Design of an autonomous sentry robot prototype and realization of target tracking control with LQR based fuzzy-neural networks

Protecting border lines or sensitive security areas is an important problem today. The use of sentry robots for this purpose provides some potential advantages. It is critical to precisely track the target detected by the sentry robot and the systems evaluated in this category. Permanent magnet synchronous motors (SMSM) used in the drive of axis movements were preferred in this study due to their high efficiency, high power density, and robust structure. However, the nonlinear dynamics and load torque disturbances of these motors restrict the precise control of these motors. In this thesis study, a sentry robot prototype that can track targets in two axes and uses SMSMs in the axis drive was designed and sensitive with a LQR based ANFIS controller. position control has been carried out. In addition, in order to minimize distortions caused by load torque, an ANFIS-based load torque observer has been designed and the control system has been made more durable. ANFIS-based controller and torque observer designs were realized with the data obtained from simulation studies in Matlab/SIMULINK environment for different LQR controller structures. In order to evaluate the target tracking performance of the system, axis movements were examined according to the reference position information entered in two axes. Subsequently, the location information of the target was obtained in real time through the face tracking algorithm of the Matlab/SIMULINK Open CV plug-in, and the sentry robot was successfully able to track the target with high precision according to the location information.