Iterative Identification Method For Two-input Output Error Bilinear-in-parameter System

With the expansion of modern production scale,the modeling,identification and control problems of large-scale complex systems that need to be dealt with have become increasingly prominent.The identification of a kind of special structure nonlinear system has received widespread attention.This class of nonlinear system is a combination of static nonlinear block and linear dynamic subsystem,which can form a nonlinear system with parallel,series,and feedback structures.About the bilinear-in-parameter system identification theory and method,that is,about the block structure nonlinear system identification theory and method.This paper discusses the calculation and optimization problems of large-scale nonlinear systems,and obtains the following main research results:(1)For the two-input output error bilinear-in-parameter system,combining the hierarchical identification principle,the multi-innovation identification theory and the gradient-based search optimization principle.The bilinear-in-parameter system decomposition-based identification model is obtained.The bilinear-in-parameter model decomposition auxiliary model multi-innovation gradient-based iterative identification algorithm is proposed to improve the parameter estimation accuracy.(2)For the two-input nonlinear system with autoregressive noise.The filtered identification submodel is obtained by means of the filtering technique and decomposition technique.The over-parameterization filtered two-stage gradient-based iterative identification algorithm is proposed,which can reduce the impact of the noise to the parameter estimates,thereby improving the parameter estimation accuracy.(3)For two-input nonlinear output error systems with autoregressive noise,using the key term separation principle,the key term separation based identification model is obtained for the two-input output error bilinear-in-parameter system.The key term separation auxiliary model three-stage gradient-based iterative algorithm is proposed to reduce the computational burden;the key term separation auxiliary model three-stage multi-innovation gradient-based iterative algorithm is proposed,which can provide more accurate parameter estimates.The paper gives the corresponding numerical simulations of the iterative parameter estimation algorithms for the two-input output error bilinear-in-parameter system.