This paper present a novel mechanism based on a characteristic of physical agent named "degrees of situation" that aids to improve the coordination among heterogeneous intelligent agents. These systems can b...
详细信息
This paper present a novel mechanism based on a characteristic of physical agent named "degrees of situation" that aids to improve the coordination among heterogeneous intelligent agents. These systems can be represented by means of the "physical agent" paradigm. One typical implementation of physical agents is autonomous mobile cooperative robots. In our approach the physical agents have different automatic controllers to generate dynamical diversity in the team-work. Here, dynamic means dynamic temporal evolution of continuous variables of the controlled system. Therefore, the multi-agent systems can be considered as a team of heterogeneous intelligent agents with different capabilities that work together to fulfil some cooperative tasks. In particular, this paper is related to studies about how the team-work improves by the "degrees of situation" management. Result and conclusions are shown, emphasizing contributions of the approach in the improvement of the cooperative team-work
This paper presents a learning system that uses Q-learning with a resource allocating network (RAN) for behavior learning in mobile robotics. The RAN is used as a function approximator, and Q-learning is used to learn...
详细信息
This paper presents a learning system that uses Q-learning with a resource allocating network (RAN) for behavior learning in mobile robotics. The RAN is used as a function approximator, and Q-learning is used to learn the control policy in 'off-policy' fashion that enables learning to be bootstrapped by a prior knowledge controller, thus speeding up the reinforcement learning. Our approach is verified on a PeopleBot robot executing a visual servoing based docking behavior in which the robot is required to reach a goal pose. Further experiments show that the RAN network can also be used for supervised learning prior to reinforcement learning in a layered architecture, thus further improving the performance of the docking behavior
There is much to gain from providing walking machines with passive dynamics, e.g. by including compliant elements in the structure. These elements can offer interesting properties such as self-stabilization, energy ef...
详细信息
There is much to gain from providing walking machines with passive dynamics, e.g. by including compliant elements in the structure. These elements can offer interesting properties such as self-stabilization, energy efficiency and simplified control. However, there is still no general design strategy for such robots and their controllers. In particular, the calibration of control parameters is often complicated because of the highly nonlinear behavior of the interactions between passive components and the environment. In this article, we propose an approach in which the calibration of a key parameter of a walking controller, namely its intrinsic frequency, is done automatically. The approach uses adaptive frequency oscillators to automatically tune the intrinsic frequency of the oscillators to the resonant frequency of a compliant quadruped robot. The tuning goes beyond simple synchronization and the learned frequency stays in the controller when the robot is put to halt. The controller is model free, robust and simple. Results are presented illustrating how the controller can robustly tune itself to the robot, as well as readapt when the mass of the robot is changed. We also provide an analysis of the convergence of the frequency adaptation for a linearized plant, and show how that analysis is useful for determining which type of sensory feedback must be used for stable convergence. This approach is expected to explain some aspects of developmental processes in biological and artificial adaptive systems that "develop" through the embodied system-environment interactions
We proposed and evaluated a network learning method called self-organizing network elements (SONE). Autonomous exploration of effective output, the use of simple external parameters, and low calculation costs were fun...
详细信息
We proposed and evaluated a network learning method called self-organizing network elements (SONE). Autonomous exploration of effective output, the use of simple external parameters, and low calculation costs were functions achieved for a robot system with this method. However, there is the need to improve performance against noises for learning more complicated tasks. Therefore, we propose a technique to adjust thresholds in a self-organizing logic circuit based on SONE. In our experiments, performing 3-bit operation tasks, controlling a Khepera robot, and generating new elements was appropriately controlled with this technique. Also, network performance against noises was improved as a result of using the proposed method
This paper discusses walking control of Theta, an innovative humanoid walking robot equipped with a double spherical hip joint and a knee joint switchable between active and passive operation mode, which has been been...
详细信息
This paper discusses low level control of the humanoid UT-Theta which has been been developed at the University of Tokyo. This innovative humanoid walking robot is equipped with knee joint switchable between active an...
详细信息
Using a robotic vehicle to inspect overhead power lines has many attractions. The concept of a small, rotorcraft which draws its power from the lines is introduced here. Some of its advantages compared with a free-fly...
详细信息
The purpose of this paper is to analyze the stability of Passive Dynamic Walking using an approximate analytical Poincaré map of the impact point, which is the state of walking robot at the collision between a sw...
详细信息
Exploiting the body dynamics to control the behavior of robots is one of the most challenging issues, because the use of body dynamics has a significant potential in order to enhance both complexity of the robot desig...
详细信息
This study discusses a learning algorithm for autonomous robots that has five characteristics including autonomous exploration of effective output, low calculation costs, capability for multi-tasking, reusing past kno...
详细信息
暂无评论