Adaptive optics (AO) can compensate for the aberrating effects of atmospheric turbulence which degrade the performance of high energy laser (HEL) weapon systems and, as such, is an enabling technology for effective deployment of HEL weapon systems. A key component in an HEL AO system is the deformable mirror (DM). However, mechanical constraints in currently available DMs limits AO system performance in stressing engagements. Guidestar teamed with UCLA to develop methods that can provide AO performance with a constrained DM that is comparable to performance with an ideal DM that does not have these physical constraints. Guidestar developed a conjugate gradient method that determined an optimal DM command without violating hardware limitations in a woofer-tweeter control architecture. This method tied together within an adaptive Model Predictive Control (MPC) framework developed by UCLA. In this proposed Phase II program, Guidestar and UCLA will advance the development of these approaches, creating a complete control architecture to provide optimal AO correction using a DM that is constrained by total stroke saturation and inter-actuator stroke limits. Our approach will be validated through modeling and simulation, as well as scaled laboratory testing of the DM control methods.