] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments 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. Wherein: Figure 1 This is a schematic diagram of the overall process of the charging facility continuous deviation verification risk early warning method based on state transition probability proposed in this invention; Figure 2 This is a trend diagram of the continuous out-of-tolerance risk transformation of the charging pile in the preferred embodiment of the present invention; Figure 3 This is a schematic diagram of the module of the charging facility continuous deviation verification risk early warning device based on state transition probability proposed in this invention. Detailed Implementation

[0019] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0020] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0021] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0022] Example 1 Reference Figure 1As an embodiment of the present invention, a method for risk early warning of continuous out-of-tolerance review of charging facilities based on state transition probability is provided. Its core concept is to transform the statutory condition in the JJG code, "more than two consecutive out-of-tolerance records triggering review," into a state transition probability problem, and to quantify the risk of triggering review in the near future through this probability value, thereby achieving early warning. This method includes the following steps: Step 1: Obtain the historical metering records of the charging facilities and calculate the metering error line by line according to the error calculation formula of JJG (《Online Remote Verification Procedure for Electric Vehicle Charging Facilities》: JJG (Anhui) 138—2026). Specifically, the historical metering records of the charging facilities are obtained after being uploaded to the cloud platform by the online verification module installed on the charging pile. The cloud platform database saves each order. The error calculation formula of JJG is detailed in the aforementioned publicly available 《Online Remote Verification Procedure for Electric Vehicle Charging Facilities》. Those skilled in the art can implement it based on this procedure, and it will not be elaborated further here.

[0023] Step 2: Compare the measurement error of each measurement record obtained in Step 1 with the error limit specified by JJG. If it exceeds the limit, it is judged as an out-of-tolerance state (recorded as 1); otherwise, it is judged as a normal state (recorded as 0), forming a binary state sequence arranged in time sequence.

[0024] Step 3: Set the sliding window size N, and select the latest N records from the binary state sequence obtained in Step 2 as the observation window. The observation window is updated in real time as new measurement records are added.

[0025] Step 4: Count the state transition events of adjacent records within the sliding window, and calculate the transition probability P of remaining out of tolerance after exceeding the tolerance. 11 :P 11 = X UU / (X UN + X UU ); Among them, X UU This represents the number of out-of-tolerance states, or the number of consecutive out-of-tolerance states, X. UN This represents the number of times the error condition oscillated between out-of-tolerance and normal states, or the number of times the error condition immediately returned to normal. In summary, X... UU and X UN These are counts within a window size N, capped by a fixed window size N. Their values ​​reflect the density and persistence of out-of-range events in the most recent N records, thus allowing the sliding window design to make the risk indicator sensitive only to recent behavior.

[0026] Step 5: Transform the JJG review conditions into a probability problem. Specifically: The JJG procedure stipulates that on-site review is required when there are more than two consecutive out-of-tolerance records. In other words, this procedure is equivalent to the occurrence of three consecutive out-of-tolerance states in a state sequence.

[0027] Therefore, in the method of this invention, the aforementioned statutory procedure is transformed into a probabilistic expression: Given that the latest record within the sliding window has been determined to be out of tolerance, for a future situation to result in three consecutive out-of-tolerance states (i.e., satisfying the review condition), the event transition from two consecutive out-of-tolerance states to an out-of-tolerance state must be satisfied. Therefore, under the first-order Markov assumption, the probability of this event is: P risk =P 11 ×P 11 = P 11 2 ; Therefore, P risk The physical meaning of this indicator is: given that the current deviation is already exceeded, the probability of actually triggering the JJG review conditions in the near future. The value range of this indicator is [0,1], and the larger the value, the higher the risk of triggering the review.

[0028] Step Six: Calculate P in Step Five. risk Compared with the preset threshold T, when P risk >T outputs a warning signal; the warning signal includes the device identifier and the current P risk The threshold value, warning level, and recommended measures are provided for use by regulatory platforms or operations and maintenance personnel. The setting of the threshold T has a clear regulatory mapping logic: due to P... risk The threshold T directly represents the probability of triggering the JJG review conditions, and the threshold T corresponds to the acceptable level of risk of triggering the review.

