论文部分内容阅读
为了在实时情况下解代数方程线性系统问题,本文提出了一种具有单片自适应学习算法的新一类简化的低成本模拟人工神经网络。本文提出的学习算法是解线性最小二乘(LS)、总量小二乘(TLS)和数据最小二乘(DLS)问题传统算法的修改和扩展。传统方法包括:kaczmarz行动作投影算法和/或LMS(Adaline)widrow-Hoff算法。该算法可应用于能公式化为线性回归问题的任何问题。所提出的神经网络的正确性和高性能可用广泛的计算机仿真结果来说明。
In order to solve the linear system of algebraic equations in real-time, a new type of simplified low-cost artificial neural network with monolithic adaptive learning algorithm is proposed in this paper. The learning algorithm proposed in this paper is a modification and extension of the traditional algorithms for solving linear least squares (LS), total squares (TLS) and data least squares (DLS) problems. Traditional methods include: kaczmarz line action projection algorithm and / or LMS (Adaline) widrow-Hoff algorithm. The algorithm can be applied to any problem that can be formulated as a linear regression problem. The correctness and high performance of the proposed neural network can be illustrated by a wide range of computer simulation results.