Advanced Materials Research Vols. 433-440

Abstract: Pseudo inverse learning rule and new activation unction performance will be evaluated and compared with the primitive learning rule, Hebb rule. Comparisons are made between these three rules to see which rule is better or outperformed other rules in the aspects of computation time, memory and complexity. From the computer simulation that has been carried out, the new activation function performs better than the other two learning methods.

Abstract: This research is focused on the integration of multi-layer Artificial Neural Network (ANN) and Q-Learning to perform online learning control. In the first learning phase, the agent explores the unknown surroundings and gathers state-action information through the unsupervised Q-Learning algorithm. Second training process involves ANN which utilizes the state-action information gathered in the earlier phase of training samples. During final application of the controller, Q-Learning would be used as primary navigating tool whereas the trained Neural Network will be employed when approximation is needed. MATLAB simulation was developed to verify and the algorithm was validated in real-time using Team AmigoBotTM robot. The results obtained from both simulation and real world experiments are discussed.

Abstract: Image processing is widely used in various fields of study including manufacturing as product inspection. In compact disc manufacturing, image processing has been implemented to recognize defect products. In this research, we implemented image processing technique as pre-processing processes. The aim is to acquire simple image to be processed and analyzed. In order to express the object from the image, the features were extracted using Invariant Moment (IM). Afterward, neural network was used to train the input from IM’s results. Thus, decision can be made whether the compact disc is accepted or rejected based on the training. Two experiments have been done in this research to evaluate 40 datasets of good and defective images of compact discs. The result shows that accuracy rate increased and can identify the quality of compact discs based on neural network training.

Abstract: Cost, quality and technology leadership are no longer sufficient for businesses to secure critical advantage. Instead, differentiations are being provided through the supply of innovative services, which can rapidly develop into a firm’s unique selling proposition. However, although services can provide additional competitive advantage, the inherent differences between product and service can cause difficulty in the effective integration of the two processes. Services are dominated by intangible elements which can be difficult to perceive and quantify. Within the medical device industry, the growing focus on usability, patient safety and increasing regulatory requirements has further complicated the already complex development process. In order to meet regulations, development is undertaken within strict boundaries to produce tightly controlled outputs. It can be seen that there is an incongruity between the nature of medical device development and the service development processes. This paper explores the constraints and inhibitors of service creation within the context of the medical device industry. Service innovation and its application is discussed. Difficulties in the addition of a service element and their potential solution within a medical device context are explored.

Abstract: In this paper, the new technology of RDIF (Radio Frequency Identification) has been used in order to identify vehicles and also 3 significant parameters including the average speed of vehicles at any side of access point, the average time for waiting and the queue length. They have been used based on the data from neural network for making the best decision throughout the process of finding out duration of the cycle and percentage of green time for each of the access point. Implementation of this system is possible in the shortest time and it has a better function in any kind of weather condition, time or place compared to similar systems.

Abstract: Optimizing the complicated engineering structures has always been a huge issue. A technique for the design optimization of different components is presented using genetic algorithm and finite element method. To reduce the runtime and increase the efficiently of proposed model a new method of coupling is presented. In addition, two different problems were solved using the presented model and the results showed a great and fast convergence.

Abstract: Handling of objects with irregular shapes and that of flexible/soft objects by ordinary robot grippers is difficult. Multi fingered gripper may be a solution to such handling tasks. However, dexterous grippers will be the appropriate solution to such problems. Although it is possible to develop robotic hands which can be very closely mapped to human hands, it is sometimes not to be done due to control, manufacturing and economic reasons. The present work aims at designing and developing a dexterous robotic hand for manipulation of objects.

Abstract: The anterior cruciate ligament in the knee connects the femur to the tibia and is often torn during a sudden twisting motion, resulting in knee instability. Effective treatment is with surgery where the ligament is replaced with a piece of healthy tendon grafted into place to hold the knee joint together. Employing a novel repair device, models for the repaired and for the intact knee are developed to evaluate the efficacy of the design the device.

Abstract: The question of rock mass deformation Long-term forecast is researched base on DRNN. The construction of neural network is optimized via reconstructed chaotic phase space, and the all nodes of DRNN are interconnected, and the feedback between nodes and that of node itself is included, and mult-linkage branch is build between two nerves, and the linkage branch stands for the link weight and the time delay of regular step. So the current moment network output of node depends on not only current moment network iutput, but also the node output of Some moment before current, so the chaotic prediction sensibility to initial condition is reduced effectively. The calculating velocity and network stability is improved effectively. Examples show that results are reasonable and the long-term prediction is reasonable and feasible.

Abstract: The improved niche hybrid hierarchy genetic algorithm is presented to overcome the premature convergence which happens in genetic algorithm constructing RBF network. The niche with poor fitness of every individual is eliminated to save system resource and raise operation speed. The simulation results demonstrate the better predicted performance on the Mackey-Glass chaotic time series than other algorithms.