New energy automobile charging pile intelligent operation and maintenance management system

Abstract

The invention relates to the technical field of charging pile collaborative management, in particular to a new energy automobile charging pile intelligent operation and maintenance management system, which comprises a data acquisition and transmission unit, an edge computing unit and a cloud server unit, and is characterized in that the data acquisition and transmission unit collects multi-aspect information of a charging pile and transmits the multi-aspect information to the edge computing unit; the edge computing unit preprocesses the data, starts an emergency protection mechanism when the power grid fluctuates, and uploads the data and decision results, the cloud server unit constructs a cloud computing platform, a data storage service stores the data, a data analysis service excavates a charging rule, and an algorithm model service deploys an intelligent adaptive power distribution algorithm. The system comprises a dynamic priority determination module, a real-time power adjustment module and a cross-station cooperation power scheduling module which are respectively used for determining the vehicle priority, adjusting the power according to the power grid load and guiding the vehicle to perform cross-station charging, thereby solving the problems of unreasonable power distribution and insufficient cross-station cooperation in the operation and maintenance management of the charging pile, and improving the use efficiency of the charging pile.

Description

Technical Field

[0001] The present invention relates to the technical field of coordinated management of charging piles, and in particular to an intelligent operation and maintenance management system for charging piles of new energy vehicles. Background Art

[0002] Collaborative management of charging piles is an important technology. With the continuous increase in the number of new energy vehicles, efficient operation and maintenance of charging piles and reasonable power distribution have become key issues that need to be urgently addressed.

[0003] At some charging pile stations near busy commercial areas or transportation hubs, commercial vehicles and private cars queue up to charge together, and the power allocation strategy of existing charging piles is often fixed. During peak hours, all vehicles are evenly allocated power or charged according to a fixed power allocation mode. Commercial vehicles cannot obtain higher charging power due to their special needs, and due to the lack of intelligent coordination mechanism between charging piles, it is impossible to obtain idle information of surrounding stations in real time, which makes it impossible for queued vehicles to wait at the station for a long time, greatly wasting the owner's time. This congestion problem will affect the overall utilization efficiency of the charging piles, reduce the number of charging services per unit time, and lead to low "car-pile" coordination efficiency. In order to solve this technical problem, we provide an intelligent operation and maintenance management system for new energy vehicle charging piles. Summary of the invention

[0004] The purpose of the present invention is to provide an intelligent operation and maintenance management system for new energy vehicle charging piles to solve the problems raised in the above background technology.

[0005] To achieve the above purpose, an intelligent operation and maintenance management system for new energy vehicle charging piles is provided, including a data acquisition and transmission unit, an edge computing unit, and a cloud server unit;

[0006] The data acquisition and transmission unit deploys a sensor component on each charging pile, and the sensor component is used to collect electrical parameters, environmental parameters, and status information of the charging pile and the charging gun and cable, and transmit the collected information to the edge computing unit;

[0007] The edge computing unit uses a data analysis model to pre-process the information, and based on local real-time data, activates a local emergency protection mechanism when there is an instantaneous fluctuation in the power grid, adjusts the power of the charging pile, and synchronously uploads the pre-processed data and local decision results to the cloud server unit;

[0008] The cloud server unit constructs a cloud computing platform based on a microservice architecture, which includes multiple service modules, wherein the data storage service uses a distributed database to store charging pile data, the data analysis service uses big data processing technology to analyze the aggregated data, extracts the charging demand patterns in different regions and time periods, and provides a basis for the power allocation strategy, and the algorithm model service deploys an intelligent adaptive power allocation algorithm model. The algorithm model includes a dynamic priority determination module, a real-time power dynamic adjustment module, and a cross-station collaborative power scheduling module. The dynamic priority determination module establishes a vehicle priority evaluation model to determine the vehicle priority, and the real-time power dynamic adjustment module establishes a two-way communication link with the power grid dispatching system, and sets the power adjustment strategy in combination with the real-time acquired power grid load information and the usage status of the charging piles in the area. The cross-station collaborative power scheduling module performs grid management on the charging pile area, wherein each grid contains multiple charging pile sites, and uses an ant colony algorithm for path planning to recommend queued vehicles to go to idle sites for charging.

[0009] As a further improvement of the technical solution, the operation of the edge computing unit when running the lightweight data analysis model is as follows:

[0010] After receiving the data from the charging pile and the vehicle, the decision tree-based classification algorithm is used to preliminarily classify the data into electrical parameter category, environmental parameter category and vehicle battery information category;

[0011] For electrical parameter data, an autoencoder-based anomaly detection algorithm is used to calculate the reconstruction error. If the reconstruction error is greater than the preset electrical parameter anomaly threshold, it is determined to be an electrical anomaly and the local emergency protection mechanism is immediately activated.

