Tower-Topology of Fuzzy Optimization Bp Neural Networks and a Case Study

Fuzzy optimization neural networks combines neural networks model with fuzzy optimization model, its application must establishes fixed expression topology according to actual problem. Aiming at the experiments in which multiple stages, influence factors and treatment levels should be considered, a tower-topology is established to simulate relationships between factors and results in this paper. In the topology, useless connections must be deleted and connection weights are constrained in. Through a case study of irrigation experiments, the relationship between treatment levels and yield is presented by more precisely than a traditional model.