应用ICM(IndependentContinuous andMapping)方法,建立了以重量最小为目标函数,以连续频率带或离散点频率的简谐激励下的响应振幅为约束的拓扑优化模型。引入了对数型Heaviside近似函数作为过滤函数,并做了敏度分析,利用对偶二次规划进行优化模型的求解,并运用敏度过滤的方法处理动力响应数值不稳定的问题。数值算例比较了利用对数型函数和幂函数作为过滤函数时对拓扑结构的影响,结果显示利用对数型函数较幂函数结构优化迭代次数更少,收敛更快。 The ICM (IndependentContinuous and Mapping) method was used to establish the topology optimization model with the minimum weight as the objective function and the response amplitude under the harmonic excitation with continuous frequency band or discrete point frequency. The logarithmic Heaviside approximation function was introduced as a filtering function, and the sensitivity analysis was done. The duality programming was used to solve the optimization model. The sensitivity filtering method was used to deal with the instability of the dynamic response. Numerical examples compare the effects of logarithmic and power functions on the topological structure. The results show that the use of logarithmic and power-efficient structures has fewer iterations and faster convergence.