Invasive weed optimization algorithm optimized fuzzy logic scaling parameters in controlling a lower limb exoskeleton

This paper describes a new modified versions of invasive weed optimization algorithm with exponential seeds-spread factor. The modified invasive weed optimization algorithm (MIWO) is employed to optimize the fuzzy input-output scaling factors of lower limb exoskeleton. A fuzzy logic control (FLC) system with the (MIWO) are evolved for reference tracking control. The exoskeleton is developed to enhance and upgrade the lower limb capability and augment the torque of knee and hip of elderly people during the walking cycle. Invasive weed optimization is a bio-inspired search algorithm that mimics how weeds colonize a certain area in nature. The algorithm is modified by applying local knowledge during distribution of seeds that depends on their cost function value in each generation to narrow the accuracy and improve the local search ability. The obtained results from the modified invasive weed optimization algorithm are compared with heuristic gain values to improve the performance of the exoskeleton system. The Visual Nastran 4D software is used to develop a simulation model of the humanoid and an exoskeleton for testing and verification of the developed control mechanism. Simulation results demonstrating the performance of the adopted approach are presented and discussed.