MODEL PREDICTIVE CONTROL BASE ON AT89S52 USING ALGORITHM-2 FIXED POINT

• Main Author: PRABOWO WIDIGDO (NIM 13203085), DITTO
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/10263
• Institution: Institut Teknologi Bandung

<p align="justify">Model Predictive Control is based on the idea to produce control input as solution of real-time optimization problem. Optimization is done based on the system model. MPC as an advanced control technology has been widely implemented in industry plants which generally have slow respond. MPC is used to solve multivariable control problem which has more than one variable that may have significant effect to the process. Another advantage of MPC showed in this final project is that MPC works effectively when facing constraints in the actuator which are relatively narrow.<p align="justify"><p>The disadvantage of MPC is on its complex algorithm that needs longer time than the other controller algorithm like PID. This final project will show that a simpler Model Predictive Control algorithm may be implemented on microcontroller, i.e. MCS51 microcontroller. The algorithm used is iteration fixed point quadratic programming. This MPC algorithm will be compared to the other algorithms which have been used generally such as Proportional Integral (PI) and Proportional Integral Anti-Windup.<p align="justify"><p>There are some steps done to finish this final project. Firstly we simulate the ideal MPC using matrix based software. Then implementing the MPC algorithm on microcontroller using simulation plant in computer. The goal of this final project is to implement the MPC algorithm on microcontroller to control the speed of motor. The results of the controllers implementation show that MPC controller gives the best output response. <br />