带共轭梯度算子的爆炸搜索算法

Explosion Search Algorithm with Conjugate Gradient Operator

爆炸搜索算法作为一种全局优化算法,在迭代后期会出现收敛速度慢、精度低的问题,而传统的优化算法恰好能克服这些缺点。因此,引入一种传统算法——近似共轭梯度法,即用差商代替导数的共轭梯度法。在此基础上,提出了带共轭梯度算子的爆炸搜索算法,先引入了新的变异算子来提高算法的全局搜索能力,再运用共轭梯度法添加一个新的算子——共轭梯度算子,实现对最优炸点的局部搜索,从而提高算法的收敛速度与精度。6个常用的benchmark函数的测试结果说明,改进算法的优化结果明显优于原算法。

As a global optimization algorithm,Explosion Search Algorithm （ESA） has some problem of low convergence speed and low optimization precision in the later period of the optimization.Fortunately,some deterministic optimization algorithms can overcome these shortcomings.Therefore,a deterministic algorithm without derivate information,which is called approximate conjugate gradient algorithm using difference quotient,was added in ESA.Based on the above,an improved Explosion Search Algorithm with Conjugate Gradient Operator （CGESA） was proposed.In CGESA,a new mutation operator is introduced to enhance the global search ability.Meanwhile,a new operator is introduced namely conjugate gradient operator to improve the local search ability of the optimal burst point,so that the convergence speed and optimization precision of CGESA are improved.Experimental results of the six well-known benchmark functions indicate that CGESA achieves better performance than ESA.