Hysteresis fitting and control of test-mass-release mechanisms via adaptive stochastic resonance: Numerical and experimental studies

Space-based gravitational wave detection plays a crucial role in advancing our understanding of the universe. One of its primary tasks is to release the test mass (TM) into free-fall state with high positional accuracy and extremely low residual velocity through a grabbing, positioning, and release mechanism (GPRM) after the satellite enters orbit, and then the TM is captured by an electrode housing. During the release process, hysteresis effects occur in the piezoelectric module of the release mechanism, and a hysteresis model must be fitted on the ground to achieve in-orbit control and compensation of the TM based on the model. However, when measuring the displacement signal from the GPRM on the ground, noise interferes with the signal collected, leading to inaccurate fitting of the hysteresis effect and impacting hysteresis control and compensation. In this work, we propose a hysteresis fitting and control method of the test-mass-release mechanism via adaptive stochastic resonance (ASR), which effectively denoises the noisy signals through numerical simulation. By simultaneously introducing a quantum particle swarm optimization algorithm and using cosine similarity as the optimization index, the method achieves optimal denoising output under different voltage excitations. Compared with other denoising methods and other particle swarm optimization algorithms, the proposed approach demonstrates a superior denoising performance. In terms of control, after ASR, the controlled displacement is free of interference, and the displacement curve is smooth. This method holds significant application value for hysteresis control in the GPRM.