A non-linear MPC based motion cueing implementation for a 9 DOFs dynamic simulator platform.

The use of dynamical driving simulators is nowadays common practice in many different application fields, such as driver training, vehicle development, and medical studies. Platforms with different mechanical structure are designed, depending on the particular application and the corresponding targeted market. The effectiveness of such devices is related to their capabilities of well reproducing the driving feelings, hence it is crucial that the motion control strategies generate both realistic and feasible signals to the platform, to assure that it is kept within its limited operation space. Such strategies are called Motion Cueing Algorithms (MCAs), and they are clearly tailored to the particular mechanical structure of the device. In this paper we describe an MCA based on non linear Model Predictive Control (NMPC) techniques for a simulator of new conception, that consists of an hexapod over a flat base moved by a tripod, thus exhibiting highly non linear behaviour. The procedure is based on previous works where a linear, MPC-based MCA has been applied to a simpler device. The algorithm has been evaluated on a simulation environment, and a first implementation on the real device is in progress. Preliminary results show that a full exploitation of the working area is achieved, while managing at best all the limitations given by the particular structure and preserving the ease of tune and intuitiveness of the linear approach.