Model-Based Control with Stochastic Simulators: Building Process Design and Control Software for Advanced Materials Processing Technology

An analysis was made on the financial feasibility of a start-up company that will sell software developed for the off-line optimization and on-line control of a thin film deposition process. This analysis found some niche applications for a potential startup company that sells thin film deposition modeling and control software solutions. Due to the potential versatility of the software that was developed, other potential markets may exist. This investigation found that the startup company can be competitive over a five year time horizon with a 20% IRR. Molecular modeling software that employs the kinetic Monte Carlo method was used for the simulation of thin film growth. Due to the capability of this model to retain both surface and internal atomic structure of the thin film, this model can simulate thin film properties such as roughness and porosity. Development work was done on producing a suitable objective function to represent a set of application-imposed thin film micro-structure property requirements. This objective function was used in the generation of an optimal transient profile. A model predictive control framework was designed to control film growth based on the objective function and the optimal transient evolution of the film. The model predictive control algorithm was analyzed and shown to perform the desired control.