摘要
为了拓展移动机器人应用场景、满足室外定位的高精度需求,提出一种基于综合卡尔曼滤波的协同室外定位算法,可解决室外复杂环境下独立传感器失灵、机器人实时定位漂移的问题.首先构建GPS和超宽频非线性定位系统模型,测试不同滤波算法对该模型的预测效果,分析对比解算速度和均方根误差,从而确定适合定位系统的最优算法;然后针对GPS信号受环境遮挡导致丢失或失准的情况,构建超宽频和惯性测量单元非线性定位补偿系统,利用基于误差的卡尔曼滤波算法预测机器人位置姿态;通过融合两种非线性系统下估计得到的不同状态向量,确定机器人室外真实位置姿态,进一步提高机器人室外定位精度,保证定位系统的稳定性.试验验证表明,本文算法室外定位误差小于10 cm,在GPS信号微弱的环境下能实时估计目标位置姿态,大幅度降低障碍物干扰的影响,准确预测机器人位置.
In order to expand the application scenarios of mobile robots and meet the high-precision requirements of outdoor positioning,this paper proposes a collaborative outdoor positioning algorithm based on synthesize Kalman filter(SKF),which solves the problem that the independent sensor fails and the real-time position of robot drifts in the complex outdoor environment.Firstly,by constructing global positioning system(GPS)and ultra-wideband(UWB)nonlinear positioning system models,we tested the prediction effects of different filtering algorithms on the model,analyzed and compared the solution speed and root mean square error,to select the optimal algorithm suitable for the positioning system.Secondly,a nonlinear positioning compensation system was constructed based on UWB and inertial measurement unit for the situation that GPS satellite signals are lost or misaligned due to environmental occlusion,and the error-state Kalman filter algorithm was used to predict the position and attitude of the robot.Finally,to further improve the positioning accuracy of the outdoor robot,and ensure the stability of the positioning system,this algorithm was used to determine the actual outdoor position and attitude of the robot by fusing the different state vectors estimated under the two nonlinear systems.The experiments show that the outdoor positioning error of the multi-sensor cooperative algorithm based on SKF is less than 10 cm.Even in the weak GPS signal environment,the algorithm can always estimate the target posture in real time,greatly reduce the influence of clutter interferences,and accurately predict the robot position.
作者
汪涛
黄家才
汤文俊
高芳征
WANG Tao;HUANG Jiacai;TANG Wenjun;GAO Fangzheng(Industrial Center/School of Innovation and Entrepreneurship,Nanjing Institute of Technology,Nanjing 211167,China)
出处
《南京工程学院学报（自然科学版）》
2023年第1期9-16,共8页
Journal of Nanjing Institute of Technology(Natural Science Edition)
基金
国家自然科学基金面上项目(61873120)
江苏省重点研发计划课题(BE20210165)
江苏省自然科学基金面上项目(BK20201469)。
关键词
协同定位
多传感器融合
综合卡尔曼滤波
状态估计
quadruped robot
indoor and outdoor positioning
visual inertial odometer
tightly coupled nonlinear optimization
