Search

All complex real world problems are essentially nonlinear. Linear models are relatively simple but inaccurate to describe the nonlinear aspects of dynamic system behaviors. Denoising techniques have been broadly applied to numerous applications in the spatial domain, frequency domain, and time domain. To increase the adaptability of denoising techniques to signal processing of arbitrary nonlinear systems, kernel based nonlinear component analysis is proposed to enhance wavelet denoising. In the multilevel wavelet decomposition, the low frequency approximations and high frequency details are produced at each level. Discrete wavelet transform (DWT) will help to decompose low frequency approximations exclusively at the subsequent levels, while the wavelet packet transform decomposes both approximations and high frequency details at each level. DWT is selected for wavelet denoising in this study, where details at each level and the approximation at specified level are all subject to simplification using nonlinear component analysis. Case studies of typical nonlinear denoising problems in various domains are conducted. The results manifest strong feasibility and adaptability across diverse denoising problems of nonlinear systems.

Even though actual composition of engine exhaust gases varies across diverse types of engines, such as gasoline, diesel, gas turbine and natural gas engines, engine exhaust temperature is always a major factor with strong impact on emission levels and catalytic converting efficiency. For spark ignition engines, exhaust temperature depends on various engine parameters, such as engine speed, engine load, A/F ratio, intake air temperature, coolant temperature and spark timing, etc. Due to complexity, it is impossible to share a unique analytical model of engine exhaust temperature. Instead, it is mostly modeled as a complicated nonlinear system. The model complexity increases significantly however accuracy cannot be guaranteed. On the other hand, a simple linear model with accurate system identification could serve as a versatile alternative to represent the engine exhaust temperature, while engine parameters are subject to model identification to be adaptable across different types of engines. Combination of linear functions in terms of dominant engine parameters of engine speed and engine load is used for exhaust temperature modeling. To identify optimal parameters, Markov Chain Monte Carlo (MCMC) is applied. The discrete-time Markov chain is introduced where the stationary probability replaces posterior density in Monte Carlo integration for numerical integration. Compared with the high order nonlinear approaches, low computation cost is involved in the simplified model. Good agreement between the model prediction data and testing results is observed. The approach could be easily extended to other types of engines.

Systematic theoretical analysis on performance of various types of notch filters is presented from a control point of view. It focuses on practical implementation of notch filters in a broad frequency range. The transfer function formulation for diverse types of 2nd order notch filters and double notch filters is conducted for performance analysis. The role of the quality factor Q on the cutoff bandwidth, edge sharpness, system stability and sensitivity is also examined. Bode plots are then introduced to evaluate stability margins of the gain and phase. It provides a theoretical basis to design electrical and optical notch filters.

Generalized notch filters have various applications in electrical and optical systems, which could also be interconnected with diverse systems. Both the transfer function and state space approaches can be used to formulate 2nd order notch filters and double notch filters to conduct performance analysis on a basis of classical control and modern control theories. Theoretical design of diverse types of notch filters is conducted, covering basic type, standard type, lowpass and highpass notch filters. In addition, root locus plots are introduced to track locations of system poles of notch filters for stability analysis and potential accurate pole placement. This approach can be expanded to notch filter analysis in other systems as well.

Feature detection is a fundamental technique in broad fields of image processing, pattern recognition and computer vision. A digital image in general contains objects, edges, noises and background. Critical changes in properties of objects can be captured via detecting sharp variations in image brightness. The edges can be detected via numerous approaches on a basis of image intensity changes. Edge broken and false detection are typical problems using classical methods, which will result in information loss and feature deformity. The notion of optimization is thus introduced into edge detection. The Canny edge detector and Ant Colony Optimization (ACO) detector are among the most successful and effective approaches for edge detection. The Canny edge detector is designed to capture edges by searching local optima of the gradient of the intensity. It is susceptible to noises presenting on the raw images, so details of images could be slightly changed when Gaussian smoothing is applied. To improve accuracy, the adaptive edge tracing scheme is proposed. On the other hand, artificial intelligence has also been introduced. Being one of metaheuristic optimization approaches, the evolutionary computing oriented ACO becomes a promising approach for feature capturing without necessity of smoothing filters. Selection of maximum intensity difference as the path visibility function for ACO will contribute better to generate true edges and avoid false edges. Both the adaptive Canny edge detection and enhanced ACO are proposed in this article. Comparative studies are also conducted to evaluate the edge detection qualities. The outcomes are analyzed and evaluated from both qualitative and quantitative points of view, where merits and drawbacks of the two schemes have been indicated.

