Research On Accuracy Optimization Method For Distributed Measurement Network

The workshop Measurement Positioning System as a typical distributed measurement system can obtain the coordinates of the measured point at multiple receiving nodes at the same time through the intersection of observation among multiple transmitting nodes,forming a networked coordinate measurement method with advantages of full-circumferential measurement range and multi-tasking.The internal structural parameters and the external relative position and pose parameters of the transmitting node are the key to constructing the measurement network,often calibrated by the geometric constraint relationship formed among the receiving nodes.In the traditional wMPS’s calibration method,the distribution of constrained reference points is uneven,and the low-order approximation error of the measurement model propagates unevenly along the network nodes,which leads to over-fitting of internal and external parameters calculations and serious deterioration of measurement accuracy in non-calibrated areas.On the other hand,the centering error of receiving node reduces the accuracy of the geometric constraint relationship among the reference points,which also leads to the reduction of the accuracy of the internal and external parameters and the measurement accuracy of the full space.This paper takes wMPS(workshop Measurement Positioning System)as the research object,and analyzes the propagation law of low-order approximation error and centering error of receiving node in networked measurement method,and studies the corresponding measurement accuracy optimization methods.The research work includes:1.The main sources of low-order approximation error in the wMPS measurement model is analyzed,and their impact on the measurement accuracy of the fullcircumferential space are explained from the perspectives of mathematical model and plane geometry.The GUM(Guide to the Uncertaincy in Measurement)method is used to derive the propagation model of the uncertainty of coordinate measurement about the internal and external parameters,and the MCM(Monte Carlo Method)method is used to verify the analysis results of the GUM method.2.In order to solve the problem of serious deterioration of measurement accuracy in non-calibrated areas caused by low-order approximation error,firstly the turntableassisted calibration method is used to construct virtual constrained reference points evenly distributed in the full-circumferential space,which ensures the even propagation of low-order approximation error and improves the calculation accuracy of the internal parameters of the transmitting node.Secondly,the forward intersection constraints are added to the calibration model for constructing the solution flow of closed-loop feedback,which enhances the stability of measurement network structure relations and reduces the influence of low-order approximation error on the calculation of external parameters among the transmitting nodes.3.In order to reduce the influence of centering error of receiving node,a new internal and external parameters calibration method based on mobile cooperative target is studied.By designing a cooperative target that integrates multiple receiving nodes,there is no need to accurately determine the structural relationship among the target receiving nodes,but to use the relative invariance of the spatial position among the receiving nodes as rigid constraints to establish the calibration model of redundant optical plane intersection information,which avoids the introduction and propagation of centering error of receiving node.Meanwhile the target position can be freely moved to make it evenly distributed in the measurement space,which ensures the even propagation of low-order approximation error.4.Finally,the corresponding wMPS experimental platform is built to verify the above-mentioned error analysis and accuracy optimization methods.