基于模拟退火算法的设计结构矩阵分析优化

DSM Analysis and Optimization Based on Simulated Annealing Algorithm

研究产品设计过程效率问题,设计结构矩阵对产品设计过程进行建模,依据设计结构矩阵的原理及优化理论,通过图论方法分析优化了简单设计过程,针对强耦合设计任务块的问题,无法使用图论方法寻找耦合任务块和对耦合任务块进行优化,为了缩短生产过程的执行时间,使用模拟退火算法进行DSM分析优化。在算法的设计上对DSM进行n进制数字编码,新状态函数产生采用n进制编码两点互换法,新状态接受函数使用Metropolis重点抽样原则产生,根据实际问题的规模选取了直线降温的退火函数,仿真试验证明,合理选取终止温度作为最优解的终止条件,应用上述关键技术可以取得满意的结果,虚拟卫星太阳翼为例验证了算法的可行性。

In this paper,design structure matrix（DSM） is used to model the product design process.The weak coupling design process is analyzed and optimized by graph theory,basing on the principles of DSM and the optimization theory of DSM.However,such strong problems coupling task blocks can not been found and optimized by graph theory.DSM is analyzed and optimized creatively by Simulated Annealing Algorithm.The idea of the transformation from DSM to Simulated Annealing Algorithm is put forward.During designing the algorithms,n-D coding about DSM model is presented.The new state function using n-D coding is obtained through exchanging two positions,the new state accepting function is produced by using Metropolis focus sampling principle.Based on the actual scale of the problem,straight-line cooling function is selected.After numerous tests,a reasonable temperature is selected as the optimal solution to terminate the termination conditions.Finally,by the reasonable application of the above key technologies,the virtual satellite solar wing is adapted as an example to verify the feasibility of the application with the strong coupling.