Virtual manufacturing has gained considerable importance in the last decade. To obtain reliable predictions in a virtual environment, the factors that influence the outcome of a manufacturing operation need to be carefully modeled and integrated in a simulation platform. The dynamic behavior of the Computer Numerical Control (CNC) system, which has a profound influence on the final part geometry and tolerance integrity, is among these factors. Classical CNC drive identification techniques are usually time consuming and need to be performed by an engineer qualified in dynamics and control theory. These techniques require the servo loop or the trajectory interpolator to be disconnected in order to inject the necessary identification signals, causing downtime to the machine. Hence, these techniques are usually not practical for constructing virtual models of existing CNC machine tools in a manufacturing environment. This thesis presents an alternative strategy for constructing virtual drive models with minimal intervention and downtime to the machinery. The proposed technique, named “rapid identification”, consists of executing a short G-code experiment and collecting input/output data using the motion capture feature available on most CNC controllers. The data is then processed to reverse engineer the equivalent tracking and disturbance transfer functions and friction characteristics of the machine. It is shown that virtual drive models constructed this way can be used to predict the real machine’s contouring performance for large class of drive systems, controlled with different control techniques. In the proposed scheme, the excitation is delivered by smoothly interpolated motion commands. Hence, convergence of parameters to their true values is not guaranteed. When the real system contains pole-zero cancellations, namely due to feedforward control action, this also results in a loss of identifiability. In order to guarantee the stability of the identified drive models, the pole locations are constrained with frequency and damping ratio limits. Hence, the rapid identification task is cast as a constrained minimization problem. Two solution strategies have been developed. In the first approach, Lagrange Multipliers (LM) technique is applied, which yields successful estimation results. However, implementation of LM is computationally intensive and requires the use of a dedicated symbolic solver. This limits the portability for industrial implementation. In the second approach, a Genetic Algorithm (GA) search technique is developed, which is a more practical but slightly approximate alternative. The GA allows parameter bounds to be incorporated in a natural manner and converges to 2-3% vicinity of the LM solution in one-tenth of the computation time. The GA solution can be easily ported to different computation platforms. Both LM and GA identification techniques were validated in simulations and experiments conducted on virtual and real machine tool drives. It is shown that although the parameters estimated using the rapid identification scheme do not always match their true values, the key tracking and disturbance rejection characteristics of the drives are successfully captured in the frequency range of the CNC motion commands. Therefore, the drive models constructed with rapid identification can be used to predict the contouring accuracy of real machine tools in a virtual process planning environment. This thesis presents an alternative strategy for constructing virtual drive models with minimal intervention and downtime to the machinery. The proposed technique, named “rapid identification”, consists of executing a short G-code experiment and collecting input/output data using the motion capture feature available on most CNC controllers. The data is then processed to reverse engineer the equivalent tracking and disturbance transfer functions and friction characteristics of the machine. It is shown that virtual drive models constructed this way can be used to predict the real machine’s contouring performance for large class of drive systems, controlled with different control techniques. In the proposed scheme, the excitation is delivered by smoothly interpolated motion commands. Hence, convergence of parameters to their true values is not guaranteed. When the real system contains pole-zero cancellations, namely due to feedforward control action, this also results in a loss of identifiability. In order to guarantee the stability of the identified drive models, the pole locations are constrained with frequency and damping ratio limits. Hence, the rapid identification task is cast as a constrained minimization problem. Two solution strategies have been developed. In the first approach, Lagrange Multipliers (LM) technique is applied, which yields successful estimation results. However, implementation of LM is computationally intensive and requires the use of a dedicated symbolic solver. This limits the portability for industrial implementation. In the second approach, a Genetic Algorithm (GA) search technique is developed, which is a more practical but slightly approximate alternative. The GA allows parameter bounds to be incorporated in a natural manner and converges to 2-3% vicinity of the LM solution in one-tenth of the computation time. The GA solution can be easily ported to different computation platforms. Both LM and GA identification techniques were validated in simulations and experiments conducted on virtual and real machine tool drives. It is shown that although the parameters estimated using the rapid identification scheme do not always match their true values, the key tracking and disturbance rejection characteristics of the drives are successfully captured in the frequency range of the CNC motion commands. Therefore, the drive models constructed with rapid identification can be used to predict the contouring accuracy of real machine tools in a virtual process planning environment.

