Fuzzy PCA-Guided Robust -Means Clustering

This paper proposes a new approach to robust clustering, in which a robust $k$-means partition is derived by using a noise-rejection mechanism based on the noise-clustering approach. The responsibility weight of each sample for the $k$-means process is estimated by considering the noise degree of the sample, and cluster indicators are calculated in a fuzzy principal-component-analysis (PCA) guided manner, where fuzzy PCA-guided robust $k$-means is performed by considering responsibility weights of samples. Then, the proposed method achieves cluster-core estimation in a deterministic way. The validity of the derived cluster cores is visually assessed through distance-sensitive ordering, which considers responsibility weights of samples. Numerical experiments demonstrate that the proposed method is useful for capturing cluster cores by rejecting noise samples, and we can easily assess cluster validity by using cluster-crossing curves.      

How to Make $T$-Transitive a Proximity Relation

Three ways to approximate a proximity relation $R$ (i.e., a reflexive and symmetric fuzzy relation) by a $T$ -transitive one where $T$ is a continuous Archimedean $t$-norm are given. The first one aggregates the transitive closure $overline{R}$ of $R$ with a (maximal) $T$-transitive relation $B$ contained in $R$ . The second one computes the closest homotecy of $overline{R}$ or $B$ to better fit their entries with the ones of $R$. The third method uses nonlinear programming techniques to obtain the best approximation with respect to the Euclidean distance for $T$ the $Lstrok$ukasiewicz or the product $t$-norm. The previous methods do not apply for the minimum $t$-norm. An algorithm to approximate a given proximity relation by a ${rm Min}$-transitive relation (a similarity) is given in the last section of the paper.      

On the NP-Hardness of Checking Matrix Polytope Stability and Continuous-Time Switching Stability

Motivated by questions in robust control and switched linear dynamical systems, we consider the problem checking whether all convex combinations of $k$ matrices in $R^{n times n}$ are stable. In particular, we are interested whether there exist algorithms which can solve this problem in time polynomial in $n$ and $k$. We show that if $k= lceil n^{d} rceil $ for any fixed real $d>0$, then the problem is NP-hard, meaning that no polynomial-time algorithm in $n$ exists provided that $P ne NP$, a widely believed conjecture in computer science. On the other hand, when $k$ is a constant independent of $n$ , then it is known that the problem may be solved in polynomial time in $n$. Using these results and the method of measurable switching rules, we prove our main statement: verifying the absolute asymptotic stability of a continuous-time switched linear system with more than $n^{d}$ matrices $A_{i} in R^{n times n}$ satisfying $0 succeq A_{i} + A_{i}^{T}$ is NP-hard.      

Factor Analysis Latent Subspace Modeling and Robust Fuzzy Clustering Using $t$-Distributions

Factor analysis is a latent subspace model commonly used for local dimensionality reduction tasks. Fuzzy $c$-means (FCM) type fuzzy clustering approaches are closely related to Gaussian mixture models (GMMs), and expectation--maximization (EM) like algorithms have been employed in fuzzy clustering with regularized objective functions. Student's $t$ -mixture models (SMMs) have been proposed recently as an alternative to GMMs, resolving their outlier vulnerability problems. In this paper, we propose a novel FCM-type fuzzy clustering scheme providing two significant benefits when compared with the existing approaches. First, it provides a well-established observation space dimensionality reduction framework for fuzzy clustering algorithms based on factor analysis, allowing concurrent performance of fuzzy clustering and, within each cluster, local dimensionality reduction. Second, it exploits the outlier tolerance advantages of SMMs to provide a novel, soundly founded, nonheuristic, robust fuzzy clustering framework by introducing the effective means to incorporate the explicit assumption about Student's $t$ -distributed data into the fuzzy clustering procedure. This way, the proposed model yields a significant performance increase for the fuzzy clustering algorithm, as we experimentally demonstrate.      

$ {cal L}_{2}$ Gain of Periodic Linear Switched Systems: Fast Switching Behavior

We investigate the $ {cal L}_{2}$ gain of periodic linear switched systems under fast switching. For systems that possess a suitable notion of a time-average system, we characterize the relationship between the ${cal L}_{2}$ gain of the switched system and the ${cal L}_{2}$ gain of its induced time-average system when the switching rate is sufficiently fast. We show that the switched system ${cal L}_{2}$ gain is in general different from the average system ${cal L}_{2}$ gain if the input or output coefficient matrix switches. If only the state coefficient matrix switches, the input-output energy gain for a fixed ${cal L}_{2}$ input signal is bounded by the ${cal L}_{2}$ gain of the average system as the switching rate grows large. Additionally, for a fixed ${cal L}_{2}$ input, the maximum pointwise in time difference between the switched and average system outputs approaches zero as the switching rate grows.      

