【摘要】：近年來,國(guó)內(nèi)多地災(zāi)難性的橋梁倒塌事故頻頻發(fā)生,引發(fā)了人們對(duì)橋梁結(jié)構(gòu)安全問題的廣泛關(guān)注。橋梁結(jié)構(gòu)健康監(jiān)測(cè)也以成為當(dāng)今橋梁工程研究的熱點(diǎn),移動(dòng)荷載識(shí)別技術(shù)能進(jìn)一步完善健康監(jiān)測(cè)系統(tǒng)。然而,現(xiàn)有的多數(shù)荷載識(shí)別方法存在不同的局限性,需要實(shí)際工程的進(jìn)一步論證。分別針對(duì)單車和多車移動(dòng)荷載的高精度識(shí)別,本文提出了基于影響線并考慮荷載橫向分布的單車移動(dòng)荷載識(shí)別方法以及基于影響線和BP神經(jīng)網(wǎng)絡(luò)的多車移動(dòng)荷載識(shí)別方法,并用數(shù)值模擬和實(shí)驗(yàn)室車橋試驗(yàn)對(duì)上述方法的可行性和適用性進(jìn)行了驗(yàn)證。本文主要研究?jī)?nèi)容及結(jié)論如下：(1)為了提高移動(dòng)荷載的識(shí)別精度,提出并建立了基于影響線并考慮荷載橫向分布的移動(dòng)荷載識(shí)別方法,并結(jié)合數(shù)值模擬算例,對(duì)未考慮荷載橫向分布的方法和考慮了荷載橫向分布的方法進(jìn)行了比較,結(jié)果表明未考慮荷載識(shí)別的方法不適合解決空間問題,而考慮了荷載橫向分布的方法,對(duì)于車輛行駛于橋面上任意位置的空間問題,其識(shí)別精度都很高,且抗噪性能優(yōu)異。(2)通過實(shí)驗(yàn)室車橋試驗(yàn),進(jìn)一步研究了考慮荷載橫向分布的影響線方法在實(shí)際工程應(yīng)用中的可行性。車速識(shí)別的結(jié)果顯示,車速識(shí)別的誤差基本能控制在±5%以內(nèi),且車速識(shí)別的精度對(duì)車重識(shí)別影響較大。車重識(shí)別的結(jié)果表明,利用考慮荷載橫向分布的影響線方法進(jìn)行荷載識(shí)別,車重相對(duì)誤差可控制在±10%以內(nèi),且93%的樣本能控制在±5%以內(nèi)。(3)由于無法建立理想的數(shù)學(xué)模型對(duì)多車移動(dòng)荷載進(jìn)行識(shí)別,本文根據(jù)應(yīng)變影響線中包含車重信息,再結(jié)合BP神經(jīng)網(wǎng)絡(luò)方法,建立了基于影響線和BP神經(jīng)網(wǎng)絡(luò)的移動(dòng)荷載識(shí)別方法。然后針對(duì)該方法在實(shí)驗(yàn)室進(jìn)行了雙車移動(dòng)荷載試驗(yàn)研究,結(jié)果表明該方法能精確識(shí)別車輛所處車道位置信息,在識(shí)別車輛荷載時(shí),所有樣本的相對(duì)誤差能控制在±10%以內(nèi),且97%的樣本能控制在±5%以內(nèi)。
[Abstract]:In recent years, many catastrophic bridge collapse accidents occurred frequently in China, which caused widespread concern about bridge structure safety. Health monitoring of bridge structures has also become a hot topic in bridge engineering nowadays. Moving load identification technology can further improve the health monitoring system. However, most existing load identification methods have different limitations and need further demonstration of practical engineering. Aiming at the high precision identification of moving load of single vehicle and multi-vehicle, this paper presents a method of identification of moving load of single vehicle based on influence line and considering the transverse distribution of load, and a method of identification of moving load of multi-vehicle based on influence line and BP neural network. The feasibility and applicability of the above method are verified by numerical simulation and vehicle bridge test in laboratory. The main contents and conclusions of this paper are as follows: (1) in order to improve the accuracy of moving load identification, a moving load identification method based on the influence line and considering the transverse distribution of load is proposed and established. The method without considering the transverse distribution of load is compared with the method of considering the transverse distribution of load. The results show that the method without load identification is not suitable to solve the spatial problem, but the method of transverse distribution of load is taken into account. For the spatial problem of vehicle traveling on the bridge deck at any position, its recognition accuracy is very high, and the anti-noise performance is excellent. (2) through the test of the vehicle bridge in the laboratory, The feasibility of the influence line method considering the transverse distribution of load in practical engineering is further studied. The result of speed recognition shows that the error of speed recognition can be controlled within 鹵5%, and the accuracy of speed recognition has great influence on the recognition of vehicle weight. The result of vehicle weight identification shows that the relative error of vehicle weight can be controlled within 鹵10% by using the influence line method considering the transverse distribution of load. And 93% of the samples can be controlled within 鹵5%. (3) because it is impossible to establish an ideal mathematical model to identify the moving loads of multiple vehicles, according to the information of vehicle weight contained in the strain influence line, and then combined with the BP neural network method, A moving load identification method based on influence line and BP neural network is established. The experimental results show that the method can accurately identify the lane position information of the vehicle, and the relative error of all samples can be controlled within 鹵10% when the vehicle load is identified. And 97% of the samples could be controlled within 鹵5%.
【學(xué)位授予單位】：東南大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2015
【分類號(hào)】：U446

