Complexity of manufacturing processes has hindered methodical specification of machine setpoints for improving productivity. Traditionally in injection molding, the machine setpoints are assigned either by trial and error, based on heuristic knowledge of an experienced operator, or according to an empirical model between the inputs and part quality attributes, which is obtained from statistical design of experiments (DOE). In this paper, a Knowledge-Based Tuning (KBT) Method is presented which takes advantage of the a priori knowledge of the process, in the form of a qualitative model, to reduce the demand for experimentation. The KBT Method provides an estimate of the process feasible region (process window) as the basis of finding the suitable setpoints, and updates its knowledge-base using the data that become available during tuning. As such, the KBT Method has several advantages over conventional tuning methods: (1) the qualitative model provides a generic form of representation for linear and nonlinear processes alike, therefore, there is no need for selecting the form of the empirical model through trial and error, (2) the use of a priori knowledge eliminates the need for initial trials to construct an empirical model, so an initial feasible region can be identified as the basis of search for the suitable setpoints, and (3) the search within the feasible region leads to a higher fidelity model of this region when the input/output data from consecutive process iterations are used for learning. The KBT Method’s utility is demonstrated in production of digital video disks (DVDs).
Real-time road traffic monitoring using a fast a priori knowledge based SAR-GMTI algorithm
