Iterative learning model predictive controller of plastic sheet temperature for a thermoforming process

Properties of the thermoforming process, such as its nonlinear, time-varying dynamics and actuator constraints, make its control challenging. An iterative control technique along with model predictive control (MPC) is presented in this paper on 2D control of the thermoforming process. This approach utilizes not only incoming information from the ongoing cycle, but also the information stored from the past cycles. To deal with constraints as well as non-repetitive disturbances in the process, the MPC technique is incorporated to update the control law within the cycle. To exploit the repetitive nature of the heating phase of the process, a cycle-to-cycle iterative learning control technique direction is proposed. The iterative learning strategy is useful for achieving desired temperature despite model mismatch and disturbances. Even though the proposed multi-zone temperature controller can handle a multivariable process, the large number of computations makes it difficult to apply to large systems such as a thermoforming machine. To reduce the computational burden, the control laws are computed offline using multi-parametric programming.