[0012] For environmental parameter data, the environmental risk assessment model based on support vector machine is used to determine whether the current environment is within the normal operating range of the charging pile. If it is beyond the range, the risk coefficient is calculated based on the current usage status of the charging pile. If the risk coefficient is greater than the preset threshold, the local emergency protection mechanism is immediately activated;

[0013] For vehicle battery information data, a battery health prediction model based on a multi-layer perceptron is used to predict the battery health status. If the prediction result shows that the battery health status is lower than the preset health threshold, battery maintenance suggestions will be pushed to the owner when the owner uses the charging pile.

[0014] As a further improvement of the technical solution, the dynamic priority determination module establishes a vehicle priority evaluation model to determine the operation of vehicle priority, which is specifically as follows:

[0015] The priority is determined based on the vehicle type, the remaining battery power, and the user's urgency. Among the vehicle types, the priority of operating vehicles is higher than that of private cars, and the priority of emergency rescue vehicles is at the top.

[0016] The priority of vehicles with a remaining battery power of less than 20% is A1, and the priority of vehicles with a remaining battery power of more than 80% is A3. The system determines the user's urgency based on the owner's historical charging behavior analysis. The priority of the owner's vehicle in an emergency is A2, where A2 is less than or equal to A1 and greater than A3;

[0017] The priority of all vehicles waiting to be charged in the area is recalculated every minute based on the above factors. When there is an instantaneous fluctuation in the power grid, the convolutional neural network is used to extract real-time features of the power grid fluctuation data and the charging pile fault characteristic data, and input them into the risk assessment model built based on the long short-term memory network to determine the fault risk level. If the risk level exceeds the preset threshold, the local emergency protection mechanism is immediately activated to adjust the charging pile power.

[0018] As a further improvement of the technical solution, the operation mode of the real-time power dynamic adjustment module is as follows:

[0019] The real-time power dynamic adjustment module establishes a two-way communication link with the power grid dispatching system to obtain the load information of the power grid in real time, including the current load, load growth rate, and peak and valley period identification. Different power adjustment strategies are set according to the power grid load and the usage status of charging piles in the area.

[0020] The period from 0:00 to 6:00 at night is designated as the off-peak period. If the charging pile has not reached full load, the charging pile is allowed to run at maximum power. At the same time, preferential information on low-price charging periods is pushed to car owners. The period from 9:00 to 11:00 in the morning and from 2:00 to 4:00 in the afternoon is designated as the off-peak period. When the utilization rate of the charging pile exceeds 70%, the power of vehicles with priority A3 is reduced by no more than 20%. For vehicles with priority A2, the maximum power is maintained. The period from 7:00 to 9:00 in the evening is designated as the peak period. When the grid load reaches the warning value, the power of all non-emergency charging vehicles is gradually reduced to a maximum of 50%. In the process of power adjustment, soft start and soft stop technology are used.

[0021] As a further improvement of the technical solution, the operation of the cross-site collaborative power scheduling module is as follows:

[0022] Carry out grid management for charging pile-dense areas. Each grid contains multiple charging pile sites. Establish a communication network within the grid to achieve data sharing and collaborative work between sites.

[0023] When the charging pile utilization rate at a certain station exceeds the saturation threshold and there are queuing vehicles, the idle charging pile resources in the surrounding grid are searched and the current location failure risk level is calculated. If the risk level exceeds the preset threshold, the local emergency protection mechanism is immediately activated to adjust the charging pile power.

[0024] As a further improvement of the technical solution, the dynamic priority determination module considers the vehicle's cruising range requirements when establishing a vehicle priority evaluation model to determine the vehicle priority, as follows:

[0025] The cloud service module establishes a communication connection with the vehicle's battery management system to obtain the vehicle's battery capacity, remaining power, vehicle energy consumption information, and the owner's travel plan in real time. The vehicle energy consumption information is the average energy consumption obtained by dividing the total energy consumption of the vehicle in the past week by the total mileage. The owner's travel plan is the destination information. The cloud system converts the destination information into geographic coordinates and calculates the estimated driving distance from the current location to the destination in combination with map data.

[0026] The remaining range of the vehicle is calculated based on the obtained vehicle battery capacity, remaining power and vehicle energy consumption information, and the calculated remaining range is compared with the estimated driving distance in the owner's travel plan. If the remaining range is less than the estimated driving distance, a fixed score is added to the original priority score of the vehicle. After the vehicle priority is adjusted, the cloud service module recalculates the priority of all vehicles to be charged in the area according to the updated priority every 1 minute, and updates the queue sequence.

