Under deregulated electric market, electricity price is no longer set by the monopoly utility company rather it responds to the market and operating conditions. Offering the right amount of electricity at the right time with the right bidding price has become the key for utility companies pursuing maximum profits under deregulated electricity market. Therefore, electricity market clearing price (MCP) forecasting became essential for decision making, scheduling and bidding strategy planning purposes. However, forecasting electricity MCP is a very difficult problem due to uncertainties associated with input variables.<p> Neural network based approach promises to be an effective forecasting tool in an environment with high degree of non-linearity and uncertainty. Although there are several techniques available for short-term MCP forecasting, very little has been done to do mid-term MCP forecasting. Two new artificial neural networks have been proposed and reported in this thesis that can be utilized to forecast mid-term daily peak and mid-term hourly electricity MCP. The proposed neural networks can simulate the electricity MCP with electricity hourly demand, electricity daily peak demand, natural gas price and precipitation as input variables. Two situations have been considered; electricity MCP forecasting under real deregulated electric market and electricity MCP forecasting under deregulated electric market with perfect competition. The PJM interconnect system has been utilized for numerical results. Techniques have been developed to overcome difficulties in training the neural network and improve the training results.

Becoming a face expert takes years of learning and development. Many research programs are devoted to studying face perception, particularly given its prerequisite role in social interaction, yet its fundamental neural operations are poorly understood. One reason is that there are many possible explanations for a change in facial appearance, such as lighting, expression, or identity. Despite general agreement that the brain extracts multiple layers of feature detectors arranged into hierarchies to interpret causes of sensory information, very little work has been done to develop computational models of these processes, especially for complex stimuli like faces. The studies presented in this thesis used nonlinear generative models developed within machine learning to solve several face perception problems. Applying a deep hierarchical neural network, we showed that it is possible to learn representations capable of perceiving facial actions, expressions, and identities, better than similar non-hierarchical architectures. We then demonstrated that a generative architecture can be used to interpret high-level neural activity by synthesizing images in a top-down pass. Using this approach we showed that deep layers of a network can be activated to generate faces corresponding to particular categories. To facilitate training models to learn rich and varied facial features, we introduced a new expression database with the largest number of labeled faces collected to date. We found that a model trained on these images learned to recognize expressions comparably to human observers. Next we considered models trained on pairs of images, making it possible to learn how faces change appearance to take on different expressions. Modeling higher-order associations between images allowed us to efficiently match images of the same type according to a learned pairwise similarity measure. These models performed well on several tasks, including matching expressions and identities, and demonstrated performance superior to competing models. In sum, these studies showed that neural networks that extract highly nonlinear features from images using architectures inspired by the brain can solve difficult face perception tasks with minimal guidance by human experts.

Soft computing techniques are generally well-suited for vehicular control systems that are usually modeled by highly nonlinear differential equations and working in unstructured environment. To demonstrate their applicability, two intelligent controllers based upon fuzzy logic theories and neural network paradigms are designed for performing a wall-following task and an autonomous parking task. Based on performance and flexibility considerations, the two controllers are implemented onto a reconfigurable hardware platform, namely a Field Programmable Gate Array (FPGA). As the number of comparative studies of these two embedded controllers designed for the same application is limited in the literature, one of the main goals of this research work has been to evaluate and compare the two controllers in terms of hardware resource requirements, operational speeds and trajectory tracking errors in following different pre-defined trajectories. The main advantages and disadvantages of each of the controllers are presented and discussed in details. Challenging issues for implementation of the controllers on the FPGA platform are also highlighted. As the two controllers exhibit benefits and drawbacks under different circumstances, this research suggests as well a hybrid controller scheme as an attempt to integrate the benefits of both control units. To evaluate its performance, the hybrid controller is tested on the same pre-defined trajectories and the corresponding results are compared to that of the fuzzy logic and the neural network based controllers. For further demonstration of the capabilities of the wall-following controllers in other applications, the fuzzy logic and the neural network controllers are used in a parallel parking system. We see this work to be a stepping stone for further research work aiming at real world implementation of the controllers on Application Specified Integrated Circuit (ASIC) type of environment.

