不确定性结构系统非线性随机最优控制的鲁棒性.pdfVIP

不确定性结构系统非线性随机最优控制的鲁棒性 不确定性结构系统非线性随机最优控制的 鲁棒性 茅僰 ，应祖光，朱位秋 （浙江大学航空航天学院，浙江 杭州 310027） 摘 要：对于具有高斯分布的不确定参数的 n 层剪切型结构、承受有理渐进谱密度的宽带随机基础激励与 切向控制作用的系统，基于系统参数不确定性与激励随机性相互的独立性，建立其平均参数的名义系统； 通过模态变换解耦，并转化为2n 维的受控的、随机激励的、耗散的可积哈密顿系统；然后，运用拟可积哈 密顿系统的随机平均法与随机动态规划原理，得到其非线性随机最优控制律；再根据受控结构系统不确定 参数的概率特征，运用随机平均法与概率分析估算系统均方响应与均方控制力的均值和标准差，以及平均 控制效果与平均控制效率，进而评价拟哈密顿系统非线性随机最优控制的鲁棒性；最后，通过大量的数值 计算分析，说明不确定参数的标准差对于受控结构系统均方响应的标准差、以及平均控制效果与平均控制 效率等的影响，从而展示上述非线性随机最优控制具有很好的鲁棒性。 关键词：随机不确定性结构；随机平均法；随机动态规划；最优控制；鲁棒性 Robustness of Nonlinear Stochastic Optimal Control of Uncertain Structure Systems MAO Bo, YING Zu-guang, ZHU Wei-qiu (School of Aeronautics and Astronautics, Zhejiang University, Hangzhou 310027, China) Abstract: For the n-storey shear-type structure with Guassian-distributed uncertain parameters and subjected to transverse control forces and a wide-band stochastic support excitation with rational spectral density, the average parametric nominal system is established based on the independence of parametric uncertainty and stochastic excitations. The nominal system is transformed into a 2n-dimensional controlled, stochastically excited and dissipative integrable Hamiltonian system by modal transformation. Then, the nonlinear stochastic optimal control law is obtained by using the stochastic averaging method for quasi integrable Hamiltonian systems and stochastic dynamical programming principle. For the controlled uncertain system, the means and standard deviations of the mean-square responses and mean-square control force are calculated by using the stochastic averaging method and the probabilistic analysis, the average control effectiveness and control efficiency are estimated, and the robustness of the nonl