Search

Inspired by the self-exploring learning approach, this paper proposes an object-learning system in which a robot interacts with objects to obtain their features and construct object concepts. The system consists of three kinds of features: interaction features, visual features, and intrinsic features. When the robot interacts with an object, it observes the changes in the object to obtain its interaction features. At the same time, the robot learns the visual features of the object. The intrinsic features are the properties of the object. Models of the relationships among the three kinds of features are constructed through an Artificial Bee Colony based Random Forest algorithm and a Convolutional Neural Network. The established models help the robot to predict the properties of new objects and to make decisions. Two experiments are constructed in this paper: the service-providing task and the stacking task. In the former, the robot decides on an appropriate object, using the object concept models, to accomplish an appointed task. In the second experiment, the robot uses a Gated Recurrent Neural Network to learn the stacking sequence of various objects. All the experimental results demonstrate that the robot can build object concept models by interacting with objects, and can utilize these models to accomplish various tasks.

Machine-to-machine (M2M) communication technologies have been rapidly developed in recent years to support different kinds of applications, such as automatic monitoring from user equipments (UEs) to home devices. Unfortunately, such applications introduce heavy traffic in both the radio access networks and core networks (CNs). 3rd generation partnership project (3GPP) has defined its local IP access (LIPA) solution, which provides UEs with the ability to access home networks directly without traversing CNs. However, 3GPP LIPA can not support the universal LIPA service. Specifically, a UE's home LIPA service can not be accessed when the UE moves into the coverage area of a visited LIPA network or a macrocell. On the other hand, the recently developed traffic offload enhancements for 3GPP networks did not consider the authentication issue. This paper proposes a telecom-based universal LIPA solution, uLIPA, for 3GPP networks. To completely offload both the UE's traffic and control signaling messages from the CN, uLIPA first enhances the existing 3GPP LIPA architecture and routing path establishment procedure by incorporating the concept of local SGSN and local GGSN. Then, an effective one-pass signaling procedure is devised for mutual authentication among the UE, the visited LIPA network, and the home LIPA network. Since this authentication procedure is based on the 3GPP authentication framework, it can readily be adopted by the 3GPP and compatible mobile networks.

Advances in mechatronics and artificial intelligence make robots play important roles in modern life. Path planning is an important issue in mobile robotics. In an environment with obstacles, path planning is to find a suitable collision-free path, which satisfies certain optimization criteria, for a mobile robot to move from a start location to a target location. Most researches associated with collision-free path planning have focused on finding the shortest path, the minimum-time path, or the safest path etc., and the generated paths may be discontinued. Smoothness of the path is essential for mobile robot navigation, because non-smooth motions can cause slippage of wheels which degrades the robot's deadreckoning ability. The smoothness property of the path also is extremely important for car-like vehicles, which are constrained with their motion abilities. Therefore an approach for smooth and collision-free path planning using genetic algorithm with B-spline curve is proposed to give the mobile robot the abilities of global path planning in static environment with obstacles.

In this paper we propose a rapid-deployment service platform to facilitate the development and the distribution of communication services over IMS network. The mechanism and the architecture to achieve a rapid-deployment platform are discussed. Furthermore, we implemented the platform in an IMS application server (IP-SMS), which provides IP-based short message service over IMS networks. On top of IP-SMS, the interoperation model of add-on services and the IMS network is presented.

Applications of the network control system (NCS) improve the efficiency, the flexibility, and the reliability of large-scale systems in modern industries. In the same time, applications of NCS for remote control and monitoring reduce the time and cost of installation, reconfiguration, and maintenance. This study integrates the Ethernet, the control area network (CAN) network, and the wireless 802.11b in the present NCS. The TCP/IP and the CAN protocols are adopted as the communication gateway in a remote control system for an AC 400 W servo motor. The time delay between the application layer of the client and the application layer of the remote control target has been measured and analyzed in this study. Experimental results indicate that different sampling rates and the environments will greatly affect the delayed time. Also, results indicate that the distance will not affect the delay time in wireless communication. However, as the sampling period reduces, the delay time of the network increases dramatically. In the wireless network, the delay time may change suddenly by the environment interferences. Finally, with a serious time delay, a Smith predictor has been successfully applied to control the servo motor from Kaohsiung which is in 300 km away to verify the feasibility of the proposed NCS.

In this paper, garage parking for a PC-based model car with the design of fuzzy logic controller is explored. At first, the realization of a PC-based model car is described. The authors set up a 1/10th-scale car with front driven and steering wheels. It has four-wheels with four points of suspension. In the PC-based model car, which is completely autonomous, all sensing, computation and control is done on-board. In the second topic, the authors investigate and simulate logic control methods such that the developed model car is able to park in the garage appropriately and automatically from any initial location. The real-time motion of the fuzzy garage-parking controlled model car is implemented in the last topic. By considering practical situations, the authors modify and redesign the fuzzy logic control strategies. Both simulation and experiment demonstrate the feasibility and effectiveness of the proposed intelligent garage-parking control system.

This paper is devoted to the design and implementation of a complete intelligent mechatronic system-a mobile robot possessing autonomous parallel parking capability. In this paper, the configuration of the overall system is firstly introduced. The intelligent parallel-parking control method is addressed in the second topic, where a feasible reference path is provided for the fuzzy logic controller to maneuver the steering angle of the robot. The simulation results illustrate the effectiveness of the developed control algorithm. The authors not only investigate intelligent parallel-parking control methods but also real-time maneuvering of the developed sensor-based mobile robot. The real-time control of the parallel parking system demonstrates the feasibility of the fuzzy control scheme.