| Abstrakt: |
This study aims to improve the control of robotic knee flexion during walking, with a particular emphasis on enhancing mobility and rehabilitation for patients with mobility problems. The objective is to develop a high-performance controller by integrating the Desired Optimal Controller (DOC)-based Multivariable Model Reference Adaptive Control (MRAC) algorithm with sophisticated optimization techniques. This study notably combines the Whale Optimization Algorithm (WOA) with a novel approach called Combined WOA-KHO to precisely optimize controller parameters. The technique provides a thorough explanation of the construction of the DOC-based MRAC algorithm, which employs a second-order transfer function for the reference model. This study emphasizes the inclusion of adaptive gains, the structural characteristics of the best controller, and the implementation of a deep neural network (DNN)-PID control system utilizing a Multi-Layer Feed-Forward Neural Network (MLFNN). In addition, this text elaborates on the optimization strategies, namely the employment of the Whale Optimization Algorithm (WOA) and the Combined WOA-KHO algorithm. The simulation results clearly demonstrate the gradual improvement of the system's performance, providing evidence for the effectiveness of the suggested DOC-based MRAC algorithm and the optimization approaches. An extensive examination of the system's response characteristics, such as settling time, rising time, and steady-state error, is performed using several simulations. A performance comparison is implemented between three optimization algorithms: Genetic Algorithm (GA), Particle Swarm Optimization (PSO), and WOA. The study finds that using all three algorithms together significantly improved the gait control of a robotic knee system, outperforming the results obtained from traditional algorithm. [ABSTRACT FROM AUTHOR] |
