基于GA-BP神经网络的客船人员疏散时间预测模型的建立
Passenger Ship Evacuation Time Prediction Model Based on GA-BP Neural Network
徐镜涵 1任玉清 1王庸凯1
作者信息
- 1. 大连海洋大学航海与船舶工程学院,辽宁大连 116023
- 折叠
摘要
针对客船人员疏散问题,设置了影响疏散时间的多个参数,构建了 3×3×3×5×6=810种不同的疏散场景,利用Pathfinder人员疏散仿真软件获得了 810组人员疏散时间的模拟数据,并在此基础上建立两种不同的预测模型:一种基于BP神经网络,另一种则结合了遗传算法的GA-BP神经网络,对其进行实验性能对比.结果表明,优化后的神经网络预测准确率达到97.7622％,相较于优化前提高了 2.5394％.采用遗传算法对神经网络进行优化,训练时间有所减少,准确性大大提高,可为船舶应急处理提供科学依据和新的技术手段.
Abstract
In order to solve the problem of passenger evacuation,several parameters affecting evacuation time were set,and 810 different evacuation scenarios with 3 × 3 × 5 × 6=were constructed,the simulated data of evacuation time for 810 groups were obtained with Pathfinder evacuation simulation software.On this basis,two different prediction models are established:one was based on BP neural network,and the other was GA-BP neural network combined with genetic algorithm.The results show that the prediction accuracy of the optimized neural network is 97.7622％,which is 2.5394％higher than that before optimization.Genetic algorithm is used to optimize the neural network,which not only reduces the training time,but also improves the accuracy.This can provide scientific basis and new technical means for ship emergency treatment.
关键词
客船疏散/预测模型/BP神经网络/遗传算法Key words
Evacuation of passenger ships/Prediction model/BP neural network/Genetic algorithm引用本文复制引用
基金项目
农村农业部渔业渔政管理局资助项目(070522056)
出版年
2024
NETL
NSTL
万方数据