Improved Fuzzy Inference Network and Its Applications in Multi-targets Optimization of the High-Rise Structure Form Selection

摘要

Firstly, two kinds of improved weighted fuzzy inference networks were established, namely the multi output multi input (MIMO) one and the multi input single output (MISO) one. The newly built networks share common merits with the weighted fuzzy evaluation method, the fuzzy inference method and the traditional fuzzy inference method, laying foundation for highly and effectively expressing, memorizing and utilizing the domain experts'' knowledge and strategies used in form selection. Then, two kinds of multi-target high-rise structure form selection optimization methods were established based on the improved fuzzy inference network and the fuzzy comprehensive evaluation model; and a comparison of characteristics was made between the two kinds of multi-targets optimization models: Both of them can directly and effectively solve the form selection optimization problem for the multi-target and multi layer structures. Moreover, the reasoning process of the first model can produce clearer and more concrete results, and the reasoning process of the second model is simpler. And it has been proven that the results of the models have the same conclusion. Finally, a project example was given, which proved that the methods can meet the needs and the optimization process better resembles people''s habits, experience and strategies in form selection.