Journal of Computer and Robotics
Volume:5 Issue: 1, Winter and Spring 2012

The ball and beam system is one of the most popular laboratory setups for control education. In this paper, we design a fuzzy PD cascade controller for a ball and beam system using Asexual Reproduction Optimization (ARO) technique. The ball & beam system consists of a servo motor, a grooved beam, and a rolling ball. This system utilizes a servo motor to control ball’s position on the beam. Changing the angle of servo motor results in the movement of the beam and, subsequently, the ball rolling on it. We designed a fuzzy PD cascade and a PD cascade controller scheme which consists of the two controller loops. The first (outer) controller and the second (inner) controller are organized in a cascaded construction.

Fuzzy clustering methods are conveniently employed in constructing a fuzzy model of a system, but they need to tune some parameters. In this research, FCM is chosen for fuzzy clustering. Parameters such as the number of clusters and the value of fuzzifier significantly influence the extent of generalization of the fuzzy model. These two parameters require tuning to reduce the overfitting in the fuzzy model. Two new cost functions are developed to set the parameters of FCM algorithm properly and the two evolutionary optimization algorithms, i.e. the multi-objective simulated annealing and the multi-objective imperialist competitive algorithm, are employed to optimize the parameters of FCM according to the proposed cost functions. The multi-objective imperialist competitive algorithm is the proposed algorithm.

Due to the increasing web, there are many challenges to establish a general framework for data mining and retrieving structured data from the Web. Creating an ontology is a step towards solving this problem. The ontology raises the main entity and the concept of any data in data mining. In this paper, we tried to propose a method for applying the "meaning" of the search system, But the problem for these methods is building a knowledge base that can be used for semantic search. The previous work interprets the query in three ways:'semantic relation in ontology', 'co-occurrence in the document', and 'semantic relation from Thesaurus'. The proposed method has two parts. The first part, using domain ontology for classified web pages based on keyword and the concept in each domain and builds Fuzzy ontology as Knowledge Base and the next section offers a method for expanding the query using built fuzzy ontology. In this paper, we tried to create knowledge base with WordNet as a comprehensive dictionary and extracted Sub string (phrases include multi words) from WordNet for each keyword in each domain ontology. The created Search engine was applied to an experimental system to evaluate the "precision – Recall” and it was revealed that applying the proposed method can improve query expansion 11%better in our experiments for  precision.

Facial expressions are the most powerful and direct means of presenting human emotions and feelings and offer a window into a persons’ state of mind. In recent years, the study of facial expression and recognition has gained prominence; as industry and services are keen on expanding on the potential advantages of facial recognition technology. As machine vision and artificial intelligence advances, facial recognition has become more accessible and is now a key technique to be employed and used in creating more natural man-machine interactions, Computer vision, and health care. In this paper, we empirically evaluate facial representation based on statistical local features, Local Binary Patterns, for person-independent facial expression recognition. Different machine learning methods are systematically examined on several databases. Extensive experiments illustrate that LBP features are effective and efficient for facial expression recognition. In this paper, we proposed a face expression detection method based on the difference of a face expression and the allocated special pattern to each expression. The analysis of the image detection system locally and through a sliding window (sliding) at multiple scales, are estimated. Multiple scales are extracted aslocally binary features. Through using the change point between windows, points of face are getting a directional movement. Through using points movement of whole facial expressions and rating system that is created the superfluous points are eliminated. The classifications are taken based on the nearest neighbor. To sum up this paper, the proposed algorithms are tested on Cohn-Kanade data set and the results showed the best performance and reliability into other algorithms. We investigated LBP features for the facial skin structural changes, which is seldom addressed in the existing literature.

In this study, detection and identification of common faults in industrial gas turbines is investigated. We propose a model-based robust fault detection(FD) method based on multiple models. For residual generation a bank of Local Linear Neuro-Fuzzy (LLNF) models is used. Moreover, in fault detection step, a passive approach based on adaptive threshold is employed. To achieve this purpose, the adaptive threshold band is made by a sliding window technique to make decision whether a fault occurred or not. In order to show the effectiveness of proposed FD method, it is used to identify a simulated single-shaft industrial gas turbine prototype model, which works in various operation points. This model is a reference simulation which is used in many similar researches with the aim of fault detection in gas turbines.

Joint flexibility is a very important factor to consider in the controller design for robot manipulators if high performance is expected. Most of the research works on control of flexible-joint robots in literature have ignored the actuator dynamics to avoid complexity in controller design. The problem of designing nonlinear controller for a class of single-link flexible-joint robot manipulators whose model incorporates the effect of the electrical actuator is considered in this paper. The main control purpose followed in this research is stabilization of the system states and backstepping approach is employed to achieve this goal and find control law. The global asymptotic stabilization of the closed-loop system is achieved in the sense of Lyapunov. Finally, to demonstrate the efficiency of the designed controller in stabilization the system states, the simulation results for system dynamics and closed-loop system, are compared in different initial conditions without and in the presence of external disturbances.