Application of multi-layer perceptron neural networks to vision problems

This paper discusses the application of a class of feed-forward Artificial Neural Networks (ANNs) known as Multi-Layer Perceptrons(MLPs) to two vision problems: recognition and pose estimation of 3D objects from a single 2D perspective view; and handwritten digit recognition. In both cases, a multi-MLP classification scheme is developed that combines the decisions of several classifiers. These classifiers operate on the same feature set for the 3D recognition problem whereas different feature types are used for the handwritten digit recognition. The backpropagationlearning rule is used to train the MLPs. Application of the MLP architecture to other vision problems is also briefly discussed.     

A game theoretic fault detection filter

The fault detection process is approximated with a disturbance attenuation problem. The solution to this problem, for both linear time-varying and time-invariant systems, leads to a game theoretic filter which bounds the transmission of all exogenous signals except the fault to be detected. In the limit, when the disturbance attenuation bound is brought to zero, a complete transmission block is achieved by embedding the nuisance inputs into an unobservable, invariant subspace. Since this is the same invariant subspace structure seen in some types of detection filters, we can claim that the asymptotic game filter is itself a detection filter. One can also make use of this subspace structure to reduce the order of the limiting game theoretic filter by factoring this invariant subspace out of the state space. The resulting lower dimensional filter will then be sensitive only to the failure to be detected. A pair of examples given at the end of the paper demonstrate the effectiveness of the filter for time-invariant and time-varying problems in both full-order and reduced-order forms     

A note on the robust stability of uncertain stochastic fuzzy systems with time-delays

Takagi-Sugeno (T-S) fuzzy models are now often used to describe complex nonlinear systems in terms of fuzzy sets and fuzzy reasoning applied to a set of linear submodels. In this note, the T-S fuzzy model approach is exploited to establish stability criteria for a class of nonlinear stochastic systems with time delay. Sufficient conditions are derived in the format of linear matrix inequalities (LMIs), such that for all admissible parameter uncertainties, the overall fuzzy system is stochastically exponentially stable in the mean square, independent of the time delay. Therefore, with the numerically attractive Matlab LMI toolbox, the robust stability of the uncertain stochastic fuzzy systems with time delays can be easily checked.     

An evolutionary approach to pattern-based time series segmentation

Time series data, due to their numerical and continuous nature, are difficult to process, analyze, and mine. However, these tasks become easier when the data can be transformed into meaningful symbols. Most recent works on time series only address how to identify a given pattern from a time series and do not consider the problem of identifying a suitable set of time points for segmenting the time series in accordance with a given set of pattern templates (e.g., a set of technical patterns for stock analysis). However, the use of fixed-length segmentation is an oversimplified approach to this problem; hence, a dynamic approach (with high controllability) is preferable so that the time series can be segmented flexibly and effectively according to the needs of the users and the applications. In view of the fact that this segmentation problem is an optimization problem and evolutionary computation is an appropriate tool to solve it, we propose an evolutionary time series segmentation algorithm. This approach allows a sizeable set of pattern templates to be generated for mining or query. In addition, defining similarity between time series (or time series segments) is of fundamental importance in fitness computation. By identifying the perceptually important points directly from the time domain, time series segments and templates of different lengths can be compared and intuitive pattern matching can be carried out in an effective and efficient manner. Encouraging experimental results are reported from tests that segment both artificial time series generated from the combinations of pattern templates and the time series of selected Hong Kong stocks.     

Kinematic optimal design of a paramill: A multi‐SP device

In this paper, a manipulability analysis of a new parallel‐type rolling mill, named “paramill,” in its conceptual design stage is investigated. The paramill considered uses two Stewart platforms (SPs) in opposite directions for the generation of 6 degree‐of‐freedom motions of individual work‐rolls. The objective of this new approach is to pursue an integrated control of the strip thickness, strip shape, pair‐crossing angle, uniform wear of the rolls, and tension of the strip. The forward/inverse kinematics problems are formulated. Two main kinematic parameters, the size of the base and the acute angle made by two neighboring joints for a given size of the work‐roll, have been determined in the way that the force and moment transmission from the actuators to the work‐rolls is maximized. © 2004 Wiley Periodicals, Inc.     

Modeling and control for a Gough‐Stewart platform CNC machine

In this paper, a complete dynamic model on task space for a 6 degrees of freedom (DOF) Gough‐Stewart platform‐type computer numerical control (CNC) machine is derived. The rotation terms of the legs are included for those inertia effects cannot be negligible in the machine tool applications. The formulation derived by means of the Euler‐Lagrange method is convenient for designing the adaptive control law. Also, the average‐type force model for end milling process is derived and included in the dynamic model and control. A composite adaptive control scheme is developed by use of filtering dynamics technique. An appropriate estimator gain is designed in the parameter adaptation law that is useful for estimating the selected important cutting parameters. Experimental results verify the proposed adaptive control scheme can achieve good tracking performance. © 2004 Wiley Periodicals, Inc.     

Design of parts and manufacturing systems for reliability and maintainability

The design of parts and manufacturing systems can be viewed from different perspectives. This paper discusses the design of parts and manufacturing systems from reliability and maintainability perspectives. A convenient method of considering reliability and maintainability is through the use of design rules which are based on empirical and theoretical knowledge. Four design rules for improvement of reliability and maintainability of systems are presented. The design rules are proven and substantiated with numerical results. The relationship between the design for reliability and maintainability rules and design for manufacturing process rules is discussed.     

Decreased Expression of Plakophilin-2 and αT-Catenin in Arrhythmogenic Right Ventricular Cardiomyopathy: Potential Markers for Diagnosis

Arrhythmogenic right ventricular cardiomyopathy (ARVC) is a hereditary disease of the heart muscle. Clinical challenges remain, however, in identifying patients with ARVC in the early or concealed stages with subtle clinical manifestations. Therefore, we wanted to identify potential targets by immunohistochemical (IHC) analysis in comparison with controls. Pathogenic mutations were identified in 11 of 37 autopsied patients with ARVC. As observed from IHC analysis of the RV, expression of αT-catenin and plakophilin-2 is significantly decreased in autopsied patients with ARVC as compared to controls, and the decreased expression is consistent in patients with and without pathogenic mutations. Furthermore, ARVC specimens demonstrated a reduced localization of αT-catenin, desmocollin-2, desmoglein-2, desmoplakin, and plakophilin-2 on intercalated discs. These findings have been validated by comparing RV specimens obtained via endomyocardial biopsy between patients with ARVC and those without. The pathogenic mutation was present in 3 of 5 clinical patients with ARVC. In HL-1 myocytes, siRNA was used to knockdown CTNNA3, and western blotting analysis demonstrated that the decline in αT-catenin expression was accompanied by a significant decline in the expression of plakophilin-2. The aforementioned effect was directed towards protein degradation rather than mRNA stability. Plakophilin-2 expression decreases concurrently with the decline in CTNNA3 expression. Therefore, the expression of αT-catenin and plakophilin-2 could be potential surrogates for the diagnosis of ARVC.