Adaptable Emergency Braking Based on a Fuzzy Controller and a Predictive Model

Abstract

This work presents the implementation of an adaptable emergency braking system for low speed collision avoidance, based on a frontal laser scanner for static obstacle detection, using a D-GPS system for positioning. A fuzzy logic decision process performs a criticality assessment that triggers the emergency braking system and modulates its behavior. This criticality is evaluated through the use of a predictive model based on a kinematic estimation, which modulates the decision to brake. Additionally a critical study is conducted in order to provide a benchmark for comparison, and evaluate the limits of the predictive model. The braking decision is based on a parameterizable braking model tuned for the target vehicle, that takes into account factors such as reaction time, distance to obstacles, vehicle velocity and maximum deceleration. Once activated, braking force is adapted to reduce vehicle occupants discomfort while ensuring safety throughout the process. The system was implemented on a real vehicle and proper operation is validated through extensive testing carried out at Tecnalia facilities.