On $Delta$-Transforms

Any set of two legs in a Gough–Stewart platform sharing an attachment is defined as a $Delta$component. This component links a point in the platform (base) to a line in the base (platform). Thus, if the two legs, which are involved in a $Delta$ component, are rearranged without altering the location of the line and the point in their base and platform local reference frames, the singularity locus of the Gough–Stewart platform remains the same, provided that no architectural singularities are introduced. Such leg rearrangements are defined as $Delta$-transforms, and they can be applied sequentially and simultaneously. Although it may seem counterintuitive at first glance, the rearrangement of legs using simultaneous $Delta$-transforms does not necessarily lead to leg configurations containing a $Delta$component. As a consequence, the application of $Delta$-transforms reveals itself as a simple, yet powerful, technique for the kinematic analysis of large families of Gough–Stewart platforms. It is also shown that these transforms shed new light on the characterization of architectural singularities and their associated self-motions.      

Visual Servo Control Achieving Nanometer Resolution in $X$ –$Y$–$Z$

This paper presents a three-axis vision motion sensor and its applications to visual servo control. The vision sensor is integrated with a three-axis piezo stage to form a visual servo control system that achieves nanometer resolution in all three $x$– $y$–$z$ motion axes. Motion measurement is achieved using a single interferometer-equipped optical microscope. A real-time image-processing algorithm that processes interference fringe patterns and that achieves nanometer out-of-plane resolution is presented. Furthermore, a feedback-control scheme is introduced to control the sensor plane using an Objective-Z-Positioner to enable automatic tracking of moving objects. It expands the out-of-plane measurement range of the vision sensor beyond its inherent depth of field of several micrometers to 100 $mu$m and beyond. An integrated visual servo system is implemented and experimental results are shown.      

Decentralized $H_{infty }$ Filter Design for Discrete-Time Interconnected Fuzzy Systems

This paper describes a decentralized $H_{infty }$ filter design for discrete-time interconnected fuzzy systems based on piecewise-quadratic Lyapunov functions. The systems consist of $J$discrete-time interconnected Takagi–Sugeno (T–S) fuzzy subsystems, and a decentralized $H_infty$ filter is designed for each subsystem. It is shown that the stability of the overall filtering-error system with $H_{infty }$ performance can be established if a piecewise-quadratic Lyapunov function can be constructed. Moreover, the parameters of filters can be obtained by solving a set of linear matrix inequalities that are numerically feasible. Two simulation examples are given to show the effectiveness of the proposed approach.      

Performance Analysis for Overlay Multimedia Multicast on $r$-ary Tree and $m$-D Mesh Topologies

Without requiring multicast support from the underlying networks, overlay multicast has the advantage of implementing inter-domain multimedia multicast communications. Usually, overlay multicast protocols employ two different topologies: $r$-ary tree and $m$-D mesh. In this paper, we study the influence of topology selection on multimedia multicast performance. We present a set of theoretical results on the worst performance, the average performance, and the performance difference along the link stress, the number of overlay hops, and the number of shortest paths for $r$-ary tree-based and $m$ -D mesh-based multicast, respectively. Furthermore, through simulations in NS2, we observe and compare tree and mesh topologies along the metrics analyzed theoretically. Simulation results match our theoretical analyses. Finally we give our evaluations of and insights into these two kinds of multicast when used to transmit multimedia streams. The selection of overlay topology is application dependent. To the best of our knowledge, this is the first evaluation of multimedia multicast performances in different overlay topologies. We believe that this study is useful for protocol design of target multimedia applications and for investigating multicast functions.      

Fuzzy Subsethood for Fuzzy Sets of Type-2 and Generalized Type- ${n}$

In this paper, we use Zadeh's extension principle to extend Kosko's definition of the fuzzy subsethood measure $S(G,H)$ to type-2 fuzzy sets defined on any set $X$ equipped with a measure. Subsethood is itself a fuzzy set that is a crisp interval when $G$ and $H$ are interval type-2 sets. We show how to compute this interval and then use the result to compute subsethood for general type-2 fuzzy sets. A definition of subsethood for arbitrary fuzzy sets of type- $n ≫ 2$ is then developed. This subsethood is a type-( $n-1$) fuzzy set, and we provide a procedure to compute subsethood of interval type-3 fuzzy sets.      

A ${bm lambda}$-Cut and Goal-Programming-Based Algorithm for Fuzzy-Linear Multiple-Objective Bilevel Optimization

Bilevel-programming techniques are developed to handle decentralized problems with two-level decision makers, which are leaders and followers, who may have more than one objective to achieve. This paper proposes a ${lambda}$-cut and goal-programming-based algorithm to solve fuzzy-linear multiple-objective bilevel (FLMOB) decision problems. First, based on the definition of a distance measure between two fuzzy vectors using ${lambda}$-cut, a fuzzy-linear bilevel goal (FLBG) model is formatted, and related theorems are proved. Then, using a ${lambda}$-cut for fuzzy coefficients and a goal-programming strategy for multiple objectives, a ${lambda}$-cut and goal-programming-based algorithm to solve FLMOB decision problems is presented. A case study for a newsboy problem is adopted to illustrate the application and executing procedure of this algorithm. Finally, experiments are carried out to discuss and analyze the performance of this algorithm.      

