Automobile charging pile safety monitoring system based on multi-modal sensor fusion

Abstract

The invention relates to the technical field of data processing, in particular to an automobile charging pile safety monitoring system based on multi-modal sensor fusion, which comprises a memory, a processor and a computer program stored in the memory and running on the processor, when the processor executes the computer program, the following steps are realized: according to data differences of various monitoring indexes of any to-be-detected automobile charging pile and the automobile charging piles with normal working conditions, the electrical abnormality degree of any to-be-detected automobile charging pile is obtained; fusing and analyzing various monitoring indexes, temperature and vibration frequency of any to-be-detected automobile charging pile to obtain an abnormal confidence coefficient; according to the abnormal confidence coefficient and the electrical abnormal degree, the comprehensive abnormal degree is obtained, safety monitoring is carried out on any one to-be-detected automobile charging pile according to the similarity between the comprehensive abnormal degree of each of other to-be-detected automobile charging piles and the comprehensive abnormal degree of any one to-be-detected automobile charging pile, and the safety monitoring accuracy of the automobile charging piles is improved.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a safety monitoring system for car charging piles based on multimodal sensor fusion. Background Technology

[0002] With the increasing popularity of electric vehicles, charging stations, as crucial facilities for electric vehicle charging, are becoming a key component of modern urban infrastructure. Charging stations not only provide electricity to electric vehicles but also involve multiple aspects such as power transmission security, environmental protection, and monitoring of equipment operation status during the charging process. To ensure the safe, reliable, and efficient charging process for electric vehicles, real-time safety monitoring of charging stations is particularly important.

[0003] Traditional methods for monitoring the safety of electric vehicle charging stations mainly rely on data from single or limited types of sensors, using preset rules or simple statistical models for anomaly detection. Judging whether a charging station is abnormal based on a single sensor lacks multi-dimensional correlation analysis. Furthermore, when sensors are affected by external environmental interference, the data they collect may be inaccurate. Performing safety checks on charging stations under such conditions significantly reduces the accuracy of monitoring results, potentially leading to missed anomalies or false alarms.

[0004] Therefore, improving the accuracy of safety monitoring of electric vehicle charging stations has become an urgent problem to be solved. Summary of the Invention

[0005] In view of this, embodiments of the present invention provide a vehicle charging pile safety monitoring system based on multimodal sensor fusion to solve the problem of how to improve the accuracy of safety monitoring of vehicle charging piles.

[0006] This invention provides a vehicle charging pile safety monitoring system based on multimodal sensor fusion, including a memory, a processor, and a computer program stored in the memory and running on the processor. The system is characterized in that the processor executes the computer program to perform the following steps:

[0007] The monitoring data of various monitoring indicators of any car charging pile under test in the target area are obtained at each sampling time in the current period. The monitoring indicators include voltage and current. The data of various monitoring indicators of car charging piles under normal working conditions in the current period are recorded as the baseline data.

[0008] Based on the differences between the monitoring data of each monitoring indicator of any of the vehicle charging piles to be tested and the baseline data, the degree of electrical anomaly of any of the vehicle charging piles to be tested in the current time period is obtained.

[0009] The temperature and vibration frequency of any of the vehicle charging piles to be tested are obtained in the current time period. Based on the temperature and vibration frequency, the data fluctuation characteristics of various monitoring indicators of any of the vehicle charging piles to be tested in the current time period are analyzed to obtain the abnormal confidence level of any of the vehicle charging piles to be tested in the current time period.

[0010] Based on the anomaly confidence level, the electrical anomaly degree is weighted to obtain the comprehensive anomaly degree of any one to be tested electric vehicle charging pile. The comprehensive anomaly degree of each other to be tested electric vehicle charging pile in the target area is obtained. Based on the similarity of the comprehensive anomaly degree of each other to be tested electric vehicle charging pile to the one to be tested, safety monitoring is performed on the one to be tested electric vehicle charging pile.

[0011] The beneficial effects of the embodiments of the present invention compared with the prior art are as follows:

[0012] This invention acquires monitoring data of various indicators of any EV charging pile under test in a target area at each sampling time within the current time period. The monitoring indicators include voltage and current. Data of the various monitoring indicators of a normally functioning EV charging pile within the current time period are recorded as baseline data. Based on the difference between the monitoring data of the various monitoring indicators of the EV charging pile under test and the baseline data, the degree of electrical anomaly of the EV charging pile under test within the current time period is obtained. The temperature and vibration frequency of the EV charging pile under test within the current time period are acquired. Based on the temperature and vibration frequency, the data fluctuation characteristics of the various monitoring indicators of the EV charging pile under test within the current time period are analyzed to obtain the anomaly confidence level of the EV charging pile under test within the current time period. Based on the anomaly confidence level, the degree of electrical anomaly is weighted to obtain the comprehensive anomaly level of the EV charging pile under test. The comprehensive anomaly level of each other EV charging pile under test in the target area is acquired. Based on the similarity of the comprehensive anomaly level of each other EV charging pile under test to the comprehensive anomaly level of the EV charging pile under test, safety monitoring is performed on the EV charging pile under test. Specifically, the method involves several key steps. First, by comparing the monitoring data of various indicators of the charging pile under test with baseline data, the degree of electrical anomaly is determined. This allows for a preliminary assessment of whether the charging pile's circuitry is malfunctioning, reducing false alarms or missed detections when using pre-defined rules or simple statistical models to detect electrical anomalies and improving the accuracy of safety monitoring. Second, by integrating and analyzing the voltage, current, temperature, and vibration frequency of the charging pile under test, the degree of electrical anomaly is weighted to obtain a comprehensive anomaly degree, more accurately reflecting the true anomaly situation of the charging pile. Third, by combining the similarity of the comprehensive anomaly degrees of all charging piles under test, safety monitoring is performed to determine whether the anomaly is caused by grid interference or an inherent malfunction, further improving the accuracy of safety monitoring. Attached Figure Description

[0013] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0014] Figure 1 This is a flowchart of a method for safety monitoring of car charging piles based on multimodal sensor fusion, provided in Embodiment 1 of the present invention. Detailed Implementation

[0015] Embodiments of this disclosure are described in detail below, with examples of these embodiments illustrated in the accompanying drawings. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain this disclosure, and should not be construed as limiting it.

