Process modeling and advanced control methods for Exposure Controlled Projection Lithography

The Exposure Controlled Projection Lithography (ECPL) is a photopolymerization-based additive manufacturing process, which cures a 3D part by projecting patterned UV light from beneath a stationary and transparent resin chamber. Its application in microfabrication is limited by the current open-loop process control method. The overall research goal is to develop some closed-loop control system for the black / grey box ECPL process. In this paper, we presented two process modeling and control systems for ECPL - a lumped parameters model with an Kalman filter equipped Evolutionary Cycle to Cycle (EC2C) control scheme, and a novel backstepping dynamics model with Adaptive Neural Network Backstepping (ANNB) control method. EC2C and ANNB methods adopt different process models, control approaches and algorithms, and can be applied under different development stages and application scenarios of the ECPL process. Preliminary study concluded that the EC2C and ANNB control methods are capable of tracking the process dynamics, thus are promising to be able to improve the ECPL process precision and robustness.