Adaptive actuator fault compensation for linear systems with matching and unmatching uncertainties

The problem of process systems subject to actuator faults (partial loss of actuator effectiveness) is considered. An active fault compensation control law is designed that utilizes compensation in a way that accounts for matching and unmatching uncertainties and the occurrence of actuator faults. The main idea is designing the robust compensation controller to guarantee closed-loop stability in the presence of faults, based on a neural network representation of the fault dynamics. Changes in the system due to faults are modeled as unknown nonlinear functions. The updating control law is derived such that all the parameters of the closed-loop system are bounded. An output feedback controller is used to the “healthy” system and the adaptive feedback controller is used to compensate for the effect of the dynamics caused by the fault. The advantage of fault compensation is the dynamics caused by faults can be accommodated online. The proposed design method is illustrated on a three-tank system.     

An adaptive navigation support system for conducting context-aware ubiquitous learning in museums

In context-aware ubiquitous learning, students are guided to learn in the real world with personalized supports from the learning system. As the learning resources are realistic objects in the real world, certain physical constraints, such as the limitation of stream of people who visit the same learning object, the time for moving from one object to another, and the environmental parameters, need to be taken into account. Moreover, the values of these context-dependent parameters are likely to change swiftly during the learning process, which makes it a challenging and important issue to find a navigation support mechanism for suggesting learning paths for individual students in real time. In this paper, the navigation support problem for context-aware ubiquitous learning is formulated and two navigation support algorithms are proposed by taking learning efficacy and navigation efficiency into consideration. From the simulation results of learning in a butterfly museum setting, it is concluded that the innovative approach is helpful to the students to more effectively and efficiently utilize the learning resources and achieve better learning efficacy.     

Mechanical behavior of metallic glasses Mg–Cu–Y using nano-indentation

The mechanical properties of amorphous bulk metallic glassy (BMG) alloy, Mg₅₈Cu₃₁Y₁₁, are examined using nano-indentation scratching test. This study investigates the influences of different scratching conditions on the mechanical properties such as the friction force and the friction coefficient (μ) to understand the abrasive behavior of the BMG. The scratching conditions include applied normal load, depth of scratch, scratching velocity, and scratching temperature. The experimental results of the friction force, friction coefficient, hardness, and scratching morphology of BMG are characterized. The result shows that the friction force is nearly proportional to the normal load; and the friction force exhibits a slightly dependent on the scratching temperature. Then, regression analysis method is utilized to establish a formula to fit the scratching condition of BMGs. The regression analysis can be applied to model the mathematical relationship between the scratching parameters. The regression result shows a good agreement with experimental one.     

Reconstruction based fault prognosis for continuous processes

In this paper, a multivariate fault prognosis approach for continuous processes with hidden faults is proposed based on statistical process monitoring methods and multivariate time series prediction. It is assumed that the fault is a slowly time-varying autocorrelated process and can be completely reconstructed. Fault magnitude is estimated first via reconstruction, then predicted by a vector AR model with wavelet based denoising. Given the fault direction, a new index is proposed to detect the fault, which integrates fault detection and prognosis together. Case studies on a continuous stirred tank reactor and the Tennessee Eastman process demonstrate the effectiveness of the proposed approaches.     

Performance-based data distribution for data mining applications on grid computing environments

Effective data distribution techniques can significantly reduce the total execution time of a program on grid computing environments, especially for data mining applications. In this paper, we describe a linear programming formulation for the data distribution problem on grids. Furthermore, a heuristic method, named Heuristic Data Distribution Scheme (HDDS), is proposed to solve this problem. We implement two types of data mining applications, Association Rule Mining and Decision Tree Construction, and conduct experiments on grid testbeds. Experimental results show that data mining programs using the proposed HDDS to distribute data could execute more efficiently than traditional schemes could.     

Planning and Implementing Kerberos for Large Distributed Systems Part 1

ABSTRACT

For maintaining the consistency of database, the recovery algorithms traditionally depend on complete rollback to a consistent checkpoint. The recovery problem from committed malicious transactions can be solved by determining the dependencies between the transactions in window of vulnerability. Since the size of transactional log may grow very large, recovery becomes a complex and time-consuming process. In this paper, we propose an approach which incorporates application specific information to determine transactional dependencies. The approach is applied to column based transaction dependency to obtain better performance. The system is implemented at application layer where SQL queries are generated. In recovery phase, we consider only affected and malicious transactions for rollback and skip the good transactions.     

A Novel Approach to $H_{bm infty}$ Decentralized Fuzzy-Observer-Based Fuzzy Control Design for Nonlinear Interconnected Systems

In general, due to the interactions among subsystems, it is difficult to design an $H_{infty }$-decentralized output-feedback controller for nonlinear interconnected systems. This study introduces $H_{infty }$-decentralized fuzzy-observer-based fuzzy control design, where the premise variables depend on the state variables estimated by a fuzzy observer, for nonlinear interconnected systems via T--S fuzzy models. The fuzzy control design for this case is more flexible but much more complex than that for the case where the premise variables depend on the state variables only. A novel decoupled method is proposed in this study to transform the non-linear matrix inequality (non-LMI) conditions into some LMI forms. By the proposed decoupled method, the problem of $H_{infty }$-decentralized fuzzy-observer-based fuzzy control design for nonlinear interconnected systems is characterized in terms of solving an eigenvalue problem (EVP) with five prespecified scalars for each subsystem. In general, it is a difficult task to solve the EVP with five prespecified scalars. Fortunately, this special EVP can be easily solved by using a genetic algorithm and an LMI-based optimization method. Finally, a simulation example is given to illustrate the design procedure and robust performance of the proposed methods.      

Integrated Mining of Visual Features, Speech Features, and Frequent Patterns for Semantic Video Annotation

To support effective multimedia information retrieval, video annotation has become an important topic in video content analysis. Existing video annotation methods put the focus on either the analysis of low-level features or simple semantic concepts, and they cannot reduce the gap between low-level features and high-level concepts. In this paper, we propose an innovative method for semantic video annotation through integrated mining of visual features, speech features, and frequent semantic patterns existing in the video. The proposed method mainly consists of two main phases: 1) Construction of four kinds of predictive annotation models, namely speech-association, visual-association, visual-sequential, and statistical models from annotated videos. 2) Fusion of these models for annotating un-annotated videos automatically. The main advantage of the proposed method lies in that all visual features, speech features, and semantic patterns are considered simultaneously. Moreover, the utilization of high-level rules can effectively complement the insufficiency of statistics-based methods in dealing with complex and broad keyword identification in video annotation. Through empirical evaluation on NIST TRECVID video datasets, the proposed approach is shown to enhance the performance of annotation substantially in terms of precision, recall, and F-measure.