Proportional integral control approach for controlling the temperatures of multi-steel cylinders barrel in an injection molding machine

This work presents a three dimensional prototype of the surface model to demonstrate the working of the controller. The three dimensional surface model is a prototype of the Injection Molding Machine (IMM), wherein the temperature of the system needs to be gradually increased to a very high temperature to melt plastics uniformly, despite the non-linear process. This process tends to be non-linear mainly due to thermal loss, heat transfer between the three cylinders and atmospheric disturbances. This paper presents a unique approach for designing a nonlinear surface model-based for controlling the temperature of multi steel cylinders in an injection molding machine. The innovation of this strategy is that the controller structure uses the nonlinear model to update the process variables at every sampling instant while the closed-loop control is executed. In this work, a new optimization routine will be used to minimize the errors between the model and process outputs. This allows the determination of the system's variables resulting in a new set of the Proportional Integral (PI) controller parameters with every sampling instant. The most important feature of 3-dimensional surface model control strategy is that it uses the process' variables to construct the surface model which are used in calculation of the control actions. In the meantime, this surface model is constructed offline by conducting several open-loop tests using different input signals and profiles while the measured output of the process is recorded. From these data, the main dynamic parameters of the process (process gain and time constant) are extracted to construct the surface model. Finally, the control law of the PI approach is updated every sampling instant to counteract the nonlinear effects of the system. In order to achieve a good control response for this process, an accurate model has to be developed to design a robust controller that can follow the reference trajectory and track the setpoint changes smoothly. Also, the model has to be of the adaptive form so that the controller has the ability to reject any disturbances or noisy feedback.