基于人工螢火蟲算法優(yōu)化gm(1,1)模型在變形監(jiān)測數(shù)據(jù)處理中的應(yīng)用研究

1、 摘要 畢業(yè)論文題目：基于人工螢火蟲算法優(yōu)化 GM(1 ,1)模型在變形監(jiān)測數(shù)據(jù)處 理中的應(yīng)用研究 大地測量學(xué)與測量工程專業(yè) 2010 級碩士生姓名 郭旭平 指導(dǎo)教師（姓名、職稱） ： 吳良才 教授

2、 摘 要 目前，如何實時、科學(xué)、準(zhǔn)確地分析和預(yù)報建筑物、構(gòu)筑物的變形，已成為科學(xué)研究以及指導(dǎo)實踐的一個重大課題。本文通過變形監(jiān)測數(shù)據(jù)處理及分析的預(yù)測模型——GM(1,1)模型的國內(nèi)外研究現(xiàn)狀，以及一種新型的智能優(yōu)化算法——人工螢火蟲算法的國內(nèi)外的方研究現(xiàn)狀，搜集大量資料，對比分析，構(gòu)思自己研究的方法，采用人工螢火蟲算法將灰色GM(1,1)模型進(jìn)行優(yōu)化，應(yīng)用于變形監(jiān)測預(yù)報與分析工作中，其主要研究內(nèi)容包括：

3、本文利用最小一乘法建立目標(biāo)函數(shù)，再通過人工螢火蟲算法優(yōu)化，求解目標(biāo)函數(shù)，從而使GM(1,1)模型得到優(yōu)化，隨后給出了優(yōu)化模型的實施步驟以及基本流程。在對預(yù)測效果檢驗時，用該優(yōu)化模型計算參考文獻(xiàn)里數(shù)據(jù)，并與文獻(xiàn)中的其它模型相比較，證明優(yōu)化模型的可行性及優(yōu)越性。把人工螢火蟲算法優(yōu)化的GM(1,1)模型應(yīng)用到變形監(jiān)測中，對沉降數(shù)據(jù)進(jìn)行處理、分析。并從殘差、相對誤差、發(fā)展系數(shù)、后驗方差值c四個方面比較和驗證優(yōu)化模型，利用實驗結(jié)果證明研究方法的可

4、行性，優(yōu)化模型的優(yōu)越性。實例結(jié)果顯示，人工螢火蟲算法優(yōu)化的GM(1,1)模型在殘差、相對誤差、發(fā)展系數(shù)、后驗方差值c等方面的確優(yōu)越GM(1,1)模型，本文采用的模型進(jìn)一步擴大了灰色GM(1,1)模型應(yīng)用的范圍。 關(guān)鍵詞： 關(guān)鍵詞：變形監(jiān)測；灰色系統(tǒng)理論；人工螢火蟲算法；GM(1,1)模型；最小一乘法 Abstract THESIS: Based on glowworm swarm optimization GM (1, 1) mode

5、l Settlement data application study in Settlement Monitoring SPECIALIZATION: Geodesy and Survey Engineering POSTGRADUATE: Xuping-Guo

6、 MENTOR: Professor Liangcai-WU Abstract Nowadays, how can real-timely, scientifically and accurately analysis and forecasting in the deformation

7、 of buildings, and structures. It has become a major topic in the practice of scientific research and guidance. In this paper, deformation monitoring data processing and analysis , prediction model- GM (1,1) model resear

8、ch status at home and abroad as well as a new type of intelligent optimization algorithm-glowworm swarm optimization square Research,collect large amounts of data , comparative analysis .the glowworm swarm optimization o

9、f gray GM (1,1) model is applied to deformation monitoring and forecasting and analysis, it mainly includes, as the follows: This article builds objective function with the Logarithm solves the objective function with th

10、e glowworm swarm optimization. So that the GM (1,1) model has been optimized. At the same time, it given the implementation steps as well as the basic processes of the optimization model .Tested the predicted effect, the

11、 optimization model for calculating the reference literature data and compared with other models in the literature, to prove the feasibility and superiority of the optimization model. After the GM (1,1) model of the glow

12、worm swarm optimization is applied to the deformation monitoring, on settlement data processing, analysising.From the residuals, the relative error, the development coefficient the posterior variance c value of the four

13、aspects of comparison and verification optimization model, which showed the superiority of the optimization model. The experimental results show the feasibility of the research methods, indicating the superiority of the

14、optimization model. The results of examples show glowworm swarm optimization optimized GM (1,1) model is indeed superior in terms of residuals , relative error , the coefficient of development , after the prescription di

15、fference c in the GM (1,1) model , The model used in this paper expand the range of gray GM (1,1) model application . Keywords: deformation monitoring, gray system theory, GM (1,1) model, logarithm, glowworm swarm optimi