论文部分内容阅读
何振亚教授是国家“八五”攀登重大项目首席科学家。他系统研究了自适应信号处理与神经智能信息处理.从一维到多维,线性到非线性,提出一系列的新理论和新技术。首次提出3种阶递归二维自适应最小二乘算法,适用于二维不同支持域的联合自回归模型参数估计难题:根据相关理论导出准确时序递归最小二乘RLS算法,解决了长期数值稳定性难题;提出一种可调参数范围很大的混沌与超混沌神经网络模型;提出一种混沌
Professor He Zhenya is the chief scientist of the major “85” climbing major project in the country. He systematically studied adaptive signal processing and neural intelligent information processing, and proposed a series of new theories and new technologies from one dimension to many dimensions and from linear to nonlinear. For the first time, three kinds of order-recursive two-dimensional adaptive least squares algorithms are proposed, which are suitable for the parameter estimation of joint autoregressive models in two-dimensional support domains. According to the related theory, accurate time-sequence recursive least square RLS algorithm is derived to solve the long-term numerical stability A chaotic and hyperchaotic neural network model with a large range of adjustable parameters is proposed; a chaos