Comparison of Fail-Operational Strategies to Make Cooperative Adaptive Cruise Control Resilient to an Electronic Control Unit Failure

Abstract

With the tremendous increase of vehicles and limited infrastructure for its smooth movement, traffic jams have become a significant problem these days. To address this issue, significant research and development in the field of Intelligent Transportation System (ITS) is currently being carried out. One such technological development that allows grouping of vehicles into platoons, controlled by one leading vehicle has proved to be a fruitful and potential solution for the existing issue of traffic jams.
The state of the art technology that enables vehicle platooning is called Cooperative Adaptive Cruise Control (CACC). In CACC, the individual vehicles are grouped into platoons and allowed to automatically adjust their speeds using on-board sensors and vehicle to vehicle communication to maintain a desired and safe inter-vehicle distance. The specialty of CACC is that it enables to have small inter-vehicle distance between the vehicles which increases road throughput and reduces air drag on the vehicle. Consequently, traffic jams and fuel emissions are reduced.
However, a crucial challenge arrives due to its high dependency on mechatronic devices, as a fault in such devices could lead to unsafe conditions for the vehicles in a platoon. Hence, there is a need of making CACC application Fail-Operational, which means that the vehicles would continue to function safely under faults and failures of those devices. TNO in its EcoTwin III project has identified Electronics Control Unit (ECU) to be a very critical component and uses the concept of redundancy to provide the Fail-operational capability that can tolerate one failure of ECU. However, this implementation involves a transition period caused due to fault detection and switching of ECU. This transition period could jeopardize the safety of vehicles under conditions where a nominal ECU would have kept the vehicles safe.
Thus the objective of this MSc thesis is to address the problem of the transition period of a homogeneous platoon under ideal communication network conditions. In the first part, a model is proposed that captures the effect of the transition from primary ECU to the standby ECU on the platoon dynamics. This is done for both warm and hot standby strategies which depend on the functionality of the standby ECU. We see how this transition period affects the overall safety of the system under a worst-case scenario (emergency braking). In the second part, to improve the safety and exclude the effect of the transition period, a new method of implementation of existing control law is proposed. This technique exploits the benefit of using communicated data from the preceding vehicle for generating control input during the transition period.
Simulations are conducted for a two-vehicle platoon. The results show the effect of the transition period in a warm standby strategy will lead to collisions. The simulations also show that the hot standby strategy outperforms the warm standby one, though it cannot avoid collisions at large transition periods. On the other hand, the proposed new strategy shows the potential of using the communicated signal in preventing collision and shows significant improvements in comparison to other two strategies at the larger transition period.