Driver-truck models for software-in-the-loop simulations

Abstract: By using vehicle-to-vehicle communication, vehicles can cooperate in many waysby sending positions and other relevant data between each other. One popularexample is platooning where many, especially heavy vehicles, drive on a trailwith short distances resulting in a reduction of air resistance. To achieve a goodefficiency of the platooning it is required that vehicle fleets are coordinated, sothat the percentage of time for driving in platoon is maximized without affectingthe total driving time and distance too much. For large fleets, this is a complexoptimization problem which would be difficult to solve by only using the realworld as the test environment. To provide a more adaptable test environment for the communication and platooningcoordination, an augmented reality with virtual vehicles (“Ghost trucks")with relevant communication abilities are developed. In order to realise the virtualtesting environment for trucks, Scania initiated a project that could be dividedinto the workload of three master theses. This thesis involved the part ofdeveloping the virtual vehicles, which include the development of a truck modeland a driver model. The developed truck model consists of a single track vehicle model and severalpowertrain models of different complexity provided by Scania. Additionally, thedriver model consists of steering wheel and speed controls in order to keep thetruck on a safe distance from the lead truck and stay on a preferred lane. The keyfeature of the driver-truck model is its modular design, which provides great flexibilityin selecting the level of detail for each component. The driver-truck modelcan be duplicated and simulated together in real time and performs platooningwith each other in a road system based on the real world. As the driver-truckmodel is module based, it can easily be extended for future purposes with morecomplex functions. The driver-truck model is implemented in Simulink and the simulation performancefor different model complexity is evaluated. It is demonstrated that theflexibility of the developed model allows a balanced decision to be made betweenrealistic truck behavior and simulation speed. Furthermore, multi-truck simulationsare performed using the model, which demonstrate the effectiveness of themodel in the evaluation of truck platooning operations.