Learning distributed object grasp by a group of robots with redundant members is the main focus of this paper. In Elahibakhsh, A. H., et al. (2004), we tackled the problem of learning form closure grasp for planar con...
详细信息
Learning distributed object grasp by a group of robots with redundant members is the main focus of this paper. In Elahibakhsh, A. H., et al. (2004), we tackled the problem of learning form closure grasp for planar convex objects by multiple non-communicating robots without any information about the shape of objects. In this paper, the problem in presence of redundant agents is investigated. Agents' states and actions are designed such that the group learns grasping different objects using Q-learning method. As the environment is not intelligent enough to assess each agent's effect on the team performance, a credit assignment algorithm based on knowledge evaluation is designed. The proposed method considers the environment credit for the team, number of redundant agents, and the expertness level of each agent in its credit assignment. Applicability of the designed approach is verified through a set of simulations. It is shown that the team learns grasping different objects. Therefore, it is expected that the proposed method can be extended for distributed grasp of deformable objects
A systematic methodology for generating multi-folded torus chaotic attractors from a simple three-dimensional piecewise-linear system is presented in this paper. Theoretical analysis shows that multi-folded torus chao...
详细信息
A systematic methodology for generating multi-folded torus chaotic attractors from a simple three-dimensional piecewise-linear system is presented in this paper. Theoretical analysis shows that multi-folded torus chaotic attractors can be created via alternative switching between two piecewise-linear systems. A novel circuit diagram is designed for physically creating multi-folded torus chaotic attractors. This is the first time in the literature to experimentally verify a maximum 9-folded torus chaotic attractor that is generated by an analog circuit.
LEARN2control is a new computer-based learning system, that can be used to deepen basic knowledge in control engineering by self-study in a project-oriented environment The didactic concept aims at teaching the depend...
详细信息
The global and local stability of process systems in generalized Lotka-Volterra form is studied in this paper using entropy-like and quadratic Lyapunov function candidates. The global stability check for LV models is ...
详细信息
The paper proposes the estimation of unmeasured variables of traffic systems for traffic control purposes. The presented method proposes constrained state estimation of unmeasured traffic variable such as turning rate...
详细信息
ISBN:
(纸本)9634208754
The paper proposes the estimation of unmeasured variables of traffic systems for traffic control purposes. The presented method proposes constrained state estimation of unmeasured traffic variable such as turning rates. The weighted constrained state approach uses moving horizon along the state trajectory, which permits to handle equality and inequality constraints belonging to nature of the traffic model. A numerical example illustrates the importance of constrained estimation and shows a possible switch control system between two control light stages.
The paper deals with the LPV systems in respect to fault detection and isolation. The FPRG method is extended to use with affine LPV system where one could guarantee the disturbance attenuation. The paper presents a m...
详细信息
ISBN:
(纸本)9634208754
The paper deals with the LPV systems in respect to fault detection and isolation. The FPRG method is extended to use with affine LPV system where one could guarantee the disturbance attenuation. The paper presents a method for the robust detection filter design based on geometric approach for Linear Parameter Varying systems. The filter is using the general Fundamental Problem in Residual Generation method in presence of external noise, while disturbance decoupling from failure signitures in LPV system is not possible. The disturbance attenuation is achieved with induced L2 norm minimization concepts for LPV model class. A numerical example is solved to demonstrate the effectiveness of robust LPV fault detection filter.
In the intelligent transportation systems, the automatic navigation system is the active research domain. The lane lines reconstruction based on the computer vision is the kernel technique. This paper describes the mu...
详细信息
In the intelligent transportation systems, the automatic navigation system is the active research domain. The lane lines reconstruction based on the computer vision is the kernel technique. This paper describes the multi-lane line reconstruction with a single view for highway application. We calibrate the camera and get the model of the road. Then, we detect the white lane lines and reconstruct the multi lane lines. This paper also analysis the basic way in keeping the road based on the vision model and the lane line model. We have experimented the system with the algorithms on highway at 150 km/h in Sichuan province and Chongqing city in China. The result shows that the algorithms can work perfectly.
Guidance through waypoinis is common for small autonomous marine vehicles. Guidance by the line of sight, which turns the vehicle directly towards the next waypoint without any reference path calculation, is computati...
详细信息
Guidance through waypoinis is common for small autonomous marine vehicles. Guidance by the line of sight, which turns the vehicle directly towards the next waypoint without any reference path calculation, is computationally the simplest form of waypoint guidance. However, the basic algorithm gives rather poor guidance due to the missed waypoint problem, lack of sea current compensation and abrupt transitions between the consecutive waypoinis. Significant path deviations and even deadlocks are possible due to these problems. Therefore, more complex algorithms are usually used in real world applications. The research reported in this paper aims to demonstrate that significant improvements of the basic line-of-sight guidance algorithm can be achieved by several intuitive, simple corrections and additions. The simplicity of the basic line-of-sight guidance algorithm is not compromised. In particular, missed waypoint detection is performed by monitoring the distance between the vehicle and the waypoint. Introduced reference heading corrections are based upon the location of the next waypoint after the one the vehicle is currently approaching, and upon the sea current direction and intensity. The results of these corrections are shown in several simulation examples. In addition, the paper includes a short discussion about the lineof-sight guidance in the diving plane. Copyright 2004 IFAC.
Guidance through waypoints is common for small autonomous marine vehicles. Guidance by the line of sight, the basic waypoint guidance algorithm, is very simple, computationally inexpensive, and flexible to sudden chan...
详细信息
Guidance through waypoints is common for small autonomous marine vehicles. Guidance by the line of sight, the basic waypoint guidance algorithm, is very simple, computationally inexpensive, and flexible to sudden changes in the desired path specification. On the other side, the basic algorithm achieves unoptimal vehicle trajectory, ignores possible side disturbances as sea currents, and can even lead to deadlock situations. Therefore it should be improved to use in the real world applications. This paper introduces several such improvements: missed waypoint detection to prevent deadlocks, novel reference heading correction method to give smoother vehicle path in turns at waypoints, and sea current compensation method to cancel its influence. This improvements address the mentioned flaws of the basic algorithm, as is demonstrated by the simulations, and yet keep it simpler than is the case for the most used waypoint-derived guidance methods.
暂无评论