Wall wetting is a typical phenomenon in Port Fuel Injection (PFI) engines. The injected fuel is atomized into tiny droplets so as to form a mixture of the vaporized fuel and air. Certain portion of the vaporized fuel or fuel droplet would condense on the wall of the intake manifold and intake value, forming the fuel puddle, which has a significant impact on the air to fuel ratio (A/F ratio) of the cylinder. Detrimental A/F ratio excursion could occur during engine transient operations, giving rise to excessive exhaust emissions and extra fuel consumption, as well as the driveability and catalyst efficiency problems. Hence, some advanced control schemes should be applied. The model predictive control methodology is proposed to compensate for A/F ratio excursion during engine transient operations at throttle tip in or tip out. In light of the fact that additional computational complexity involves in nonlinear intelligent control approaches and big approximation error occurs in simple linear approaches, the relatively simple but powerful nonlinear Markov Chain Monte Carlo (MCMC) is introduced to solve this problem. MCMC is a general approach for obtaining random samples at the stationary probability to substitute arbitrary type of the posterior density. The discrete-time Markov chain is introduced whose state reaches the desired distribution after large numbers of iterations. It is combined with Monte Carlo integration to implement numerical integration. Excellent match between the model prediction data and actual experimental data is observed using the MCMC approach with the low computation cost.

To expand the dynamic range of intensity levels in actual scenes, high dynamic range imaging is needed which is able to distinguish detail information in the scenes accurately throughout highlight areas and shadow areas at the same time. Digital image fusion is the potential solution to capture critical information from scenes covering both lightest and darkest areas. Wavelet analysis is one of leading techniques for image fusion, which decomposes digital images by means of a set of basis functions. At each level of 2D discrete wavelet transform (DWT), the source image is decomposed into four images of a quarter size at a coarser scale, resulting in one approximation coefficient and three detail coefficients. Two source images can be merged by simple fusion operations such as averaging, minimizing and maximizing. To further enhance image fusion quality across a broader dynamic range, support vector machine (SVM) is proposed which is used to extract a well-trained hyperplane for binary decision making in the fusion process. This integration approach demonstrates the efficiency and effectiveness on data retrieval. Visual appealing depicts remarkable improvement accordingly. In addition, the quantitative metrics of output fusion images have demonstrated the advantages of fusion integration.

Biometric verification is one of the leading security issues. Being the unique signature (e.g., fingerprint) of human beings, biometric pattern recognition serves as an interesting branch of study in disciplines of digital image processing. Due to the fact that the accuracy of pattern recognition determines the correctness of decision making, two nonlinear approaches have been introduced, where independent component analysis (ICA) and discrete wavelet transform (DWT) are among the most effective methods. After sensing information is digitized into image matrix files, data matrices could be analyzed using ICA algorithms. 2D DWT is capable of compressing fingerprints and then reconstructing patterns by reorganizing approximation components and revised components of horizontal detail, vertical detail and diagonal detail via thresholding. Satisfactory results can be obtained from both approaches, however, it is hard to show which approach could be better performed from visual appealing. In this case, one set of quantitative measures is presented to evaluate outcomes of biometric verification on a basis of ICA and 2D DWT, where the objective measures of the discrete entropy, discrete energy, relative entropy, contrast, homogeneity, dissimilarity and mutual information are used to illustrate drawbacks and merits of these two dominating biometric recognition approaches.

Feature extraction involves detecting and separating fundamental portions of digital images, as well as simplifying a complex input data set into a set of features in a reduced order. It has a variety of practical applications in virtually all areas of pattern recognition and image processing disciplines. Dimensionality reduction is a major goal of feature extraction to retain the relevant information and eliminate the redundant and noisy information. To accurately describe a huge set of complex data with a lower dimensional space representation, optimality should be introduced into feature extraction, which includes both linear and nonlinear approaches, among which, both Principal Components Analysis (PCA) and Independent Component Analysis (ICA) are the most successful ones. Merits and drawbacks of these two linear and nonlinear approaches can be obtained via comparative analysis. The RGB model is an additive color model which displays a color by mixing intensity levels of red, green and blue components. The grayscale value is set when all three primary colors have the same intensity. Most digital images are displayed in RGB format, so optimal feature extraction should be conducted according to three individual components in the RGB model. In order to evaluate the two leading schemes of PCA and ICA for feature extraction, the quantitative metrics based on the information theory have been introduced so as to examine the qualities of optimal feature extraction approaches from both subjective and objective points of view.