[0027] As a further improvement of the technical solution, the real-time power dynamic adjustment module is combined with a fuzzy control algorithm when using soft start and soft stop technology, as follows:

[0028] Through the communication interface between the charging pile and the vehicle battery management system, the key status parameters of the vehicle battery, including battery voltage, battery temperature and current charging current, are collected in real time, and the collected data is filtered;

[0029] Obtaining the load change rate of the power grid from a two-way communication link established with the power grid dispatching system, and determining two input variables and one output variable of the fuzzy control algorithm, wherein the input variables are the battery state and the power grid load change rate, and the output variable is the power change rate;

[0030] For the battery status, the fuzzy set is defined as {low, medium, high}, corresponding to the battery being in a poor, general and good state respectively. For the grid load change rate, the fuzzy set is defined as {negative large, negative small, zero, positive small, positive large}, reflecting the downward or upward trend and degree of the grid load. For the output variable power change rate, the fuzzy set is defined as {very slow, slow, moderate, fast, very fast}, and then the fuzzy reasoning rules are formulated based on historical experience;

[0031] According to the currently collected and fuzzified values ​​of battery status and grid load change rate, their memberships to their respective defined fuzzy sets are calculated respectively, the memberships of input variables are matched with fuzzy reasoning rules, the activated rules are found, and the rule strength of each activated rule is calculated. According to the conclusion part and rule strength of the activated rules, the membership of the output variable to the fuzzy set to which it belongs is calculated, and the membership of all activated rules is calculated, and finally the comprehensive membership distribution of the output variable on each fuzzy set is obtained;

[0032] Finally, the centroid method is used to convert the fuzzy set membership of the output variable into a power change rate value, and the output power of the charging pile is adjusted in real time according to the power change rate obtained by defuzzification.

[0033] As a further improvement of the technical solution, the cross-station collaborative power scheduling module uses an ant colony algorithm to perform path planning when searching for idle charging pile resources in the surrounding grids, as follows:

[0034] The charging pile stations are regarded as nodes in the graph. Each node contains the idle status information and location information of the charging pile. The road connections between nodes are regarded as edges. The weight of the edge is determined by the distance, traffic congestion level and estimated arrival time.

[0035] Set the number of ants, pheromone volatility coefficient, pheromone heuristic factor, and expected heuristic factor, initialize the pheromone concentration on each edge, and randomly place A ants at the charging station where there are currently queued vehicles, that is, the starting station. When each ant is at a certain node, it selects the next node according to the probability formula. The ant selects the next node according to the above probability until it reaches an idle charging station, that is, the target node, completing a path construction;

[0036] After all ants have completed the path construction, the pheromone of each edge is volatilized and updated. Finally, the pheromone added by all ants on the edge is accumulated to obtain the new pheromone concentration of the edge. Then the maximum number of iterations is set. If the current number of iterations reaches the maximum number of iterations, the algorithm is terminated. The optimal path length obtained in each iteration is recorded. If the change in the optimal path length for n consecutive iterations is less than a threshold, the algorithm is considered to have converged and terminated.

[0037] After the algorithm terminates, the optimal path obtained in all iterations is selected as the optimal path from the current station to the idle station, and the path information is provided to the queued vehicles to guide the vehicles to charge at the idle station, while locking the parking space at the idle station for the vehicles.

[0038] Compared with the prior art, the present invention has the following beneficial effects:

[0039] In the intelligent operation and maintenance management system of new energy vehicle charging piles, the data analysis service of the cloud server unit uses big data processing technology to deeply explore the charging demand patterns in different regions and time periods, provide a scientific basis for the power allocation strategy, and improve the utilization efficiency of resources. The dynamic priority determination module in the algorithm model service comprehensively considers vehicle type, remaining battery power, user urgency and range demand factors to accurately determine vehicle priority, ensuring that vehicles in urgent need of charging receive priority service. The real-time power dynamic adjustment module communicates bidirectionally with the power grid dispatching system and flexibly adjusts power according to the power grid load and the usage status of the charging pile, which not only ensures the stability of the power grid but also reduces the charging cost of users. The cross-station collaborative power dispatching module adopts grid management and ant colony algorithm to guide vehicles to idle sites, effectively alleviating site congestion and improving the overall utilization efficiency of charging piles. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 It is an overall block diagram of the present invention.

[0041] The meaning of each number in the figure is:

[0042] 1. Data acquisition and transmission unit; 2. Edge computing unit; 3. Cloud server unit; 31. Dynamic priority determination module; 32. Real-time power dynamic adjustment module; 33. Cross-site collaborative power scheduling module. DETAILED DESCRIPTION

[0043] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0044] The present invention provides an intelligent operation and maintenance management system for new energy vehicle charging piles. Figure 1 As shown, it includes a data acquisition and transmission unit 1, an edge computing unit 2 and a cloud server unit 3;

[0045] The data acquisition and transmission unit 1 deploys a sensor component on each charging pile. The sensor component is used to collect the electrical parameters, environmental parameters and status information of the charging gun and cable of the charging pile. The electrical parameter data includes voltage, current and power, and the environmental parameters include temperature, humidity and air quality. The collected information is transmitted to the edge computing unit 2.

