DNA double helix based hybrid genetic algorithm for the gasoline blending recipe optimization problem

Jili Tao,Ning Wang
DOI: https://doi.org/10.1002/ceat.200700322
IF: 2.2148
2008-01-01
Chemical Engineering & Technology
Abstract:A hybrid genetic algorithm is proposed for heavily nonlinear constrained optimization problems by utilizing the global exploration and local exploitation characteristics, and the convergence rate of the proposed algorithm is analyzed. In the global exploration phase, a DNA double helix structure is used to overcome Hamming cliffs and DNA computing based operators are applied to improve the global searching capability. When the feasible domains are located, the sequential quadratic programming (SQP) method is performed to quickly find the local optimum and improve the solution accuracy. The comparison results of typical numerical examples and the gasoline blend recipe optimization problem are employed to demonstrate the reliability and efficiency of the proposed algorithm.
What problem does this paper attempt to address?