Optimal dispatching of electric vehicles based on optimized deep learning in IoT

Recent years have witnessed a growing trend in the utilization of Electric Vehicles (EVs), however with the increased usage of EVs, appropriate strategies for supporting the charging demands has not garnered much attention. The absence of adaptable plans in charging may result in minimized participation; further, the charging demands have to be addressed with utmost care for ensuring reliability and efficiency of the grid. In this paper, an efficient EV charging technique based on blockchain based user transaction and smart contract is devised. Here, charge scheduling is performed by acquiring the information the charging demand of the EV over Internet of things. In case the EV does not have sufficient power to reach the target, nearest Charging Station (CS) with the minimal electricity price is identified. The CS is selected considering various factors, such average waiting time, distance, power, traffic, and so on. Here, power prediction is performed using the Deep Maxout Network (DMN), whose weights are adapted based on the devised Exponentially Snake Optimization (ESO) algorithm. Furthermore, the efficacy of the devised ESO-DMN is examined considering metrics, like average waiting time, distance, and number of EVs charged and power and is found to have attained values of 1.937 s, 13.952 km, 55 and 2.876 J.