您好,欢迎访问三七文档
当前位置:首页 > 办公文档 > 理论文章 > 3-5附有参数的条件平差
1§3-5附有参数的条件平差2学时一、平差原理设条件平差中有观测值n个,必要观测值t个,多余观测数r个,取u个非观测量作为参数(设为xˆ),则要列出的条件方程数为c=r+u(3-5-1)附有参数的条件平差的函数模型为1,1,1,,1,,0~ccuucnncWxBA(3-5-2)用Δ和x~的估值v和xˆ代替,则附有参数的条件平差法的平差值条件方程及改正数条件方程分别为1,1,01,,1,,0ˆˆccuucnncAXBLA(3-5-3)1,1,1,,1,,0ˆccuucnncWxBVA(3-5-4)其中)(00ABXALW(3-5-6)条件平差的随机模型为nnnnnnPQD,120,20,(3-5-7)还应按照函数极值的拉格朗日乘数法,先组建函数)ˆ(2WxBAVKPVVTT(3-5-8)为求Φ的极小值,将Φ分别对V和xˆ求一阶导数,并令其为零022)(2)(AKPVVAVKVPVVdVdTTTT02ˆ)ˆ(2ˆBKxxBKxddTT以上两式转置,得KAPVT(3-5-9)0KBT(3-5-10)由(3-5-9)式得改正数方程KQAKAPVTT1(3-5-11)从而可得附有参数的条件平差的基础方程为200ˆ1,,1,,,11,1,1,1,,1,,cTcucTcnnnnccuucnncKBKAPVWxBVA(3-5-12)将改正数方程代入条件方程后,得1,1,1,,1,,,1,0ˆccuuccTcnnnncWxBKAPA(3-5-13)取TaaAAPN1,不难知道cARAAPRNRTaa)()()(1,Naa为对称可逆方阵,上式写为0ˆWxBKNaa(3-5-14)则)ˆ(1xBWNKaa(3-5-15)上式代入(3-5-10),得0)ˆ(1xBWNBaaT0ˆ11xBNBWNBaaTaaT得0ˆ11WNBxBNBaaTaaT(3-5-16)取BNBNaaTbb1,易知R(Nbb)=u,Nbb为对称可逆方阵。(3-5-16)式写为0ˆ1WNBxNaaTbb(3-5-17)解上式,得WNBNxaaTbb11ˆ(3-5-18)则上式代入(3-5-15)式,可计算出K,或者,将(3-5-15)式代入(3-5-11),得)ˆ(111xBWNAPKAPVaaTT(3-5-19)即可直接计算出观测值的改正数V。再由VLLˆ,xXXˆˆ0分别计算出观测值平差值和非观测量的最或是值。二、精度评定1.单位权中误差计算附有参数的条件平差的单位权方差和中误差的计算,仍使用下述公式计算3rPVVT20ˆ(3-1-17)rPVVT0ˆ(3-5-18)2.协因数阵首先写出各基本向量的表达式:L=L)()(0000ABXALABXALWWNBNXxXXaaTbb1100ˆˆxBNWNKaaaaˆ11KAPVT1VLLˆ写出有关协因数阵:111111111ˆˆbbbbaaaaaaTbbbbaaWWaaTbbXXNBNNNNBNBNNQNBNQ(3-5-19)11111111111111111111)()()()(aaTbbaaaaaaaaTbbaaTaaaaTbbaaTaaaaTbbaaTaaaaTbbaaKKNBBNNNNNBBNNAQANNBBNNNNBBNNAQANNBBNNQAQQWLTLWQAQAQNBNQNBNQaaTbbWLaaTbbLX1111ˆ(3-5-20)1111ˆbbaaTbbaaLWXLBNNQABNNQQ(3-5-21)AQQAQNBBNNAQNBQNQNQKKaaTbbaaaaLxaaWLaaKL1111ˆ11KKTaaTbbaaTaaTaaTxLaaLWLKQQANBBNNQANQANBQNQQ11111ˆ1AQQQAQKKTVVQQLLaaTWWNAQAQ4VVKKTLKLVQAQQQAAQQQVVKKTKLTVLQAQQQAQQAQAQNBBNNNQAQQQQQQQQQQQQaaTbbaaaaTVVVVVVVVVVVLLVLLLL)(1111ˆˆ(3-5-22)3.平差值函数中误差计算同条件平差一样,在附有参数的条件平差中,要评定一个量的精度,首先要将该量表达成关于观测量平差值和参数平差值的函数形式,再依据协因数传播律,计算该量的协因数,最后计算出其方差或中误差。