Files

Download

Download Full Text (87 KB)

Faculty Advisors

Richter, Hanz

Description

The objectives of this research project are to design and build an instrumented shoe to measure the vertical ground reaction force (GRF) associated with a person walking or running. Sensor outputs are calibrated to actual GRF with an artificial neural network. Currently, GRF measurements require special equipment such as force plates or scientific treadmills. A force plate measures GRF over a limited area. A shoe insole fitted with sensors was identified as a good solution that allows free-range walking over arbitrary surfaces. Piezoelectric film sensors were chosen due to their low cost, flexibility and for being self-powered. Eight sensors were bonded to a conventional insole and wires attached. A data acquisition interface was prepared using a dSPACE MicroLabBox system, which contained digital filters for noise removal. Training and validation data were collected using a force-sensing treadmill available at the Parker-Hannifin Human Motion and Control Lab at CSU. A 3-layer feedforward network was successfully trained to approximate the training data. A separate data set was used to validate the trained network. A normalized root mean square error associated with training was 1.01, while the error in validation was 2.78.

Publication Date

2016

College

Waskewicz College of Engineering

Ground Reaction Force Measurement with a Piezoelectric Insole

Share

COinS