For improving the estimation accuracy and the convergence speed of the unscented Kalman filter(UKF),a novel adaptive filter method is proposed.The error between the covariance matrices of innovation measurements and t...For improving the estimation accuracy and the convergence speed of the unscented Kalman filter(UKF),a novel adaptive filter method is proposed.The error between the covariance matrices of innovation measurements and their corresponding estimations/predictions is utilized as the cost function.On the basis of the MIT rule,an adaptive algorithm is designed to update the covariance of the process uncertainties online by minimizing the cost function.The updated covariance is fed back into the normal UKF.Such an adaptive mechanism is intended to compensate the lack of a priori knowledge of the process uncertainty distribution and to improve the performance of UKF for the active state and parameter estimations.The asymptotic properties of this adaptive UKF are discussed.Simulations are conducted using an omni-directional mobile robot,and the results are compared with those obtained by normal UKF to demonstrate its effectiveness and advantage over the previous methods.展开更多
With the increasing use of humanoid robots in several sectors of industrial automation and manufacturing, navigation and path planning of humanoids has emerged as one of the most promising area of research. In this pa...With the increasing use of humanoid robots in several sectors of industrial automation and manufacturing, navigation and path planning of humanoids has emerged as one of the most promising area of research. In this paper, a navigational controller has been developed for a humanoid by using fuzzy logic as an intelligent algorithm for avoiding the obstacles present in the environment and reach the desired target position safely. Here, the controller has been designed by careful consideration of the navigational parameters by the help of fuzzy rules. The sensory information regarding obstacle distances and bearing angle towards the target are considered as inputs to the controller and necessary velocities for avoiding the obstacles are obtained as outputs. The working of the controller has been tested on a NAO humanoid robot in V-REP simulation platform. To validate the simulation results, an experimental platform has been designed under laboratory conditions, and experimental analysis has been performed.Finally, the results obtained from both the environments are compared against each other with a good agreement between them.展开更多
Based on the design of a docking mechanism,this paper thoroughly investigates the space automatic doc- king of self-reconfiguration modular exploration robot system(RMERS).The method that leads robot to achieve space ...Based on the design of a docking mechanism,this paper thoroughly investigates the space automatic doc- king of self-reconfiguration modular exploration robot system(RMERS).The method that leads robot to achieve space docking by using two-dimensional PSD is put forward innovatively for the median size robot system.At the same time,in order to enlarge the detecting extension and the precision of PSD and reduce its dependence on light- ing signal,the PSD was remade by increasing the optical device over its light-sensitive surface.The emission board and LED light scheduling were designed according to docking arithmetic,and the operating principle of docking process was analyzed based on these.The simulation experiments were carried out and their results are presented.展开更多
A new kind of optimal fuzzy PID controller is proposed, which contains two parts. One is an on line fuzzy inference system, and the other is a conventional PID controller. In the fuzzy inference system, three adjustab...A new kind of optimal fuzzy PID controller is proposed, which contains two parts. One is an on line fuzzy inference system, and the other is a conventional PID controller. In the fuzzy inference system, three adjustable factors x p, x i , and x d are introduced. Their functions are to further modify and optimize the result of the fuzzy inference so as to make the controller have the optimal control effect on a given object. The optimal values of these adjustable factors are determined based on the ITAE criterion and the Nelder and Mead′s flexible polyhedron search algorithm. This optimal fuzzy PID controller has been used to control the executive motor of the intelligent artificial leg designed by the authors. The result of computer simulation indicates that this controller is very effective and can be widely used to control different kinds of objects and processes.展开更多
In this paper, we study autonomous landing scene recognition with knowledge transfer for drones. Considering the difficulties in aerial remote sensing, especially that some scenes are extremely similar, or the same sc...In this paper, we study autonomous landing scene recognition with knowledge transfer for drones. Considering the difficulties in aerial remote sensing, especially that some scenes are extremely similar, or the same scene has different representations in different altitudes, we employ a deep convolutional neural network(CNN) based on knowledge transfer and fine-tuning to solve the problem. Then, LandingScenes-7 dataset is established and divided into seven classes. Moreover, there is still a novelty detection problem in the classifier, and we address this by excluding other landing scenes using the approach of thresholding in the prediction stage. We employ the transfer learning method based on ResNeXt-50 backbone with the adaptive momentum(ADAM) optimization algorithm. We also compare ResNet-50 backbone and the momentum stochastic gradient descent(SGD) optimizer. Experiment results show that ResNeXt-50 based on the ADAM optimization algorithm has better performance. With a pre-trained model and fine-tuning, it can achieve 97.845 0% top-1 accuracy on the LandingScenes-7dataset, paving the way for drones to autonomously learn landing scenes.展开更多