[0016] It should be noted that the terms "first," "second," etc., used in this disclosure and the accompanying drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this disclosure described herein can be implemented in orders other than those illustrated or described herein. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this disclosure. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this disclosure.

[0017] To illustrate the technical solution of the present invention, specific embodiments are described below.

[0018] The specific scenario addressed by this invention is as follows: Safety monitoring methods for car charging piles mainly rely on data from a single or limited number of sensors, using preset rules or simple statistical models for anomaly detection. These methods determine whether a charging pile is abnormal based on a single sensor, lacking multi-dimensional correlation analysis. Furthermore, when the sensor is affected by external environmental interference, the data collected may be inaccurate. Performing safety monitoring on charging piles under these conditions significantly reduces the accuracy of the monitoring results, potentially leading to missed or false alarms. Therefore, this invention improves the accuracy of safety monitoring of car charging piles by integrating monitoring data from multiple types of sensors and analyzing the anomalies of other charging piles in the same area.

[0019] This invention provides a vehicle charging pile safety monitoring system based on multimodal sensor fusion, including a processor and a memory. The processor executes a computer program stored in the memory to implement a vehicle charging pile safety monitoring method based on multimodal sensor fusion, such as... Figure 1 As shown, the method includes the following steps:

[0020] Step S101: Obtain the monitoring data of various monitoring indicators of any car charging pile under test in the target area at each sampling time in the current time period. The monitoring indicators include voltage and current. Record the data of various monitoring indicators of car charging piles under normal working conditions in the current time period as the baseline data.

[0021] Any centralized car charging area is designated as the target area. Considering that the electrical safety of car charging piles directly depends on the stability of current and voltage, when the circuit of a car charging pile malfunctions, its current and voltage will change significantly. Therefore, in this embodiment of the invention, voltage and current are used as the main monitoring indicators for the safety monitoring of car charging piles.

[0022] Taking the i-th car charging pile to be tested in the target area as an example, assuming that the i-th car charging pile to be tested is a DC charging pile, firstly, the current sensor is installed on the positive and negative cables of the DC output bus (close to the output end of the charging module), and the voltage sensor is connected in parallel between the positive and negative terminals of the DC output and ground. The data acquisition frequency of the current sensor and the voltage sensor is set to 1kHz to obtain the monitoring data of the current and voltage of the i-th car charging pile to be tested at each sampling moment within the current 1s (i.e., within 1s before the current moment, including the current moment). There is no restriction here. The implementer can set the data acquisition frequency and the time period length according to the specific scenario. The current sampling moment is the last sampling moment within the current 1s, which is also the current moment.

[0023] Traditionally, electrical anomalies in car charging stations are detected based on monitoring data from various indicators using preset rules or simple statistical models. However, sensors are susceptible to environmental interference, which may lead to inaccurate monitoring data for various indicators obtained through sensors. If preset rules or simple statistical models are used to detect electrical anomalies in car charging stations, there may be cases of missed or false alarms, which greatly reduces the accuracy of the detection results.

[0024] Therefore, in this embodiment of the invention, the current and voltage data of a normally functioning car charging pile within the current 1 second are obtained as the reference data of the current and voltage of the car charging pile within the current 1 second. The normally functioning car charging pile has the same working mode as the i-th car charging pile to be tested within the current 1 second. Based on the difference between the monitoring data of the i-th car charging pile to be tested and the reference data of various monitoring indicators (current and voltage), it is preliminarily determined whether the circuit of the i-th car charging pile to be tested is abnormal at the current moment, thereby improving the accuracy of safety monitoring of car charging piles.

[0025] Step S102: Based on the difference between the monitoring data of each monitoring indicator of any vehicle charging pile to be tested and the benchmark data, the degree of electrical abnormality of any vehicle charging pile to be tested in the current time period is obtained.

[0026] Step S101 obtains the monitoring data and baseline data of various monitoring indicators of the i-th car charging pile to be tested. Further, by analyzing the differences between the monitoring data of various monitoring indicators of the i-th car charging pile to be tested and the baseline data, it is preliminarily determined whether the circuit of the i-th car charging pile to be tested is abnormal at the current moment, so as to improve the accuracy of safety monitoring of car charging piles.

[0027] Taking current as an example, the monitoring data of the current of the i-th charging pile to be tested obtained in step S101 at each sampling time within the current 1 second are combined into a monitoring data sequence. The reference data of the current of a charging pile under normal operating conditions at each sampling time within the current 1 second are combined into a reference data sequence. Based on the difference between the monitoring data sequence corresponding to the current and the reference data sequence, the degree of abnormality of the current of the i-th charging pile to be tested within the current 1 second is obtained. Specifically:

[0028] Calculate the deviation from the mean for each data point in the baseline data sequence, record the mean of all deviations as the baseline data fluctuation characteristic value, and obtain the monitoring data fluctuation characteristic value based on the deviation from the mean for each data point in the monitoring data sequence.

[0029] The absolute value of the difference between the fluctuation characteristic value of the reference data and the fluctuation characteristic value of the monitoring data is calculated to obtain the first degree of deviation of the current;

[0030] Calculate the standard deviation of the monitoring data for all data in the monitoring data sequence and the standard deviation of the benchmark data for all data in the benchmark data sequence. Calculate the absolute value of the difference between the standard deviation of the monitoring data and the standard deviation of the benchmark data to obtain the second degree of deviation of the current.

[0031] Calculate the product between the first deviation degree and the second deviation degree to obtain the abnormality degree of the current of the i-th car charging pile to be detected in the current 1 second.

[0032] In one embodiment, the formula for calculating the degree of anomaly of the current of the i-th vehicle charging station to be detected within the current 1 second is:

[0033]

[0034] Among them, D i,I This represents the degree of anomaly in the current of the i-th car charging pile under test within the current 1 second, where m represents the number of data points in the monitoring data sequence corresponding to the current. The number of data points in the monitoring data sequence corresponding to the current is the same as the number of data points in the baseline data sequence. j I represents the j-th monitoring data in the monitoring data sequence corresponding to the current. avg I' represents the mean of all data in the monitoring data sequence corresponding to the current. bI' represents the b-th reference data in the reference data sequence corresponding to the current. avg θ represents the mean of all data in the reference data sequence corresponding to the current. I θ' represents the standard deviation of all data in the monitoring data sequence corresponding to the current. I The standard deviation of the reference data for all data in the reference data sequence corresponding to the current is represented by ||, where || represents the absolute value sign.

