A research on the intelligent multi-objective optimization problem based on wavelet theory and neural networks
Department of Mathematic and Computational Science, Langfang Teachers University, Langfang 065000, China
Aiming to solve the multi-objective optimization problem caused by wavelet multiresolution a1nalysis (MRA), the thesis improves the original multi-objective non-dominated genetic algorithm. After fast non-dominated sorting, the evolution of population is achieved through particle swarm optimization (PSO). In this way, the thesis realizes a more effective, organic combination of the multi-objective optimization problem and neural networks. MRA is a natural fit for the multi-objective optimization problem. The ability of neural networks to deal with complex errors is improved through error decomposition based on different wavelet decomposition scales.