Optimizing Charging Strategies for Electric Vehicles in Residential Areas Based on Neiderreit Sequence Initialized Tasmanian Devil Optimization

Electric vehicles (EV) have attracted the attention of many consumers due to their energy-saving and environmentally friendly characteristics. However, as the number of EVs continues to increase, a large number of EVs, when randomly connected to the power distribution network through charging stations, will cause a large-scale uncertain electricity load burden on the power distribution network. Therefore, it is meaningful to study the proper way to manage the increasing charging behavior of EVs and achieve their orderly charging. In recent years, scholars have been particularly inclined to describe the optimization of EV charging strategies as a multi-objective optimization problem. However, a considerable number of EV charging strategy models have an objective function that is essentially L1 norm, which is a non-smooth optimization problem. In order to effectively optimize its parameters, in this paper, an algorithm that does not rely on gradient information for global optimization has been proposed, which is called Neiderreit sequence initiated tasmanian devil optimization (NSTDO). NSTDO can overcome the shortcomings of commonly used coordinate descent (CD) method and sub gradient descent (SubGD) method, which are prone to falling into the trap of local minima and unable to find the global optimal solution. Via numerical experiment, the effectiveness of the proposed NSTDO algorithm has been verified.