Model optimization based on Hybrid Pneumatic Artificial Muscle Group and Its Genetic algorithm PID Control

Pneumatic Artificial Muscle (PAM) can be classified as mono-articular tendon and biarticular tendon in humanoid robot limb. Biarticular PAM acts on two or more joins simultaneous, while monoarticular PAM only acts on single joint. Biarticular PAM and monoarticular PAM mixed with each other and can be worked as Hybrid Pneumatic Artificial Muscle Group(HPAMG), in which two monoarticular PAM are series connected first, and then parallel connected with biarticular PAM. It's important to know how biarticular PAM and monoarticular PAM are orderly coordinated with each other in trajectory tracking with load. Optimization algorithms including ant colony algorithm optimization (ACO)and ant colony genetic algorithm optimization (ACGAO) were proposed, integrating with Proportional-Integral-Derivative (PID) control algorithm to track both monoarticular PAM and monoarticular PAM trajectories. Genetic algorithm PID (GAPID) was specifically adapted and subjected to experimental, to validate its superiority compared with PID. Experiment results indicate that ACO+PID and ACGAO+PID exhibited significant efficacy compared with the exclusive utilization of PID control algorithm. Furthermore, the ant colony genetic algorithm surpassed the performance of the ant colony algorithm. Notably, genetic PID trajectory tracking demonstrated distinct advantages over conventional PID control. The synergistic application of ACGAO and GAPID demonstrated a robust capability in achieving precise trajectory tracking.