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 宋海良;;國(guó)內(nèi)橋梁垮塌事故的分析與反思[J];交通世界(建養(yǎng).機(jī)械);2012年08期

2 張世英;;鋼筋銹蝕對(duì)超載鋼筋混凝土梁式橋的裂縫影響分析[J];交通世界(建養(yǎng).機(jī)械);2012年07期

3 余本國(guó);;BP神經(jīng)網(wǎng)絡(luò)局限性及其改進(jìn)的研究[J];山西農(nóng)業(yè)大學(xué)學(xué)報(bào)(自然科學(xué)版);2009年01期

4 譚冬蓮;;基于影響線理論應(yīng)用監(jiān)測(cè)信息反演橋上車輛荷載[J];力學(xué)與實(shí)踐;2008年02期

5 袁向榮;;梁振動(dòng)響應(yīng)曲線滑動(dòng)擬合法及在移動(dòng)荷載識(shí)別中的應(yīng)用[J];噪聲與振動(dòng)控制;2006年03期

6 于秀娟;余有龍;張敏;廖延彪;賴淑蓉;;鈦合金片封裝光纖光柵傳感器的應(yīng)變和溫度傳感特性研究[J];光電子·激光;2006年05期

7 譚金華;陳惟珍;程飛;;基于運(yùn)營(yíng)狀態(tài)監(jiān)測(cè)數(shù)據(jù)識(shí)別過橋車輛荷載[J];橋梁建設(shè);2006年01期

8 張曉文,楊煜普,許曉鳴;神經(jīng)網(wǎng)絡(luò)傳遞函數(shù)的功能分析與仿真研究[J];計(jì)算機(jī)仿真;2005年10期

9 馬翔;陳新楚;王劭伯;;均勻設(shè)計(jì)法在RBF神經(jīng)網(wǎng)絡(luò)樣本優(yōu)選中的應(yīng)用[J];模式識(shí)別與人工智能;2005年02期

10 詹亞歌,蔡海文,耿建新,瞿榮輝,向世清,王向朝;鋁槽封裝光纖光柵傳感器的增敏特性研究[J];光子學(xué)報(bào);2004年08期

相關(guān)博士學(xué)位論文 前1條

1 王寧波;非路面式橋梁動(dòng)態(tài)稱重理論與試驗(yàn)研究[D];中南大學(xué);2013年

相關(guān)碩士學(xué)位論文 前6條

1 張劍超;關(guān)于橋梁荷載橫向分布系數(shù)的研究[D];武漢理工大學(xué);2011年

2 韓清海;中小跨徑橋梁荷載橫向分布系數(shù)計(jì)算方法的研究及其應(yīng)用[D];吉林大學(xué);2009年

3 陳修輝;基于神經(jīng)網(wǎng)絡(luò)的橋梁移動(dòng)荷載識(shí)別[D];西南交通大學(xué);2009年

4 劉軍;神經(jīng)網(wǎng)絡(luò)學(xué)習(xí)算法研究[D];江西師范大學(xué);2009年

5 王波;基于BP神經(jīng)網(wǎng)絡(luò)的橋上移動(dòng)荷載識(shí)別[D];天津大學(xué);2006年

6 張劍飛;貝葉斯網(wǎng)絡(luò)學(xué)習(xí)方法和算法研究[D];東北師范大學(xué);2005年

，

本文編號(hào)：2268251