Multiple Classifier Based Highly Accurate Emergency Vehicles Management System for Real-time Traffic Control and Monitoring Applications

Abstract

Emergency vehicle management (EVM) plays a major role in daily life from lifesaving to military purposes. Emergency vehicles (EV) should reach the destination in very little time for better services in emergencies. The transition time of an emergency vehicle is directly proportional to the complexity of the transition path. Decision making in a real-time environment is a difficult task due to computational complexity and nonlinearity in the route parameters that make delay in the transmission of parameters to the central processing unit such as cloud servers. Similarly, decision making should be performed at the right time so that the EV can smoothly select the best path.  In this work, an emergency vehicle management by voting-based decision-making using multiple classification systems (EVMVBDM-MCS) is implemented to improve the performance of decision making in the real-time environment. Different route parameters such as distance between source to destination, traffic density, road width, and other linear and nonlinear route parameters also considered selecting the effective path of an emergency vehicle. EVMVBDM-MCS is implemented and tested using a real-time route parameter and found high accuracy when compared to the classical path prediction techniques. Support vector machine (SVM), Extreme Learning Machine (ELM), K-nearest neighbor (KNN) are used to predict the better route parallelly for the voting. Around 98% of accuracy is obtained using EVMVBDM-MCS for a real-time dataset for a particular route.