摘要
本文提出了一个解(0,1)规划的新的数学方法,即(0,1)规划的二进制数法,并将该法应用于离散变量结构优化设计,针对结构优化设计的特点,提出了约束条件中的设计变量系数皆为正,且约束方程皆不小于零的线性约束情况下的(0,1)规划的另一修正解法。这两个解法都不需要解任何方程,从而大大节省机时。针对本文的方法,编写了Fortran语言的计算机程序,并在微机上计算了一些算例,结果充分说明了本法的有效性。
In this paper, a new algorithm is proposed for solving a zero-one programming (the objective is linear function of the design variables, and the constraints may be arbitrary functions of the design variables) .It is named binary number combinatorial algorithm.For the discrete optimum des.ign problems of structures, at first, the authors translate the problem linto a linear 0-1 programming, and then the above algorithm is applied and modified based on the character of the structure design,No matter which of the original or the modified algorithm provides an efficient method for solving the corresponding problems,because in these two algorithms,it is not needed to solve equations, and so the efforts for mathematical calculations are greatly saved.The authors have programmed the problem in computer code with Fortran 77 language. Some numerical results show that these algorithm are more efficient than the others which have been seen so far,
