面向核事故应急处置的遥操作系统设计与实验研究

针对核事故现场的物理屏蔽严重、应急处置作业条件复杂、作业对象不明确等特点,研究并设计了一款基于动态自组网的机器人遥操作系统。在自组织网络实时Mesh保持良好通讯的情况下,利用叠加在真实回传视频中的虚拟模型,提出一种基于空间梯度的人工势场力觉引导算法。操作者可通过观察视觉交互系统中模型的变化和感受施加在自研主手的引导力,提高时延条件下的操作效率和临场感。实验表明,核辐射条件下遥操作系统130 m的通讯时延在30 ms以下,可满足视频传输的带宽要求。力觉引导下的物体避障抓取实验验证了算法的可行性和遥操作系统的工程实用性。因此,本文设计的基于动态自组网的机器人遥操作系统,可使操作者以更加直观自然的方式参与到遥操作系统中,有效提高操作员完成复杂遥操作任务的安全性和操作性能。

An improved potential field method for mobile robot navigation

In order to overcome the inherent oscillation problem of potential field methods(PFMs) for autonomous mobile robots in the presence of obstacles and in narrow passages,an enhanced potential field method that integrates Levenberg-Marquardt(L-M) algorithm and k-trajectory algorithm into the basic PFMs is proposed and simulated.At first,the mobile robot navigation function based on the basic PFMs is established by choosing Gaussian model.Then,the oscillation problem of the navigation function is investigated when a mobile robot nears obstacles and passes through a long and narrow passage,which can cause large computation cost and system instability.At last,the L-M algorithm is adopted to modify the search direction of the navigation function for alleviating the oscillation,while the k-trajectory algorithm is applied to further smooth trajectories.By a series of comparative experiments,the use of the L-M algorithm and k-trajectory algorithm can greatly improve the system performance with the advantages of reducing task completion time and achieving smooth trajectories.

Magnetometer calibration algorithm based on ellipsoid constraint

Calibration of magnetometer is an essential part to obtain high measurement precision.However,the existing calibration methods are basically the calibration of all attitudes,which means tough work when the magnetometer is applied in strapdown inertial navigation system(SINS).So a quick,easy and effective calibration algorithm is developed based on the ellipsoid constraint to calibrate magnetometers.In this paper,the measuring principle and error characteristic of the magnetometer are analysed to study its magnetic interference.During the process,a magnetometer calibration model is set up to convert the calibration to ellipsoid fitting based on the characteristic of hard magnetic interference and soft magnetic interference.Then the algorithm is tested by mimic experiment.The result shows that measurement precision is improved after the calibration,and then the magnetometer is installed in a control cabin of an underwater robot which is designed and developed by us,and actual magnetometer calibration experiments are conducted to further verify the validity of the algorithm.

Semantic categorization of indoor places using CNN for mobile robot exploration

The ability of achieving a semantic understanding of workspaces is an important capability for mobile robot. A method is proposed to categorize different places in a typical indoor environment by using a Kinect sensors for mobile robot exploration. At first,the invariant feature based images stitching approach is adopted to form a panoramic image according to Kinect visual information,and the translation between Kinect depth information and obstacle distance information is performed to obtain virtual LIDAR data. Then,the semantic classifier is designed by using convolutional neural networks( CNN) for indoor place categorization based on Kinect visual observations with panoramic view. At last,a frontier-based exploration method is applied to carry out indoor autonomous exploration of mobile robots,which integrates the CNN-based categorization approach. The proposed method has been implemented and tested on a real robot,and experiment results demonstrate the approach effectiveness on solving the semantic categorization problem for mobile robot exploration.

Initial alignment for low-cost attitude and heading reference systems

MEMS(micro electro mechanical systems) inertial navigation system(MINS) has been wildly used in robots for its low-cost.The MINS and magnetometers are commonly the component parts of the attitude and heading reference systems(AHRS),which provide pitch and roll angles relative to the earth gravity vector,and heading angle relative to the north.However,the performance of sensors with low cost AHRS is not so good.The gyros are not sensitive enough to observe the earth angular velocity,so the traditional technique like alignment algorithm is invalid.The measurements of gyros become useless to determine the initial attitude matrix from navigation frame to body frame.The alignment algorithm is computed by the accelerometers and magnetometers.The process is established as an optimization problem of finding the maximum eigenvector.Meanwhile the sensitive analysis with respect to the biases of accelerometers is proposed.Then the recursive least squares algorithm(RLSA) is introduced.The comparison between the proposed method and RLSA is provided.The results demonstrate its accuracy favorably and verify the feasibility of the proposed algorithm.