A systematic approach for spectral calibration, extraction and characterization is presented. It involves the simple technology integration of the linear and nonlinear methodologies. Raman spectra are essentially weak signals whose features are sensitive to a variety of noises, while the background spectra (e.g., fluorescence spectra) act as another primary intensity component in the raw Raman spectrograph. Thus the calibrated Raman spectra from measurements are subject to detrend and denoising so as to extract the intrinsic spectra, which contain solely the unique spectral signatures of the individual biomedical samples. By removing the slowly varying baseline background spectra and rapidly varying noises, intrinsic Raman spectra could be extracted using a combination of linear and nonlinear methodologies. The goal is to capture the distinguishable wave number information on the Raman shift for feature identification. This approach is crucial for spectra analysis to associate Raman spectra with the medical diagnosis. Considering the slowly varying nature of background spectra such as fluorescence, linear least squares estimation is applied to detrend the background spectra from the measured spectra being calibrated. Since most noises are rapidly varying, nonlinear discrete wavelet transform (DWT) denoising is applied to separate the intrinsic Raman spectra from noises for spectral identification. This approach is proposed so as to differentiate the normal spectra from the abnormal spectra of the fresh mice lung samples. Numerical simulation outcomes include the calibrated Raman spectra, detrended spectra free of background effects and the extracted intrinsic spectra after nonlinear DWT denoising. The satisfactory results indicate that the proposed simple linear and nonlinear technology integration provides a useful approach for biomedical sample characterization. In addition, it reduces computational complexity compared with artificial intelligence approaches, which has no technical difficulty in real time medical diagnosis implementation. This serves as a fundamental step for further data clustering and accurate decision making among numerous diverse samples.

Digital image compression has numerous practical applications to effectively utilize storage capacities. Two basic techniques are lossless and lossy compression. The former employs probabilistic models for lossless storage on a basis of statistical redundancy occurred in digital images. However, it has limitation in compression ratio. Thus, the latter has been widely applied to further improve storage capacities, covering various fundamental digital image processing approaches. It has been documented that most lossy compression schemes will provide perfect visual perception under an exceptional compression ratio, among which discrete wavelet transform (DWT), discrete Fourier transform (DFT) and statistical optimization compression schemes (e.g., principal component analysis or independent component analysis) are dominant approaches. It is necessary to evaluate these compression and reconstruction schemes objectively. Using a well defined set of quantitative metrics by means of Information Theory, a comparative study on several typical image compression and reconstruction schemes will be made in this research.

The motion of typical aerodynamic systems can be simplified as the rigid multibody dynamics, such as applications on the vehicle dynamics and flight dynamics. The multibody dynamics represents the interaction among the individuals with large translational and rotational displacements. In the presence of parametric and nonparametric uncertainties, the stability issue is extremely critical to nonlinear control design of the multibody dynamic systems. Backstepping control acts as an engineering technique to stabilize controls for nonlinear dynamical systems, which is regardless of operating points. It decomposes complex nonlinear systems into a series of subsystems, where candidate Lyapunov functions and virtual controls can be employed. As the approaches for designing the controllers vary significantly, this research is intended to derive and simplify the fundamental approach for aerodynamic nonlinear systems via the nonlinear backstepping approach.

Data compression is essentially a technical solution to confront the needs of the storage capacity and data redundancy. As a typical multimedia technique, compression on digital images and videos plays an important role in data compression. Some lossy image compression methods are acceptable for achieving a substantial reduction in bit rate, with minor information loss. From the point of view of compression, wavelet transform has emerged as a dominant technology in the fields of signal and image processing, such as the discrete wavelet transform and wavelet packet decomposition. However, a regular transform is limited by its wavelet bases which downsize by a power of two towards the low frequencies, thus it may not produce best results. Being a stochastic population based evolutionary algorithm, the particle swarm optimization (PSO) simulates social optimization, where the swarm is modeled by particles in multidimensional space together with a position and a velocity. In this case, it has been proposed to enhance the quality of the wavelet based image compression. Objective measures are also presented to evaluate the integration of the wavelet expansion technologies and PSO optimization scheme.