[0035] It should be noted that, The larger |θ I -θ' I The larger the value of |, the greater the difference in current fluctuation between the i-th tested car charging pile and a normally functioning car charging pile under the same operating mode, and thus D i,I The larger the value, the greater the probability that the circuit of the i-th car charging pile to be tested will malfunction within the current 1 second.

[0036] Similarly, the degree of voltage anomaly of the i-th car charging station to be tested within the current 1 second is obtained and denoted as D. i,U The average of the abnormality levels of current and voltage is calculated, and this average is taken as the electrical abnormality level of the i-th tested car charging pile in the current 1 second, denoted as D'. i D' i The larger the value, the greater the probability that the circuit of the i-th car charging pile to be tested will malfunction within the current 1 second, that is, the greater the probability that the i-th car charging pile to be tested will have a safety hazard within the current 1 second.

[0037] Thus, the electrical anomaly level of the i-th car charging pile to be tested within the current 1 second is obtained.

[0038] Step S103: Obtain the temperature and vibration frequency of any of the vehicle charging piles to be tested in the current time period. Based on the temperature and vibration frequency, analyze the data fluctuation characteristics of various monitoring indicators of any of the vehicle charging piles to be tested in the current time period to obtain the abnormal confidence level of any of the vehicle charging piles to be tested in the current time period.

[0039] Considering that the data collected by the current sensor and voltage sensor are easily affected by environmental interference and may deviate, the electrical anomaly level obtained in step S102 may not accurately reflect the true abnormal state of the i-th car charging pile to be tested. Therefore, it is also necessary to determine the accuracy of the electrical anomaly level of the i-th car charging pile to be tested within the current 1 second.

[0040] When an abnormality occurs in the circuit of a car charging station, it is generally classified as a transient abnormality or a continuous abnormality. A transient abnormality is a short circuit, while a continuous abnormality is mainly an overload fault. Specifically, when a short circuit occurs in the circuit of a car charging station, the current increases instantaneously, the voltage drops sharply, and the temperature rises accordingly. At the same time, an electric arc discharge occurs, causing high-frequency mechanical vibration. When an overload occurs in the circuit of a car charging station, the current remains higher than the rated value, the voltage drops slightly, the temperature rises slowly and remains almost unchanged for a short period of time, and low-frequency mechanical vibration occurs, with the vibration frequency almost zero.

[0041] Therefore, in this embodiment of the invention, a temperature sensor is installed on the heat sink surface of the charging module of the i-th car charging pile to be tested, and a vibration sensor is installed near the mechanical locking mechanism of the charging gun (such as the area where the charging gun connects to the vehicle socket, the tail of the gun body, or the plug shell). The temperature sensor and the sampling frequency of the temperature sensor are set to 1Hz to obtain the temperature and vibration frequency of the i-th car charging pile to be tested in the current 1s, which is also the temperature and vibration frequency of the i-th car charging pile to be tested at the current sampling time. Based on the temperature and vibration frequency, the data fluctuation characteristics of the voltage and current of the i-th car charging pile to be tested in the current 1s are analyzed to obtain the abnormal confidence level of the i-th car charging pile to be tested in the current 1s, which is used to characterize the accuracy of the electrical abnormality obtained in step S102.

[0042] The steps to obtain the anomaly confidence level of the i-th car charging pile to be detected within the current 1 second are as follows:

[0043] (1) Based on the rate of change of the current and voltage of the i-th vehicle charging pile under test in the current 1s, and the temperature and vibration frequency of the i-th vehicle charging pile under test at the current sampling time, obtain the instantaneous abnormality of the i-th vehicle charging pile under test in the current 1s.

[0044] Specifically, the monitoring data sequences corresponding to the current and voltage of the i-th car charging pile to be tested are denoted as the current monitoring data sequence and the voltage monitoring data sequence, respectively.

[0045] Obtain the rated current of the i-th vehicle charging pile to be tested, calculate the product between the rated current and the preset short-circuit current multiple to obtain the short-circuit current, calculate the proportion of the maximum current value in the current monitoring data sequence in the short-circuit current, and obtain the short-circuit probability index of the i-th vehicle charging pile to be tested.

[0046] In the current monitoring data sequence, the minimum current value preceding the maximum current value is obtained, the time interval between the minimum current value and the maximum current value is obtained, and the product between the short circuit probability index and the reciprocal of the time interval is calculated to obtain the instantaneous abnormal current characteristic value of the i-th vehicle charging pile to be detected in the current time period.

[0047] The sampling times corresponding to the minimum current value and the maximum current value are obtained respectively, and are denoted as the first sampling time and the second sampling time. In the voltage monitoring data sequence, the difference between the voltage monitoring data at the first sampling time and the voltage monitoring data at the second sampling time is calculated to obtain the instantaneous voltage anomaly characteristic value of the i-th car charging pile to be detected in the current 1 second.

[0048] Calculate the product between the temperature and vibration frequency of the i-th vehicle charging pile to be tested at the current sampling time to obtain the instantaneous correlation feature value of the i-th vehicle charging pile to be tested in the current 1 second;

[0049] Using the instantaneous correlation feature value as a weight, the sum of the instantaneous anomaly feature values ​​of the current and the instantaneous anomaly feature values ​​of the voltage is weighted to obtain the comprehensive instantaneous anomaly feature value of the i-th vehicle charging pile to be detected in the current 1 second. The negative of the comprehensive instantaneous anomaly feature value is used as the independent variable of an exponential function with the natural constant as the base to obtain the first function value. The constant 1 is subtracted from the first function value to obtain the instantaneous anomaly degree of the i-th vehicle charging pile to be detected in the current 1 second.

