| With great material abundance,social technology has entered an era of rapid development,and people's living standards are also improving,prompting research in fields related to artificial intelligence to advance rapidly.Robots have entered various fields of our life,in the outdoors,robots can replace humans into dangerous,narrow and other hard to reach locations for search and rescue,exploration and other professional work;indoors,the popularity of robots,to provide convenience for human life,robots can achieve cargo handling,medical care,large areas of guidance and other functions.However,with the continuous development of robot technology,the environment in which robots work is becoming more and more complex,and the requirement for precise control of robots is becoming higher and higher.This has led to the emergence of autonomous mobile robots,whose core technologies are divided into three areas: perception,analysis,and execution.The basis for autonomous mobile robots is perception,which is like the human eye,allowing the robot to determine its position in the environment.In complex and crowded environments,such as hospitals,airports,exhibition halls,warehouses,and restaurants,it is difficult for robots to determine their own location efficiently and accurately,which limits the application of autonomous robots.At present,there are two types of indoor positioning methods for robots: mechanical positioning methods,such as course projection positioning technology,and sensor-based positioning methods.Radio sensors are widely used because they can meet all indoor scenarios and are simple to operate.Common wireless positioning techniques are susceptible to noise interference and also suffer from multipath effects and non-line-of-sight propagation problems.For example,Wi-Fi,Bluetooth,infrared,ultrasonic,RF signals and other wireless positioning technologies are limited by the band and bandwidth,so that the indoor positioning tracking system based on these technologies can not reach a satisfactory positioning accuracy.The characteristics of the ultra-wideband signal itself can obtain high-precision distance measurement error,and has anti-interference ability,strong penetration,good anti-multipath effect,so it has broad prospects for development.This paper is based on the research of ultra-wideband radar for high-precision positioning methods,and the specific work of the paper is as follows.(1)Research on indoor mobile robot positioning methods,firstly,several common indoor mobile robot positioning methods are introduced,and then focus on beacon-based sensor positioning technology,introducing the positioning principles of several common sensors and analyzing several different positioning models in detail.(2)Starting from the definition of ultrawideband radar,we introduce the pulse forms of ultrawideband radar.The indoor 3D positioning model of ultra-wideband radar is established,and the nonlinear least squares method is applied to solve the model to obtain the positioning results.The error analysis of the indoor three-dimensional positioning model found that the factors affecting the indoor three-dimensional positioning results of ultra-wideband radar are mainly the sensor clock interference,multipath effect and the signal non-line-of-sight propagation error.(3)A multi-method fusion of ultra-wideband radar ranging data processing method is proposed for the causes of the error: 1)extracting high-frequency measurements from the measured values by adaptive parameter DBSCAN clustering algorithm;2)using high-frequency measurements and distance truth value,training linear regression model training,error correction of high-frequency measurements;3)setting the confidence interval of the measured values according to the training regression model thresholds,and the Gaussian function is used for outlier detection and elimination.(4)The paper examines the multi-method fusion data processing method for the analysis of this measurement data processing put indoor positioning experiment,and the error of the measurement data estimate obtained is significantly smaller than the pre-processing distance error.The experimental results show that the indoor positioning accuracy of the robot is significantly improved by the method of this paper. |
PDF Full Text Request