在这份报纸，形成控制和障碍回避问题被处理一统一了控制算法，它允许追随者当维持从领导人的需要的相对适用或相对距离时，避免障碍。In the known 领导人追随者机器人形成控制文学，领导人机器人的绝对运动状态被要求控制追随者，它...在这份报纸，形成控制和障碍回避问题被处理一统一了控制算法，它允许追随者当维持从领导人的需要的相对适用或相对距离时，避免障碍。In the known 领导人追随者机器人形成控制文学，领导人机器人的绝对运动状态被要求控制追随者，它不能在一些环境是可得到的。在这研究，领导人追随者机器人形成以在领导人和追随者机器人之间的相对运动状态被建模并且控制。领导人机器人的绝对运动状态没在建议形成控制器被要求。而且，研究基于察觉到在机器人和障碍之间的相对运动被扩大了到一个新奇障碍回避计划。试验性的调查用平台被进行了由活动机器人和计算机视觉系统，和结果表明了的三 nonholonomic 组成了建议方法的有效性。展开更多
Previously, researchers raised the accuracy for a robot′s hand to track a specified path in Car-tesian space mainly through increasing the number of knots on the path and the number of the path′s segments, which res...Previously, researchers raised the accuracy for a robot′s hand to track a specified path in Car-tesian space mainly through increasing the number of knots on the path and the number of the path′s segments, which results in the heavier online computational burden for the robot controller. Aiming at overcoming this drawback, the authors propose a new kind of real-time accurate hand path tracking and joint trajectory planning method. Through selecting some extra knots on the specified hand path by a certain rule and introducing a sinusoidal function to the joint displacement equation of each segment, this method can greatly raise the path tracking accuracy of robot′s hand and does not change the number of the path′s segments. It also does not increase markedly the computational burden of robot controller. The result of simulation indicates that this method is very effective, and has important value in increasing the application of industrial robots.展开更多
In this paper,we study scene image recognition with knowledge transfer for drone navigation.We divide navigation scenes into three macro-classes,namely outdoor special scenes(OSSs),the space from indoors to outdoors o...In this paper,we study scene image recognition with knowledge transfer for drone navigation.We divide navigation scenes into three macro-classes,namely outdoor special scenes(OSSs),the space from indoors to outdoors or from outdoors to indoors transitional scenes(TSs),and others.However,there are difficulties in how to recognize the TSs,to this end,we employ deep convolutional neural network(CNN)based on knowledge transfer,techniques for image augmentation,and fine tuning to solve the issue.Moreover,there is still a novelty detection prob-lem in the classifier,and we use global navigation satellite sys-tems(GNSS)to solve it in the prediction stage.Experiment results show our method,with a pre-trained model and fine tun-ing,can achieve 91.3196%top-1 accuracy on Scenes21 dataset,paving the way for drones to learn to understand the scenes around them autonomously.展开更多
基金Supported by National High Technology Research and Development Program of China(863 Program)Hi-Tech Research and Development Program of China(2003AA421020)
文摘For improving the estimation accuracy and the convergence speed of the unscented Kalman filter(UKF),a novel adaptive filter method is proposed.The error between the covariance matrices of innovation measurements and their corresponding estimations/predictions is utilized as the cost function.On the basis of the MIT rule,an adaptive algorithm is designed to update the covariance of the process uncertainties online by minimizing the cost function.The updated covariance is fed back into the normal UKF.Such an adaptive mechanism is intended to compensate the lack of a priori knowledge of the process uncertainty distribution and to improve the performance of UKF for the active state and parameter estimations.The asymptotic properties of this adaptive UKF are discussed.Simulations are conducted using an omni-directional mobile robot,and the results are compared with those obtained by normal UKF to demonstrate its effectiveness and advantage over the previous methods.
文摘With the increasing use of humanoid robots in several sectors of industrial automation and manufacturing, navigation and path planning of humanoids has emerged as one of the most promising area of research. In this paper, a navigational controller has been developed for a humanoid by using fuzzy logic as an intelligent algorithm for avoiding the obstacles present in the environment and reach the desired target position safely. Here, the controller has been designed by careful consideration of the navigational parameters by the help of fuzzy rules. The sensory information regarding obstacle distances and bearing angle towards the target are considered as inputs to the controller and necessary velocities for avoiding the obstacles are obtained as outputs. The working of the controller has been tested on a NAO humanoid robot in V-REP simulation platform. To validate the simulation results, an experimental platform has been designed under laboratory conditions, and experimental analysis has been performed.Finally, the results obtained from both the environments are compared against each other with a good agreement between them.
