多激励识别的自适应延迟逆系统方法

摘要

An adaptive delayed inverse model of mechanical system was adopted to determine time histories of multiple input forces aiming at improving the shortcoming that the existing inverse system method relies on the state-space model.An adaptive algorithm was applied to identify the delayed inverse model,followed by time domain force estimation using operational response data combined with the inverse model of mechanical system.The multi-input forces of a simply supported beam were reconstructed by using accelerations without and with noise disturbances.The simulation results illustrate that the reconstruction accuracy decreases under noise condition,yet is still relatively satisfied and the method for multi-input identification is feasible.Furthermore,the method was validated experimentally.%针对现有逆系统方法依赖状态空间模型的缺点提出采用系统自适应延迟逆模型识别多激励时间历程。采用自适应算法对系统进行延迟逆模型辨识，结合逆模型利用工作状态响应数据识别时域载荷。以简支梁为研究对象用无噪声干扰、有噪声干扰加速度响应识别多点稳态激励及多点非稳态激励。仿真结果表明，噪声干扰时识别精度有所降低，但较满意，验证该方法对多激励源识别的可行性。