Study Of Multi-sensors Data Fusion Method In Measure And Control System

Multi-sensor data fusion (MSDF) is a new subject, which is widely used in many areas, such as radar, sonar, navigation, automatic control and robotics. The purpose of multi-sensor data fusion is making the utmost of the redundant information and the complementary information which come from multiple sensors, in order to product more accurate, better and more reliable information. Depending on these information, we can reduce the uncertainty and ambiguity coming from the observation information in order to improve the ability of target detection, tracking and recognition. The background of this thesis is the measurement and control system. In this thesis, we carried on a research about the data fusion methods to the muti-sensor system. The main research and innovation contents are as following:(1) Some models of the data fusion are provided in this thesis, which include the data fusion model about flash ranging and GPS, the data fusion model about radar and GPS, the data fusion model about flash ranging, radar and GPS. In addition, the method to determine optimal weights and the way of non-linear parameter estimation are given.(2) The parameters of the track are shown in spline function. And the models and methods of multi-source data fusion are also introduced in this thesis. Constant, slow variable and other major system errors are diagnosed and estimated.(3) Through the joint treatment of data which are coming from multi-station or multi-type device, the measurement data can carry through self-calibration and reciprocity calibration. According to the results, we may analyze the device’s tracking status and measurement accuracy.(4) For the case of incomplete measurements, the corresponding data processing methods are provided in this thesis.(5) At the end of this thesis, the simulation results are summarized about the data coming from different treatment options and measurements methods. According to the simulation results, we may analyse that combination of different measurements, different measurement geometry and different data processing methods have an effect on parameter estimation accuracy, the main error estimation precision of the tracking system and the weight of each measuring unit. Through the analysis results the best one may be found and we may obtain the best scheme of data processing.