[0050] In one embodiment, the formula for calculating the instantaneous anomaly level of the i-th vehicle charging station to be detected within the current 1 second is:

[0051]

[0052] Among them, A i I represents the instantaneous anomaly level of the i-th car charging station to be detected within the current 1 second. max This represents the maximum current value in the current monitoring data sequence, α represents the preset short-circuit current multiple, and RC represents the rated current of the i-th vehicle charging pile to be tested. This represents the time interval between the maximum current value and the minimum current value preceding the maximum current value in the current monitoring data sequence. This represents the voltage monitoring data at the sampling time corresponding to the minimum current value preceding the maximum current value in the current monitoring data sequence. T represents the voltage monitoring data at the sampling time corresponding to the maximum current value in the current monitoring data sequence. i F represents the temperature of the i-th car charging station to be tested at the current sampling time. i Let represent the vibration frequency of the i-th car charging pile to be tested at the current sampling time, and exp() represent an exponential function with the natural constant as the base.

[0053] It should be noted that when a short circuit occurs in the circuit, the current can reach 10-20 times the rated current. Therefore, in this embodiment of the invention, α = 10 is set. There is no limitation here, and implementers can set it according to the specific scenario. The larger the value, the greater the current of the i-th car charging station to be tested, and thus A. i The larger the value, the greater the probability that the circuit of the i-th car charging pile to be tested will short-circuit; The smaller the value, the more it indicates a sudden increase in the current of the i-th car charging station being tested, which is more consistent with the characteristics of a short circuit, and thus A i The larger the value, the greater the probability that the circuit of the i-th car charging pile to be tested will short-circuit; The larger the value, the more it indicates a sudden voltage drop in the i-th tested car charging station, which is more consistent with the characteristics of a short circuit, thus A... i The larger the value of T, the greater the probability of a short circuit in the circuit of the i-th car charging station to be tested; i ×F i The larger the value of A, the higher the temperature of the i-th tested car charging station, accompanied by high-frequency vibration. i The larger the value, the greater the probability that the circuit of the i-th car charging pile to be tested will short-circuit.

[0054] (2) Obtain the monitoring data of the voltage, current, temperature and vibration frequency of the i-th car charging pile to be tested within 10 seconds after the current 1 second. Based on the changes in the monitoring data of the voltage, current, temperature and vibration frequency of the i-th car charging pile to be tested within 10 seconds after the current 1 second, obtain the degree of continuous abnormality of the i-th car charging pile to be tested.

[0055] When the circuit of a car charging station experiences overload or other faults, it will manifest as a continuous abnormality, specifically as the current continuously exceeding the rated value, the voltage slightly decreasing, the temperature slowly rising and remaining almost unchanged for a short period of time, and at the same time, it will cause low-frequency mechanical vibration, with the vibration frequency being almost 0.

[0056] Therefore, in this embodiment of the invention, the monitoring data of the voltage, current, temperature, and vibration frequency of the i-th car charging pile to be tested within 10 seconds after the current 1 second (including the current sampling time) are obtained to obtain the voltage data sequence, current data sequence, temperature data sequence, and vibration data sequence to be analyzed. Based on the changes in the monitoring data of voltage, current, temperature, and vibration frequency within 10 seconds after the current 1 second, the degree of continuous abnormality of the i-th car charging pile to be tested is obtained. Specifically:

[0057] The current of a normally functioning car charging pile is obtained within 10 seconds after the current 1 second. The current data sequence to be analyzed and the reference current data sequence are integrated respectively to obtain the integration results of the current data sequence to be analyzed and the reference current data sequence. The difference between the integration results of the current data sequence to be analyzed and the reference current data sequence is calculated to obtain the continuous abnormal current characteristic value of the i-th car charging pile to be detected.

[0058] Obtain the historical voltage data sequence of the i-th vehicle charging pile to be tested within 10 seconds before the current 1 second (including the current sampling time), calculate the average value of the data in the historical voltage data sequence and the voltage data sequence to be analyzed respectively, obtain the historical voltage average value and the voltage average value to be analyzed, calculate the difference between the historical voltage average value and the voltage average value to be analyzed, and obtain the degree of voltage drop.

[0059] The product of the reciprocal of the standard deviation of all data in the voltage data sequence to be analyzed and the degree of voltage drop is calculated to obtain the characteristic value of the continuous voltage anomaly of the i-th car charging pile to be detected.

[0060] In the temperature data sequence to be analyzed, the difference between each data point other than the first data point and the first data point is calculated. The average of all differences is then calculated to obtain the degree of temperature rise, denoted as G. Where nT represents the number of all data points in the temperature data sequence to be analyzed, and x g Let x represent the g-th data point in the temperature data sequence to be analyzed, and x1 represent the 1-th data point in the temperature data sequence to be analyzed. g The smaller the value of -x1), the smaller the temperature rise of the i-th car charging pile under test within 10 seconds after the current 1 second, indicating that the temperature of the i-th car charging pile under test is rising slowly.

[0061] If the temperature rise is less than 0, i.e., the temperature rise is negative, it indicates that the temperature of the i-th charging pile under test shows a decreasing trend within 10 seconds after the current 1 second, meaning that the temperature of the i-th charging pile under test is not abnormal. Therefore, the constant 0 is taken as the temperature characteristic value of the i-th charging pile under test, denoted as . Right now If the temperature rise is greater than or equal to 0, i.e., the temperature rise is non-negative, then the reciprocal of the sum of the temperature rise and a preset constant is calculated to obtain the temperature characteristic value of the i-th vehicle charging pile to be detected, denoted as . Right now Where G represents the degree of temperature rise, and ε represents a preset constant used to prevent the denominator from being zero. In this embodiment of the invention, ε is set to 0.01. There is no limitation here, and implementers can set it according to specific scenarios.

[0062] In the vibration sequence data sequence to be analyzed, the duration of vibration frequency non-zero is obtained, that is, the first non-zero data in the vibration data sequence to be analyzed is obtained. If there is zero data in the data after the first non-zero data in the vibration data sequence to be analyzed, the first zero data after the first non-zero data is obtained. The time interval between the first non-zero data and the first zero data is calculated to obtain the vibration duration of the i-th car charging pile to be detected.

[0063] If there is no zero data in the data after the first non-zero data, then the time interval between the first non-zero data and the last data in the vibration data sequence to be analyzed is obtained to obtain the vibration duration of the i-th car charging pile to be detected.

