Examination of scale-invariant characteristics of epileptic electroencephalograms using wavelet-based analysis

There is evidence that biological and physiological systems including the brain exhibit can exhibit fractal characteristics that can be used to identify the state of the system. In this study, wavelet-based fractal analysis is used to examine self-similar or scale-invariant characteristics of intracranial EEG data in terms of the spectral exponent. The intracranial EEG data were recorded from subjects with epilepsy during non-seizure period and during epileptic seizure activity. From the computational results, it is observed that the self-similar or scale-invariant characteristics of the intracranial EEG data obtained during these two periods are significantly different. The actual value of the estimated spectral exponent depends on the wavelet bases used for the computations.     

Inverter statistics for online detection of stator asymmetries in inverter-fed induction motors

A simulation model of an induction machine that enables the analysis of the influence of stator fault conditions on the machine control of inverter-fed drives is presented. Based on this model, the influence of the breakdown of the stator windings insulation on the behavior of the machine and especially on the current control scheme is shown. A new online method to detect such asymmetries caused, for example, by an interturn insulation failure in the stator windings is proposed and investigated. This new method utilizes the influence of stator asymmetries on the inverter current control scheme. By evaluating the statistical distribution of the different inverter switching states and switching times, asymmetries in the stator can be detected and isolated. The measurements required to implement this method are already available in modern inverter fed drives as they are used to realize the current control loop. Thus, no additional sensors are necessary. The practical realization of the fault detection algorithm is demonstrated in combination with a predictive single step current controller. Measurements performed on a drive test stand verify the applicability of the proposed online method to detect and isolate stator faults.     

Generation of trajectory sensitivities: minimal-order realization

The problem of obtaining a minimal-order realization of a trajectory sensitivity model is investigated for a linear multi-input multi-output system. The dimension of the minimal realization is a function of some structural parameters of the system model and can be computed in a closed form. Several examples are given to illustrate the results.     

Fault detection and diagnosis of rotating machinery

A model-based approach to the detection and diagnosis of mechanical faults in rotating machinery is studied in this paper. For certain types of faults, for example, raceway faults in rolling element bearings, an increase in mass unbalance, and changes in stiffness and damping, algorithms suitable for real-time implementation are developed and evaluated using computer simulation     

Entropy-based measures for quantifying sleep-stage transition dynamics: relationship to sleep fragmentation and daytime sleepiness

We present two novel entropy-based measures that quantify sleep-stage transition dynamics (sleep structure) from polysomnogram derived hypnograms: Walsh spectral entropy (WSE) and Haar spectral entropy (HSE). These measures quantify patterns of temporal regularity of a categorical time series without requiring numerical encoding (scaling) of the (categorical) sleep stages. Additionally, we show that conditional entropy (CE) is well suited for quantifying predictability of the hypnogram. The relationship of those measures with traditional sleep fragmentation indices (arousal index, total sleep time, and sleep efficiency) is explored for a 394 participant sample of the Cleveland Family Study, an epidemiologic study in which standardized single-night polysomnogram data were collected. The new entropy-based sleep structure measures (WSE, HSE, and CE) are positively correlated (moderate to weak) with the traditional sleep fragmentation indices. Because the sleep structure measures developed in this paper provide direct information related to temporal patterns of sleep that is not contained in traditional sleep fragmentation measures, the correlation between these new alternative sleep structure measures and the traditional sleep fragmentation measures is less important. Our goal is not to develop alternative measures that correlate highly with traditional measures of sleep fragmentation, but rather to provide methods to quantify sleep structure by examining other (e.g., dynamic sleep-stage transition) properties of the hypnogram. Additionally, the relationship of the new entropy-based and traditional measures with daytime sleepiness as quantified by the Epworth sleepiness scale (ESS) is investigated. Multiple linear regression analysis shows that WSE has a stronger relationship with ESS than the traditional measures, even after both are adjusted for common confounders (age, race, gender, and body mass index). This further suggests that the entropy-based measures, especially WSE, are capturing additional temporal patterns of sleep not captured in the traditional sleep fragmentation measures, and have a relationship with daytime sleepiness.     

Optimal control of unknown parameter systems

The problem is discussed of finding a cost functional for which an adaptive control law is optimal. The system under consideration is a partially observed linear stochastic system with unknown parameters. It is well known that an optimal finite-dimensional filter for this problem can be derived when the parameters belong to a finite set. Since the optimal filter involves the evaluation of a finite set of a posteriori probabilities for each of the parameter values given the observations, a natural adaptive control scheme is: (i) develop the optimal linear feedback law given each parameter; (ii) use the a posteriori probabilities to form the weighted average (convex combination) of the individual control policies; and (iii) use the weighted average as the control law. A quadratic cost functional is devised for which this strategy is optimal, in a general case, and it is shown that the probing effect identified with dual control problems is inherent in the standard linear-quadratic-Gaussian problem with parameter uncertainty     

On the realization of input-decentralized representations for large-scale systems

The input-decentralization procedure proposed by Siljak and Vukcevic is analyzed in this note. It is shown that the procedure is valid on an open and dense subset of the parameter space of all large-scale systems with controllable subsystems. The examples which have appeared recently for which the proposed procedure fails are isolated points in the entire parameter space.     

New estimates for solutions of Lyapunov equations

Results for estimating the solution of differential and algebraic Lyapunov matrix equations are obtained, and some of the well-known results are generalized     

Speed control of electrical machines: unknown load torque case

The problem of specifying a desired torque trajectory to achieve speed tracking in passivity-based control of induction motors is addressed. This paper presents a solution to the problem that does not require an acceleration measurement nor knowledge of the load torque. To prove the main result a variant of Sontag's input to state stability is used