基金Supported by the National High Technology Research and Development Program of China(2002AA422130)
文摘Based on the design of a docking mechanism,this paper thoroughly investigates the space automatic doc- king of self-reconfiguration modular exploration robot system(RMERS).The method that leads robot to achieve space docking by using two-dimensional PSD is put forward innovatively for the median size robot system.At the same time,in order to enlarge the detecting extension and the precision of PSD and reduce its dependence on light- ing signal,the PSD was remade by increasing the optical device over its light-sensitive surface.The emission board and LED light scheduling were designed according to docking arithmetic,and the operating principle of docking process was analyzed based on these.The simulation experiments were carried out and their results are presented.
文摘A new kind of optimal fuzzy PID controller is proposed, which contains two parts. One is an on line fuzzy inference system, and the other is a conventional PID controller. In the fuzzy inference system, three adjustable factors x p, x i , and x d are introduced. Their functions are to further modify and optimize the result of the fuzzy inference so as to make the controller have the optimal control effect on a given object. The optimal values of these adjustable factors are determined based on the ITAE criterion and the Nelder and Mead′s flexible polyhedron search algorithm. This optimal fuzzy PID controller has been used to control the executive motor of the intelligent artificial leg designed by the authors. The result of computer simulation indicates that this controller is very effective and can be widely used to control different kinds of objects and processes.
基金supported by the National Natural Science Foundation of China (62103104)the China Postdoctoral Science Foundation(2021M690615)。
文摘In this paper, we study autonomous landing scene recognition with knowledge transfer for drones. Considering the difficulties in aerial remote sensing, especially that some scenes are extremely similar, or the same scene has different representations in different altitudes, we employ a deep convolutional neural network(CNN) based on knowledge transfer and fine-tuning to solve the problem. Then, LandingScenes-7 dataset is established and divided into seven classes. Moreover, there is still a novelty detection problem in the classifier, and we address this by excluding other landing scenes using the approach of thresholding in the prediction stage. We employ the transfer learning method based on ResNeXt-50 backbone with the adaptive momentum(ADAM) optimization algorithm. We also compare ResNet-50 backbone and the momentum stochastic gradient descent(SGD) optimizer. Experiment results show that ResNeXt-50 based on the ADAM optimization algorithm has better performance. With a pre-trained model and fine-tuning, it can achieve 97.845 0% top-1 accuracy on the LandingScenes-7dataset, paving the way for drones to autonomously learn landing scenes.
基金Supported in part by National High Technology Research and Development Program of P.R.China(2001AA422140)
文摘在这份报纸，形成控制和障碍回避问题被处理一统一了控制算法，它允许追随者当维持从领导人的需要的相对适用或相对距离时，避免障碍。In the known 领导人追随者机器人形成控制文学，领导人机器人的绝对运动状态被要求控制追随者，它不能在一些环境是可得到的。在这研究，领导人追随者机器人形成以在领导人和追随者机器人之间的相对运动状态被建模并且控制。领导人机器人的绝对运动状态没在建议形成控制器被要求。而且，研究基于察觉到在机器人和障碍之间的相对运动被扩大了到一个新奇障碍回避计划。试验性的调查用平台被进行了由活动机器人和计算机视觉系统，和结果表明了的三 nonholonomic 组成了建议方法的有效性。
基金Foundation of the Robotics Laboratory, Chinese Academy of Sciences (No: RL200002)
文摘Previously, researchers raised the accuracy for a robot′s hand to track a specified path in Car-tesian space mainly through increasing the number of knots on the path and the number of the path′s segments, which results in the heavier online computational burden for the robot controller. Aiming at overcoming this drawback, the authors propose a new kind of real-time accurate hand path tracking and joint trajectory planning method. Through selecting some extra knots on the specified hand path by a certain rule and introducing a sinusoidal function to the joint displacement equation of each segment, this method can greatly raise the path tracking accuracy of robot′s hand and does not change the number of the path′s segments. It also does not increase markedly the computational burden of robot controller. The result of simulation indicates that this method is very effective, and has important value in increasing the application of industrial robots.
基金supported by the National Natural Science Foundation of China(62103104)the Natural Science Foundation of Jiangsu Province(BK20210215)the China Postdoctoral Science Foundation(2021M690615).
文摘In this paper,we study scene image recognition with knowledge transfer for drone navigation.We divide navigation scenes into three macro-classes,namely outdoor special scenes(OSSs),the space from indoors to outdoors or from outdoors to indoors transitional scenes(TSs),and others.However,there are difficulties in how to recognize the TSs,to this end,we employ deep convolutional neural network(CNN)based on knowledge transfer,techniques for image augmentation,and fine tuning to solve the issue.Moreover,there is still a novelty detection prob-lem in the classifier,and we use global navigation satellite sys-tems(GNSS)to solve it in the prediction stage.Experiment results show our method,with a pre-trained model and fine tun-ing,can achieve 91.3196%top-1 accuracy on Scenes21 dataset,paving the way for drones to learn to understand the scenes around them autonomously.
