Human-in-the-loop optimization of knee-joint biomechanical energy harvester to maximize power generation with minimal user effort

Abstract

Knee-joint harvesters of biomechanical energy promise to supply usable electric energy and reduce metabolic cost of walking for field workers, but few have succeeded. Optimizing the power generation pattern for each user based on the power generation of the harvester and metabolic cost of the user was assumed to be able to solve the problem. We have developed human-in-the-loop optimization for identifying the power generation pattern that can maximize power generation with minimal metabolic cost during walking. The experimental results showed that walking with the harvesters using the optimized power generation pattern reduced net metabolic cost by 2.36 ± 2.52% compared to walking without the harvester and by 5.81 ± 2.76% compared to walking with the harvesters using the static pattern. The harvester using the optimized power generation pattern can generate 2.75 ± 0.33 W electric power, which was similar to the power generation of the harvester using the static power generation pattern and was 11 times more than the power generation of the harvester that can reduce metabolic cost.