中文摘要：

针对传统经验方法用于确定空间钢结构临时支撑体系布置方案时的不合理性,提出一种基于小生境遗传算法的优化算法,该算法以指数尺度变换后的结构总应变能为优化目标函数,采用格雷编码作为染色体的编码方式,对于规则的搜索区域,以连续的空间位置作为遗传算法的决策变量;对于不规则的搜索区域,以离散的空间节点作为决策变量,在进化过程中引入自适应的遗传算子和改进的最优个体保存策略;模拟＂鸟巢＂结构的一榀主桁架,验证了算法的有效性;研究矩形平板网架和某异形网壳的最优临时支撑布置方案.分析表明：应变能综合反映了结构的受力性能,该应变能跟踪算法对空间钢结构施工过程中临时支撑体系方案确定提供了科学的依据.

英文摘要：

Due to the irrationality of traditional methods for layout of falsework systems in spatial steel structures,an algorithm based on niche genetic algorithm was proposed.The strain energy of the structural system was exponentially rescaled and taken as optimization objective function.Chromosome was represented by Gray code.When the search region was regular,continuous spatial location was taken as decision variable;otherwise,discrete node was taken as decision variable.Self-adaptive genetic operator and improved elitist strategy were introduced into the evolution.The effectiveness of the proposed algorithm was verified by a simulation example on a main truss of＂bird nest＂.Analysis of the optimum falsework layout of plane grid structures and deformed latticed shell structures indicates that the strain energy represents the stress of structures.