A daytime optimal stochastic energy management for EV commercial parking lots by using approximate dynamic programming and hybrid big bang big crunch algorithm

The design of optimal energy management systems has been an important problem, in particular for daytime plug-in electric vehicles' (PEVs') parking lots at workplaces and commercial buildings, where the number of vehicles, their arrival and departure times are typically unknown and time-varying. The paper addresses this problem by introducing a two-stage optimization based on Approximate Dynamic Programing (ADP) and Hybrid Big Bang Big Crunch (HBB-BC) algorithm considering a Multi-Layer Perceptron Artificial Neural Network (MLP-ANN) which predicts the electricity price. The proposed optimal energy management minimizes the cost of parking lot owner with respect to the Time of Use (TOU) Demand Response (DR) program without reducing the welfare of EV owners. The two-stage optimization manages the charge scheduling subject to short and long term data coming from the MLP-ANN and intelligent transportation system. The stochastic features of the commercial parking are fully addressed into the problem. The effectiveness of the proposed charging technology is assessed and discussed by using real electricity data from ERCOT, under various stochastic conditions. The results demonstrate optimal energy management during peak and off time periods at the minimum cost.