Virtual manufacturing has gained considerable importance in the last decade. To obtain reliable predictions in a virtual environment, the factors that influence the outcome of a manufacturing operation need to be carefully modeled and integrated in a simulation platform. The dynamic behavior of the Computer Numerical Control (CNC) system, which has a profound influence on the final part geometry and tolerance integrity, is among these factors. Classical CNC drive identification techniques are usually time consuming and need to be performed by an engineer qualified in dynamics and control theory. These techniques require the servo loop or the trajectory interpolator to be disconnected in order to inject the necessary identification signals, causing downtime to the machine. Hence, these techniques are usually not practical for constructing virtual models of existing CNC machine tools in a manufacturing environment. This thesis presents an alternative strategy for constructing virtual drive models with minimal intervention and downtime to the machinery. The proposed technique, named “rapid identification”, consists of executing a short G-code experiment and collecting input/output data using the motion capture feature available on most CNC controllers. The data is then processed to reverse engineer the equivalent tracking and disturbance transfer functions and friction characteristics of the machine. It is shown that virtual drive models constructed this way can be used to predict the real machine’s contouring performance for large class of drive systems, controlled with different control techniques. In the proposed scheme, the excitation is delivered by smoothly interpolated motion commands. Hence, convergence of parameters to their true values is not guaranteed. When the real system contains pole-zero cancellations, namely due to feedforward control action, this also results in a loss of identifiability. In order to guarantee the stability of the identified drive models, the pole locations are constrained with frequency and damping ratio limits. Hence, the rapid identification task is cast as a constrained minimization problem. Two solution strategies have been developed. In the first approach, Lagrange Multipliers (LM) technique is applied, which yields successful estimation results. However, implementation of LM is computationally intensive and requires the use of a dedicated symbolic solver. This limits the portability for industrial implementation. In the second approach, a Genetic Algorithm (GA) search technique is developed, which is a more practical but slightly approximate alternative. The GA allows parameter bounds to be incorporated in a natural manner and converges to 2-3% vicinity of the LM solution in one-tenth of the computation time. The GA solution can be easily ported to different computation platforms. Both LM and GA identification techniques were validated in simulations and experiments conducted on virtual and real machine tool drives. It is shown that although the parameters estimated using the rapid identification scheme do not always match their true values, the key tracking and disturbance rejection characteristics of the drives are successfully captured in the frequency range of the CNC motion commands. Therefore, the drive models constructed with rapid identification can be used to predict the contouring accuracy of real machine tools in a virtual process planning environment. This thesis presents an alternative strategy for constructing virtual drive models with minimal intervention and downtime to the machinery. The proposed technique, named “rapid identification”, consists of executing a short G-code experiment and collecting input/output data using the motion capture feature available on most CNC controllers. The data is then processed to reverse engineer the equivalent tracking and disturbance transfer functions and friction characteristics of the machine. It is shown that virtual drive models constructed this way can be used to predict the real machine’s contouring performance for large class of drive systems, controlled with different control techniques. In the proposed scheme, the excitation is delivered by smoothly interpolated motion commands. Hence, convergence of parameters to their true values is not guaranteed. When the real system contains pole-zero cancellations, namely due to feedforward control action, this also results in a loss of identifiability. In order to guarantee the stability of the identified drive models, the pole locations are constrained with frequency and damping ratio limits. Hence, the rapid identification task is cast as a constrained minimization problem. Two solution strategies have been developed. In the first approach, Lagrange Multipliers (LM) technique is applied, which yields successful estimation results. However, implementation of LM is computationally intensive and requires the use of a dedicated symbolic solver. This limits the portability for industrial implementation. In the second approach, a Genetic Algorithm (GA) search technique is developed, which is a more practical but slightly approximate alternative. The GA allows parameter bounds to be incorporated in a natural manner and converges to 2-3% vicinity of the LM solution in one-tenth of the computation time. The GA solution can be easily ported to different computation platforms. Both LM and GA identification techniques were validated in simulations and experiments conducted on virtual and real machine tool drives. It is shown that although the parameters estimated using the rapid identification scheme do not always match their true values, the key tracking and disturbance rejection characteristics of the drives are successfully captured in the frequency range of the CNC motion commands. Therefore, the drive models constructed with rapid identification can be used to predict the contouring accuracy of real machine tools in a virtual process planning environment.

