12 DOF Biped Robot Gait Dynamic Analysis and Optimization using Genetic Algorithm

This paper offers a unique method for trajectory optimization to increase the energy economy of bipedal robots. For a 12 -DOF biped robot, we created a thorough simulation framework that uses spline interpolation and genetic algorithms to produce the best walking patterns. The strategy aims to maintain stability throughout the walking cycle while establishing smooth hip and foot trajectories between control points. We computed joint angles during the gait cycle using inverse kinematics, and we analyzed dynamic forces and torques using a NewtonEuler recursive technique. By using Zero Moment Point (ZMP) restrictions, stability was guaranteed. A notable $33.10 \%$ decrease in energy usage, from 101.30 joules to 67.77 joules each walking cycle, was shown by our MATLAB implementation.