Multi-Objective Optimization and Experimental Validation of Assistive Strategies for a Hip Exoskeleton

To address the limited assistive performance and insufficient individual adaptability of hip exoskeletons, a multi-objective optimization-based assistive strategy is proposed. A parameterized assistive torque model is constructed based on human gait characteristics, with the objectives of reducing joint load and improving human–robot interaction coordination. The NSGA-II (Non-dominated Sorting Genetic Algorithm II) is employed to optimize the assistive parameters, and the optimized results are implemented in a gait phase-based control method to achieve synchronized torque output over the gait cycle. Experimental validation is conducted on a hip exoskeleton platform using motion capture and electromyography measurements. The results demonstrate that the proposed method effectively reduces hip joint torque, decreases muscle activation levels, and enhances human–robot interaction performance.