A model based predictive compensation for ionic polymer metal composite sensors for displacement measurement

• A nonlinear auto-regressive and moving average model with exogenous input (NARMAX) is used to describe the behavior of IPMC sensor.
• The corresponding model based d-step-ahead predictive compensator is constructed to compensate for the effect of hysteresis, dynamics and time-delay inherent in IPMC.
• Experimental results and comparison with neural network based compensation scheme are illustrated.

It is known that IPMC sensors can be used for the measurement of displacement and vibration. However, the IPMC sensors are also involved with both hysteresis and dynamics which significantly affect the performance of the measurement. In this paper, a model based compensation scheme to degrade the effect of hysteresis and dynamics is proposed. Firstly, a modeling method to describe the characteristics of an ionic polymer metal composite (IPMC) sensor is proposed. As the hysteresis in IPMC is a non-smooth nonlinearity with multi-valued mapping, an expanded input space with hysteresis operator is introduced to transform the multi-valued mapping of hysteresis to a single-valued mapping. To describe the dynamic and hysteretic characteristics, a nonlinear auto-regressive and moving average model with exogenous input (NARMAX) is used to describe the behavior of the IPMC sensor. By considering the case that time-delay exists in IPMC sensor, a d-step-ahead nonlinear predictor based on the obtained model is developed. Subsequently, the corresponding model based predictive compensator is constructed to compensate for the effect of hysteresis, dynamics and time-delay inherent in IPMC sensor. In this scheme, the compensator is developed based on inverse model of the sensor on the expanded input space consisting of the output of sensor and the output of inverse hysteretic operator. Finally, the experimental results are demonstrated to validate the proposed model based compensation scheme for IPMC sensors.