An Integral Enhanced Coevolutionary Neural Network Optimization Algorithm for Solving NonConvex Problem in Noisy Environments

This paper addresses nonconvex optimization problems in noisy environments by proposing an integral enhanced coevolutionary neural network (IECNN) optimization algorithm. The IECNN optimization algorithm integrates an integral enhanced recurrent neural network (IERNN) model with noise tolerance characteristics and an improved particle swarm optimization (PSO) algorithm. It enhances global search capabilities and avoids local optima through adaptive inertia weight and stochastic mutation strategies. Experimental results demonstrate that the IECNN optimization algorithm performs excellently in solving the Griewank function and its application to angle-of-arrival localization problems. And it can obtain the optimal solution with fewer iterations and maintains high robustness and accuracy under various noise conditions.