Multi hot runners system is a newly booming technique widely used in injection molding industry to improve productivity as it can enable simultaneous production of multiple parts within one molding cycle. The temperatures synchronization control performance of its subsystems is critical to the product quality consistency of injection molding processes. However, the control of multi hot runners system's heating process suffers from the difficulties caused by model-plant mismatch, periodic disturbance and setpoint online adjustment during operation of injection molding process and large number of sub runners. Thus, the controller should meet following demands: 'immediate' synchronization, fast convergence in periodic nature handling and dynamic transition, small computation burden...[ Read more ]

Multi hot runners system is a newly booming technique widely used in injection molding industry to improve productivity as it can enable simultaneous production of multiple parts within one molding cycle. The temperatures synchronization control performance of its subsystems is critical to the product quality consistency of injection molding processes. However, the control of multi hot runners system's heating process suffers from the difficulties caused by model-plant mismatch, periodic disturbance and setpoint online adjustment during operation of injection molding process and large number of sub runners. Thus, the controller should meet following demands: 'immediate' synchronization, fast convergence in periodic nature handling and dynamic transition, small computation burden.

With above motivations, first develop Master-Slave Generalized Predictive Synchronization Control (M-S GPSC) to achieve the 'instantaneous' synchronization in idle case. Then Repetitive Control (RC), which is a well known methodology for handling periodic nature in continuous process, is integrated into the prediction model of Generalized Predictive Control (GPC), which leads to the Generalized Predictive Repetitive Control (GPRC). Moreover, an improvement scheme for GPRC is proposed to improve the convergence speed. Last but not the least, an effort on reduction of computation burden is made by designing a simpler integration of RC and GPC: Direct GPRC (DGPRC), which has less robust stability than GPRC but can still be expected to have satisfactory performance in large scale of industrial applications.

The effectiveness of control algorithms developed was tested on the MATLAB simulation, and M-S GPSC's performance is verified on an industrial-sized multi hot runners system under idle case. The control requirements are all met in both the simulations and experimental test.