TOU Price based Optimal Scheduling of EV Clusters

The widespread adoption of Electric Vehicles (EVs) and their uncoordinated simultaneous charging puts additional stress on the grid due to an increase in peak load. The charging schedule of EVs can be coordinated by an EV Aggregator (EVA) through appropriate price signals. This paper presents an optimal scheduling framework for EV clusters using Time-of-use (TOU) price. Firstly, the EVA aggregates the total demand corresponding to the vehicle's arrival and departure times for optimal power scheduling using clustering algorithms. The accuracy of cluster-based aggregation plays a vital role, hence this study adopts the advanced clustering technique: spectral clustering algorithm, to accurately cluster the EVs. However, the optimal EV scheduling is motivated by dynamic prices like Real-Time Price (RTP) and Time of Use (TOU) but the acceptance rate of RTP is quite less due to its highly volatile nature. Hence, the proposed work focuses on the optimal scheduling of EV clusters based on TOU prices. For this, the historical data of RTPs are considered for TOU price design using hierarchical clustering. For the case study, 500 EVs are aggregated using spectral clustering and compared with the traditional k-means. The aggregation results are analyzed using Principal Component Analysis (PCA) decomposition; highlighting the increased accuracy of aggregation in terms of time. Further, the optimal scheduling of EV clusters is achieved based on the proposed pricing and aggregation strategies.