In this paper, we developed a lightweight, self-sensing electret-based dynamic vibration absorber (ESDVA) for micro-vibration suppressions. We modeled the electromechanical coupling procedure of the ESDVA based on the first principles and proposed a sensing model based on support vector regression machine (SVR). The SVR algorithm helps to linearize the original voltage generated by the electret for precise vibration sensing. A prototype of the ESDVA is fabricated, and the theoretical model and SVR algorithms are verified by experiments. According to experimental results, the ESDVA successfully reduced primary structure vibration amplitudes by up to 50% with a mass burden of 1.4% of the primary structure. The proposed sensing model achieve an accuracy rate of over 93.5% for vibration sensing and the robustness of the model was also assessed. Moreover, the advantages of the proposed electret-based sensing method over classical methods are discussed.

在本文中，我们开发了一种轻质、自感应驻极体动态吸振器 (ESDVA)，用于抑制微振动。我们根据第一原理对 ESDVA 的机电耦合过程进行了建模，并提出了基于支持向量回归机 (SVR) 的传感模型。SVR 算法有助于线性化驻极体产生的原始电压，以实现精确的振动传感。制作了ESDVA原型机，并通过实验验证了理论模型和SVR算法。根据实验结果，ESDVA 成功地将主结构振动幅度降低了 50%，主结构的质量负担为 1.4%。所提出的传感模型的振动传感准确率超过 93.5%，并且还评估了模型的鲁棒性。