Application of PID algorithms to control robotic platform using Simulink

Abstract

The usage of robots in the industry, research and hobby has increased during the years. In the world of academia robots present a perfect test bench for researchers. Many fields like robotics and automatics use them to solve many specific problems. The research area of robots is toward the implementation of new control techniques, new control and machine learning algorithms. Autonomous robots are the main trend in the modern research and industry area. This project is related to the use of MATLAB Simulink in the process of rapid prototyping control algorithms for a robotic platform. The Alphabot2-Pi is used as a research platform. The control algorithms are implemented as software processes started in real-time in the operating system environment of the robot. In this study, Simulink models are developed using standard and non-standard blocks. The system is implemented as the hardware, software and communication tools are used. Simulink toolbox like Embedded Coder is used for automatic code generation, compilation and deploying to the robotic platform. The standard and nonstandard (Fuzzy) PID algorithms are tested to perform and implement the control of the platform steering angle to travel along the specified path. The improvement of PID algorithms is done by varying the coefficients using some heuristic methods. The implementation, configuration and testing of the proposed system capabilities are investigated in a real environment. The study shows that the system, with varying success, is capable of performing the proposed tasks using different control parameters. The results obtained with standard and nonstandard (fuzzy) PID algorithms are satisfactory. Both algorithms solved the proposed task, but the standard PID control algorithm turned out to be better for high-speeds than the fuzzy PID.