Dissertation > Excellent graduate degree dissertation topics show

Technology Research of the Multi-objective Detection and Tracking for Intelligent Vehicle Based on Radars

Author: ZhaoWanLi
Tutor: CaiZiXing
School: Central South University
Course: Electronic Science and Technology
Keywords: millimeter wave radar lidar the least_square method Joint probabilistic data association algorithm
CLC: TN957.51
Type: Master's thesis
Year: 2011
Downloads: 100
Quote: 1
Read: Download Dissertation

Abstract

With the intelligent traffic gradually rising in the world, unmanned vehicles began to become a research hotspot.In general,unmanned vehicles are equipped with digital cameras, laser radar, millimeter wave radar, GPS, inertial navigation systems and other sensors. As the focus of this thesis, the main work is to use laser radar and millimeter-wave radar sensor to get more the target information in front of unmanned vehicles and fuse target data of two sensors as a basis for decision-making board computer.The main contents of this paper can be divided into the following sections:(1) In the second chapter, on the millimeter-wave radar program comes with a lot of changes to conform to the requirements of our projects, and analyzes the functions of each module of the millimeter-wave radar acquisition program and the associated algorithm, the program commissioning are summarized.(2) In the third chapter, the main work is to calculate the road range by lidar. As the highway is a very regulated environment, there are railings on both sides of the road. Therefore, it is feasible to obtain lane both sides of the fence, and then use the least squares method to calculate the road range, lidar returns data but he exception of some of the data occurs, it takes bad impact on the projected scope of the road,the least squares are made corresponding improvement. The main improvement is that firstly the variance calculation of all lidar data to the lane edge line which has been fitted, then the threshold is set to half of the average of all the variance, lidar data point of the lane edge line that is greater than the threshold will be discarded, the rest of the edge lane has been re-fited, getting a new, more consistent with real-lane edge line.(3) In the fourth chapter, the main innovation of this paper is as follows: the result of Chapter III has been used to filter out the target detected by the millimeter-wave radar outside the road lane edge line and track only within the lane edge line, significantly saving in the efficiency of multi-target tracking. Unmanned vehicles realized the multi-target track in front of the projections, analysis of the simplest principles of the nearest neighbor method and disadvantages, and using joint probabilistic data association algorithm while using laser radar data to calculate multi-target track In order to ensure real-time system, the algorithm related improvement made, the experimental results show that the joint probability using the algorithm and associated data fusion and laser radar is used to solve multi-objective reckoning with a good prospect.

Related Dissertations

CLC: > Industrial Technology > Radio electronics, telecommunications technology > Radar > Radar equipment,radar > Radar receiving equipment > Radar signal detection and processing
© 2012 www.DissertationTopic.Net Mobile