自适应布谷鸟算法及其在化工优化中的应用

Adaptive cuckoo search algorithm and its application to chemical engineering optimization problem

动态优化问题广泛存在于化工自动控制过程中,对其求解是化工过程工业发展的一个不可忽视的环节。群智能算法求解此类优化问题时不可避免地存在后期收敛速度慢、求解精度的不高等不足,这一直是一个研究热点。针对新兴的布谷鸟算法与以上问题,提出一种变步长自适应布谷鸟搜索算法(VSACS),将基本布谷鸟搜索(CS)算法中的随机步长改进成根据迭代次数自适应调整的步长。通过15个标准测试函数的测试,结果验证了改进的算法有较快的收敛速度和较高的求解精度。最后将改进的算法用于批示反应器、管式反应器、生物反应器等3个典型的化工动态优化问题中,获得了满意的实验结果,同时也进一步表明该算法的有效性。

Dynamic optimization is widely exists in the chemical automatic control process, it is also an inevitable trend of the industrial development of chemical process. What＇s more, it is considerable to find the solution of dynamic optimization problems. Recent years, it attracts more and more researchers＇ attention that swarm intelligence algorithms simulating some natural phenomenon play an important part in solving optimization problems. In the meantime, there exist some defects such as slow convergence rate in global search and low accuracy in local search. Therefore, variable step adaptive cuckoo search （VSACS） is proposed to cope with the problem of slow convergence and low precision of the cuckoo search （CS） algorithm. The step is decreased dynamically along with the increase of iteration to make sure the algorithm can get better solution. Experimental results of 15 standard test functions show faster convergence rate and higher accuracy. At last, the experimental results testify the algorithm is a valid method to solve chemical process dynamic optimization problem.