[0064] Calculate the average value of all data in the vibration data sequence to be analyzed, and take the product of the reciprocal of the average value and the vibration duration as the vibration feature value. Calculate the sum between the temperature feature value and the vibration feature value to obtain the continuous correlation feature value of the i-th vehicle charging pile to be detected.

[0065] Using the continuous correlation feature value as a weight, the sum between the continuous current anomaly feature value and the continuous voltage anomaly feature value is weighted to obtain the comprehensive continuous anomaly feature value of the i-th car charging pile to be detected. The negative of the comprehensive continuous anomaly feature value is used as the independent variable of an exponential function with the natural constant as the base to obtain the second function value. The continuous anomaly degree of the i-th car charging pile to be detected is obtained by subtracting the second function value from the constant 1.

[0066] In one embodiment, the formula for calculating the degree of continuous abnormality of the i-th vehicle charging station to be detected is:

[0067]

[0068] Among them, B i This indicates the degree of continuous abnormality of the i-th car charging station to be tested. This represents the sequence of current data to be analyzed for the i-th car charging station to be tested. I represents the integration result of the current data sequence to be analyzed. ref This represents the reference current data sequence of a car charging station under normal operating conditions, ∫I ref This represents the integration result over the reference current data sequence. This represents the historical average voltage of the i-th car charging station to be tested. Let θ represent the average voltage to be analyzed for the i-th car charging station to be tested. U,tbrThis represents the standard deviation of all data in the voltage data sequence to be analyzed. F represents the temperature characteristic value of the i-th car charging station to be tested. avg This represents the average value of all data in the vibration data sequence to be analyzed. Let represent the duration of vibration of the i-th car charging pile to be tested, and exp() represents an exponential function with the natural constant as the base.

[0069] It should be noted that, The larger the value, the greater the degree of current exceeding the limit of the i-th tested car charging pile within 10 seconds after the current 1 second. The current is continuously higher than the current under normal operating conditions, which is more consistent with the characteristics of a continuous abnormality, and thus B... i The larger the value, the greater the likelihood that the circuit of the i-th car charging pile to be tested will experience overload or other persistent abnormalities; θ is used to characterize the probability that the voltage of the i-th tested car charging station will decrease and remain stable within 10 seconds after the current 1 second. U,tbr The smaller the value, the more stable the voltage of the i-th car charging station to be tested will be within 10 seconds after the current 1 second, and thus... The larger B is i The larger the value, the more the circuit of the i-th car charging pile to be tested conforms to the characteristics of a continuous anomaly, and the greater the possibility that the circuit of the i-th car charging pile to be tested will experience overload or other continuous anomalies. The larger the value, the more slowly the temperature of the i-th tested car charging pile rises within 10 seconds after the current 1 second, which is more consistent with the characteristics of a continuous anomaly, and thus B... i The larger the value, the greater the likelihood that the circuit of the i-th car charging pile to be tested will experience overload or other persistent abnormalities; The larger F is avg The smaller the value, the more likely it is that the i-th tested car charging station experiences low-frequency vibration within 10 seconds after the current 1 second, and the longer the duration, the more consistent it is with the characteristics of a persistent anomaly, thus B i The larger the value, the greater the likelihood that the circuit of the i-th car charging station to be tested will experience overload or other persistent abnormalities.

[0070] (3) Based on the instantaneous anomaly level and the continuous anomaly level, obtain the anomaly confidence level of the i-th vehicle charging pile to be detected in the current time period.

[0071] Specifically, since a real short-circuit anomaly will directly trigger the circuit breaker, causing the i-th tested car charging station to stop working, the continuous anomaly level will be 0. Therefore, if the continuous anomaly level is 0, i.e., B... i If the value is 0, then the constant 1 is taken as the anomaly confidence level of the i-th car charging pile to be detected in the current 1 second, denoted as Z. i Z i=1;

[0072] If the degree of persistent abnormality is not 0, i.e., B i If the value is not equal to 0, then the mean between the sustained anomaly level and the instantaneous anomaly level is calculated to obtain the anomaly confidence level of the i-th test vehicle charging pile in the current 1 second, denoted as Z. i ,Right now Among them, A i B represents the instantaneous anomaly level of the i-th car charging station to be tested. i This indicates the degree of continuous abnormality of the i-th car charging station to be tested.

[0073] Thus, the anomaly confidence level of the i-th vehicle charging pile to be tested within the current 1 second is obtained, which is used to characterize the accuracy of the degree of electrical anomaly obtained in step S102.

[0074] Step S104: Based on the anomaly confidence level, the electrical anomaly degree is weighted to obtain the comprehensive anomaly degree of any vehicle charging pile to be tested; the comprehensive anomaly degree of each other vehicle charging pile to be tested in the target area is obtained; and based on the similarity of the comprehensive anomaly degree of each other vehicle charging pile to be tested with that of any vehicle charging pile to be tested, safety monitoring is performed on any vehicle charging pile to be tested.

[0075] After obtaining the anomaly confidence level of the i-th EV charging pile to be tested in step S103, the electrical anomaly level obtained in step S102 is weighted according to the anomaly confidence level to obtain the comprehensive anomaly level of the i-th EV charging pile to be tested, denoted as W. i That is, W i =Z i ×D' i Among them, Z i D' represents the anomaly confidence level of the i-th car charging station to be tested. i This indicates the degree of electrical anomaly of the i-th car charging station to be tested.

[0076] Considering that grid interference can also cause malfunctions in car charging stations, and that grid interference can cause malfunctions in all car charging stations within the target area, this embodiment of the invention obtains the comprehensive malfunction level of all car charging stations within the target area according to the method for obtaining the comprehensive malfunction level of the i-th car charging station to be tested. Based on the similarity between the comprehensive malfunction level of the i-th car charging station to be tested and that of each other car charging station to be tested within the target area, the final malfunction level of the i-th car charging station to be tested is obtained. This is used to determine the cause of the malfunction of the i-th car charging station to be tested, thereby improving the efficiency of safety monitoring of car charging stations and reducing false alarms and missed alarms. If the similarity between the comprehensive malfunction level of the i-th car charging station to be tested and that of each other car charging station to be tested within the target area is high, it indicates that the malfunction of the i-th car charging station to be tested is caused by grid interference; otherwise, it indicates that the i-th car charging station to be tested itself has malfunctioned.