[0046] The edge computing unit 2 uses a data analysis model to pre-process the information, and based on local real-time data, activates a local emergency protection mechanism when there is an instantaneous fluctuation in the power grid, adjusts the power of the charging pile, and synchronously uploads the pre-processed data and local decision results to the cloud server unit 3.

[0047] The operations of the edge computing unit 2 when running the lightweight data analysis model are as follows:

[0048] After receiving the data from the charging pile and the vehicle, the decision tree-based classification algorithm is used to preliminarily classify the data into electrical parameter category, environmental parameter category and vehicle battery information category;

[0049] For electrical parameter data, an autoencoder-based anomaly detection algorithm is used to calculate the reconstruction error. The autoencoder can learn data features without a large amount of labeled data and can adapt to the dynamic changes of electrical parameters. If the reconstruction error is greater than the preset electrical parameter anomaly threshold, it is judged as an electrical anomaly and the local emergency protection mechanism is immediately activated to reduce the risk of equipment damage and ensure the safe and stable operation of charging piles and vehicles.

[0050] For environmental parameter data, the environmental risk assessment model based on support vector machine is used to determine whether the current environment is within the normal operating range of the charging pile. The operation of the charging pile is greatly affected by environmental factors. Different environmental conditions may affect the performance of the charging pile and even cause failures. If it exceeds the range, the risk factor is calculated in combination with the current use status of the charging pile. If the risk factor is greater than the preset threshold, the local emergency protection mechanism is immediately activated to warn of environmental risks in advance, avoid damage to the charging pile caused by long-term operation in harsh environments, improve the reliability and service life of the charging pile, and ensure the safety of the charging process;

[0051] For vehicle battery information data, a battery health prediction model based on a multi-layer perceptron is used to predict the battery health status. If the prediction result shows that the battery health status is lower than the preset health threshold, battery maintenance suggestions will be pushed to the owner when the owner uses the charging pile, helping the owner to understand the battery health status in a timely manner, take corresponding maintenance measures, and extend the battery life.

[0052] The cloud server unit 3 constructs a cloud computing platform based on a microservice architecture, which includes multiple service modules, among which the data storage service uses a distributed database to store charging pile data, the data analysis service uses big data processing technology to analyze the aggregated data, extracts the charging demand patterns in different regions and time periods, and provides a basis for the power allocation strategy, and the algorithm model service deploys an intelligent adaptive power allocation algorithm model. The algorithm model includes a dynamic priority determination module 31, a real-time power dynamic adjustment module 32, and a cross-station collaborative power scheduling module 33. The dynamic priority determination module 31 establishes a vehicle priority evaluation model to determine the vehicle priority, and the real-time power dynamic adjustment module 32 establishes a two-way communication link with the power grid dispatching system, and sets the power adjustment strategy in combination with the real-time acquired power grid load information and the usage status of the charging piles in the area. The cross-station collaborative power scheduling module 33 performs grid management on the charging pile area, wherein each grid contains multiple charging pile sites, and uses an ant colony algorithm for path planning to recommend queued vehicles to go to idle sites for charging.

[0053] The dynamic priority determination module 31 establishes a vehicle priority evaluation model to determine the operation of vehicle priority, which is specifically as follows:

[0054] The priority coefficients of emergency rescue vehicles, operating vehicles and private cars are set. The priority of vehicles with a remaining battery power of less than 20% is A1, and the priority of vehicles with a remaining battery power of more than 80% is A3. The system judges the user's urgency based on the owner's historical charging behavior analysis. The priority of the owner's vehicle in an emergency is A2, where A2 is less than or equal to A1 and greater than A3. The priority of all vehicles to be charged in the area is recalculated every 1 minute based on the above factors. When there is an instantaneous fluctuation in the power grid, the convolutional neural network is used to extract real-time features of the power grid fluctuation data and the charging pile fault feature data, and the key features of the power grid fluctuation and the charging pile fault are extracted. The extracted features are input into the risk assessment model constructed based on the long short-term memory network to judge the fault risk level. The output layer outputs the predicted result of the fault risk level. If the risk level exceeds the preset threshold, the local emergency protection mechanism is immediately activated to adjust the charging pile power.

[0055] The dynamic priority determination module 31 establishes a vehicle priority evaluation model to determine the vehicle priority, taking into account the vehicle's cruising range requirements, as follows:

[0056] The cloud service module 3 establishes a connection with the battery management system of the vehicle through a communication protocol, and reads the battery capacity and remaining power of the vehicle in real time from the battery management system. For the vehicle energy consumption information, the total energy consumption and total mileage of the vehicle in the past week are obtained from the vehicle's driving computer, and the average energy consumption is calculated. The owner enters the destination information through the charging pile APP or the vehicle's built-in system. The cloud system converts the text description of the destination into geographic coordinates, and uses the path planning algorithm in combination with the map data to calculate the estimated driving distance from the current vehicle position to the destination. The remaining cruising range of the vehicle is calculated based on the obtained vehicle battery capacity, remaining power and vehicle energy consumption information, and the calculated remaining cruising range is compared with the estimated driving distance in the owner's travel plan. If the remaining cruising range is less than the estimated driving distance, a fixed score is added to the original priority score for the vehicle. When the remaining cruising range of the vehicle cannot meet the owner's travel plan, it means that the vehicle is in urgent need of charging. In order to give priority to the charging needs of such vehicles, their priority needs to be increased. After completing the adjustment of the vehicle priority, the cloud service module 3 recalculates the priority of all vehicles to be charged in the area every 1 minute and updates the queue sequence.

