SBIR-STTR Award

This work will address the problem of deviations of unmanned vehicle convoy followers by developing a method to estimate the parameters of the disturbances that cause lateral deviations from the desired vehicle trajectory. These parameters will be estimated each time a vehicle passes over a section of terrain. These spatially dependent estimates of the disturbance parameters will then be passed to the next vehicle in the convoy line. This next vehicle will then use the estimated disturbances to better drive the current section while also refining the estimated disturbance parameters. Each time a vehicle passes over the same section of terrain, the parameter estimations will be improved and the lateral deviation of each successive follower will be reduced. This will not only reduce the lateral deviations of the rear vehicles in the convoy but will also ensure that the vehicles in the rear of the convoy have better parameter estimates than those that went before.

Keywords:
Autonomous Convoy, Autonomous Vehicle, Path Controller, Terrain Disturbance Estimation, Model Predictive Control, Kalman Filter, Terrain Modeling, Automotive Sensors

Unmanned automated convoys provide an efficient means for transporting the cargo of many vehicles across the same terrain to reach a desired location. As more vehicles are added, the efficiency of the convoy can increase; however, if each successive vehicle uses the previous vehicle for path following, the convoy accuracy will gradually degrade. In order to improve the path following of successive vehicles, the phase I of this project showed that use of disturbance and off-path error information can improve the accuracy of successive vehicles by passing this information to the following vehicle. The phase II of this project will continue to build on this idea, utilizing more general disturbance estimations. Additionally, improving the dynamic models of the vehicles and using preview information of the path with AMAS sensors will improve prediction steps of the Kalman filter and path control in general. This project will utilize ASIs large assortment of equipment and vehicles with expertise from Auburn University and their AMAS sensor suite to ensure success of this project. Successful completion of this project will provide benefits to both manned and unmanned ground vehicles. Commercialization efforts will be an important part of this project.