基于改进遗传算法优化的BP神经网络水利调度优化
DOI:
作者:
作者单位:

1.重庆邮电大学 软件工程学院;2.重庆邮电大学 计算机科学与技术学院

作者简介:

通讯作者:

中图分类号:

TP181

基金项目:

国家自然科学基金面上项目(6172099);重庆市高校创新团队建设项目(CXTDG201602010);重庆市“三百”科技创新领军人才支持计划(CSTCCXLJRC201917);重庆市高校优秀成果转化资助项目(KJZH17116);重庆市人工智能技术创新重大主题专项(CSTC2017rgzn - zdyf0140);重庆市创新创业示范团队培育计划(CSTC2017kjrc - cxcytd0063);重庆市技术创新与应用示范重大主题专项项目(CSTC2018JSZX - CYZTZX0178,CSTC2018JSZX - CYZTZX0185);重庆市基础科学与前沿技术研究项目(CSTC2017jcyjAX0270, CSTC2018jcyjA0672, CSTC2017jcyjAX0071)


Optimization of BP Neural Network Water Conservancy DispatchingBased on Improved Genetic Algorithm Optimization
Author:
Affiliation:

1.School of Software Engineering,Chongqing University of Posts and Telecommunications;2.PRChina;3.Schoolof Computer Science and Technology,Chongqing University of Posts and Telecommunications

Fund Project:

National Natural Science Foundation of China (6172099); Chongqing University Innovation Team Building Project (CXTDG201602010); Chongqing "300" Science and Technology Innovation Leader Talents Support Plan (CSTCCXLJRC201917); Chongqing University Excellent Achievements Transfer Subsidy Project (KJZH17116); Chongqing Artificial Intelligence Technology Innovation Major Theme Project (CSTC2017rgzn - zdyf0140); Chongqing Municipal Innovation and Entrepreneurship Demonstration Team Cultivation Program (CSTC 2017kjrc-cxcytd0063); Chongqing Technical Innovation and Application Demonstration Major Theme Project (CSTC 2018JSZX-CYZTZX0178, CSTC 2018JSZX-CYZTZX0185); Chongqing Basic Science and Frontier Technology Research Project (CSTC2017jcyjAX0270, CSTC2018jcyA0672, CSTC2017jcyAX0071)

  • 摘要
  • |
  • 图/表
  • |
  • 访问统计
  • |
  • 参考文献
  • |
  • 相似文献
  • |
  • 引证文献
  • |
  • 资源附件
  • |
  • 文章评论
    摘要:

    随着现代水利调度规模的扩大,内涵深化和要求的不断提高,水利调度领域日益面临着来自各个方向的困难与挑战,其中包括水利调度时的动态变化,水资源的流量不可控,高纬度的决策向量以及多目标,计算量庞大等困难。面对如此纷繁的问题本文只在提出一种基于粒子群算法改进的遗传算法优化BP神经网络,通过重构变异算子克服遗传算法易陷入局部极值的缺陷,再通过BP神经网络采用改进遗传算法输出的最优阈值,权值训练模型。达到一个良好的水利调度优化效果。

    Abstract:

    With the expansion of the scale of modern water dispatching, the deepening of its connotation and the continuous improvement of its requirements, the field of water dispatching is increasingly faced with difficulties and challenges from various directions, including the dynamic changes in water dispatching, the uncontrollable flow of water resources, the decision vector of high latitude, multi-objective and huge calculation. Faced with such a complex problem, this paper only proposes an improved genetic algorithm based on particle swarm optimization to optimize BP neural network. By reconstructing mutation operator to overcome the shortcomings of genetic algorithm easily falling into local extremum, and then through BP neural network to adopt the optimal threshold of improved genetic algorithm output, weight training model. It achieves a good effect of water dispatching optimization.

    参考文献
    相似文献
    引证文献
引用本文
分享
文章指标
  • 点击次数:
  • 下载次数:
  • HTML阅读次数:
  • 引用次数:
历史
  • 收稿日期:2019-07-17
  • 最后修改日期:2019-10-27
  • 录用日期:2019-10-29
  • 在线发布日期:
  • 出版日期: