| With the development of science and technology,the technology of multisensor data fusion is attracting people's attention gradually. It is applied to the military field and relevance fields such as intelligence robot,remote sense,medical treatment diagnose, automation surveillance,image treatment,principles of pattern recognition and complicated industry process control mainly. The purpose that the technology of multisensor data fusion studies is how to be also the main body of this paper research subject by the fact that the information of the multi information source gains the estimation and reasoning being especially more accurate than depending on any single information source.Multisensor data fusion on resolving problem such as probing, tracking and target identification has anti-interference, the strong survival ability , large space-time coverage range, and it has increased the measurement dimension , strengthened the ability of probing, and enhanced the reliability and the accuracy of the system , and it has many merit such as the strong space resolution ability, keen reaction , every sensor advantage function complementation ,so the sensor probing system which is composed by multisensor can not be replaced by the single sensor. In multisensor system , the radar is the main sensor on guarding the aerial target part ,and it is used to sound the floor target and other targets, and it is also used to track landform , measure height. A parameter of the radar measuring gives coordinates about target distance , bearing and angle of elevation; Infrared sensor is passive generally , it uses many wave bands to measure target's much frequency spectrum characteristic ,and carries out classification according to the frequency spectrum characteristic ,so infrared sensor does not provide independent range measurement .Furthermore, infrared sensor has many merits such as accuracy angular measure and the strong capacity to target discernment.This paper has studied the data fusion algorithm based on radar and infrared sensor.It has been divided into four sections for data fusion technology: 1, sums up; 2, sensor summary and the current situation of multisensor data fusion; 3, fusion technology based on radar and infrared sensor; 4, observation.The first chapter is the summarization. It introduces the problem's purpose and significance, arranges the studying content and structure in this paper.The second chapter is the sensor summary and current situation of the technology of multisensor data fusion. It has mainly introduced knowledge in the sensor and data fusion aspect, and the technology of different sensor data pretreatment. First we introduce that sensor summary; then we introduce the current situation on data fusion , concept and model and their main method, data fusion method as well as the ideas model , function model and structural model has been introduced and analyzed.In the actually multisensor data fusion system ,different sensor lies on different interface .Different sensor gets different physical signals ,so the structure of data fusion is usually compound structure which is combined sensor's fusion structure and central's fusion structure .After sensor gets data , it must dispose data in a given quantity .This is different sensor data pretreatment . We have mainly introduced the space and time matched accurate fundamental method , the error source , and the matched accurate basis in the final chapter .To different sensor's matched technology of system error ,we have preferred an different sensor's error matched accurate algorithm on generalized measured function .On basis of time and space matching ,we have achieved accurate estimate of system error ,improved fusion and track precision .The third chapter is the technology of radar and infrared sensor data fusion. This paper has chiefly solved different sensor system's data fusion .We have preferred an multisensor data fusion algorithm based on radar and infrared in the concentrated noise wave environment . And we have led into the accurate target angle information coming from infrared (IR) sensor acquisition. Because it hasn't any distance information in the infrared providing data ,we have made use of infrared information angulations' linearity approaching in the structural algorithm . In orthogonal coordinate system , we calculate the target's forecast vector ,and fuse radar data . Simulated result indicates this algorithmic validity . Contrasted single radar or infrared tracking algorithm , accuracy and precision has improved .The fourth chapter has brought forward the observation of the target state .The front chapter has discussed that infrared trackes the target and it fuses radar data .If infrared doesn't track the target ,it will affect data fusion .So we discuss the observation condition of the target state .On basis about that Nardone S C and Hammel S E and others has studied which the two-dimensional space uniform motion target may be observed ,specifically for the two-dimensional space uniform linear or uniformly accelerated motion target , we have analyzed the relation between the target's observed value and moving state; then we have changed nonlinearity angle observation equation to the linear frame, making use of triangle function's shift means; finally , making use of some characteristic of entirely matrix, we have gained the necessary condition that the system can observe the target .This is that for single infrared sensor, we can adopt measure station moving to get target's observation.Sum up full text working finally.This paper carries out a little beneficial research about sounding technology , to the angle of data fusion . During the qcourse ,we have mainly studied three aspect .First ,we have achieved accurate estimate of system error ,improved fusion and tracking precision in the pretreatment of data fusion ;Second ,in radar and infrared sensor data fusion ,we have brought out an improved algorithm and got anticipated purpose ;Finally ,we have studied the condition that single passive sensor can be observed .The purpose of this paper is that we can improve the decision-making problem ,improve the accuracy of sounding result and adapt the environment . Being to manufacture and develop sounding , navigation and so on ,we must establish basis .But as a whole, the research in multisensor data fusion is still in the primary stage , and there are many problems and difficulty that have not been solved , we must do further research in some detail problem. |
PDF Full Text Request