Muscle group activation estimation in human leg during gait using recursive least squares embodying Hill's muscle model

This paper presents a novel estimation method for extracting the activation rate of the human leg muscles using the recursive least squares algorithm. It is shown that the output force of each leg muscle group can be simply estimated from the reconstructed real measurement data. The estimation result turned out to be fairly comparable to those from electromyography, yet much simpler and faster. Considering the importance of the knowledge regarding the activation and deterioration state of each leg muscle group for rehabilitation, the proposed method is expected to contribute to the progress in the fields of biomechanics, by providing a simple, accurate, and fast estimation data to the developers, which will lead to the controller design of adaptive type walking assist devices.