[0029] In one embodiment, a multi-level threshold strategy can be employed, dividing a preset threshold into a first threshold, a second threshold, and a third threshold, wherein the first threshold is less than the second threshold, and the second threshold is less than the third threshold. When P risk When the value is less than the first threshold, it is judged as an occasional deviation, which has low risk and does not require attention; when P risk If P is greater than or equal to the first threshold and less than the second threshold, it is judged as a Level 1 warning, requiring attention to risk; when P risk If the value is greater than or equal to the second threshold and less than the third threshold, it is judged as a level two warning, and it is recommended to arrange an inspection; when P risk If the value is greater than or equal to the third threshold, it is judged as a level three warning, and it is recommended to review it immediately.

[0030] The first, second, and third thresholds mentioned above can be adjusted based on factors such as regulatory requirements, equipment importance level, and tolerance for false alarm rates. For example, lower thresholds can be used to improve sensitivity for public charging stations in core areas; while thresholds can be appropriately increased in remote areas to reduce false alarms.

[0031] Furthermore, the embodiments also provide a P-based risk Establish a work order priority processing mechanism, which includes: if the current sliding window has triggered the out-of-tolerance state 3 times consecutively, and P risk If the value is greater than or equal to the preset threshold, the processing unit will be given the first processing priority. If the current sliding window does not trigger the out-of-tolerance state 3 times consecutively, but P risk If the value is greater than or equal to the preset threshold, the processing unit will be given the second priority; if the current sliding window has triggered the out-of-tolerance state 3 times consecutively, and P risk If the error is less than the preset threshold, the work unit will be processed at the third priority; if the current sliding window has not triggered the out-of-tolerance state 3 times consecutively, but P risk If the value is less than the preset threshold, the processing unit will be placed in the fourth priority category.

[0032] To further facilitate understanding of the technical solution of the present invention, a preferred embodiment will be described in detail below. This preferred embodiment takes an actual operating charging pile (number 90000241005) connected to a provincial online remote verification platform as an example to illustrate the complete implementation process and technical effects of the present invention.

[0033] I. Data Source and Preprocessing: All charging order records generated by this charging pile between September 3, 2025 and June 17, 2026 were selected. A total of 1,943 original data entries were collected. After cleaning, invalid data was removed, and 1,942 valid records with measurement results of qualified or unqualified were retained.

[0034] According to JJG (Anhui) 138-2026, this charging pile is a level-two charging pile with an error limit of ±2%. For each record, the power error is calculated using the error formula specified in JJG and compared with the 2% limit. The judgment result is binarized: exceeding the limit is recorded as "1" (out of tolerance), and passing is recorded as "0" (normal). Finally, a state sequence of length 1942, arranged in chronological order, is obtained.

[0035] 2. Set the sliding window size N=50; generate the first window starting from the 50th record, and slide it once for each new record, resulting in a total of 1893 valid analysis windows (1942-50+1=1893).

[0036] For each window, count the state transition events of 50 adjacent records within the window (denoted as NN, NU, UN, UU): NN: Normal → Normal state transition event, the number of transitions is denoted as X. NN ; NU: Normal → Out-of-tolerance state transition event, the number of transitions is denoted as X. NU ; UN: Out-of-tolerance → Normal state transition event, the number of transitions is denoted as X. UN ; UU: Out-of-tolerance → Out-of-tolerance state transition event, the number of transitions is denoted as X. UU ; Calculate the probability of exceeding the tolerance: P 11 = X UU / (X UN + X UU ); It should be noted that when X UN + X UU When P = 0, it indicates that no out-of-tolerance events occurred at the charging facility within the observation window—that is, the conditional event "currently in an out-of-tolerance state" did not occur. At this time, there is no statistical basis for "continuing to exceed tolerances after exceeding tolerances," and P = 0. 11 It is directly 0.

