The large-scale distributed measurement system is a measurement network based on multiple measurement fusion. The network structure is the key to improve the network performance, which has an important impact on the measurement accuracy, efficiency and even cost. To solve the network structure reconstruction caused by the changing of obstacles, measurement objects and requirements, based on the coverage performance of the network nodes, the method to determine the lost-of-light issue based on the fast collision detection algorithm is first investigated and the problem of blind spot judgement is solved. Then, the network reconstruction algorithm based on Next-Best-View (NBV) is studied. At the same time, in order to improve the networking efficiency, the improved grey wolf optimization algorithm is used as the best location search algorithm to rearrange the numbers and positions of nodes, and thus efficient networking is realized and the reconstruction accuracy is improved. Finally, with the workshop Measurement Positioning System (wMPS) as the verification platform and from three aspects of measurement coverage, accuracy and networking efficiency, the effectiveness of the proposed method is confirmed by changing measurement conditions and requirements.