Advanced Materials Research Vols. 403-408

Abstract: Sensors of robots that act in unstructured environment sometimes do not provide complete observation, due to occlusion or limitation of sensing range. This paper presents a motion generation method for robot with multiple sensors with limited sensing ranges. The proposed method introduce extension of the action- observation mapping to outside of the sensing range of a sensor, based on the diffusion-based learning of Jacobian matrices between control input and observation variable. Multiple observation spaces can be integrated by finding correspondence between the virtual observation spaces. When a target observation variable is given to the robot, it can generate a motion from an observation space to- ward the target with another observation space using the extended observation space. The proposed framework is verified by two robot tasks, reaching motion toward the floor with a manipulator and navigation of mobile robot around the wall. In both cases, observation space by camera with limited view was extended and appropriate motion trajectories were obtained.

Abstract: Robot applications are mostly first developed on a computer and thereafter loaded onto the robot. However, in many situations, developing applications directly on the robot may be more effective. For instance, children who have not learned using a computer yet and who develop their robot applications while playing. Or for instance in the robots' operating environment, where there is no computer available. In this contribution we present the properties of the software tool becerik, for developing applications on-board a robot and for running them in multi-tasking mode concurrently. Furthermore, we introduce the programming language of the applications that has the same name becerik, which consists of only 6 commands.

Abstract: There are several tasks in which humans require learning such as reading, writing, drawing etc. This paper gives an idea to make HOAP-2 (Humanoid Open Architecture Platform) robot capable of writing text of any particular language. It can also draw special symbol in a specific coordinate frame. For writing any alphabet the robot follows a trajectory defined in terms of intermediate points. Three types of interpolation: Straight, Clockwise and anticlockwise provides exact direction and smoothening to a particular pattern of the alphabet. On the basis of these interpolated results robot usable joint values are generated. The simulator utilises the updated robot joint values for writing the sentence. We have tested our method on Webots (a simulated framework for HOAP-2) in which the robot can write any sentence of English characters correct and precise.

Abstract: Intrusion detection is the task of detecting, preventing and possibly reacting to the attacks and intrusions in a network based computer system. The neural network algorithms are popular for their ability to ’learn’ the so called patterns in a given environment. This feature can be used for intrusion detection, where the neural network can be trained to detect intrusions by recognizing patterns of an intrusion. In this work, we propose and compare the three different Relevant Features Hybrid Neural Networks based intrusion detection systems, where in, we first recognize the most relevant features for a connection record from a benchmark dataset and use these features in training the hybrid neural networks for intrusion detection. Performance of these three systems are evaluated on a well structured KDD CUP 99 dataset using a number of evaluation parameters such as classification rate, false positive rate, false negative rate etc.

Abstract: Effective agent communication is always been an important modern area of research. This paper focuses on achieving greater precision in common by improving agent-human communication with the help of visual attention and auditory localization based on a simple model of the superior colliculus in the human brain system. The model receives individual visual and auditory sensory stimuli and combines them to generate an integrated stimulus predicting the location of the sound source. This integrated stimulus is used to generate a motor saccade of the visual system to attend to the sound. The computational model is based on a neural network approach with learning and is explored in experiments reflecting varied conditions to determine whether it mimes the performance of superior colliculus in auditory and visual stimuli integration. Finally with a evaluation strategy carried between unimodal and multimodal data, the efficiency of the computational model of Superior Colliculus is determined. Performance of the neural network based computational model has proven effective in terms of learning, the better performance of the integrated response over unimodal response and providing a realistic communication experience.

Abstract: This paper addresses further enhancements of the earlier developed Smooth Nearness-Diagram Navigation (SND) method for mobile robots moving in complex and cluttered environments. The enhanced method, entitled SSND, improves the safety of paths generated by the SND and solves the problem of trapping the robot in narrow corridors, where the difference in the number of threats on its sides is high. This is achieved by adjusting the difference in the number of obstacles on the two sides of the robot heading direction. The power of our method is demonstrated by simulation results.

Abstract: DTMF (Dual Tone Multi Frequency) is a system of signal tones used for telecommunications. DTMF uses two tones to represent each key on the touch pad. DTMF data transfer technique has advantages such as high reliability, constant speed, and high signal to noise ratio, low cost and optimal utilization of existing resources. These features make DTMF an attractive Data Transfer Technique. It finds application in home and car security systems, robot control, SMS and voice call controlled industrial applications. In this paper, we discuss the use of DTMF data transfer in a voice call to control a toy car. Cell phone 1(CP 1), which is at user end, makes a call to cell phone 2(CP 2) at the machine end and establishes a connection. A key is pressed at the user side. Two tones corresponding to one key are encoded and sent through the cell phone network. Both tones are tapped through the earphone jack of cell phone at machine end and are decoded. The output is fed to a Micro-controller. The Micro-controller is connected to the remote control unit of the toy car, which in turn controls the motion of the car. The car moves in various directions according to the key pressed user. The electronic circuit is divided into 2 parts. The transmitter side consisting of Cell phone 1 with an inbuilt encoder and the receiver side consisting of Cell Phone 2, 8870 DTMF decoder and Atmega 16 micro-controller. The programming has been carried out on AVR Bascom®.

Abstract: In this paper, a multi-objective uniform-diversity genetic algorithm (MUGA) is used for Pareto optimum design of fuzzy fractional-order PID controllers for plants with parametric uncertainties. Two conflicting objective functions have been used in Pareto design of the fuzzy fractional-order PID controller. The results clearly show that an effective trade-off can be compromisingly achieved among the different fuzzy fractional-order PID controllers obtained using the methodology of this work and to achieve a robust design against the plant’s uncertainties.

Abstract: The paper describes the dynamics and control aspects of an arm wheel based mobile robot for stair navigation with such features as skid steered motion on flat surface, navigation on staircase i.e. ascending and descending stairs keeping payload carrier platform parallel to the ground. A feedback linearization control scheme is adopted to decipher the skid steer phenomena. The nearness of the experimental data to corresponding simulated data validates design and analysis of the present system.

Abstract: A neural network approach has been proposed & analyzed to control the motion of robot. This involves study of dynamics of a manipulator, which is highly non linear, study & design of neural network to solve the problem. The designed neural network has given the correct output joint angles, joint velocities according to given input torque initial joint angles & joint velocities. Feasibility of the approach is demonstrated through simulation using matlab using the physical parameters of SCARA robot for open loop as well as close loop.