Identification of Parameters of a Model of a Movable Motion Platform

Abstract

We use specialized simulators to train drivers of vehicles difficult to drive. The environment created by such simulator for drivers must be as close to reality as possible. To simulate the overloads that occur during the motion of the simulated object, movable platform stands are used. Usually simulator software is developed for a particular imitation stand, which limits the use of such software for other equipment configuration. In addition, often the sensors in the mechanisms of the platform are either not precise enough or absent altogether, and the geometric parameters of the platform change with time due to deformations due to constant loads. These factors negatively affect the management of the platform and, therefore, the accuracy of dynamic simulation and the quality of coordination with visual and other types of simulation.The purpose of this paper is to show that identification of platform parameters and platform positioning can be constructed without using measurements from internal sensors located in the mechanisms of the platform. To solve the problem, the authors developed an algorithm of semi-automatic identification of parameters of a platform model with a system of video analysis. Upon identification of parameters of a platform it is possible to monitor its angular motions without video analysis. Assessment of platform orientation is performed with angular velocity sensors (AVS) and accelerometers. The use of the suggested algorithms enables quick adaptation of stimulator software to any motion platform.