[0077] The specific method for obtaining the final anomaly level of the i-th car charging pile to be tested is as follows:

[0078] Among all the overall anomaly levels of the tested electric vehicle charging piles in the target area, the probability that at least two have completely identical overall anomaly levels is very small. If the difference between the overall anomaly levels of any two tested electric vehicle charging piles is small, such as 0.01 or 0.02, the overall anomaly levels of these two tested electric vehicle charging piles can also be considered similar. Therefore, according to the preset decimal counting unit, the overall anomaly level of each tested electric vehicle charging pile in the target area is adjusted to obtain the rounded value of the overall anomaly level of each tested electric vehicle charging pile. In this embodiment of the invention, the preset decimal counting unit is set to one-tenth, that is, the overall anomaly level of each tested electric vehicle charging pile in the target area is rounded to one decimal place. There is no restriction here, and the implementer can set the decimal counting unit according to the specific scenario.

[0079] Calculate the probability and information content of the rounded-down value of the overall anomaly level of the i-th EV charging station to be detected among the rounded-down values ​​of the overall anomaly level of all EV charging stations to be detected in the target area, where the probability of occurrence is denoted as P(W'). i The information content is denoted as -logP(W'). i ), W' i Let S(W') represent the rounded value of the overall anomaly level of the i-th EV charging station to be detected. The product of the occurrence probability and the information content is calculated to obtain the entropy contribution value of the rounded value of the overall anomaly level of the i-th EV charging station to be detected, denoted as S(W'). i ), that is, S(W' i )=-P(W' i )×logP(W'i The opposite of the entropy contribution value is used as the independent variable of an exponential function with the natural constant as the base to obtain the abnormal similarity between the i-th car charging pile to be detected and other car charging piles to be detected. The occurrence probability and information content are existing technologies and will not be elaborated here.

[0080] Subtracting the anomaly similarity from the constant 1 yields the anomaly correction coefficient, which corrects the overall anomaly level of the i-th EV charging pile to be tested. The product of the anomaly correction coefficient and the overall anomaly level of the i-th EV charging pile to be tested is then calculated to obtain the final anomaly level of the i-th EV charging pile to be tested.

[0081] In one embodiment, the formula for calculating the final anomaly level of the i-th vehicle charging station to be detected is:

[0082] Y i ={1-exp[-S(W' i )]}×W i

[0083] Among them, Y i W' represents the final anomaly level of the i-th car charging station to be tested. i S(W') represents the rounded value of the overall anomaly level of the i-th car charging station to be tested. i W represents the entropy contribution value of the rounded value of the overall anomaly degree of the i-th car charging pile to be detected. i represents the overall anomaly level of the i-th car charging station to be tested, and exp() represents an exponential function with the natural constant as the base.

[0084] It should be noted that S(W' i The larger the value of Y, the more similar the overall anomaly level of the i-th tested charging pile is to that of every other tested charging pile in the target area. This indicates that the anomaly observed in the i-th tested charging pile is more likely to be caused by grid interference, and thus Y... i The smaller the value, the lower the probability that the i-th charging station to be tested is itself abnormal; that is, the overall abnormality level W of the i-th charging station to be tested is... i The less credible it is.

[0085] Furthermore, based on the principle of rounding, a preset anomaly threshold of 0.5 is set. This is not a restriction; implementers can set it according to specific scenarios. If the final anomaly degree Y of the i-th car charging pile to be detected... i If the value is ≥0.5, it indicates that the i-th charging pile under test has malfunctioned. In this case, an alarm needs to be triggered for the i-th charging pile under test so that relevant staff can be notified in a timely manner for handling.

[0086] In summary, this invention acquires monitoring data of various monitoring indicators of any EV charging pile under test in the target area at each sampling time within the current time period. These monitoring indicators include voltage and current. Data of the various monitoring indicators of a normally functioning EV charging pile within the current time period are recorded as baseline data. Based on the difference between the monitoring data of each monitoring indicator of the EV charging pile under test and the baseline data, the degree of electrical anomaly of the EV charging pile under test within the current time period is obtained. The temperature and vibration frequency of the EV charging pile under test within the current time period are acquired. Based on the temperature and vibration frequency, the data fluctuation characteristics of the various monitoring indicators of the EV charging pile under test within the current time period are analyzed to obtain the anomaly confidence level of the EV charging pile under test within the current time period. Based on the anomaly confidence level, the degree of electrical anomaly is weighted to obtain the comprehensive anomaly level of the EV charging pile under test. The comprehensive anomaly level of each other EV charging pile under test in the target area is acquired. Based on the similarity of the comprehensive anomaly level of each other EV charging pile under test to the comprehensive anomaly level of the EV charging pile under test, safety monitoring is performed on the EV charging pile under test. Specifically, the method involves several key steps. First, by comparing the monitoring data of various indicators of the charging pile under test with baseline data, the degree of electrical anomaly is determined. This allows for a preliminary assessment of whether the charging pile's circuitry is malfunctioning, reducing false alarms or missed detections when using pre-defined rules or simple statistical models to detect electrical anomalies and improving the accuracy of safety monitoring. Second, by integrating and analyzing the voltage, current, temperature, and vibration frequency of the charging pile under test, the degree of electrical anomaly is weighted to obtain a comprehensive anomaly degree, more accurately reflecting the true anomaly situation of the charging pile. Third, by combining the similarity of the comprehensive anomaly degrees of all charging piles under test, safety monitoring is performed to determine whether the anomaly is caused by grid interference or an inherent malfunction, further improving the accuracy of safety monitoring.