[0057] The operation mode of the real-time power dynamic adjustment module 32 is specifically as follows:

[0058] The real-time power dynamic adjustment module 32 establishes a two-way communication link with the power grid dispatching system to obtain the load information of the power grid in real time, including the current load, load growth rate and peak and valley period identification, and defines the period from 0:00 to 6:00 at night as the valley period. If the charging pile has not reached full load, the charging pile is allowed to operate at maximum power. At the same time, preferential information on low-price charging periods is pushed to car owners to reduce energy waste. 9:00 to 11:00 in the morning and 2:00 to 4:00 in the afternoon are defined as off-peak periods. When the utilization rate of the charging pile exceeds 70%, the power is reduced for vehicles with priority A3 by no more than 20%. For vehicles with priority A2, the maximum power is maintained. 7:00 to 9:00 in the evening is defined as the peak period. When the power grid load reaches the warning value, the power is gradually reduced for all vehicles except A1, and the maximum can be reduced to 50%. In the power adjustment process, soft start and soft stop technology are used to improve system reliability and safety.

[0059] The real-time power dynamic adjustment module 32 uses soft start and soft stop technology in combination with a fuzzy control algorithm, as follows:

[0060] The battery voltage, battery temperature and current charging current of the vehicle battery are collected in real time through the communication interface between the charging pile and the vehicle battery management system. The collected data is filtered, the load change rate of the power grid is obtained from the two-way communication link, and two input variables and one output variable of the fuzzy control algorithm are determined. The input variable is the battery status. and grid load change rate , the output variable is the power change rate , for battery status , define the fuzzy set as {low, medium, high}, corresponding to the battery being in a poor, general and good state, respectively. , define the fuzzy set as {negative large, negative small, zero, positive small, positive large}, reflecting the decreasing or increasing trend and degree of the power grid load, and the output variable power change rate , define the fuzzy set as {extremely slow, slow, moderate, fast, extremely fast}, and then formulate fuzzy reasoning rules based on historical experience. According to the currently collected and fuzzified values ​​of battery status and grid load change rate, calculate their membership to the fuzzy sets defined respectively, match the membership of the input variables with the fuzzy reasoning rules, find out the activated rules, calculate the rule strength for each activated rule, calculate the membership of the output variable to the fuzzy set to which it belongs according to the conclusion part and rule strength of the activated rules, and calculate the membership of all activated rules, and finally obtain the comprehensive membership distribution of the output variables on each fuzzy set, as follows:

[0061] For the currently collected battery status and grid load change rate, substitute their respective membership functions, calculate their membership to each fuzzy set, match the membership of the input variable with the fuzzy inference rule, find out the rules whose input variable membership in the premise is not zero, that is, the activated rule, for the activated rule, use the smaller operation to calculate the rule strength, and calculate the membership of the output variable to the fuzzy set to which it belongs according to the conclusion part and rule strength of the activated rule, which provides a reliable basis for subsequent precise control, and finally use the centroid method to convert the fuzzy set membership of the output variable into the power change rate value. , the calculation formula is as follows: ;in is the representative value of the power change rate fuzzy set, yes For The membership degree of a fuzzy set is is the number of fuzzy sets, is the power change rate, and the power change rate obtained by defuzzification is , the output power of the charging pile Real-time adjustment is performed during the soft start phase, from the initial power Start by following Gradually increase the power. During the soft stop phase, follow the Gradually reduce the power, The output power of the charging pile is adjusted in real time at time intervals to ensure the stability and safety of the charging process.

[0062] The operation of the cross-site collaborative power scheduling module 33 is as follows:

[0063] The charging pile area is managed in a grid manner, with each grid containing multiple charging pile sites. A communication network is established within the grid to achieve data sharing and collaborative work between sites, improve the utilization rate of charging piles, and enhance users' charging experience.

