Collision avoidance system for vehicles applying model predictive control theory

One of the most important issues in controlling vehicles automatically is collision avoidance. A collision avoidance system (CAS) should cause the vehicle to stop reliably when a preceding vehicle stops suddenly on the road or when the CAS detects an obstacle. The paper proposes a collision, avoidance algorithm applying model predictive control (MPC) theory. The CAS proposed has the following features: (1) prediction of the future behavior of a vehicle, (2) good ride comfort preventing unnecessary brake reference change. A simple and high-performance CAS can be achieved using the proposed algorithm. Computer simulations have confirmed its high performance.