"After decades of development, Intelligent Tutoring Systems (ITSs) have become a common learning environment for learners of various domains and academic levels. ITSs are computer systems designed to provide instruction and immediate feedback, which is customized to individual students, but without requiring the intervention of human instructors. All ITSs share the same goal: to provide tutorial services that support learning. Since learning is a very complex process, it is not surprising that a range of technologies and methodologies from different fields is employed. Student modeling is a pivotal technique used in ITSs. The model observes student behaviors in the tutor and creates a quantitative representation of student properties of interest necessary to customize instruction, to respond effectively, to engage students¡¯ interest and to promote learning. In this dissertation work, I focus on the following aspects of student modeling. Part I: Student Knowledge: Parameter Interpretation. Student modeling is widely used to obtain scientific insights about how people learn. Student models typically produce semantically meaningful parameter estimates, such as how quickly students learn a skill on average. Therefore, parameter estimates being interpretable and plausible is fundamental. My work includes automatically generating data-suggested Dirichlet priors for the Bayesian Knowledge Tracing model, in order to obtain more plausible parameter estimates. I also proposed, implemented, and evaluated an approach to generate multiple Dirichlet priors to improve parameter plausibility, accommodating the assumption that there are subsets of skills which students learn similarly. Part II: Student Performance: Student Performance Prediction. Accurately predicting student performance is one of the most desired features common evaluations for student modeling. for an ITS. The task, however, is very challenging, particularly in predicting a student¡¯s response on an individual problem in the tutor. I analyzed the components of two common student models to determine which aspects provide predictive power in classifying student performance. I found that modeling the student¡¯s overall knowledge led to improved predictive accuracy. I also presented an approach, which, rather than assuming students are drawn from a single distribution, modeled multiple distributions of student performances to improve the model¡¯s accuracy. Part III: Wheel-spinning: Student Future Failure in Mastery Learning. One drawback of the mastery learning framework is its possibility to leave a student stuck attempting to learn a skill he is unable to master. We refer to this phenomenon of students being given practice with no improvement as wheel-spinning. I analyzed student wheel-spinning across different tutoring systems and estimated the scope of the problem. To investigate the negative consequences of see what wheel-spinning could have done to students, I investigated the relationships between wheel-spinning and two other constructs of interest about students: efficiency of learning and ¡°gaming the system¡±. In addition, I designed a generic model of wheel-spinning, which uses features easily obtained by most ITSs. The model can be well generalized to unknown students with high accuracy classifying mastery and wheel-spinning problems. When used as a detector, the model can detect wheel-spinning in its early stage with satisfying satisfactory precision and recall. "

This study investigates the main factors of production, their interaction and influence on cycle time efficiency for shovel-truck systems on surface mines. The main factors are truck payload, cycle time and operator proficiency. It is now routine that shoveltruck cycles are analysed using simulation methods. The Elbrond, FPC, Talpac, Arena and Machine Repair simulation models are discussed to explain how their model characteristics contribute to the differences in their reported cycle efficiency as indicated by productivity results. The Machine Repair Model based on Markov chains is adapted for shovel-truck systems and examined for calculating shovel-truck cycle times. The various probability distributions that can be use to model particular cycle time variables and some methods in selecting the “best” fit are examined. Truck cycle time variable sensitivity is examined by using the Excel® add-on program @Risk (Palisade Corp.) in determining their respective weighting or contribution within the total cycle time variability. The analysis of cycle efficiency leads ultimately to sizing of a shovel-truck system. When determining a fleet size for a particular surface operation the planning engineers will tend to use one and to a lesser extent perhaps two separate simulation models. This study calculates the productivity (tonnes per hour) for a “virtual mine” with a variable number of trucks, variable cycle distances and variable truck loading times. The study also includes a separate analysis of cycle time variables and their probability distributions for the Orapa diamond mine in Botswana, to show possible distributions for various cycle variables. The study concludes with a calculation of the truck fleet size using the Elbrond, FPC, Talpac and Arena and Machine Repair models for the Optimum Colliery coal mine and then compares the results and their correlation. The main findings are that the calculation of waiting time is different for the various models, each model yields a unique fleet sizing solution and any solution in effect represents a range of results.