A Selection Model for Optimal Fuzzy Clustering Algorithm and Number of Clusters Based on Competitive Comprehensive Fuzzy Evaluation

Fuzzy $c$-means (FCM) and its variants suffer from two problems---local minima and cluster validity---which have a direct impact on the formation of final clustering. There are two strategies---optimization and center initialization strategies---that address the problem of local minima. This paper proposes a center initialization approach based on a minimum spanning tree to keep FCM from local minima. With regard to cluster validity, various strategies have been proposed. On the basis of the fuzzy cluster validity index, this paper proposes a selection model that combines multiple pairs of a fuzzy clustering algorithm and cluster validity index to identify the number of clusters and simultaneously selects the optimal fuzzy clustering for a dataset. The promising performance of the proposed center-initialization method and selection model is demonstrated by experiments on real datasets.      

Divide and Conquer: EKF SLAM in $O(n)$

In this paper, we show that all processes associated with the move-sense-update cycle of extended Kalman filter (EKF) Simultaneous Localization and Mapping (SLAM) can be carried out in time linear with the number of map features. We describe Divide and Conquer SLAM, which is an EKF SLAM algorithm in which the computational complexity per step is reduced from $O(n^2)$ to $O(n)$, and the total cost of SLAM is reduced from $O(n^3)$ to $O(n^2)$. Unlike many current large-scale EKF SLAM techniques, this algorithm computes a solution without relying on approximations or simplifications (other than linearizations) to reduce computational complexity. Also, estimates and covariances are available when needed by data association without any further computation. Furthermore, as the method works most of the time in local maps, where angular errors remain small, the effect of linearization errors is limited. The resulting vehicle and map estimates are more precise than those obtained with standard EKF SLAM. The errors with respect to the true value are smaller, and the computed state covariance is consistent with the real error in the estimation. Both simulated experiments and the Victoria Park dataset are used to provide evidence of the advantages of this algorithm.      

Simultaneous $H^{infty}$ Control for Nonlinear Systems

This note develops a novel method for designing simultaneous $H^{infty}$ state feedback controllers for a collection of single-input nonlinear systems. Based on the Kalman—Yakubovich—Popov Lemma, necessary and sufficient conditions for the existence of simultaneous $H^{infty}$ controllers are derived by the control storage function approach. A universal formula for constructing continuous, time-invariant, simultaneous $H^{infty}$ state feedback controllers is presented.      

Post-CMOS-Compatible Aluminum Nitride Resonant MEMS Accelerometers

This paper describes the development of aluminum nitride (AlN) resonant accelerometers that can be integrated directly over foundry CMOS circuitry. Acceleration is measured by a change in resonant frequency of AlN double-ended tuning-fork (DETF) resonators. The DETF resonators and an attached proof mass are composed of a 1- $muhbox{m}$ -thick piezoelectric AlN layer. Utilizing piezoelectric coupling for the resonator drive and sense, DETFs at 890 kHz have been realized with quality factors $(Q)$ of 5090 and a maximum power handling of 1 $muhbox{W}$. The linear drive of the piezoelectric coupling reduces upconversion of $1/f$ amplifier noise into $1/f^{3}$ phase noise close to the oscillator carrier. This results in lower oscillator phase noise, $-$96 dBc/Hz at 100-Hz offset from the carrier, and improved sensor resolution when the DETF resonators are oscillated by the readout electronics. Attached to a 110-ng proof mass, the accelerometer microsystem has a measured sensitivity of 3.4 Hz/G and a resolution of 0.9 $hbox{mG}/surdhbox{Hz}$ from 10 to 200 Hz, where the accelerometer bandwidth is limited by the measurement setup. Theoretical calculations predict an upper limit on the accelerometer bandwidth of 1.4 kHz.$hfill$ [2008-0190]      

Serially Connected Weight-Balanced Digital Actuators for Wide-Range High-Precision Low-Voltage-Displacement Control

This paper presents the design, fabrication, and characterization of a new serial digital actuator, achieving an improvement in range-to-precision and range-to-voltage performance. We propose a weight-balanced design for the serial actuators with serpentine springs with a serial arrangement of unit digital actuators. We have measured the displacement range, precision, and drive voltage of unit and serial actuation at 1 Hz. The serial digital actuators produce a full-range displacement of $28.44 pm 0.02 muhbox{m}$ , accumulating unit displacements of $2.8 pm 0.5 muhbox{m}$ at an operating voltage of 4.47 $pm$ 0.07 V. In addition, the serial digital actuators that have a displacement precision of 37.94 $pm$ 6.26 nm do not accumulate the displacement errors of the unit actuators, i.e., 36.0 $pm$ 17.7 nm. We experimentally verify that the serial digital actuators achieve a range-to-squared-voltage ratio of 1.423 $muhbox{m/V}^{2}$ and a range-to-precision ratio of 749.6.$hfill$ [2009-0020]      