Although nonlinear model predictive control (NMPC) might be the best choice for a nonlinear plant, it is still not widely used. This is mainly due to the computational burden associated with solving online a set of nonlinear differential equations and a nonlinear dynamic optimization problem in real time. This thesis is concerned with strategies aimed at reducing the computational burden involved in different stages of the NMPC such as optimization problem, state estimation, and nonlinear model identification. A major part of the computational burden comes from function and derivative evaluations required in different parts of the NMPC algorithm. In this work, the problem is tackled using a recently introduced efficient tool, the automatic differentiation (AD). Using the AD tool, a function is evaluated together with all its partial derivative from the code defining the function with machine accuracy. A new NMPC algorithm based on nonlinear least square optimization is proposed. In a first–order method, the sensitivity equations are integrated using a linear formula while the AD tool is applied to get their values accurately. For higher order approximations, more terms of the Taylor expansion are used in the integration for which the AD is effectively used. As a result, the gradient of the cost function against control moves is accurately obtained so that the online nonlinear optimization can be efficiently solved. In many real control cases, the states are not measured and have to be estimated for each instance when a solution of the model equations is needed. A nonlinear extended version of the Kalman filter (EKF) is added to the NMPC algorithm for this purpose. The AD tool is used to calculate the required derivatives in the local linearization step of the filter automatically and accurately. Offset is another problem faced in NMPC. A new nonlinear integration is devised for this case to eliminate the offset from the output response. In this method, an integrated disturbance model is added to the process model input or output to correct the plant/model mismatch. The time response of the controller is also improved as a by–product. The proposed NMPC algorithm has been applied to an evaporation process and a two continuous stirred tank reactor (two–CSTR) process with satisfactory results to cope with large setpoint changes, unmeasured severe disturbances, and process/model mismatches. When the process equations are not known (black–box) or when these are too complicated to be used in the controller, modelling is needed to create an internal model for the controller. In this thesis, a continuous time recurrent neural network (CTRNN) in a state–space form is developed to be used in NMPC context. An efficient training algorithm for the proposed network is developed using AD tool. By automatically generating Taylor coefficients, the algorithm not only solves the differentiation equations of the network but also produces the sensitivity for the training problem. The same approach is also used to solve online the optimization problem of the NMPC. The proposed CTRNN and the predictive controller were tested on an evaporator and two–CSTR case studies. A comparison with other approaches shows that the new algorithm can considerably reduce network training time and improve solution accuracy. For a third case study, the ALSTOM gasifier, a NMPC via linearization algorithm is implemented to control the system. In this work a nonlinear state–space class Wiener model is used to identify the black–box model of the gasifier. A linear model of the plant at zero–load is adopted as a base model for prediction. Then, a feedforward neural network is created as the static gain for a particular output channel, fuel gas pressure, to compensate its strong nonlinear behavior observed in open–loop simulations. By linearizing the neural network at each sampling time, the static nonlinear gain provides certain adaptation to the linear base model. The AD tool is used here to linearize the neural network efficiently. Noticeable performance improvement is observed when compared with pure linear MPC. The controller was able to pass all tests specified in the benchmark problem at all load conditions.

Becoming a face expert takes years of learning and development. Many research programs are devoted to studying face perception, particularly given its prerequisite role in social interaction, yet its fundamental neural operations are poorly understood. One reason is that there are many possible explanations for a change in facial appearance, such as lighting, expression, or identity. Despite general agreement that the brain extracts multiple layers of feature detectors arranged into hierarchies to interpret causes of sensory information, very little work has been done to develop computational models of these processes, especially for complex stimuli like faces. The studies presented in this thesis used nonlinear generative models developed within machine learning to solve several face perception problems. Applying a deep hierarchical neural network, we showed that it is possible to learn representations capable of perceiving facial actions, expressions, and identities, better than similar non-hierarchical architectures. We then demonstrated that a generative architecture can be used to interpret high-level neural activity by synthesizing images in a top-down pass. Using this approach we showed that deep layers of a network can be activated to generate faces corresponding to particular categories. To facilitate training models to learn rich and varied facial features, we introduced a new expression database with the largest number of labeled faces collected to date. We found that a model trained on these images learned to recognize expressions comparably to human observers. Next we considered models trained on pairs of images, making it possible to learn how faces change appearance to take on different expressions. Modeling higher-order associations between images allowed us to efficiently match images of the same type according to a learned pairwise similarity measure. These models performed well on several tasks, including matching expressions and identities, and demonstrated performance superior to competing models. In sum, these studies showed that neural networks that extract highly nonlinear features from images using architectures inspired by the brain can solve difficult face perception tasks with minimal guidance by human experts.