设:平差后一个量关于观测值与参数平差值的函数为)ˆ,,ˆ,ˆ,ˆ,,ˆ,ˆ()ˆ,ˆ(ˆ2121unXXXLLLfXLfF(3-5-23)对其全微分,得权函数式XdFLdFFdTxTlˆˆˆ(3-5-24)其中nXXLLnXXLLXXLLTlLdLfLdLfLdLfFˆˆˆˆˆˆ000ˆˆ2ˆˆ21ˆˆ1uXXLLuXXLLXXLLTxXdXfXdXfXdXfFˆˆˆˆˆˆ000ˆˆ2ˆˆ21ˆˆ1TnLdLdLdLdˆˆˆˆ21,TuXdXdXdXdˆ,,ˆ,ˆˆ21根据协因数传播律,得函数的协因数为xXXTxlLXTxxXLTllLLTlFFFQFFQFFQFFQFQˆˆˆˆˆˆˆˆˆˆ(3-5-25)式中LLQˆˆ、XLQˆˆ、LXQˆˆ、XXQˆˆ等,均可参照(3-5-19)、(3-5-20)、(3-5-21)、(3-5-22)等式计算。函数Fˆ的中误差为FFFFQDˆˆ20ˆˆ即2ˆ20ˆˆ202ˆ1ˆˆˆFFFFPQ2ˆ0ˆˆ0ˆ1ˆˆˆFFFFPQ(3-5-26)5三、例题例[3-2]如图3-15所示三角网,A,B为已知点,其坐标为A(1000.00,0.00),B(1000.00,1732.00)(单位:m),BD边的边长为SBD=1000.0m。各角值均为等精度观测(取QLL=E),观测值分别为:L1=60°00′03″L2=60°00′02″L3=60°00′04″L4=59°59′57″L5=59°59′56″L6=59°59′59″取∠BAD的最或是值为未知数Xˆ。试用附有参数的条件平差法对该网进行平差,并求∠CAB平差后最或是值的中误差。解:本题中,总观测数n=6,必要观测数t=4,多余已知值p=1,附加一个未知参数u=1,则r=n+p–t=3,c=n+u–r=4可以写出图形条件2个、极条件1个、固定边条件1个,分列出最或是值条件方程如下:01)ˆˆsin(ˆsin01ˆsinˆsin)ˆˆsin(ˆsin)ˆˆsin()ˆˆsin(0180ˆˆˆ0180ˆˆˆ535424531654321LLSXSXLLLLLLXLLLLLLLBDAB取xXXˆˆ0,由固定边条件可计算其近似值X0=30°00′00″将最或是值条件方程中的非线性式线性化,并计算出改正数条件方程:00507.6ˆ7321.15774.05774.00196.5ˆ1546.15774.01547.15774.05773.0732.108095354321654321xvvxvvvvvvvvvvv05774.005774.00005774.01547.15774.05773.0732.1111000000111A,7321.11546.100B,C2416ABx35D图3-15667.067.358.058.067.334.3858.046.358.058.00.3058.046.300.31TaaAAPNBNBNaaTbb1=27.5743WNBNxaaTbb11ˆ=2.8193)ˆ(1xBWNKaa=[-2.56622.7286–0.96610.6440]TKQAVT=[-4.23–3.12-1.631.643.662.73]TVLLˆ=[59°59′58.8″59°59′58.9″60°00′02.4″59°59′58.6″59°59′59.7″60°60′01.7″]TxXXˆˆ0=30°00′02.8″平差值函数式011ˆˆˆˆXxLXLFFl=[100000]T,Fx=[-1]T1ˆˆbbXXNQ=[0.0363]AQNBNQaaTbbLX11ˆ=[-0.06560.0935-0.0279-0.0391-0.08090.1199]11ˆbbaaTXLBNNQAQ=[-0.06560.0935-0.0279-0.0391-0.08090.1144]T6660.03445.03215.00153.00014.00139.03445.04881.01436.02455.00223.02232.03215.01436.04651.02608.00237.02371.00153.02455.02608.03292.03027.00265.00014.00223.00237.03027.06639.03612.00139.02232.02371.00265.03612.03877.0)(1111ˆˆAQNBBNNNQAQQaaTbbaaaaTLL则xXXTxlLXTxxXLTllLLTlFFFQFFQFFQFFQFQˆˆˆˆˆˆˆˆˆˆ=0.55510507.6196.589W7FFFQˆˆ0ˆˆˆ=±2.7″
本文标题:3-5附有参数的条件平差
链接地址:https://www.777doc.com/doc-2926601 .html