<div><br></div><div><p>In many complex domains, the input data are often not suited for the typical vector representations used in deep learning models. For example, in knowledge representation, relational learning, and some computer vision tasks, the data are often better represented as graphs or sets. In these cases, a key challenge is to learn a representation function which is invariant to permutations of set or isomorphism of graphs. </p><p>In order to handle graph isomorphism, this thesis proposes a subgraph pattern neural network with invariance to graph isomorphisms and varying local neighborhood sizes. Our key insight is to incorporate the unavoidable dependencies in the training observations of induced subgraphs into both the input features and the model architecture itself via high-order dependencies, which are still able to take node/edge labels into account and facilitate inductive reasoning. </p><p>In order to learn permutation-invariant set functions, this thesis shows how the characteristics of an architecture’s computational graph impact its ability to learn in contexts with complex set dependencies, and demonstrate limitations of current methods with respect to one or more of these complexity dimensions. I also propose a new Self-Attention GRU architecture, with a computation graph that is built automatically via self-attention to minimize average interaction path lengths between set elements in the architecture’s computation graph, in order to effectively capture complex dependencies between set elements.</p><p>Besides the typical set problem, a new problem of representing sets-of-sets (SoS) is proposed. In this problem, multi-level dependence and multi-level permutation invariance need to be handled jointly. To address this, I propose a hierarchical sequence attention framework (HATS) for inductive set-of-sets embeddings, and develop the stochastic optimization and inference methods required for efficient learning.</p></div>

Space weather has become an international issue due to the catastrophic impact it can have on modern societies. Solar flares are one of the major solar activities that drive space weather and yet their occurrence is not fully understood. Research is required to yield a better understanding of flare occurrence and enable the development of an accurate flare prediction system, which can warn industries most at risk to take preventative measures to mitigate or avoid the effects of space weather. This thesis introduces novel technologies developed by combining advances in statistical physics, image processing, machine learning, and feature selection algorithms, with advances in solar physics in order to extract valuable knowledge from historical solar data, related to active regions and flares. The aim of this thesis is to achieve the followings: i) The design of a new measurement, inspired by the physical Ising model, to estimate the magnetic complexity in active regions using solar images and an investigation of this measurement in relation to flare occurrence. The proposed name of the measurement is the Ising Magnetic Complexity (IMC). ii) Determination of the flare prediction capability of active region properties generated by the new active region detection system SMART (Solar Monitor Active Region Tracking) to enable the design of a new flare prediction system. iii) Determination of the active region properties that are most related to flare occurrence in order to enhance understanding of the underlying physics behind flare occurrence. The achieved results can be summarised as follows: i) The new active region measurement (IMC) appears to be related to flare occurrence and it has a potential use in predicting flare occurrence and location. ii) Combining machine learning with SMART¿s active region properties has the potential to provide more accurate flare predictions than the current flare prediction systems i.e. ASAP (Automated Solar Activity Prediction). iii) Reduced set of 6 active region properties seems to be the most significant properties related to flare occurrence and they can achieve similar degree of flare prediction accuracy as the full 21 SMART active region properties. The developed technologies and the findings achieved in this thesis will work as a corner stone to enhance the accuracy of flare prediction; develop efficient flare prediction systems; and enhance our understanding of flare occurrence. The algorithms, implementation, results, and future work are explained in this thesis.

Space weather has become an international issue due to the catastrophic impact it can have on modern societies. Solar flares are one of the major solar activities that drive space weather and yet their occurrence is not fully understood. Research is required to yield a better understanding of flare occurrence and enable the development of an accurate flare prediction system, which can warn industries most at risk to take preventative measures to mitigate or avoid the effects of space weather. This thesis introduces novel technologies developed by combining advances in statistical physics, image processing, machine learning, and feature selection algorithms, with advances in solar physics in order to extract valuable knowledge from historical solar data, related to active regions and flares. The aim of this thesis is to achieve the followings: i) The design of a new measurement, inspired by the physical Ising model, to estimate the magnetic complexity in active regions using solar images and an investigation of this measurement in relation to flare occurrence. The proposed name of the measurement is the Ising Magnetic Complexity (IMC). ii) Determination of the flare prediction capability of active region properties generated by the new active region detection system SMART (Solar Monitor Active Region Tracking) to enable the design of a new flare prediction system. iii) Determination of the active region properties that are most related to flare occurrence in order to enhance understanding of the underlying physics behind flare occurrence. The achieved results can be summarised as follows: i) The new active region measurement (IMC) appears to be related to flare occurrence and it has a potential use in predicting flare occurrence and location. ii) Combining machine learning with SMART's active region properties has the potential to provide more accurate flare predictions than the current flare prediction systems i.e. ASAP (Automated Solar Activity Prediction). iii) Reduced set of 6 active region properties seems to be the most significant properties related to flare occurrence and they can achieve similar degree of flare prediction accuracy as the full 21 SMART active region properties. The developed technologies and the findings achieved in this thesis will work as a corner stone to enhance the accuracy of flare prediction; develop efficient flare prediction systems; and enhance our understanding of flare occurrence. The algorithms, implementation, results, and future work are explained in this thesis.