Abstract
A Cooperative Adaptive Cruise Control (CACC) system normally consists of different operation modes such as CC (in case no object is detected), ACC and/or CACC. The objective of this study is to find a method for control reconfiguration which ensures a stable, safe and comfortable switching between these operation modes. The approach of mixing control outputs is based on a Multiple-Model Adaptive Control with Mixing (MMACM) method which reconfigures the control of an uncertain plant by making a linear combination of candidate controllers’ signals; the weighting factors for the linear combination are dependent on the estimated unknown plant parameters. This approach is adapted to implement control reconfiguration between stabilizing controllers for a known plant. The stability of this method is proven for a convex combination of the stabilizing controllers’ signals. The stability is investigated for constant and time-varying weighting factors in the controller combination. Boundary conditions for stability are found. Mixing functions are defined as weighting factors such that these fulfil the boundary conditions for stability and such that these enable a smooth transition. The transition time is used to introduce the notions of safety and comfort. The method is evaluated with experiments. The experiments demonstrate the control reconfiguration for a vehicle cut-in scenario which represents a transition from CC to CACC. Test results show a smooth acceleration profile which fulfils the safety and comfort requirements.
KEYWORDS – cooperative driving, fault tolerance, control reconfiguration, safety, comfort.