混合进化策略算法在多准则决策中的应用研究

On the Application of the Hybrid Evolution Strategy Algorithm in the Multi-Criteria Decision Making

针对多变量、强约束的复杂非线性规划优化问题,本文将进化策略引入其中,并结合动态及模拟退火罚函数给出了一种快速、易实现的混合求解算法,并成功应用于多准则决策问题中,解决了以往计算难度大的问题。最后通过实例进行验证,结果表明该混合算法效率高、优化性能好。此算法亦可用于其他领域中的类似问题,具有一定借鉴作用和较强的实用性。

For the complex nonlinear programming optimization problem under diversified variables and strong constraints, this paper proposes an effective hybrid algorithm which combines the evolution strategy with the dynamic penalty function and the simulated annealing penalty function. It successfully solves the multi-criteria decision model and greatly reduces the calculating difficulty in the past. At last, a practical example is indicated to verify its feasibility, effectiveness and efficiency. This algorithm can also be used in many similar or relative fields for reference.