A mathematical model for the validation of the Ground Reaction Force Sensor in Human Gait Analysis
A sensor validation scheme investigates a system that requires control based on monitored sensor readings in order to allow immediate corrective actions in the case of aberrations from a desired value. In this paper, the proposed sensor validation strategy is based on a mathematical model for the calculation and classification of some white noise or faults occurring in the biomechanical system during walking on an instrumented treadmill. This strategy attempts to build a predictive model from measurements to determine whether the actual values are within an expected range. Particular attention is focused on the system used to measure vertical ground reaction force (vGRF) during normal walking on the instrumented treadmill. The aim of the study is to perform sensor validation to improve the resolution and accuracy of the acquired sensor data in order to provide reliable, repeatable, reproducible information for decision making.
The absence of sensor validation will exhibit major obstacles in the efficient acquisition and resolution of measured data that may corrupt the data of the human gait analysis. In this paper, the authors introduce a new sensor validation scheme based on the method of Autoregressive Moving Averages (ARMAs) to assure the quality of the acquired vGRF data. This methodology facilitates the process for determining the validity of the acquired sensor signals, for evaluating the levels of noise and for providing a timely warning from the expected signals. The experimental results show that the mathematical model for vGRF data is a robust and efficient method for the operator in making correct decisions based on the segment and continuous signal validation results.