【摘要】 结合小波变换的压缩和去噪能力,研究了在FIR模型辨识过程中求得更精简非参数模型的一种方法.在给出FIR模型辨识的具体步骤后,通过一个化工实例的具体计算,进一步比较与一般FIR建模方法以及基于不同小波基函数进行压缩去噪后辨识的效果.结果表明该技术可以压缩模型结构参数,降低了运算量,有利于实现在线辨识.更多还原

【Abstract】 Combining the compression and noise-filted capability of wavelet shift, this paper studies an approach of establishing a more concise non-parameter model in the process of FIR-model identification.This paper presents detailed wavelet compression technology that is popular in the theory of modern control, and apply it to the identification of FIR model, describes the realization process, and obtaines the expected result. The result of computer simulation indicated that this technology could compress the parameters of model structure, reduce the amounts of mathematics operation in the model identification, and make model identification online possible.更多还原