The objective of image denoising is to remove the noises and to retain important image features as much as possible. Linear approaches could be effective for some simple cases with slowly varying noises, but not for other slowly varying noise cases and rapidly varying noise cases. As a nonlinear wavelet based technique, the wavelet thresholding is effective to denoise blurring aerial images. Either the discrete wavelet transform or wavelet packets technique can be employed using wavelet decomposition. At each level of wavelet decompositions, the digital image is split into four subbands, representing approximation (low frequency feature) and three details (high frequency features) in horizontal, vertical and diagonal directions. The proposed soft thresholding wavelet decomposition at multiple levels is a simple and efficient method for reduction of noises. For multiple level decompositions in terms of both the discrete wavelet transform and wavelet packets techniques, the approximation component will always be decomposed at each level. If the detail components are further decomposed as well similar to that of the approximation, it is the wavelet packet approach, otherwise it is the discrete wavelet transform. On a basis of the proposed thresholding technique at different levels for wavelet denoising, objective metrics can be introduced also to evaluate and compare the denoising effects of the discrete wavelet transform and wavelet packets quantitatively rather than qualitative observation, such as the metrics of the discrete entropy, energy and mutual information.

The research field of wireless sensor networks is a challenging area of engineering and science. In contrast to most traditional networks, sensor nets of wireless sensor networks encompass numerous small battery powered autonomous devices which are referred to as the sensor nodes. These devices have data processing and short range radio communication capabilities. Each node relies on the wireless channels for transmitting data and receiving data from other nodes. The deployment of sensor nodes is subject to various types of harsh environments. Wireless sensor networks differ in network densities, from sparse to dense deployments, which requires the adaptive protocols. On the other hand, small scale and low energy wireless devices are needed for communication. The combination of computing, sensing and communication technologies makes it possible. With the dense network deployment, broadcasting by simple flooding will lead to information redundancy. It gives rise to the application of adaptive topology control to the wireless sensor network. The maximum distance of the individual node could communicate with each other is characterized by the communication units on the sensor node. The location performance and energy usage, based on communication distance, can be evaluated by the Kalman filter which exploits the different selection approaches. In this article, topology control of the wireless sensor networks via integration of Kalman filters and adaptive estimation has been proposed for the design of wireless sensor networks.

Data aggregation is an important area of wireless sensor networks. The importance of efficient data aggregation in wireless sensor networks is due to the fact of the networks which are volatile and ad hoc in nature. Problems associated include efficient energy consumption and routing techniques. Zone Efficient Localized Data Aggregation (Zelda) Algorithm has been proposed. This algorithm employs the data funneling technique and zonal nodes in order to send information over a wireless sensor network thereby reducing amounts of energy and traffic required. The energy consumption is estimated at the setup phase of the algorithm, which can be compared with data funneling. The results show that the proposed scheme is a better candidate for applications using the pseudo stationary wireless sensor networks. In addition, the Kalman filter and feedback control scheme is proposed for potential optimization of wireless sensor network systems. Wireless sensor networks topology control is proposed using Kalman filters and feedback control schemes as well.

This study is to investigate image segmentation approaches using software programming from diverse perspectives. The outcomes will be evaluated with intuitive visualization qualitatively and objective metrics quantitatively. Watershed segmentation is applied for pattern recognition. The watershed transform locates watershed ridges and catchment basins in an image, assuming that bright pixels are high and dark pixels are low on the surface. Segmentation can be improved using markers for foreground objects and background locations. Another approach to separate groups of objects is clustering. K-means segmentation assigns each object a space location via content-based clustering. Through K-means clustering using distance metric criteria, aerial images can be processed and analyzed. Matlab programming is a useful software tool for image processing while visualization effects of two approaches can be observed directly from visual appeals. To evaluate image segmentation objectively, quantitative metrics should be taken into account. Thus, the notions of the discrete entropy, gray level energy, relative entropy and mutual information are used to measure effects of image segmentation approaches.

Image enhancement and image segmentation are both among most practical approaches in image processing. There are a wide variety of applications on medical diagnosis, homeland security, weather forecasting and environment protection. Two typical technologies out of these two categories are contrast enhancement and watershed segmentation. Under improper illumination and unexpected disturbances, adaptive image enhancement can be conducted, which adapts to the intensity distribution within that image. For the trimulus true color system, each of the three color components takes an independent role. Watershed segmentation is an effective approach to automatically separate different objects, where watershed lines separate catchment basins. Erosion and dilation operations are implemented in the segmentation process. Foreground and background markers can be applied to avoid over-segmentation. To represent effects of image enhancement and image segmentation, quantitative measures should be taken into account together with the qualitative observations. Based on gray level quantity measures and the expanded true color component quantity measures, component energy, discrete entropy and relative entropy are calculated to indicate the impact of adaptive image contrast enhancement technique and watershed segmentation techniques. Thus useful results on image processing are obtained.

