A Driving Decision Strategy (DDS) Based on Machine learning for an autonomous vehicle

Abstract

Currently, an independent car's driving method is chosen based on external criteria (pedestrian crossings, road surfaces, etc.) without considering the car's interior state. “A Driving Decision Approach (DDS) Based on Machine Learning for an Autonomous Vehicle” predicts the proper approach for an autonomous vehicle by searching outside and inside factors. The DDS trains a genetic set of rules that develops an autonomous car's best use method using cloud-based sensor information. The proposed DDS with rules compares to Random Forest and MLP (multilayer perceptron set of rules). Precise DDS beats random forest and MLP. This study compared DDS to MLP and RF neural community models. The DDS had a 5% lower loss rate than conventional car gateways in the study, and it computed Revolutions per minute, speed, direction angle, and converting lanes 40% faster than the MLP and 22% faster than the RF neural networks. DDS provides sensor records to a genetic collection of rules, which chooses the most acceptable value for extra unique prediction.