一类半无限规划问题的神经网络算法

Neural network model for solving semi-infinite problem

考虑了一类带有不等式和等式混合约束的半无限规划问题。通过运用极大熵方法,将多个约束条件的问题转化为单个约束条件的非线性规划模型,并提出了求解它的一个神经网络模型,严格证明了该模型是Lyapunov稳定的,并且在有限时间内收敛到原问题的一个精确解。数值实验表明,新模型不仅可行而且有效。

The paper considers the semi-infinite problem with inequalities and equality constraints. By using maximal entropy method, it converts many constraints problem into single constraint nonlinear programming. Then it proposes a new neural net- work for solving it. It is shown to be Lyapnuov stable, and convergent to an exact solution of the problem in finite time. Illustra- tive examples show the feasibility and efficiency of the network.