A Method for Rapid Self-Calibration of Wearable Soft Strain Sensors

Compared to rigid sensors, soft stain sensors are more suitable for measuring human joint motion because soft sensors are more comfortable to wear. In order to perform more precise measurements, the soft stain sensor should be re-calibrated before performing each measurement in order to eliminate error from donning and doffing. Current calibration methods are often performed by optical motion capture systems (OMCSs). However OMCSs are large and cumbersome, and there is no directly applicable calibration method for soft stain sensors in the measurement of human joint motion. Calibration should be able to be performed quickly and without the need of large equipment such as OMCSs. This paper proposes a calibration method for wearable soft strain sensors that can be done quickly and automatically. The soft sensor structure for self-calibration is parallel and partitioned, and there are integrated primary and redundant sensors. The basic idea of this method is to use redundant sensors to re-calibrate the primary sensor, and only a simple calibration action is required for the self-calibration. After self-calibration, the average errors of measurement were all less than 5 degrees, and the relative errors were all less than 4% for sensor and clothing donned and doffed and long-term sensor migration while wearing. Self-calibration can be performed in only 10 seconds. The results presented that the proposed calibration method is feasible for engineering applications.