关键词: 充电站选址, 可靠性, 中断情景, 带精英策略的免疫优化算法, 灵敏度分析

Abstract: With the popularization of electric vehicles, the reliability of charging station construction has become an important factor affecting the user experience and market development of electric vehicles. Reasonable planning of charging station location and improving the reliability of charging station services are key issues that need to be addressed in the layout planning of electric vehicle charging stations. This paper considers the scenario where charging stations have the risk of random interruptions and users have emergency charging needs, and constructs an optimization model for reliable location of charging stations to explore reliable location allocation schemes.
Firstly, charging stations are stratified according to their distance from demand points, with the charging scenarios of users divided into emergency and non-emergency types. It is assumed that the proportion of emergency users will not exceed a specific limit, thus the number of users in emergency charging situations is constrained by setting the proportion of emergency situations. Secondly, we consider the random interruption risk of charging stations, meaning that the interruption probability of each charging station is a random value within a specified range. Taking into account the reliability of charging stations under interruption scenarios, the diversity of user charging scenarios and the economic feasibility of construction costs, a reliability-based location optimization model for charging stations is constructed. Accordingly, reliable charging stations are searched based on the hierarchical relationship between demand points and corresponding charging stations: when users at demand points are in non-emergency charging scenarios, if the currently assigned charging station is interrupted, the next hierarchical level charging station will be searched, and if the demand point is not assigned to any hierarchical level charging station, it is considered a failed allocation; when users at demand points are in emergency charging scenarios, the charging station assigned to the demand point cannot exceed two hierarchical levels, and if the demand point is not assigned in the first two hierarchical levels, it is considered a failed allocation. Then, this paper designs an immune optimization algorithm with an elite strategy to solve the model. The algorithm introduces an elite strategy to improve search performance and convergence speed. The fitness value of individuals measures the optimization objective, which is the total cost. The elite individuals are those with lower fitness values, representing good feasible solutions with lower total costs. The iterative process of feasible solution updates highlights the process of survival of the fittest, ultimately determining the optimal location-allocation scheme. Finally, the comparative analysis of the location-allocation of demand points and charging stations at different scales is conducted, and a sensitivity analysis of the interruption probability and emergency situation proportion parameters is performed.
This paper validates the effectiveness of the location model and algorithm for charging stations through case analysis. The study solves the location model for different demand points and alternative charging station capacities, obtaining reliable optimal location-allocation schemes. The experimental results demonstrate that, in scenarios where charging stations face random interruption risks and users have urgent charging needs, the immune optimization algorithm with an elite strategy can quickly find reliable location-allocation schemes while reducing total costs as much as possible to meet all user demands. In different scale experiments, the trend of the optimal fitness value is consistent, initially decreasing and then reaching a balanced state. The sensitivity analysis results indicate that the interruption probability is an important influencing parameter for the optimal fitness value and charging station location-allocation. As it increases or decreases, the optimal fitness value also increases or decreases accordingly. However, the variation of the emergency situation ratio does not show a clear impact on the results.

Key words: charging station location, reliability, disruption scenarios, immune optimization algorithm with elite strategy, sensitivity analysis