用神经网络解非线性方程组

提出了一种解非线性方程组的神经网络模型，并在非线性方程组有惟一实根、有限个实根以及无穷多个实根情形下严格地证明了该模型的稳定性，然后，给出了一个模拟算法，该算法不仅可以用来解非线性方程组，而且还可以用来解多元非线性方程及线性方程组，数值试验结果表明，该算法是有效的。

Neural network for solvingsystems of nonlinear equations

A neural network model for solving systems of nonlinear equations is proposed, which is strictly proved to be stable whether a system of nonlinear equations has only one real solution, finite real solutions or infinite ones. Then a simulation algorithm is given, which can be used to solve not only systems of nonlinear equations, but also nonlinear multivariable equations and systems of linear equations. Finally, the algorithm is illustrated to be effective with some numerical tests.