[0064] When the charging pile utilization rate of a station exceeds the saturation threshold and there are queued vehicles, it means that the charging resources of the station are already tight. The idle charging pile resources in the surrounding grid can be searched, and the queued vehicles can be guided to other stations with idle resources to achieve optimal resource allocation and set the saturation threshold of the charging pile utilization rate. , real-time monitoring of the charging pile usage rate at each station ,in It is a site The number of charging stations in use, It is a site The total number of charging piles at the site of When there are queuing vehicles, the station is used as the center to search for grids within a certain range around it, and the number of idle charging piles at each station in the surrounding grid is obtained through the data sharing platform. , Indicates the number of surrounding sites, filter out The sites are used as available idle charging pile resources to reduce the queuing time of vehicles, the path from the current site to the idle site is calculated, and traffic congestion and estimated arrival time are taken into account. The geographical location information of the current site and the idle site and the traffic congestion of the road are obtained by using map data and real-time traffic data. The path planning algorithm is used to calculate the path from the current site to each idle site. When calculating the path, the distance of the road and the traffic congestion coefficient are used as weight factors. According to the weight of the path and the driving speed of the vehicle, the estimated arrival time is calculated, and the path with the shortest estimated arrival time is selected as the optimal path. The path information is sent to the queued vehicles, so that the vehicles can reach the idle site for charging faster, further shortening the total charging time and improving user satisfaction. The fault risk level of the power grid is monitored in real time. If the risk level exceeds the preset threshold, the local emergency protection mechanism is immediately activated, the charging pile power is adjusted, and a communication connection is established with the power grid dispatching system. The relevant parameters of the power grid, including voltage, current and frequency, are obtained in real time. A fault risk assessment model is constructed, the power grid parameters are analyzed, and the fault risk level of the power grid is calculated. , and set the preset threshold of the fault risk level ,when When the local emergency protection mechanism is activated immediately, the power of all charging piles is reduced by a certain proportion. Suppose the current charging pile power is , then the adjusted power ,in It is the power reduction ratio that ensures the safety of users and equipment and reduces economic losses.

[0065] When searching for idle charging pile resources in the surrounding grid, the cross-station collaborative power scheduling module 33 uses the ant colony algorithm for path planning, as follows:

[0066] In order to use the ant colony algorithm for path planning, it is necessary to abstract the actual charging pile sites and road connections into a graph structure, regard the charging pile sites as nodes in the graph, and the road connections between nodes as edges. The weight of the edge is determined by the distance, traffic congestion level and expected arrival time. Set the number of ants A and the pheromone volatility coefficient , pheromone heuristic factors and the expectation heuristic factor , initialize each edge Pheromone concentration , Expressed as Charging station, each node Contains the number and location coordinates of idle charging piles, Expressed as There are charging stations, and ant A is randomly placed at the starting station. When ants At the node When the ants follow the probability Select the next node until you reach an idle charging station , complete a path construction, after all ants complete the path construction, for each edge , update pheromone concentration ,in is the current iteration number. Finally, the pheromones added by all ants on the edge are accumulated to obtain the new pheromone concentration of the edge. Then the maximum number of iterations is set. If the current number of iterations reaches the maximum number of iterations, the algorithm is terminated. The optimal path length obtained in each iteration is recorded. If the change in the optimal path length for n consecutive iterations is less than a threshold, the algorithm is considered to have converged and terminated.

[0067] After the algorithm terminates, the optimal path obtained in all iterations is selected as the optimal path from the current station to the idle station. The path information is provided to the queuing vehicles to guide the vehicles to charge at the idle stations. At the same time, the parking spaces at the idle stations are locked for the vehicles, which optimizes the allocation of charging pile resources and improves the service quality of the entire charging system.

[0068] In the present invention, various information of the charging pile is collected by the data acquisition and transmission unit 1 and transmitted to the edge computing unit 2. The edge computing unit 2 pre-processes the data, starts the emergency protection mechanism when the power grid fluctuates, and uploads the data and decision results. The cloud server unit 3 builds a cloud computing platform, and its data storage service stores data, the data analysis service mines the charging rules, and the algorithm model service deploys an intelligent adaptive power allocation algorithm, which includes a dynamic priority determination module 31, a real-time power dynamic adjustment module 32 and a cross-station collaborative power scheduling module 33, which are respectively used to determine the vehicle priority, adjust the power according to the power grid load, and guide the vehicle to charge across stations, which solves the problems of unreasonable power allocation and insufficient cross-station coordination in the operation and maintenance management of charging piles, and improves the utilization efficiency of charging piles.

