Dual-loop optimization of digital accelerometers based on advanced dual-degree-of-freedom Smith predictor

Abstract

In digital accelerometers, the setpoint loop is designed to balance the controlled object, and the disturbance loop is used to measure external inputs. Traditional closed-loop control structure of digital accelerometers is optimized only for the disturbance loop, failing to optimize the performance of the dual loops simultaneously. An advanced dual-degree-of-freedom (ADOF) Smith predictive control is proposed herein, structurally isolating the setpoint loop from the disturbance loop. Firstly, the principle of the closed-loop digital accelerometer is elaborated, and the contradiction between the performance optimization of dual loops is analyzed. Secondly, an ADOF Smith predictive control is proposed and compared with other control structures. Furthermore, the setpoint and disturbance controllers in the structure are designed using the Dahlin algorithm and H $\infty$ optimal control methods for parameter tuning, with the controllers being determined by a single parameter. To quantitatively describe the relationship between the control parameters and robust performance, the relationship between the control parameters and the margin indicators is derived, and the control performance is evaluated. Finally, an experimental platform of a digital closed-loop accelerometer is constructed to validate the proposed ADOF control structure and parameter tuning methods. The experimental results show that the performance of the proposed ADOF control structure significantly outperforms traditional digital closed-loop control and filtered Smith predictive control, and compared with the dual-degree-of-freedom control structure, it exhibits better nominal performance under low robustness parameters due to delay compensation.