灾情评估的结果既是制定防洪减灾决策的基本依据,也是评价防洪工作成效的重要指标。提出了一种基于免疫遗传算法优化的神经网络模型(IGA—NN)并用于洪水灾情的评估,给出了IGA—NN的设计方法和优化步骤。同时还将ICA—NN与标准遗传算法优化的神经网络模型进行了比较,显示了IGA—NN的优越性。最后通过实例,验证了IGA—NN的实用性。
The result of disaster evaluation is not only the basis for decision-making to flood control but also the important index of judging flood control effect. In this paper, an immune genetic algorithm based on neural network model (IGA-NN) for flood disaster evaluation is proposed. The design method and optimization procedures of the IGA-NN is provided. Also, this paper demonstrates superiority of the IGA-NN by comparing it with the neural network optimized by standard genetic algorithm. Finally, the practicality of the IGA-NN is verified by practical application.