Scholarlink

<p>This paper investigates the stochastic dynamics of a nonlinear hybrid vibrational energy harvester (VEH) under Gaussian white noise excitation. Accurate quantification of stiffness and damping coefficient is challenging for high-dimensional nonlinear VEHs because the corresponding Fokker–Planck–Kolmogorov equation for the response is complex. To solve this problem, a new dynamical analysis framework has been proposed to enable rapid and accurate analysis of system responses. First, a data-driven approach is used to develop a lumped parameter model of the electromechanical coupling for the VEH system. Second, machine learning techniques are employed to solve the system's dynamical equations, enabling the efficient computation of the system's steady-state response. The effectiveness and efficiency of this method are validated through comparing with numerical results. It should mention that the computation time of the proposed method is only 5% of what the Monte Carlo simulation requires. Finally, the effects of damping coefficient and noise intensity on the output power and power conversion efficiency are analyzed, and a comparison is made with the performance of a piezoelectric VEH (PEVEH) model. The results demonstrate that the power conversion efficiency of the hybrid VEH is twice that of the PEVEH. As the linear stiffness coefficient increases, the output power of hybrid VEH exhibits a non-monotonic variation due to the deepening of the potential well. An optimal combination of coupling coefficients is found to maximize the output power.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Stochastic response and performance analysis of a high-dimensional nonlinear hybrid energy harvester

  • 期刊论文

  • 发表日期：

    2025年5月3日

查看全文
  • Chengjia Sun,
  • Yanfei Jin

摘要

This paper investigates the stochastic dynamics of a nonlinear hybrid vibrational energy harvester (VEH) under Gaussian white noise excitation. Accurate quantification of stiffness and damping coefficient is challenging for high-dimensional nonlinear VEHs because the corresponding Fokker–Planck–Kolmogorov equation for the response is complex. To solve this problem, a new dynamical analysis framework has been proposed to enable rapid and accurate analysis of system responses. First, a data-driven approach is used to develop a lumped parameter model of the electromechanical coupling for the VEH system. Second, machine learning techniques are employed to solve the system's dynamical equations, enabling the efficient computation of the system's steady-state response. The effectiveness and efficiency of this method are validated through comparing with numerical results. It should mention that the computation time of the proposed method is only 5% of what the Monte Carlo simulation requires. Finally, the effects of damping coefficient and noise intensity on the output power and power conversion efficiency are analyzed, and a comparison is made with the performance of a piezoelectric VEH (PEVEH) model. The results demonstrate that the power conversion efficiency of the hybrid VEH is twice that of the PEVEH. As the linear stiffness coefficient increases, the output power of hybrid VEH exhibits a non-monotonic variation due to the deepening of the potential well. An optimal combination of coupling coefficients is found to maximize the output power.