求解一类不可微优化问题极大熵微粒群混合算法

A Maximum-Entropy Particle Swarm Optimization for a Class of Non-Differentiable Optimization Problems

针对一类不可微优化问题,本文提出了一个新的算法—极大熵微粒群混合算法.首先利用极大熵方法把带约束的不可微优化问题转换成无约束的单目标最优化问题,然后利用微粒群算法对其进行求解.利用4个测试函数对其进行测试并于其它算法进行比较,计算结果表明,本文提出算法在求解的准确性和有效性方面均优于其它算法.

To solve a class of non-differentiable optimization problems, this paper proposed a new method called maximum-entropy particles swarm optimization algorithm. First, using the maximum-entropy function, constrained non-differentiable optimization problem can be transformed to the approximation unconstrained differentiable optimization problem. Then using the particles swarm optimization to solve this problem. Four examples were used to demonstrate the validity of the method and the results were compared with the ones of other methods. It is showed that the proposed method is more accurate and effective.