本文设计了一个实际电路,用它来实现一种可以求解矩阵方程Au=B的神经元网络模型,并着重研究了该电路的硬件实现。经过大量实际测试及仿真实验证明:该电路确实能够用来求解矩阵方程,而且由于Hopfield网络具有极强的快速计算能力,因而被解方程的阶次越高,用神经元网络硬件代替软件计算这一方法的有效性越大。 In this paper, a practical circuit is designed to implement a neural network model that solves the matrix equation Au = B and focuses on the hardware implementation of the circuit. After a large number of actual tests and simulation experiments show that: This circuit can indeed be used to solve the matrix equation, and because of the Hopfield network has a very fast computational ability, so the higher the order of the solution to the equation, using neural network hardware instead of software to calculate this The more effective a method is.