Position estimation for timing belt drives of precision machinery using structured neural networks

This paper focuses on a viable position estimation scheme for timing-belt drives using artificial neural networks. In this study, the position of a carriage (load) is calculated via a structured neural network topology accepting input from a position sensor on the actuator side of the timing belt. The paper presents a detailed discussion on the source of transmission errors. The characteristics of the error in different operation regimes are exploited to construct different network topologies. That is, a relevant neural network model is developed by the sketchy guidance of a priori knowledge on the process. The resulting structured neural network is shown to estimate the error of the carriage quite accurately whereas generic recurrent neural networks fail to capture the dynamics of the system under investigation altogether. Extensive testing demonstrates the effectiveness of proposed method when the drive system is not subjected to external loads while the operating conditions such as ambient temperature and belt tensions do not deviate from the experimental conditions.

► Position estimation scheme for timing-belt drives using artificial neural networks is focused. ► The characteristics of the error pattern in different operation regimes are exploited to construct different network topologies. ► A number of relevant neural network models are structured utilizing a priori knowledge on the process. ► The resulting structured neural network is shown to estimate the error of the carriage accurately. ► It is seen that a generic neural network's performance is inferior to that of the presented network.