A Driving Decision Strategy (DDS) by Using Genetic Algorithm for an Autonomous Vehicle

Abstract

A modern-day self-sustaining car determines its driving method by means of thinking about solely exterior factors (Pedestrians, street conditions, etc.) barring thinking about the interior circumstance of the vehicle. To overcome above problems, in this paper author proposed a new strategy i.e A Driving Decision Strategy(DDS) Based on Machine learning for an autonomous vehicle” Analysis of both external and internal factors determines the optimal strategy for an autonomous vehicle (consumable conditions, RPM levels etc.). To implement this, the project author has introduced and algorithm called DDS (Driving Decision Strategy) algorithm which is based on genetic algorithm to choose optimal gene values which helps in taking better decision or prediction. DDS algorithm obtained input from sensor and then passes to genetic algorithm to choose optimal value which helps in faster and efficient prediction. Propose DDS with genetic algorithm performance is comparing with existing machine learning algorithm such as Random Forest and MLP (multilayer perceptron algorithm.). Propose DDS shows better prediction accuracy compare to random forest and MLP.