[0069] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments. The above embodiments and descriptions are only preferred examples of the present invention and are not intended to limit the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. Intelligent operation and maintenance management system for new energy vehicle charging piles, characterized by: It comprises a data collection and transmission unit (1), an edge computing unit (2) and a cloud server unit (3); The data acquisition and transmission unit (1) deploys a sensor component on each charging pile, wherein the sensor component is used to collect electrical parameters, environmental parameters, and status information of the charging pile and the charging gun and the cable, and transmits the collected information to the edge computing unit (2); The edge computing unit (2) uses a data analysis model to pre-process information, and based on local real-time data, activates a local emergency protection mechanism when an instantaneous fluctuation occurs in the power grid, adjusts the power of the charging pile, and synchronously uploads the pre-processed data and the local decision results to the cloud server unit (3); The cloud server unit (3) constructs a cloud computing platform, which includes multiple service modules, wherein the data storage service uses a distributed database to store charging pile data, the data analysis service uses big data processing technology to analyze the data and extract the charging demand patterns in different areas and time periods, and the algorithm model service deploys an intelligent adaptive power allocation algorithm model, the algorithm model includes a dynamic priority determination module (31), a real-time power dynamic adjustment module (32) and a cross-station collaborative power scheduling module (33). The dynamic priority determination module (31) establishes a vehicle priority evaluation model to determine the vehicle priority, the real-time power dynamic adjustment module (32) establishes a two-way communication link with the power grid dispatching system, sets a power adjustment strategy, and the cross-station collaborative power scheduling module (33) performs grid management on the charging pile area and uses an ant colony algorithm for path planning to recommend queued vehicles to go to idle stations for charging.

2. The intelligent operation and maintenance management system for new energy vehicle charging piles according to claim 1 is characterized by: The operation of the edge computing unit (2) when running the lightweight data analysis model is as follows: After receiving the data from the charging pile and the vehicle, the decision tree-based classification algorithm is used to preliminarily classify the data into electrical parameter category, environmental parameter category and vehicle battery information category; For electrical parameter data, an autoencoder-based anomaly detection algorithm is used to calculate the reconstruction error. If the reconstruction error is greater than the preset electrical parameter anomaly threshold, it is determined to be an electrical anomaly and the local emergency protection mechanism is immediately activated. For environmental parameter data, the environmental risk assessment model based on support vector machine is used to determine whether the current environment is within the normal operating range of the charging pile. If it is beyond the range, the risk coefficient is calculated based on the current usage status of the charging pile. If the risk coefficient is greater than the preset threshold, the local emergency protection mechanism is immediately activated; For vehicle battery information data, a battery health prediction model based on a multi-layer perceptron is used to predict the battery health status. If the prediction result shows that the battery health status is lower than the preset health threshold, battery maintenance suggestions will be pushed to the owner when the owner uses the charging pile.

3. The intelligent operation and maintenance management system for new energy vehicle charging piles according to claim 1 is characterized in that: The dynamic priority determination module (31) establishes a vehicle priority evaluation model to determine the operation of vehicle priority, specifically as follows: The priority is determined based on the vehicle type, the remaining battery power, and the user's urgency. Among the vehicle types, the priority of operating vehicles is higher than that of private cars, and the priority of emergency rescue vehicles is at the top. The priority of vehicles with a remaining battery power of less than 20% is A1, and the priority of vehicles with a remaining battery power of more than 80% is A3. The system determines the user's urgency based on the owner's historical charging behavior analysis. The priority of the owner's vehicle in an emergency is A2, where A2 is less than or equal to A1 and greater than A3; The priority of all vehicles waiting to be charged in the area is recalculated every minute based on the above factors. When there is an instantaneous fluctuation in the power grid, the convolutional neural network is used to extract real-time features of the power grid fluctuation data and the charging pile fault characteristic data, and input them into the risk assessment model built based on the long short-term memory network to determine the fault risk level. If the risk level exceeds the preset threshold, the local emergency protection mechanism is immediately activated to adjust the charging pile power.

4. The intelligent operation and maintenance management system for new energy vehicle charging piles according to claim 1 is characterized in that: The operation mode of the real-time power dynamic adjustment module (32) is specifically as follows: The real-time power dynamic adjustment module (32) establishes a two-way communication link with the power grid dispatching system to obtain the load information of the power grid in real time, including the current load, load growth rate and peak and valley time period identification, and sets different power adjustment strategies according to the power grid load situation and the use status of the charging piles in the area; The period from 0:00 to 6:00 at night is designated as the off-peak period. If the charging pile has not reached full load, the charging pile is allowed to run at maximum power. At the same time, preferential information on low-price charging periods is pushed to car owners. The period from 9:00 to 11:00 in the morning and from 2:00 to 4:00 in the afternoon is designated as the off-peak period. When the utilization rate of the charging pile exceeds 70%, the power of vehicles with priority A3 is reduced by no more than 20%. For vehicles with priority A2, the maximum power is maintained. The period from 7:00 to 9:00 in the evening is designated as the peak period. When the grid load reaches the warning value, the power of all non-emergency charging vehicles is gradually reduced to a maximum of 50%. In the process of power adjustment, soft start and soft stop technology are used.

