﻿ 一种用于跟踪不连续运动目标的视觉伺服方案

# 一种用于跟踪不连续运动目标的视觉伺服方案A Visual Servoing Scheme for Tracking Discontinuous Moving Targets

Abstract: The main reason for the limited tracking speed in visual servo is the delay caused by image acqui-sition, image processing and speed estimation. Prediction algorithm is a solution to deal with delay, but the disadvantage of prediction algorithm is that the prediction behavior of discontinuity in moving target is poor. In this paper, a visual servo scheme to overcome this problem is proposed. Firstly, a new feed forward-feedback control scheme is designed, in which the Kalman filter in the control system is set before the motion controller, so as to obtain the tracking target position in-formation at the current time, and then reduce the value range of the task function to control the robot motion. Monitor using Predictive Monitor if discontinuity is detected, the prediction quality of adaptive Kalman filter is switched to the appropriate steady-state Kalman filter, which is better than adaptive Kalman filter in dealing with discontinuity. This new prediction algorithm can obtain good prediction quality of smooth motion and intermittent motion.

1. 引言

2. 控制方案说明

Figure 1. Visual servo block diagram

3. 滤波器设计

(1)

(2)

(3)

(4)

(5)

(6)

(7)

(8)

(9)

,

(10)

(11)

(12)

(13)

(14)

(15)

(16)

(17)

(18)

(19)

(20)

(21)

(22)

1) 令,初始化机器人的初始状态,状态预测误差协方差矩阵,噪声协方差矩阵Q、R以及参数,由于两个滤波器在同一条件下进行工作,故将参数设置成相同的值。

2) 由式(3)、(8)分别计算出,以及系数矩阵

3) 由式(15)计算出,联合式(18)、(20)计算出渐消矩阵

4) 由式(17)计算出,由式(5)、(6)计算,

5) 由式(9)、(7)分别更新状态估计和误差协方差矩阵

4. 预测监视器

(23)

PM的结构如图3所示。其中定义了允许的预测误差带的宽度,而是这个误差带的中心。

(24)

(25)

(26)

(27)

,而不是来调整。

Figure 2. IAM values after PM performs initialization

Figure 3. Structure of predictive monitor PM

5. 仿真实验与结果分析

Figure 4. Shows the tracking target of acceleration and deceleration motion

Figure 5. (a) Th = 0.04 sec; Th = 0.002 sec

6. 结论

NOTES

*通讯作者。

[1] 钟建冬, 苏建波. 基于视觉预测的运动目标实时跟踪系统[J]. 机器人, 2010, 32(4): 516-521.

[2] Chroust, S. and Vincze, M. (2003) Improvement of the Prediction Quality for Visual Servoing with a Switching. International Journal of Robotics Research, 22, 905-922.
https://doi.org/10.1177/027836490302210008

[3] Fang, J.C. and Yang, S. (2011) Study on Innovation Adaptive EKF for In-flight Alignment of Airborne POS. IEEE Transactions on Instrumentation & Measurement, 60, 1378-1388.
https://doi.org/10.1109/TIM.2010.2084710

[4] Tsai, C.Y., Song, K.T., Dutoit, X., Van Brussel, H. and Nuttin, M. (2009) Robust Visual Tracking Control System of a Mobile Robot Based on a Dual-Jacobian Visual Interaction Model. Robotics and Au-tonomous Systems, 57, 652-664.
https://doi.org/10.1016/j.robot.2009.01.001

[5] 穆邵硕, 乔川. 小型机载光电平台中基于扩展卡尔曼滤波的地面目标定位算法[J]. 光学学报, 2019, 39(5): 1-11.

[6] 陆可, 肖建. 基于强跟踪延迟滤波算法的互馈双电机联合状态估计[J]. 中国电机工程学报, 2008, 28(36): 80-86.

[7] 徐祥, 史凡伟, 等. 基于参数识别的SINS/DVL初始对准方法[J]. 中国惯性技术学报, 2019, 27(2): 176-180.

[8] 刘志林, 李国胜, 张军. 有横摇约束的欠驱动船舶航迹跟踪预测控制[J]. 哈尔滨工程大学学报, 2019, 40(2): 312-317.

[9] 燕必希, 朱立夫, 董明利, 等. 卡尔曼滤波单目相机运动目标定位研究[J]. 仪器仪表学报, 2018, 39(8): 220-229.

[10] Corke, P.I. and Good, M.C. (1996) Dynamic Effects in Visual Closed-Loop Systems. IEEE Transactions on Robotics and Automation, 12, 671-653.
https://doi.org/10.1109/70.538973

Top