Position control and simulation in mobile robots

Today, the PID development of technology has increased the use of mobile robots in various areas such as health, agriculture, industry and logistics. For the successful implementation of these systems, position control is of critical importance in terms of the robot's environmental interactions and task performance. In particular, providing accurate and efficient control allows mobile robots to perform complex tasks reliably. In this thesis, the location control problem, which is the cornerstone of autonomous locomotion capability of mobile robots, is reconsidered at both a theoretical and applied level, and AI-based control strategies are systematically compared under different simulation scenarios. Firstly, the kinematic and dynamic behaviors of mobile robots are analytically modeled, and various control algorithms (PID, PD, fuzzy logic, DDPG and DQN) are applied on the nonlinear model derived for differential drive systems. The effectiveness of the control systems is generally compared on the positioning error, orientation accuracy, timing and stability measures, and the parameter sensitivity and adaptation capacity of each method are analyzed. This thesis not only consists of re-implementing existing control algorithms, but also contributes to the generalization of control architecture and the development of adaptive strategies against dynamic environmental variations. In particular, the first steps towards the design of hybrid control systems have been taken by integrating deep reinforcement learning algorithms with classical controllers. The results show that, although machine learning based control strategies offer higher accuracy and environmental adaptation capability compared to classical methods, new challenges such as computational cost, data starvation and security challenges are introduced. In this context, the original contribution of the thesis is its methodological framework that includes multi-dimensional analysis of different control strategies and the holistic approach that considers the transferability of simulation results to the real world. September 2025, 81 pages. Keywords: Mobile robot, position control, PID control, simulation, fuzzy logic