[0087] 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 the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A vehicle charging pile safety monitoring system based on multimodal sensor fusion, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it performs the following steps: The monitoring data of various monitoring indicators of any car charging pile under test in the target area are obtained at each sampling time in the current period. The monitoring indicators include voltage and current. The data of various monitoring indicators of car charging piles under normal working conditions in the current period are recorded as the baseline data. Based on the differences between the monitoring data of each monitoring indicator of any of the vehicle charging piles to be tested and the baseline data, the degree of electrical anomaly of any of the vehicle charging piles to be tested in the current time period is obtained. The temperature and vibration frequency of any of the vehicle charging piles to be tested are obtained in the current time period. Based on the temperature and vibration frequency, the data fluctuation characteristics of various monitoring indicators of any of the vehicle charging piles to be tested in the current time period are analyzed to obtain the abnormal confidence level of any of the vehicle charging piles to be tested in the current time period. Based on the anomaly confidence level, the electrical anomaly degree is weighted to obtain the comprehensive anomaly degree of any one to be tested electric vehicle charging pile. The comprehensive anomaly degree of each other to be tested electric vehicle charging pile in the target area is obtained. Based on the similarity of the comprehensive anomaly degree of each other to be tested electric vehicle charging pile to the one to be tested, safety monitoring is performed on the one to be tested electric vehicle charging pile.

2. The vehicle charging pile safety monitoring system based on multimodal sensor fusion according to claim 1, characterized in that, The method of determining the degree of electrical anomaly of any tested electric vehicle charging pile in the current time period based on the differences between the monitoring data of various monitoring indicators and the baseline data includes: For any monitoring indicator, the monitoring data of the monitoring indicator are combined into a monitoring data sequence, and the benchmark data of the monitoring indicator are combined into a benchmark data sequence. Calculate the deviation from the mean for each data point in the baseline data sequence, record the mean of all deviations as the baseline data fluctuation characteristic value, and obtain the monitoring data fluctuation characteristic value based on the deviation from the mean for each data point in the monitoring data sequence. Calculate the absolute value of the difference between the fluctuation characteristic value of the benchmark data and the fluctuation characteristic value of the monitoring data to obtain the first degree of deviation of any monitoring indicator; Calculate the standard deviation of the monitoring data for all data in the monitoring data sequence and the standard deviation of the benchmark data for all data in the benchmark data sequence. Calculate the absolute value of the difference between the standard deviation of the monitoring data and the standard deviation of the benchmark data to obtain the second degree of deviation of any monitoring indicator. Calculate the product between the first degree of deviation and the second degree of deviation to obtain the degree of abnormality of any monitoring indicator of any vehicle charging pile to be tested in the current time period; The degree of abnormality of each monitoring indicator of any of the vehicle charging piles to be tested is obtained, the average value of the degree of abnormality of all monitoring indicators is calculated, and the degree of electrical abnormality of any of the vehicle charging piles to be tested in the current time period is obtained.

3. The vehicle charging pile safety monitoring system based on multimodal sensor fusion according to claim 1, characterized in that, The process involves fusing and analyzing the data fluctuation characteristics of various monitoring indicators of any tested electric vehicle charging pile within the current time period based on temperature and vibration frequency to obtain the anomaly confidence level of any tested electric vehicle charging pile within the current time period, including: The monitoring data of the current of any vehicle charging pile to be tested at each sampling time within the current time period are combined into a current monitoring data sequence. The last monitoring data in the current monitoring data sequence is the monitoring data of the current of any vehicle charging pile to be tested at the current sampling time. The current sampling time is the last sampling time within the current time period. The voltage monitoring data sequence of any vehicle charging pile to be tested within the current time period is obtained. Based on the rate of change of data in the current monitoring data sequence and the voltage monitoring data sequence, as well as the temperature and vibration frequency of any vehicle charging pile to be tested at the current sampling time, the instantaneous abnormality of any vehicle charging pile to be tested in the current time period is obtained. The system acquires monitoring data of the voltage, current, temperature, and vibration frequency of any of the vehicle charging piles to be tested within a preset time period after the current time period. Based on the changes in the monitoring data of the voltage, current, temperature, and vibration frequency of any of the vehicle charging piles to be tested within the preset time period, the system acquires the degree of continuous abnormality of any of the vehicle charging piles to be tested. Based on the instantaneous anomaly level and the sustained anomaly level, the anomaly confidence level of any vehicle charging station to be detected in the current time period is obtained.

4. The vehicle charging pile safety monitoring system based on multimodal sensor fusion according to claim 3, characterized in that, The step of obtaining the instantaneous anomaly degree of any tested electric vehicle charging pile in the current time period based on the rate of change of data in the current monitoring data sequence and the voltage monitoring data sequence, as well as the temperature and vibration frequency of any tested electric vehicle charging pile at the current sampling time, includes: Obtain the rated current of any of the vehicle charging piles to be tested, calculate the product between the rated current and the preset short-circuit current multiple to obtain the short-circuit current, calculate the proportion of the maximum current value in the current monitoring data sequence in the short-circuit current, and obtain the short-circuit probability index of any of the vehicle charging piles to be tested. In the current monitoring data sequence, the minimum current value preceding the maximum current value is obtained, the time interval between the minimum current value and the maximum current value is obtained, and the product between the short circuit probability index and the reciprocal of the time interval is calculated to obtain the instantaneous abnormal current characteristic value of any vehicle charging pile to be tested in the current time period. The sampling times corresponding to the minimum current value and the maximum current value are obtained respectively and denoted as the first sampling time and the second sampling time. In the voltage monitoring data sequence, the difference between the voltage monitoring data at the first sampling time and the voltage monitoring data at the second sampling time is calculated to obtain the instantaneous voltage anomaly characteristic value of any vehicle charging pile to be detected in the current time period. Based on the temperature and vibration frequency of any vehicle charging pile to be tested at the current sampling time, the instantaneous correlation feature value of any vehicle charging pile to be tested in the current time period is obtained. Using the instantaneous correlation feature value as a weight, the sum of the instantaneous anomaly feature values ​​of the current and the instantaneous anomaly feature values ​​of the voltage is weighted to obtain the comprehensive instantaneous anomaly feature value of any vehicle charging pile to be tested in the current time period. The negative of the comprehensive instantaneous anomaly feature value is used as the independent variable of an exponential function with the natural constant as the base to obtain the first function value. The first function value is subtracted from the constant 1 to obtain the instantaneous anomaly degree of any vehicle charging pile to be tested in the current time period.

5. The vehicle charging pile safety monitoring system based on multimodal sensor fusion according to claim 4, characterized in that, The step of obtaining the instantaneous correlation feature value of any test vehicle charging pile in the current time period based on its temperature and vibration frequency at the current sampling time includes: Calculate the product between the temperature and vibration frequency of any of the vehicle charging piles to be tested at the current sampling time to obtain the instantaneous correlation feature value of any of the vehicle charging piles to be tested in the current time period.