Biometric pattern is the unique signature of people. Biometric verification is a very complicated procedure involving technologies of pattern recognition, signal processing and image processing. In most cases it is necessary to employ artificial intelligence based approaches. This work is focused on intelligent control applications on biometric verification. The actual sensing information is digitized into image matrix files and then data matrices are analyzed using advanced algorithms. Thus, the real patterns of fingerprints will be captured. The information is indicated by plaintexts containing inherent signatures. After transforming plaintext data into ciphertext data, each of three RGB intensity components of the trimulus color system is computed and analyzed individually. Then principal component analysis (PCA) and independent component analysis (ICA) are applied for decision making. Its assumption is evident and very effective that biometric fingerprints can be effectively recognized by utilizing advanced artificial intelligence technologies.

A novel catalyst converter system has been developed for NOx emission aftertreatment of lean burn gasoline engines. The goal is to investigate its impact on emission characteristic and Break Specific Fuel Consumption (BSFC) across a wide range of engine speed and load operating regions, subject to several arrangement schemes for this catalyst converter. It has been indicated from experimental results that the upstream placement of TWC (Three Way Catalyst) ahead of the NOx Adsorber Catalyst is the best solution, which gives rise to the highest converting efficiency to reduce the NOx emission level of the lean burn gasoline engine. The effects of engine speed on exhaust emissions and BSFC are also reflected by operating time of lean mode and rich mode as well as the time ratio between the two using adsorber-reduction catalyst converters. Engine load is in fact the major factor in affecting exhaust characteristics and BSFC of lean burn engines.Copyright © 2007 by ASME

IGCC plants are potentially the most competitive and economical concepts to produce electricity and thermal energy which will significantly reduce emission levels through advanced control methodologies. Considering sustainability issues on natural resource and environment, an IGCC plant is by far the best one with regard to the power generation. However, the technical difficulty prevents its widespread commercial usage. A typical IGCC plant consists of the air separation unit, gasification system, gas cleanup processes, gas and steam turbine combined cycle. Modeling and control of multiple ICCC components and technology integration play a leading role. In this article, some practical control approaches for individual components of the ICCC plants are explored, such as air blown control and oxygen blown control under conditions of startup, loading, steady state, unloading and shutdown operations throughout the compressor trip, gasifier trip, gas turbine trip and steam turbine trip. In addition, coordinate control and system integration in terms of IGCC performance and environmental objectives will also be investigated.

The gasification process of IGCC plants enables the production of electricity as main product and production of heat and some chemicals (e.g., sulfur, slag) as by-products. In order to testify its feasibility for large scale production, IGCC control and optimization methodologies are needed to enhance its characteristic performance and environmental sustainability via IGCC system integration. Thermoeconomic optimization has been proposed to coordinate multiple components of IGCC plants (compressor, air separation unit, gasifier, gas cleanup system, gas turbine cycle and steam turbine cycle) using advanced control technologies. IGCC technology allows the environment-friendly use of coal. It converts coal to synthesis gas and sends it to the gas turbine system. When gas turbine is fired on a gas fuel emitted from gasification of solid carbonaceous material, this cycle becomes an integrate cycle. In the mean while, control oriented thermoeconomic optimization takes an important role in this potentially widespread commercial technology.

One-step dual-shape memory polymers (SMPs) recover their original (permanent) shape upon small variation of environmental conditions such as temperature, electric field, light, magnetic field, and solvent/chemicals. For advanced applications such as aerospace and medical devices, complicated, multiple-step, spatially controllable, and two-way shape memory effects (SMEs) are required. In the past decade, researchers have devoted great effort to improve the versatility of the SME of SMPs to meet the needs of advanced applications. This paper is intended to review the up-to-date research endeavors on advanced SMEs. The problems facing the various SMPs are discussed. The challenges and opportunities for future research are discussed. (Some figures may appear in colour only in the online journal)

Wireless sensor networks (WSNs) use battery-powered sensor nodes for sensing, thus the energy efficiency is critical to extend the lifespan. The performance depends on the trade-off among energy consumption, latency and reliability. Data aggregation is a fundamental approach to eliminate redundancy and minimize transmission cost so as to save energy. Dynamic clustering based routing is proposed to achieve good performance via adaptive algorithms. The generalized Ant Colony Optimization (ACO) is applied to increase the reliable lifespan of sensor nodes with energy constraints. Each sensor node is modeled as an artificial ant and dynamic routing is modeled as ant foraging. The ant pheromone is released when an energy efficient channel from the source to sink is secured. Route discovery, data aggregation and information loss are modeled as the processes of pheromone diffusion, accumulation and evaporation. Each sensor node estimates the residual energy and dynamically calculates probabilities to select an optimal channel to extend the lifespan of WSNs.