A tip-tilt mirror (TTM) control method is designed to enhance the control bandwidth and ensure the rejection performance of the adaptive optics (AO) tip-tilt correction system. Optimized with the Smith predictor and filter, linear active disturbance rejection (LADRC) is adopted to achieve the tip-tilt correction. An AO tip-tilt correction experimental platform was built to validate the method. Experimental results show that the proposed method improves the control bandwidth of the system by at least 3.6 times compared with proportional–integral (PI) control. In addition, under the same control bandwidth condition, compared with the Smith predictor and proportional–integral (PI–Smith) control method, the system is more capable of rejecting internal and external disturbances, and its dynamic response performance is improved by more than 29%.

Stabilization of optical beams has always been a key factor affecting the performance of many optical systems. The Adaptive optics (AO) beam stabilization system requires further development to cope with increasingly complex application scenarios and challenges. Motivated by this, a new filter-based off-policy policy iteration (FB-OPPI) control scheme is proposed and experimentally verified in this paper to provide AO systems with a flexible beam stabilization method. The FB-OPPI is based on the policy iteration, a model-free controller design principle. To address the challenges such as convergence speed, data requirements and control stability that it faces in practice, we have proposed an implicit state reconstruction method based on the Kalman filter and introduced the adaptive transverse filter technology. Additionally, the off-policy learning mechanism is deployed to simplify the optimization process. An AO beam stabilization system was constructed to verify the effectiveness of the proposed method. Experimental results show that the FB-OPPI method features simple design and fast training, releases the requirement for additional sensors or model recognition. The FB-OPPI method is superior to traditional integral controllers, effectively handling high-frequency narrowband and complex beam jitters. Despite not requiring model identification, it is on par with the advanced Linear Quadratic Gaussian (LQG) control.

The tip–tilt mirror (TTM) is an important component of adaptive optics (AO) to achieve beam stabilization and pointing tracking. In many practical applications, the information of accurate TTM dynamics, complete system state, and noise characteristics is difficult to achieve due to the lack of sufficient sensors, which then restricts the implementation of high precision tracking control for TTM. To this end, this paper proposes a new method based on noisy-output feedback Q-learning. Without relying on neural networks or additional sensors, it infers the dynamics of the controlled system and reference jitter using only noisy measurements, thereby achieving optimal tracking control for the TTM system. We have established a modified Bellman equation based on estimation theory, directly linking noisy measurements to system performance. On this basis, a fast iterative learning of the control law is implemented through the adaptive transversal predictor and experience replay technique, making the algorithm more efficient. The proposed algorithm has been validated with an application to a TTM tracking control system, which is capable of quickly learning near-optimal control law under the interference of random noise. In terms of tracking performance, the method reduces the tracking error by up to 98.7% compared with the traditional integral control while maintaining a stable control process. Therefore, this approach may provide an intelligent solution for control issues in AO systems.

Effective control of large-aperture tip-tilt mirrors (TTMs) has been a pressing concern in beam stabilization due to the influence of their internal mechanical resonance. To the best of our knowledge, this Letter reports a new high-efficiency actor-critic (HEAC) scheme to effectively eliminate the impact of mechanical resonance on jitter suppression stabilization and performance. Moreover, HEAC features rapid deployment and training that rely solely on online noisy measurements through the establishment of a specific tracking control problem, the decoupling of the unknown system parameters, and the design of a quadratic homogeneous polynomial basis function network. The experimental results demonstrate that the HEAC has successfully achieved stable and efficient suppression of jitter near the resonance peak or close to the Nyquist frequency for a $\Phi ~405$ mm aperture TTM.