[0037] III. Calculation of Risk Indicator for Continuous Exceedance Verification; The JJG procedure requires on-site verification when "more than two consecutive exceedance records" occur, which is equivalent to three consecutive exceedances (UUU) in the state sequence. This invention transforms this legal condition into a probabilistic problem: assuming the current observation window is already in an exceedance state, the formation of a UUU chain requires two more consecutive "U→U" transitions. Under the first-order Markov assumption, the approximate probability of this event is: P risk =P 11 ×P 11 = P 11 2 P here risk This is the continuous excess review risk indicator, which represents the probability of actually triggering the JJG review conditions in the short term, given that the current deviation has already occurred.

[0038] IV. Threshold Setting and Early Warning Output; This embodiment sets three levels of early warning thresholds: a first threshold of 0.2, a second threshold of 0.3, and a third threshold of 0.4, all based on the tolerance for the probability of triggering the review condition; the specific early warning levels are as follows: Level 1 Warning (Attention Level): 0.2≤P risk <0.3, meaning the probability of triggering a review exceeds 20%; Level II Warning (Alert Level): 0.3≤P risk <0.4, meaning the probability of triggering a review exceeds 30%; Level 3 Warning (Severe): P risk≥0.4 means the probability of triggering a review exceeds 40%.

[0039] For each window, calculate P. risk Compared with the above thresholds, record the highest warning level reached by this window (if multiple thresholds are met simultaneously, only the highest level is taken). When P risk When any threshold is exceeded, the system outputs a warning signal that includes the device identifier, the current risk value, the warning level, and recommended measures.

[0040] V. Early Warning Results and Analysis; The analysis results for all 1893 windows are attached. Figure 2 As shown in Table 1, the main statistics are as follows; Table 1 Statistical table of analysis results Total number of valid records 1942 Effective number of analysis windows 1893 <![CDATA[Initial window P 11 > 0.0000 <![CDATA[Initial window P risk > 0.0000 Number of Level 1 Warnings 114 times Number of Level II warnings 124 times Number of Level 3 Warnings 108 times Note: Warning counts are based on mutually exclusive statistics of the highest warning level reached in each window. A Level 1 warning refers to a window that exceeds 0.2 (inclusive) but does not reach 0.3; a Level 2 warning refers to a window that exceeds 0.3 (inclusive) but does not reach 0.4; and a Level 3 warning refers to a window that exceeds 0.4 (inclusive). More precisely, P... risk = 0.4→P 11 = 0.632. The out-of-tolerance behavior is very persistent, and the equipment has most likely already experienced three consecutive out-of-tolerance events—at this point, the warning is only a reactive notification. P risk = 0.2→P 11 =0.447. The out-of-tolerance behavior is just beginning to show, and the equipment is at a critical stage of evolving from "intermittent" to "continuous". At this point, an alert is issued, giving maintenance personnel a time window to respond.

[0041] Through append Figure 2 It can be seen that P throughout the entire monitoring period 11 With P risk The dynamic changes can be clearly divided into two stages.

[0042] Phase 1: Period of Occasional Out-of-Potential Losses (September 2025 to April 2026). During this phase, P... 11 P fluctuates below 0.15 for a long period of time. risk The reading remained consistently below 0.02, far from the 0.2 warning threshold. The system did not issue any warnings. This indicates that while there were occasional single instances of exceeding the tolerance during this period, the proportion of instances where the reading immediately returned to normal (UN) was far higher than the proportion of consecutive exceedances (UU). The exceedances were merely occasional occurrences, and the measurement module had not yet experienced sustained accuracy degradation.

[0043] Phase Two: Period of Sustained Excess (May to June 2026).

[0044] After entering May 2026, P 11A rapid rise and breakthrough of 0.5 indicates that the out-of-range condition has become quite persistent; once an out-of-range condition occurs, the probability of it continuing in the next trade exceeds 50%. P risk The value then rises sharply, frequently crossing the 0.2, 0.3, and even 0.4 threshold lines, triggering level one, two, and three warnings in succession. The dense area of ​​red dots in Figure 2 corresponds to this stage.

[0045] Actual verification: Shortly after the charging pile issued a series of warnings, it generated more than three out-of-tolerance records in succession, which actually triggered the on-site verification conditions stipulated in the JJG regulations, thus confirming the accuracy of the warnings issued by this invention.

