Parameter identification study of a trilayer conjugated polymer actuator curvature model

This article investigates the effect of two uncertain parameters on a recent new model of conjugated polymer actuators. These uncertain parameters are the diffusion coefficient (D) and the double-layer capacitance (C_dl), which are difficult to measure directly. The model sensitivity to these parameters is analysed and a parameter estimation study is performed using artificially generated data. The parameter estimation method used in this article is based on a Bayesian cost function, and gives us an insight on how much the estimation can be trusted, which is useful information for the design of controllers. Results indicate that for controllers to be designed effectively using this model, the double-layer capacitance is the best known parameter and should therefore be designed for with greater confidence in its value, while the controller should be much more robust with respect to the diffusion coefficient, which should be treated as a stochastic variable for a certain range of possible values.