6. The vehicle charging pile safety monitoring system based on multimodal sensor fusion according to claim 3, characterized in that, The step of obtaining the degree of continuous abnormality of any one of the tested electric vehicle charging piles based on the changes in monitoring data of voltage, current, temperature, and vibration frequency within the preset time period includes: The monitoring data of voltage, current, temperature and vibration frequency of any of the vehicle charging piles to be tested within the preset time period are respectively composed into a voltage data sequence to be analyzed, a current data sequence to be analyzed, a temperature data sequence to be analyzed and a vibration data sequence to be analyzed; A reference current data sequence of a normally functioning car charging pile is obtained within a preset time period. The current data sequence to be analyzed and the reference current data sequence are integrated respectively to obtain the integration results of the current data sequence to be analyzed and the reference current data sequence. The difference between the integration results of the current data sequence to be analyzed and the reference current data sequence is calculated to obtain the continuous abnormal current characteristic value of any car charging pile to be tested. Obtain the historical voltage data sequence of any of the vehicle charging piles to be tested within a preset time period before the current time period, calculate the average value of the data in the historical voltage data sequence and the voltage data sequence to be analyzed, obtain the historical voltage average value and the voltage average value to be analyzed, calculate the difference between the historical voltage average value and the voltage average value to be analyzed, and obtain the degree of voltage drop. The product of the reciprocal of the standard deviation of all data in the voltage data sequence to be analyzed and the degree of voltage drop is calculated to obtain the characteristic value of the continuous voltage anomaly of any vehicle charging pile to be detected. Based on the data changes in the temperature data sequence and the vibration data sequence to be analyzed, the continuous correlation characteristic value of any of the vehicle charging piles to be detected is obtained; Using the continuous correlation feature value as a weight, the sum between the continuous current anomaly feature value and the continuous voltage anomaly feature value is weighted to obtain the comprehensive continuous anomaly feature value of any vehicle charging pile to be tested. The negative of the comprehensive continuous anomaly feature value is used as the independent variable of an exponential function with the natural constant as the base to obtain the second function value. The continuous anomaly degree of any vehicle charging pile to be tested is obtained by subtracting the second function value from the constant 1.

7. The vehicle charging pile safety monitoring system based on multimodal sensor fusion according to claim 6, characterized in that, The step of obtaining the continuous correlation feature value of any vehicle charging pile to be detected based on the data changes in the temperature data sequence and the vibration data sequence to be analyzed includes: In the temperature data sequence to be analyzed, the difference between each data point other than the first data point and the first data point is calculated. The average of all differences is calculated to obtain the temperature rise. If the temperature rise is less than 0, the constant 0 is used as the temperature characteristic value of any vehicle charging pile to be tested. If the temperature rise is greater than or equal to 0, the reciprocal of the sum of the temperature rise and the preset constant is calculated to obtain the temperature characteristic value of any vehicle charging pile to be tested. Obtain the first non-zero data in the vibration data sequence to be analyzed. If there is zero data in the data after the first non-zero data in the vibration data sequence to be analyzed, then obtain the first zero data after the first non-zero data. Calculate the time interval between the first non-zero data and the first zero data to obtain the vibration duration of any vehicle charging pile to be tested. If there is no zero data in the data following the first non-zero data, then the time interval between the first non-zero data and the last data in the vibration data sequence to be analyzed is obtained to obtain the vibration duration of any vehicle charging pile to be detected. Calculate the average value of all data in the vibration data sequence to be analyzed, and use the product of the reciprocal of the average value and the vibration duration as the vibration feature value. Calculate the sum between the temperature feature value and the vibration feature value to obtain the continuous correlation feature value of any vehicle charging pile to be detected.

8. The vehicle charging pile safety monitoring system based on multimodal sensor fusion according to claim 3, characterized in that, The step of obtaining the anomaly confidence level of any tested electric vehicle charging station in the current time period based on the instantaneous anomaly level and the sustained anomaly level includes: If the degree of continuous anomaly is 0, then a constant 1 is used as the anomaly confidence level of any of the vehicle charging piles to be detected in the current time period; If the sustained anomaly level is not 0, the mean between the sustained anomaly level and the instantaneous anomaly level is calculated to obtain the anomaly confidence level of any vehicle charging pile to be detected in the current time period.

9. The vehicle charging pile safety monitoring system based on multimodal sensor fusion according to claim 1, characterized in that, The step of conducting safety monitoring on any one of the vehicle charging piles under test based on the similarity of the overall anomaly level between each other vehicle charging pile under test and the any one vehicle charging pile under test includes: Based on the preset decimal counting unit, the overall anomaly level of each vehicle charging pile to be tested in the target area is adjusted to obtain the rounded value of the overall anomaly level of each vehicle charging pile to be tested. Calculate the probability and information content of the rounded value of the comprehensive anomaly degree of any to-be-detected electric vehicle charging pile among the rounded values ​​of the comprehensive anomaly degree of all to-be-detected electric vehicle charging piles in the target area. Calculate the product between the probability of occurrence and the information content to obtain the entropy contribution value of the rounded value of the comprehensive anomaly degree of any to-be-detected electric vehicle charging pile. Use the negative of the entropy contribution value as the independent variable of an exponential function with the natural constant as the base to obtain the anomaly similarity between any to-be-detected electric vehicle charging pile and other to-be-detected electric vehicle charging piles. Subtracting the anomaly similarity from the constant 1 yields an anomaly correction coefficient that corrects the overall anomaly level of any of the vehicle charging piles to be tested. The product of the anomaly correction coefficient and the overall anomaly level of any of the vehicle charging piles to be tested is calculated to obtain the final anomaly level of any of the vehicle charging piles to be tested. Safety monitoring is performed on any of the vehicle charging piles to be tested based on the final degree of anomaly.

10. The vehicle charging pile safety monitoring system based on multimodal sensor fusion according to claim 9, characterized in that, The step of conducting safety monitoring on any one of the vehicle charging piles to be tested based on the final degree of abnormality includes: If the final abnormality level is greater than or equal to the preset abnormality threshold, an abnormality alarm will be triggered for any of the vehicle charging piles to be tested.

Images