【摘要】：在利用測量數(shù)據(jù)確定舊路平面線形擬合參數(shù)初始值的問題上,相關(guān)理論技術(shù)還不夠完善。而舊路平面線形擬合參數(shù)初始值在舊路改造工程中,對于舊路平面線形調(diào)整、提高舊路平面線形擬合的準(zhǔn)確性等都起著比較重要的作用。本文在長沙理工大學(xué)劉蘇教授“舊路平面線形計(jì)算機(jī)輔助擬合技術(shù)研究”的基礎(chǔ)上,結(jié)合G321線改擴(kuò)建工程對舊路平面線形擬合參數(shù)初始值進(jìn)行進(jìn)一步研究。舊路平面線形擬合參數(shù)初始值包括直線擬合參數(shù)初始值和曲線擬合參數(shù)初始值。本文利用一元線性回歸最小二乘擬合得到各直線方程,確定出直線擬合參數(shù)初始值。曲線擬合參數(shù)初始值確定包括圓曲線半徑、圓心坐標(biāo)和緩和曲線長度的確定。緩和曲線長度在圓曲線半徑和圓心坐標(biāo)確定的基礎(chǔ)上,通過相應(yīng)的函數(shù)關(guān)系唯一確定,相對簡單。因此,確定平面線形擬合參數(shù)初始值的關(guān)鍵在于如何求取圓曲線擬合參數(shù)初始值(包括圓曲線半徑和圓心坐標(biāo))。本文主要針對圓曲線擬合點(diǎn)數(shù)量大于3個(gè)的情況進(jìn)行分析,提出了一種新的圓曲線擬合方法一多組合平均值法,以提高圓曲線擬合參數(shù)初始值的精度。在確定圓曲線擬合點(diǎn)的前提下,分別采用不同組合方法取平均值求取圓曲線擬合參數(shù)初始值,每種平均值組合都有對應(yīng)的圓曲線擬合中誤差,將中誤差最小的組合對應(yīng)的圓曲線擬合參數(shù)初始值作為最終的圓曲線擬合參數(shù)初始值。同時(shí)將多組合平均值法與其它方法作比較,驗(yàn)證它的適用性。利用VB語言編寫程序?qū)崿F(xiàn)了多組合平均值法在平面線形擬合中的應(yīng)用。通過多個(gè)路段圓曲線擬合CAD圖論證了圓曲線擬合中誤差的大小取決于圓曲線擬合點(diǎn)的離散度大小,分析了離散度的影響因素,確定路面分中誤差是引起路面分中點(diǎn)離散度的主要原因,通過采用相應(yīng)辦法,提高路面分中的準(zhǔn)確性。本文結(jié)合321國道舊路改造項(xiàng)目進(jìn)行實(shí)例分析研究,利用VB程序(其結(jié)合了一元線性回歸最小二乘擬合原理和多組合平均值法)確定了全路段的平面線形擬合參數(shù)初始值,最后根據(jù)規(guī)范及對稱緩和曲線的原則,對計(jì)算得到的舊路平面線形擬合參數(shù)初始值進(jìn)行調(diào)整,得到調(diào)整后的舊路平面線形擬合參數(shù)初始值。
[Abstract]:On the problem of determining the initial values of the plane linear fitting parameters of the old road by using the measured data, the relevant theoretical techniques are not perfect enough. The initial values of the plane alignment fitting parameters of the old road play an important role in the old road reconstruction project, such as adjusting the plane alignment of the old road and improving the accuracy of the plane alignment fitting of the old road. On the basis of Professor Liu Su of Changsha University of Science and Technology "Research on computer aided fitting Technology of plane alignment of Old Road", this paper further studies the initial value of plane alignment fitting parameter of old road combined with G321 line reconstruction and expansion project. The initial values of the old plane linear fitting parameters include the initial values of the linear fitting parameters and the initial values of the curve fitting parameters. In this paper, the linear equations are obtained by linear regression least square fitting, and the initial values of the fitting parameters are determined. The initial values of curve fitting parameters include the radius of circle curve, the coordinate of circle center and the length of transition curve. On the basis of the radius of circle curve and the coordinate of circle center, the length of transition curve is determined only by the corresponding function relation, which is relatively simple. Therefore, the key to determine the initial values of plane linear fitting parameters is how to obtain the initial values of circular curve fitting parameters, including the radius of circular curve and the coordinate of circle center. In this paper, the number of fitting points of circular curve is more than 3, and a new method of circular curve fitting-multi-combination average method is proposed to improve the accuracy of initial value of circular curve fitting parameters in order to improve the accuracy of the initial value of the fitting parameters of circular curve. On the premise of determining the fitting point of the circular curve, the initial values of the fitting parameters of the circular curve are calculated by means of different combination methods, and each combination of the average values has the corresponding error in the fitting of the circular curve. The initial value of the fitting parameter of the circular curve corresponding to the combination with the minimum error is taken as the initial value of the final fitting parameter of the circular curve. At the same time, the multi-combinatorial average method is compared with other methods to verify its applicability. The application of multi-combination average method in plane linear fitting is realized by VB programming. Through the CAD diagram of circle curve fitting of several sections, it is proved that the error of circle curve fitting depends on the discreteness of the fitting point of circle curve, and the influencing factors of the degree of dispersion are analyzed. In order to improve the accuracy of road surface grading, the main reason for the dispersion degree of road surface center point is to determine the error of road surface centralization, and adopt the corresponding method to improve the accuracy of road surface grading. Based on the case study of the old road reconstruction project of National Highway 321, the initial values of plane linear fitting parameters of the whole road are determined by using the VB program (which combines the principle of linear regression least square fitting and the multi-combination average method). Finally, according to the criterion and the principle of symmetry transition curve, the initial values of the calculated old road plane alignment fitting parameters are adjusted, and the original values of the adjusted old road plane alignment fitting parameters are obtained.
【學(xué)位授予單位】：長沙理工大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2014
【分類號】：U412;U418.8

【參考文獻(xiàn)】

相關(guān)期刊論文 前1條

1 廖若宇;陳雨人;;基于數(shù)字圖像的道路平面線形恢復(fù)的研究[J];交通信息與安全;2011年02期

，

本文編號：2443932