Application Of Iterative Learning Control In The Injection And Packing Control Of An Injection Molding Machine

Injection Molding Machine (IMM) is one of the most important equipment in plastic industry. Since it appeared in the 3rd decade of last century, unlike the tiny change in the structure and transmission principle, the control technology of the IMM has renovated and developed rapidly. So research on the control method of IMM has significant meaning for both the quality improvement of plastic product and the progress of plastic industry.As a typical batch process, molding process is made up of injection, packing, plasticization and cooling phases, among which the injection phase, inject velocity/packing pressure (V/P) control transition and packing phase are the most important parts which decide the product quality. This dissertation is just aiming at this process. Base on the IMM named HTF58X2-B made in Ningbo Haitian Company, the dissertation presents an Iterative Learning Control (ILC) scheme for the entire process of injection and packing, and validates the scheme with experiments.First of all, the dissertation does a deeply research on the technologic process of injection molding through two aspects, one is the dynamic state and rheological characteristic of the polymer, and the other is the composing structure and working principle of the IMM. Then it summarizes the multi working procedures, time varying, nonlinearity and uncertain characteristic of the injection molding manufacture process.These characteristics show that it is difficult to set up the accurate mathematical model of the injection molding process.Afterward,the dissertation discusses the ILC's concept, algorithm, and development. ILC method doesn't depend on the exact mathematical model of the dynamic system, can deal with system with high uncertain degree, needs little previous knowledge, has good adaptability, and can be implemented easily. All the above advantages make the algorithm especially suitable for the application in batch process with complicated dynamic characteristics such as injection molding process.At the end, experiments are carried out to validate the effectiveness of ILC control scheme in injection molding process. For the injection and packing phases, curve setting of optimum injection speed and packing pressure is firstly researched. Then based on the analysis of the instability problem of ILC caused by sampling delay, the open-loop P-type control scheme dealing with the delay factor is implemented, and the complete tracking of injection speed and packing pressure is realized after lots of simulations and experiments.As for the V/P control switchover part, existing transition modes are firstly discussed, and then according to the actual experiment condition, a new mode adopting the cavity pressure to control switch has been presented. This mode not only improves the control accuracy, but also guarantees the initial constrained condition of ILC on packing pressure. Base on all the above researches, this dissertation finally implements the ILC scheme in the entire injection phase-V/P control switchover-packing phase. The experiment result shows that the proposed method realizes the smooth Fill-to-Pack transition, and has such advantages as less adjusting time, easier parameter selection, better dynamic tracking performance, and higher steady state accuracy compared with normal PID control method.