An Intelligent Decoupling Control Scheme for Vacuum Casting Process

The vacuum casting process, which has strong ability to produce thin wall and transparent parts, is time variable, nonlinear and strong-coupled. In order to control the vacuum casting process effectively, the intelligent decoupling control scheme that uses artificial neural networks (ANNs) embedded within internal model control (IMC) structure is employed. By using artificial neural networks trained based on inputs/outputs data which were taken from experiment, the process model which can descript the relation between the key process variables and the machine variables is firstly developed. Then the decoupling controller is derived based on the error between the outputs of process and the model. As a result, the system set-point tracking response is well decoupled from the system disturbance rejection response. Since there is an open-loop control for the nominal set-point tracking, the nominal system stability can be easily identified. At the same time, the system is robust because of the employed internal control structure.