6H-SiC JFETs for 450 $^{circ}hbox{C}$ Differential Sensing Applications

N-channel 6H-SiC depletion-mode junction field-effect transistors (JFETs) have been fabricated, and characterized for use in high-temperature differential sensing. Electrical characteristics of the JFETs have been measured and are in good agreement with predictions of an abrupt-junction long-channel JFET model. The electrical characteristics were measured across a 2-in wafer for temperatures from 25 $^{ circ}hbox{C}$ to 450 $^{circ}hbox{C}$, and the extracted pinchoff voltage has a mean of 11.3 V and a standard deviation of about 1.0 V at room temperature, whereas pinchoff current has a mean of 0.41 mA with standard deviation of about 0.1 mA. The change in pinchoff voltage is minimal across the measured temperature range, whereas pinchoff current at 450 $^{circ}hbox{C}$ is about half its value at room temperature, consistent with the expected change in the $nmu_{n}$ product. The characterization of differential pairs and hybrid amplifiers constructed using these differential pairs is also reported. A three-stage amplifier with passive loads has a differential voltage gain of 50 dB, and a unity-gain frequency of 200 kHz at 450 $^{circ}hbox{C}$, limited by test parasitics. A two-stage amplifier with active loads has reduced sensitivity to off-chip parasitics and exhibits a differential voltage gain of 69 dB with a unity-gain frequency of 1.3 MHz at 450 $^{circ}hbox{C}$.$hfill$[2009-0029]      

Computation of Diagnosable Fault-Occurrence Indices for Systems With Repeatable Faults

Certain faults, such as intermittent or non-persistent faults, may occur repeatedly. For discrete-event systems prone to repeated faults, Jiang, Kumar, and Garcia in 2003 introduced the notion of $ell $-diagnosability requiring the diagnosis of the $ell$ th occurrence of a fault within a bounded delay. The present paper studies the identification of the set of all indices $ell $ for which the system is not $ell $-diagnosable. (These are precisely the occurrence indices for which a repeatable-fault cannot be diagnosed.) We present an algorithm of cubic complexity to determine whether a system is diagnosable for every fault-occurrence index. For those systems for which the answer is negative, we show that the set of indices for which the system is non-diagnosable possesses the property that it is either finite or consists of a finite number of eventually periodic sets.      

Real-Time Temperature Compensation of MEMS Oscillators Using an Integrated Micro-Oven and a Phase-Locked Loop

We present a new temperature compensation system for microresonator-based frequency references. It consists of a phase-locked loop (PLL) whose inputs are derived from two microresonators with different temperature coefficients of frequency. The resonators are suspended within an encapsulated cavity and are heated to a constant temperature by the PLL controller, thereby achieving active temperature compensation. We show repeated real-time measurements of three 1.2-MHz prototypes that achieve a frequency stability of $pm$ 1 ppm from $-20 ^{circ}hbox{C}$ to $+80 ^{circ}hbox{C}$, as well as a technique to reduce steady-state frequency errors to $pm$0.05 ppm using multipoint calibration.$hfill$[2009-0074]      

Discharge-Based Pressure Sensors for High-Temperature Applications Using Three-Dimensional and Planar Microstructures

Two versions of microdischarge-based pressure sensors, which operate by measuring the change, with pressure, in the spatial current distribution of pulsed dc microdischarges, are reported. The inherently high temperatures of the ions and electrons in the microdischarges make these devices amenable to high-temperature operation. The first sensor type uses 3-D arrays of horizontal bulk metal electrodes embedded in quartz substrates with electrode diameters of 1–2 mm and 50–100-$muhbox{m}$ interelectrode spacing. These devices were operated in nitrogen over a range of 10–2000 torr, at temperatures as high as 1000 $^{circ}hbox{C}$. The maximum measured sensitivity was 5420 ppm/torr at the low end of the dynamic range and 500 ppm/torr at the high end, while the temperature coefficient of sensitivity ranged from $-$925 to $-$550 ppm/K. Sensors of the second type use planar electrodes and have active areas as small as 0.13 $hbox{mm}^{2}$. These devices, when tested in a chemical sensing system flowing helium as a carrier gas, had a maximum sensitivity of 9800 ppm/torr, a dynamic range of 25–200 torr, and a temperature coefficient of sensitivity of approximately $-$1412 ppm/K.$hfill$ [2008-0262]