A new simulated annealing method of designing NMR biplanar shim coils

A new SA (simulated annealing) method is presented for the design of longitudinal shim coils for NMR (nuclear magnetic resonance) equipment. First, AGL (adaptive Gauss-Lobatto) quadrature used by the traditional SA method is analyzed. Two kinds of complete elliptic integrals are introduced to integrate the spatial magnetic field, which increases the computing speed by 13.3 times. Then, the system energy calculation is extended to the field of the z-axis. The results illustrate that the coil-set efficiency designed by the new SA method is 15% higher than that of the traditional one, and the field inhomogeneity is reduced from 1.20% to 0.49%. Finally, a prototype of coil assembly for NMR rock-core analyzer is built, and the experimental results are in good agreement with the theoretically calculated results.     

面向多无人机协同对地攻击的双层任务规划方法

无人机(unmanned aerial vehicles, UAVs)的任务规划包含任务分配、执行顺序确定以及航迹优化等。为了达到任务规划的全局最优,需要全盘梳理任务的各个方面,提出高效的优化策略。综合考虑任务规划过程中任务分配、执行顺序确定以及航迹优化等方面的需求和相互间影响,首先从优化框架出发, 设计了双层互耦的任务规划求解策略, 而后将任务规划模型分为上层任务分配和下层任务序列优化, 并对每一层的优化方法和优化步骤进行了详细设计。在任务分配问题中, 基于模拟退火算法, 提出了可跳出局部最优的模拟退火-撒点(simulated-annealing-shooting, SAS)算法, 并详细探讨了算法参数的设计原则。最后通过仿真分析, 验证了所提出的规划框架和SAS优化算法的有效性。     

改进遗传模拟退火算法在多峰值函数优化中的应用

采用均匀变异算子及自适应最优保存的方法对遗传模拟退火算法进行了改进,并用MATLAB实现了改进算法中各算子的编程方法．实例证明该算法能够在多峰值函数优化中较快的收敛至最优解．     

基于模拟退火Memetic算法的复合材料层合板铺层顺序优化

应用基于模拟退火局部寻优的Memetic算法对复合材料层合板的铺层顺序进行优化设计,以层合板的面内几何因子和弯曲因子为优化对象,建立Memetic算法的优化模型。通过遗传操作,搜索问题的最优解,借助于模拟退火进行局部寻优,扩大了算法的寻优范围,加速了算法的收敛速度。数值算例中给定了层合板的面内几何因子和弯曲因子,应用Memetic算法求解层合板的最佳铺层顺序。并应用标准遗传算法和基于模拟退火局部寻优Memetic算法的结果进行比较,然后用有限元进行模拟分析,证明了本文中优化模型的有效性和基于模拟退火Memetic算法的优越性。