Although higher correlation between gauge and radar at hourly or daily accumulations are reported, it is rarely observed at higher time resolution (e.g. 10 -minute). This study investigates six major rainfall events in year 2000 in the greater Winnipeg area with durations varying from four to nine hours. The correlation between gauge and radar measurements of precipitation is found to be only 0.3 at 10-minute resolution and 0.55 at hourly resolution using Marshall-Palmer’s Z-R relationship (Z=200R1.6). The rainfalls are classified into convective and stratiform regions using Steiner et al. (1995)’s algorithm and two different Z-R relationships are tested to minimize the error associated with the variability of drop-size-distribution, however no improvement is observed. The performance of the artificial neural network is explored as a reflectivity-rainfall mapping function. Three different types of neural networks are explored: the back propagation network, the radial basis function network, and the generalized regression neural network. It is observed that the neural network’s performance is better than the Z-R relationship to estimate the rainfall events which was used for training and validation (correlation 0.67). When this network is tested on a new rainfall its performance is found quite similar to that obtained from the Z-R relationship (correlation 0.33). Based on this observation neural network may be recommended as a post-processing tool but may not be very useful for operational purposes - at least as used in this study. Variability in weather and precipitation scenarios affects the radar measurements which apparently makes it impossible for the neural network or the Z-R relationship to show consistent performance at every rainfall event. To account for variability in weather and rainfall scenarios conventional correction schemes for attenuation and hail contamination are applied and a trajectory model is developed to account for rainfall advection due to wind drift. The trajectory model uses velocity obtained from the single-doppler observation. A space-time interpolation technique is applied to generate reflectivity maps at one-minute resolution based on the direction obtained from the correlation based tracking algorithm. The trajectory model uses the generated reflectivity maps having one-minute resolution which help to account for the travel time by the rainfall mass to reach to the ground. It was found that the attenuation correction algorithm adversely increases the reflectivity. This study assumes that the higher reflectivity caused by hail contaminated regions is one reason for the overestimation in the attenuation correction process. It was observed that the hail capping method applied prior to the attenuation correction algorithm helps to improve the situation. A statistical expression to account for radome attenuation is also developed. It is observed that the correlation between the gauge and the radar measurement is 0.81 after applying the various algorithms. Although Marshall-Palmer’s relationship is recommended for stratiform precipitation only, this study found it suitable for both convective and stratiform precipitation when attenuation is properly taken into account. The precipitation processing model developed in this study generates more accurate rainfall estimates at the surface from radar observations and may be a better choice for rainfall-runoff modellers. / February 2006

Fabrics are unique materials which consist of various properties affecting their performance and end-uses. A computerized fabric property evaluation and analysis method plays a crucial role not only in textile industry but also in scientific research. An accurate analysis and measurement of fabric property provides a powerful tool for gauging product quality, assuring regulatory compliance and assessing the performance of textile materials. This thesis investigated the solutions for applying computerized methods to evaluate and intelligently interpret the texture and material properties of fabric in an inexpensive and efficient way. Firstly, a method which allows automatic recognition of basic weave pattern and precisely measuring the yarn count is proposed. The yarn crossed-areas are segmented by a spatial domain integral projection approach. Combining fuzzy c-means (FCM) and principal component analysis (PCA) on grey level co-occurrence matrix (GLCM) feature vectors extracted from the segments enables to classify detected segments into two clusters. Based on the analysis on texture orientation features, the yarn crossed-area states are automatically determined. An autocorrelation method is used to find weave repeats and correct detection errors. The method was validated by using computer simulated woven samples and real woven fabric images. The test samples have various yarn counts, appearance, and weave types. All weave patterns of tested fabric samples are successfully recognized and computed yarn counts are consistent to the manual counts. Secondly, we present a methodology for using the high resolution 3D surface data of fabric samples to measure surface roughness in a nondestructive and accurate way. A parameter FDFFT, which is the fractal dimension estimation from 2DFFT of 3D surface scan, is proposed as the indicator of surface roughness. The robustness of FDFFT, which consists of the rotation-invariance and scale-invariance, is validated on a number of computer simulated fractal Brownian images. Secondly, in order to evaluate the usefulness of FDFFT, a novel method of calculating standard roughness parameters from 3D surface scan is introduced. According to the test results, FDFFT has been demonstrated as a fast and reliable parameter for measuring the fabric roughness from 3D surface data. We attempt a neural network model using back propagation algorithm and FDFFT for predicting the standard roughness parameters. The proposed neural network model shows good performance experimentally. Finally, an intelligent approach for the interpretation of fabric objective measurements is proposed using supported vector machine (SVM) techniques. The human expert assessments of fabric samples are used during the training phase in order to adjust the general system into an applicable model. Since the target output of the system is clear, the uncertainty which lies in current subjective fabric evaluation does not affect the performance of proposed model. The support vector machine is one of the best solutions for handling high dimensional data classification. The complexity problem of the fabric property has been optimally dealt with. The generalization ability shown in SVM allows the user to separately implement and design the components. Sufficient cross-validations are performed and demonstrate the performance test of the system.

No description available.

No description available.

No description available.