An Intelligent Optimization-Based Residual Negative Magnitude Shaping Scheme for Vibration Control

Abstract

With the rapid advancement of modern manufacturing, suppressing residual vibrations in flexible and underactuated systems has become a critical challenge. Input shaping (IS) has garnered attention for its effectiveness in mitigating vibrations and enhancing motion performance. However, existing input shapers typically encounter unavoidable time delays (TDs), modeling inaccuracies, ineffective multimodal suppression and poor adaptability, limiting their control performance. Targeting at overcome these critical issues, this article proposes an intelligent optimization-based residual negative magnitude (NM) shaping vibration (IRV) control scheme with two novel ideas: 1) employing a data-driven differential evolution (DE) algorithm to estimate system errors; and 2) designing a robust particle swarm optimization (PSO)-based residual negative magnitude (PR) shaper to reduce TDs and compensate for modeling inaccuracies in multimodal vibration systems, thereby enhancing control adaptability to diverse system configurations. To validate its performance, eight real-world datasets have been established and made publicly available. Empirical studies demonstrate that the proposed PR shaper outperforms state-of-the-art shapers, and the IRV scheme achieves significant vibration suppression, reducing maximum residual vibrations by at least 9.26% compared to conventional methods. These advancements substantially improve vibration control in precision systems.