论文部分内容阅读
针对整数规划问题提出了一种以植物向光性为启发式准则的智能优化算法——植物多向生长模拟算法.改进了植物生长激素的分配方式并将随机选择机制引入新枝生长方向的选择,更符合植物生长的自然机理.利用马尔可夫链描述算法迭代过程,证明了算法的收敛性.利用无约束和有约束两类具有多个全局最优解的非线性整数规划实例测试了植物多向生长模拟算法的性能,并与基本植物生长模拟算法、填充函数法、罚函数法以及基于遗传算法的混合算法进行了对比.植物多向生长模拟算法不仅提高全局寻优能力,增加解的多样性和准确性而且提高收敛速度.
Aiming at the problem of integer programming, an intelligent optimization algorithm based on plant-light property, a multi-directional plant growth simulation algorithm, is proposed to improve the allocation of plant growth hormone and to introduce the random selection mechanism into the growth direction of new branches. Which is more in line with the natural mechanism of plant growth.Using the iterative process of Markov chain description algorithm to prove the convergence of the algorithm.Unconstrained and constrained two kinds of nonlinear integer programming examples with multiple global optimal solutions were used to test plant The performance of the algorithm is simulated and compared with the basic plant growth simulation algorithm, the filling function method, the penalty function method and the hybrid algorithm based on genetic algorithm. The multi-directional plant growth simulation algorithm not only improves the global optimization ability but also increases the diversity of solutions Sexual and accuracy but also improve convergence speed.