5. The intelligent operation and maintenance management system for new energy vehicle charging piles according to claim 1 is characterized in that: The operation of the cross-site collaborative power scheduling module (33) is as follows: Carry out grid management for charging pile-dense areas. Each grid contains multiple charging pile sites. Establish a communication network within the grid to achieve data sharing and collaborative work between sites. When the charging pile utilization rate of a certain station exceeds the saturation threshold and there are queuing vehicles, the idle charging pile resources in the surrounding grid are searched, the path from the current station to the idle station is calculated, and traffic congestion and estimated arrival time are taken into account. The fault risk level of the power grid is monitored in real time. If the risk level exceeds the preset threshold, the local emergency protection mechanism is immediately activated to adjust the charging pile power.

6. The intelligent operation and maintenance management system for new energy vehicle charging piles according to claim 3 is characterized by: The dynamic priority determination module (31) establishes a vehicle priority evaluation model to determine the vehicle priority, taking into account the vehicle's cruising range requirements, as follows: The cloud service module (3) establishes a communication connection with the battery management system of the vehicle to obtain the battery capacity, remaining power, vehicle energy consumption information and the travel plan of the owner in real time. The vehicle energy consumption information is the average energy consumption obtained by dividing the total energy consumption of the vehicle by the total mileage in the past week. The travel plan of the owner is the destination information. The cloud system converts the destination information into geographic coordinates and calculates the estimated travel distance from the current location to the destination in combination with the map data. The remaining range of the vehicle is calculated based on the obtained vehicle battery capacity, remaining power and vehicle energy consumption information, and the calculated remaining range is compared with the estimated driving distance in the owner's travel plan. If the remaining range is less than the estimated driving distance, a fixed score is added to the original priority score of the vehicle. After the vehicle priority is adjusted, the cloud service module recalculates the priority of all vehicles to be charged in the area according to the updated priority every 1 minute, and updates the queue sequence.

7. The intelligent operation and maintenance management system for new energy vehicle charging piles according to claim 4 is characterized by: The real-time power dynamic adjustment module (32) combines the fuzzy control algorithm when using the soft start and soft stop technology, specifically as follows: Through the communication interface between the charging pile and the vehicle battery management system, the key status parameters of the vehicle battery, including battery voltage, battery temperature and current charging current, are collected in real time, and the collected data is filtered; Obtaining the load change rate of the power grid from a two-way communication link established with the power grid dispatching system, and determining two input variables and one output variable of the fuzzy control algorithm, wherein the input variables are the battery state and the power grid load change rate, and the output variable is the power change rate; For the battery status, the fuzzy set is defined as {low, medium, high}, corresponding to the battery being in a poor, general and good state respectively. For the grid load change rate, the fuzzy set is defined as {negative large, negative small, zero, positive small, positive large}, reflecting the downward or upward trend and degree of the grid load. For the output variable power change rate, the fuzzy set is defined as {very slow, slow, moderate, fast, very fast}, and then the fuzzy reasoning rules are formulated based on historical experience; According to the currently collected and fuzzified values ​​of battery status and grid load change rate, their memberships to their respective defined fuzzy sets are calculated respectively, the memberships of input variables are matched with fuzzy reasoning rules, the activated rules are found, and the rule strength of each activated rule is calculated. According to the conclusion part and rule strength of the activated rules, the membership of the output variable to the fuzzy set to which it belongs is calculated, and the membership of all activated rules is calculated, and finally the comprehensive membership distribution of the output variable on each fuzzy set is obtained; Finally, the centroid method is used to convert the fuzzy set membership of the output variable into a power change rate value, and the output power of the charging pile is adjusted in real time according to the power change rate obtained by defuzzification.

8. The intelligent operation and maintenance management system for new energy vehicle charging piles according to claim 5 is characterized by: The cross-station collaborative power scheduling module (33) uses an ant colony algorithm to perform path planning when searching for idle charging pile resources in the surrounding grids, as follows: The charging pile stations are regarded as nodes in the graph. Each node contains the idle status information and location information of the charging pile. The road connections between nodes are regarded as edges. The weight of the edge is determined by the distance, traffic congestion level and estimated arrival time. Set the number of ants, pheromone volatility coefficient, pheromone heuristic factor, and expected heuristic factor, initialize the pheromone concentration on each edge, and randomly place A ants at the charging station where there are currently queued vehicles, that is, the starting station. When each ant is at a certain node, it selects the next node according to the probability formula. The ant selects the next node according to the above probability until it reaches an idle charging station, that is, the target node, completing a path construction; After all ants have completed the path construction, the pheromone of each edge is volatilized and updated. Finally, the pheromone added by all ants on the edge is accumulated to obtain the new pheromone concentration of the edge. Then the maximum number of iterations is set. If the current number of iterations reaches the maximum number of iterations, the algorithm is terminated. The optimal path length obtained in each iteration is recorded. If the change in the optimal path length for n consecutive iterations is less than a threshold, the algorithm is considered to have converged and terminated. After the algorithm terminates, the optimal path obtained in all iterations is selected as the optimal path from the current station to the idle station, and the path information is provided to the queued vehicles to guide the vehicles to charge at the idle station, while locking the parking space at the idle station for the vehicles.

Images