Performance Analysis of Model Predictive Control for Electrical Quadriplegic Wheelchair

Abstract

Quadriplegic: Afflicted by partial or complete paralysis of both arms and legs, often as a result of disease or damage to the neck region of the spinal cord. A quadriplegic wheelchair that travels along a track must adhere to it exactly. The wheelchair chosen for this thesis has two front wheels and one caster situated in the back. The quadriplegic wheelchair's mathematical model includes the dynamics of two identical series DC motors, the wheelchair's chassis, and its kinematics because the model of the plant is crucial for developing the MPC. In MATLAB, the quadriplegic wheelchair model is simulated. For the purpose of improving position and orientation straight path line tracking, the influence of a change in body mass is regarded as a disturbance. This thesis in order to settled the unstability of position and orientation and trajectory track problems. Improve the performance of the quadriplegic wheelchair to track its position and orientation from the ideal track line position, this study focuses on building MPC (Model Predictive Control) for quadriplegic wheelchair operation in track. MATLAB simulates the two DC series motors located in each of the front wheels as well as the controller. The pinhole for the input component is defined. To limit the forward speed and the maximum angle or direction error, the quadriplegic wheelchair's position (x, y), orientation (a), and control signals (uL, uR) are all regulated. By simulation tests, the effectiveness of the suggested controllers for tracking position and orientation set points is assessed. The result showed that the closed loop system without consideration of the disturbance was very good to track the path with cost function =0. 653∗e -10. This shows that the controller tracks the desired path with small angular error -0.3∗10−10 degree. The other simulation result is control system that took disturbances into account. The cost function used to track the desired path was roughly j=0.653*e-10, but the angular error was -0.3*10-4 degrees. The lowest error between the desired and actual position and orientation is the lowest cost function. When working in a path line that follows the track due to track restrictions, the MPC technique is particularly favorable and exhibits improved performance.