Motion control logic for large-excursion driving simulators

A new approach to motion control logic was developed, starting with a linear analysis of the Classical Washout Algorithm. Using the analysis, a bang-bang minimum time control was designed to control tilt coordination. The minimum time control allowed the removal of the highpass filters on linear acceleration, reducing specific force false cues. The algorithm was extended to six degrees of freedom assuming a system configuration of an unlimited yaw ring, on a hexapod, on a +/−10 meter excursion x-y track. Methods of prepositioning the motion base laterally based on upcoming curves and longitudinally based on speed were developed. In addition, a method of predicting lateral acceleration 0.1 seconds into the future was formulated. The prepositioning and prediction were used to enhance the performance of the minimum time control.; The new algorithm provides higher performance in both yaw rate and specific force when compared to the Classical Algorithm. The new algorithm was tested using data collected on the Iowa Driving Simulator. Vehicle linear acceleration and angular rate data from twenty-four subjects were rerun through the new algorithm. The RMS errors between the specific forces and angular rates generated by the algorithm and the original vehicle data were calculated. RMS errors for specific force were less than 0.09 m/s/s. RMS errors for yaw angular rate were less than 0.01 rad/s. RMS errors for roll and pitch angular rate were less than 0.04 rad/s. Using models of the vestibular system, the RMS error between perceived specific forces and angular rates generated by the algorithm and perceived specific forces and angular rates of the original vehicle were calculated. RMS errors for perceived specific force were less than 0.06 m/s/s. RMS errors of perceived yaw angular rate were less than 0.001 rad/s. RMS errors of perceived pitch and roll angular rate were less than 0.005 rad/s. For 22 of the 24 runs, the motion base remained within its +/−10 meter excursion limit. For two runs, the motion base strayed up to 1.3 meters outside the excursion limit. It was found that poor longitudinal prepositioning caused the excursion limit to be exceeded in both runs.