Nonlinear multi-physical system identification of a chemical stirred tank

Abstract

This paper presents a coupled electro-thermo-mechanical lumped parameter model for a chemical stirred tank, considering the device-related thermo-mechanical losses and process-related effects. This separation can help compute the key performance factors for competing processes during the scale-up. As the model has unknown thermo-mechanical parameters, two system identification experiment sets are conducted to determine the friction torque and heat transfer coefficient. A data acquisition system is developed in this study based on the Teensy 4.1 microcontroller to stream the logged data of the installed sensors on the serial port. An extended Kalman filter and a moving horizon estimation exploit the model and measurements to estimate the unknown model parameters. The results include the coupled electro-thermo-mechanical effects in estimating the heat transfer coefficient.