一类非光滑非凸优化问题的神经网络方法

提出了解决一类带等式与不等式约束的非光滑非凸优化问题的神经网络模型。证明了当目标函数有下界时，神经网络的解轨迹在有限时间收敛到可行域。同时，神经网络的平衡点集与优化问题的关键点集一致，且神经网络最终收敛于优化问题的关键点集。与传统基于罚函数的神经网络模型不同，提出的模型无须计算罚因子。最后，通过仿真实验验证了所提出模型的有效性。

Neural network optimization method for class of nonconvex nonsmooth optimization problems

This paper proposed a novel neural network to solve nonsmooth nonconvex optimization problems with equality and inequality constraints. It was proved that when the objective function had a lower bound, the neural network converged to a feasible domain in a finite time. Meanwhile, the solution trajectory of neural network converged to optimal solution set of the corresponding optimization problems, which finally converged to critical point set of optimization problems. Comparing with traditional neural network which based on penalty function, the neural network model did not need to calculate any penalty parameters. Finally, the effectiveness of the proposed model is verified by simulation experiments.