Optimization of Secured Cluster Based Charging Dynamics and Scheduling of EV Using Deep RNN

Abstract

In the near future, Electric Vehicles (EVs) are anticipated to develop into fantastic modes of transportation. Due to their limited range and under powered batteries, EVs are crucial for lowering the use of conventional fuel. When the battery charge is about to reach a critical level, it is essential to be aware of local Charging Stations (CS). As a result, we could spot two issues: (1) Secured Cluster based CS allocation and routing to CS (2) Scheduling vehicle at CS based on delay prediction. First, a Cluster based Vacant charging slot is searched in clustered charging stations using cloud and Vehicular Adhoc Network (VANET) model, along with evolutionary Social Ski Driven (SSD) optimized algorithm using Deep Recurrent Neural Network (DRNN) as a new optimal routing for EVs to reach CS based on established fitness function computing distance, battery power and traffic congestion. Second, at CS, vehicle time scheduling is done using the DRNN approach, considering delay-based distance computation. When compared to the stochastic Particle Swarm Optimization (PSO) algorithm for routing, the proposed DRNN-SSD routing algorithm optimizes delay and traffic congestion significantly achieving better successful allocation rate of CS during On-peak and Off-peak hours.