遗传算法与神经网络组合的塔式起重机起升机构模糊优化设计

Fuzzy optimization design of hoisting mechanism for tower crane based on genetic algorithm and neural network

考虑到设计参数取值的不确定性,在满足承载能力和传动比分配要求条件下,以起升机构的减速器中心距最小为优化目标,建立该问题模糊优化的数学模型。按最大隶属度原则求出最优水平截集,将模糊优化间接转化为普通优化问题。另外,通过神经网络方法得出网络权值和阀值以拟和待求系数,应用Matlab遗传算法工具箱寻求问题最优解,从而提高设计精度和搜索效率。

Considering randomness of selecting the design parameters and meeting load - bearing capacity and distribution of transmission ratio requirements, the paper expounds how to build the fuzzy optimal mathematic model to get the optimum center distance of reducer, how indirectly to change the fuzzy optimization problem into a common optimization one and get the optimum level cut set, how to train feed - forward networks for fitting relative coefficient by means of neural networks algorithm. It comes to aconclusion that Toolbox of Matlab can be used to solve the optimization model and to enhance design accuracy and search efficiency.