Controlling a process such as injection molding can be quite complicated, due to the thermo-viscoelastic nature of polymers. Control of part quality has traditionally been provided by the machine operator who first defines “quality” for the specific part being manufactured. Second, using measurements of the quality of each part, the machine settings (or polymer state trajectories) are varied to optimize the result. Significant research and discussion has centered around improvements to this control strategy. Researchers suggest control can be simplified through the use of process control loops allowing the operator to prescribe state trajectories as control inputs to the system. They suggest that process parameters such as melt temperature and cavity pressure have a simpler, more direct relationship with product quality than the machine variables. [1] Others have treated the process as a linear multivariate system. [2] Attempts at intelligent or learning strategies include such studies as an expert system based on cavity pressure feedback [3], and a technique that uses an evolutionary strategy to search for the machine settings that optimize part quality. [4]