灰色模型和BP神经在闸坝调度管理评估中的应用

发布时间：2018-03-03 16:02

本文选题：闸坝调度管理评估　切入点：BP神经网络　出处：《中国科学技术大学》2017年硕士论文　论文类型：学位论文

【摘要】：闸坝调度管理情况反映着闸坝应对突发状况的能力,对闸坝调度管理评价进行研究具有较为重要的理论和现实指导意义。闸坝调度管理评价指标体系构建的正确和完整与否以及评价方法选择得当与否直接关系到是否能够真实全面地反映闸坝调度管理情况,科学、合理的评价指标体系与正确的评价方法在闸坝管理、闸坝应急调度等方面都起着重要的作用。论文通过深入探讨闸坝调度管理评估研究背景、意义及国内外闸坝调度管理评估研究现状,发现现有的闸坝评估体系只是考虑到了水质、水量的调度评价,考虑的只是闸坝本身而非闸坝调度管理情况及水生态相关影响。所以,现有的闸坝评估体系,存在不足。本文重新思考与定位现有的闸坝调度管理评价体系,改革原有的不考虑水生态影响和管理相关情况的评价模式,建立了科学的闸坝调度管理评估新评价体系,通过层次分析(AHP)法求解评价指标的权重,为后续基于灰色GM(1,1)模型和优化的BP人工神经网络的评价方法奠定基础。建立闸坝评估体系后,本文基于GA优化的BP人工神经网络(GA-BP)建立了闸坝调度管理评估模型,基于灰色GM(1,1)模型和优化的BP人工神经网络(GM-BP)构建了闸坝调度管理能力评估串并联组合模型并进行了实例分析。通过实例分析,本着科学性、动态性、层次性和实用性的原则,进行综合评判,检验该模型的科学程度及可操作性。同时,本文提出的评价方法与其他的评价方法,灰色关联分析法、TOPSIS法和FCE法,共7种方法进行综合对比验证,确定选取评估方法的正确性。实例结果表明,基于灰色GM(1,1)模型和优化的BP人工神经网络的方法在处理多层次的、复杂的以及信息不完全的闸坝调度管理能力评估问题方面具有有效性、可行性和参考性,为闸坝调度管理评估问题提供了有益的探索和实践。本论文提出的评估方法,不仅会在淮河流域闸坝调度管理评估工作中发挥重要作用,同时还会推广应用到全国其它类似河流。因此,本论文取得的成果在淮河流域以及全国其它类似地区都具有很好的推广和应用前景。
[Abstract]:The operation and management of the gate and dam reflect the ability of the gate and dam to deal with the sudden situation. It is of great theoretical and practical significance to study the evaluation of gate and dam dispatching management. Whether the evaluation index system of gate and dam dispatching management is correct and complete or not and whether the evaluation method is appropriate or not is directly related to. Whether it can truly and comprehensively reflect the operation and management situation of the gate and dam, Scientific and reasonable evaluation index system and correct evaluation method play an important role in gate dam management, gate dam emergency dispatch and so on. It is found that the existing assessment system only takes into account the water quality and the quantity of water, and the evaluation of water quality and quantity of water is only taken into account in the evaluation of water quality and quantity of water. Only the dam itself is considered, not the dam dispatching and management situation and the influence of water ecology. Therefore, the existing dam assessment system is inadequate. This paper reconsiders and orientates the existing dam dispatching management evaluation system. By reforming the original evaluation model which does not consider the ecological impact of water and the related situation of management, a new evaluation system for dam and gate dispatching management is established, and the weight of the evaluation index is calculated by using the analytic hierarchy process (AHP) method. In order to lay a foundation for the subsequent evaluation method based on the grey GMM-1) model and the optimized BP artificial neural network. After the establishment of the gate dam evaluation system, this paper establishes the dam dispatching management evaluation model based on the GA optimized BP artificial neural network (GA-BP). Based on the grey GMX1) model and the optimized BP artificial neural network (GM-BP1), a series-parallel combined model for evaluating the dispatching and management ability of the gate dam is constructed and an example analysis is carried out. Through the example analysis, based on the principles of scientific, dynamic, hierarchical and practicability, At the same time, the evaluation method proposed in this paper is compared with other evaluation methods, grey relational analysis method, TOPSIS method and FCE method. The validity of the selected evaluation method is determined. The results of an example show that the method based on the grey GM1 / 1) model and the optimized BP artificial neural network is used to deal with multi-level, The evaluation problem of complex and incomplete dam dispatching management ability is effective, feasible and referential, which provides a useful exploration and practice for the assessment of gate and dam dispatching management. Not only will it play an important role in the management and evaluation of dam and gate operation in the Huaihe River Basin, but it will also be popularized and applied to other similar rivers throughout the country. The results obtained in this paper have a good prospect of popularization and application in Huaihe River Basin and other similar areas in China.
【学位授予单位】：中国科学技术大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TV698;X52

【参考文献】