[0046] VII. Comparison of effectiveness with traditional post-event review mechanisms. Through retrospective comparison, this method triggered a Level 1 warning (P) for the first time. risk The time point of ≥0.2) was approximately 25 valid orders earlier than the time point when a continuous chain of errors actually formed (i.e., the JJG verification conditions were met). During this period, the charging pile averaged about 15 charges per day, which is equivalent to issuing a risk warning about 1.5 days in advance. This provided maintenance personnel with sufficient time to respond and intervene, allowing for advance scheduling of inspections or calibrations, and preventing the continued expansion of losses due to metering inaccuracies.

[0047] Reference Figure 3 This embodiment provides a charging facility continuous deviation review risk early warning device based on state transition probability. The device includes: a data acquisition module, a binary state sequence construction module, a sliding window construction module, a probability calculation module, and an output module.

[0048] The data acquisition module acquires historical metering records of charging facilities and calculates metering errors for each record. The binary state sequence construction module compares the metering error of each record with the error limit. If the error exceeds the limit, it is considered an out-of-tolerance state and recorded as 1; otherwise, it is considered a normal state and recorded as 0, forming a binary state sequence arranged chronologically. The sliding window construction module sets a sliding window of size N, extracts the latest N records in the binary state sequence as the observation window, and updates the observation window in real time with new metering records. The probability calculation module counts the state transition events of adjacent records within the sliding window, calculates the transition probability of remaining out of tolerance after exceeding the limit, and calculates the event transition probability from two consecutive out-of-tolerance states to the current out-of-tolerance state, assuming that the latest record in the sliding window has been determined to be out of tolerance. It also calculates the event transition probability from two consecutive out-of-tolerance states to the current out-of-tolerance state if three consecutive out-of-tolerance states occur. The output module compares the event transition probability from two consecutive out-of-tolerance states to the current out-of-tolerance state with a preset threshold. When P... risk When the value is greater than or equal to the preset threshold, an early warning signal is output.

[0049] It should be noted that the above modules can be functional modules or program modules, and can be implemented through software or hardware. For modules implemented through hardware, the above modules can reside in the same processor; or the above modules can be located in different processors in any combination.

[0050] The information interaction and execution process between the above modules are based on the same concept as the method embodiment of this application. It is a device corresponding to the above-mentioned charging facility continuous deviation review risk warning method based on state transition probability. All implementation methods in the above method embodiment are applicable to the embodiment of this device. For its specific functions and the technical effects it brings, please refer to the method embodiment section, which will not be repeated here.

[0051] This embodiment also provides a storage medium storing a computer program. When executed by a processor, the program implements the method for risk warning of continuous out-of-tolerance verification of charging facilities based on state transition probability, as proposed in the above embodiment. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0052] It should be noted that the above embodiments are only used to illustrate the technical solutions 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 solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for risk early warning of continuous out-of-tolerance verification of charging facilities based on state transition probability, characterized in that, Includes the following steps: Obtain historical metering records of charging facilities and calculate metering errors for each record; The measurement error of each measurement record is compared with the error limit. If it exceeds the limit, it is judged as an out-of-tolerance state and recorded as 1; otherwise, it is judged as a normal state and recorded as 0, forming a binary state sequence arranged in time sequence. Set the sliding window size N, select the latest N records in the binary state sequence as the observation window, and update the observation window in real time as new measurement records are added; Statistically analyze the state transition events of adjacent records within the sliding window, and calculate the transition probability P of remaining out of tolerance after exceeding the tolerance. 11 :P 11 = X UU / (X UN + X UU ); Among them, X UU X represents the number of consecutive deviations. UN This indicates the number of times the error immediately returned to normal. Assuming the latest record within the observation window has been determined to be out of tolerance, and considering the possibility of three consecutive out-of-tolerance states in the future, calculate the event transition probability P from two consecutive out-of-tolerance states to the current out-of-tolerance state. risk For: P risk =P 11 ×P 11 = P 11 2 ; According to the P risk Compared with a preset threshold, when P risk When the value is greater than or equal to the preset threshold, an early warning signal is output.

2. The method for risk early warning of continuous out-of-tolerance verification of charging facilities based on state transition probability according to claim 1, characterized in that: The historical metering records of the charging facilities are obtained through a cloud platform database.

3. The method for risk early warning of continuous out-of-tolerance verification of charging facilities based on state transition probability according to claim 1, characterized in that: The preset thresholds include: a first threshold, a second threshold, and a third threshold, wherein the first threshold is less than the second threshold, and the second threshold is less than the third threshold; when P risk When the value is less than the first threshold, it is determined to be an occasional deviation, at which point the risk is low, and monitoring should continue; when the P value is less than the first threshold, it is considered an occasional deviation, at which point monitoring should continue; risk If P is greater than or equal to the first threshold and less than the second threshold, the warning level is determined to be Level 1 warning; when P risk If P is greater than or equal to the second threshold and less than the third threshold, the warning level is determined to be a Level II warning; when P risk If the warning level is greater than or equal to the third threshold, the warning level is determined to be a Level 3 warning.

4. The method for risk early warning of continuous out-of-tolerance verification of charging facilities based on state transition probability according to claim 3, characterized in that: The method further includes: for each of the sliding windows, calculating P... risk Compared with a preset threshold, record the highest warning level reached by the sliding window. When P risk When any threshold is exceeded, an early warning signal is output.

5. The method for risk early warning of continuous out-of-tolerance verification of charging facilities based on state transition probability according to claim 4, characterized in that: The warning signal includes the device identifier and the current P. risk Values, warning levels, and recommended measures.

6. The method for risk early warning of continuous out-of-tolerance verification of charging facilities based on state transition probability according to claim 5, characterized in that: The warning recommendation for Level 1 is to pay attention to the risk; the warning recommendation for Level 2 is to arrange inspections; and the warning recommendation for Level 3 is to immediately review.

7. The method for risk early warning of continuous out-of-tolerance verification of charging facilities based on state transition probability according to claim 3, characterized in that: The first threshold is 0.2, the second threshold is 0.3, and the third threshold is 0.

4.

8. The method for risk early warning of continuous out-of-tolerance verification of charging facilities based on state transition probability according to claim 1, characterized in that: The method further includes: based on the P risk Establish a work order priority processing mechanism, which includes: If the current sliding window has triggered the out-of-tolerance state 3 times consecutively, and P risk If the value is greater than or equal to the preset threshold, the processing unit will be given the first processing priority. If the current sliding window does not trigger the out-of-tolerance state 3 times consecutively, but P risk If the value is greater than or equal to the preset threshold, the processing unit will be assigned to the second processing priority. If the current sliding window has triggered the out-of-tolerance state 3 times consecutively, and P risk If the value is less than the preset threshold, the processing unit will be placed in the third priority. If the current sliding window does not trigger the out-of-tolerance state 3 times consecutively, but P risk If the value is less than the preset threshold, the processing unit will be placed in the fourth priority category.

9. A charging facility continuous deviation verification risk early warning device based on state transition probability, applied to the charging facility continuous deviation verification risk early warning method based on state transition probability as described in any one of claims 1-8, characterized in that: The device includes: The data acquisition module is used to acquire historical metering records of charging facilities and calculate metering errors for each record. The binary state sequence construction module is used to compare the measurement error of each measurement record with the error limit. If the error exceeds the limit, it is judged as an out-of-tolerance state and recorded as 1; otherwise, it is judged as a normal state and recorded as 0, forming a binary state sequence arranged in time sequence. The sliding window construction module is used to set a sliding window of size N, extract the latest N records in the binary state sequence as the observation window, and the observation window is updated in real time as new measurement records are added. The probability calculation module is used to count the state transition events of adjacent records within the sliding window, calculate the transition probability of remaining out of tolerance after exceeding the tolerance, and calculate the event transition probability from two consecutive out-of-tolerance states to the out-of-tolerance state if the latest record in the observation window has been determined to be out of tolerance and three consecutive out-of-tolerance states are to be formed in the future. The output module is used to compare the event transition probability from two consecutive out-of-tolerance states with a preset threshold. When P risk When the value is greater than or equal to the preset threshold, an early warning signal is output.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the steps of the charging facility continuous deviation review risk warning method based on state transition probability as described in any one of claims 1 to 8.

Images