PRISM :: Browsing by Author "Xue, Deyi"

Browsing by Author "Xue, Deyi"

Now showing 1 - 20 of 58

Open Access
A design database representation and evolution model
(2006) Yang, Haoguang; Xue, Deyi
Open Access
A Neural Network Driven Sensor Array for Locating a Permanent Magnet
(2017) Russel, Marco; Goldsmith, Peter; Nowicki, Ed; Ramirez-Serrano, Alejandro; Xue, Deyi
This thesis explores the use of a magnetic sensor array and artificial neural networks for determining the position of a permanent magnet on a 2D plane. One of the major motivations for this research is to track the stylus of a magnetic haptic interface that uses electromagnets to provide haptic feedback to a permanent magnet on the stylus. Hence our method must be able to sense the magnet position in the presence of disruptive materials, such as the iron cores of the actuating electromagnets. As these electromagnets could saturate the sensors in the direction normal to the array, we also investigate the network’s effectiveness when only using the magnetic field components in directions parallel to the plane of the array. The networks used are multilayer perceptron networks consisting of a hidden layer and an output layer. The effectiveness of four different training methods are compared to determine the most effective method for training such a network and the best network parameters for that method. The accuracy is then compared to that of a traditional tracking method, both with and without a disruptive steel bar and/or magnetic field present. The neural networks are found to have much better accuracy than the traditional method in the presence of the interfering material and solve for position more than 6 times faster. Their solution speed is also much less variable than that of the traditional method, making them more suitable for real-time tracking applications. Of the training methods investigated, networks trained using Bayesian Regularization were found to be most accurate, with several networks achieving mean position errors of less than 1 mm. The Bayesian Regularization method was also found to be less susceptible to premature termination of training.
Open Access
ADAPTABLE DESIGN OF MECHANICAL PRODUCTS WITH ROBUST PERFORMANCE
(2014-10-23) ZHANG, JIAN; Gu, Peihua; Xue, Deyi
Adaptable product is the one that can be adapted, such as reconfigured and upgraded, during the operation stage to satisfy different requirements. Development of adaptable products can bring with both economic and environmental benefits. Adaptable design is a design paradigm that can be used for the design of adaptable products. Since quality of adaptable product can be influenced by parameter variations caused by uncertainties, product robustness has to be considered in adaptable design. This research aims at developing an adaptable design approach such that the product is adaptable to various changes in requirements, meanwhile the functional performance measures are the least sensitive to parameter variations. In this research, robust adaptable design at three levels in the embodiment design stage is considered: parameter, configuration, and architecture levels. At the parameter level, an adaptable product can be achieved through the changes of parameter values. In this research, product/operating parameters are classified into four categories: un-adaptable design parameters, adaptable design parameters, unchangeable non-design parameters, and changeable non-design parameters. Mathematical models are established to describe the relationships among functional performances, product/operating parameters, and variations of both performances and parameters. An optimization model is also developed to identify the optimal values of un-adaptable design parameters. At the configuration level, an adaptable product can be achieved through the changes of product configurations. In this research, different configuration candidates in design and different product configurations in operation stage to satisfy design requirements are modeled by a hybrid AND-OR tree. Parameters associated with configurations are also modeled in this AND-OR tree. A two-level optimization method is developed for identifying the optimal design configuration and its parameter values. At the architecture level, an adaptable product can be achieved through selection of the open architecture as the product architecture. In this research, characteristics of open architecture products are investigated. Methods for the modeling of platform, add-on modules, and open interfaces are developed. A multi-level optimization method is also developed to identify the optimal design of the open architecture products considering both product performance and variation of the performance.
Open Access
Adaptive production scheduling and control in one-of-a-kind production
(2010) Li, Wei; Tu, Yiliu (Paul); Xue, Deyi
Open Access
Adaptive production scheduling and control in one-of-a-kind production
(2005) Li, Wei; Tu, Yiliu (Paul); Xue, Deyi
Open Access
An information systems framework for computer aided one-of-a-kind production
(2009) Dean, Paul R.; Tu, Yiliu (Paul); Xue, Deyi
Open Access
An Integrated Approach for Precision Machining of Freeform Surfaces
(2014-01-17) Lasemi, Ali; Xue, Deyi; Gu, Peihua
Freeform surfaces, also called sculptured surfaces, have been increasingly used in different industries such as automotive, aerospace, and die and mold manufacturing. Increasing precision requirements for products with freeform surfaces have led to significant challenges for manufacturing companies. Multi-axis CNC machining is the primary method to manufacture these freeform surfaces. Prediction of machining errors with different geometric shapes under varying machining conditions is the key to improve quality in freeform surface manufacturing. An integrated approach for precision machining of freeform surfaces has been developed in this research. In the first step of this approach, geometric errors of machine tool and process-related errors are identified and compensated. Geometric errors of machine tool are identified through a newly developed offline identification method using the kinematic modeling of machine tool and magnetic double ball bar measurement of the machine's volumetric errors. Process-related errors are modeled as functions of machining process and freeform surface parameters, and measured using on-machine inspection in a multi-layer machining method. These identified error sources are then compensated through tool path re-planning using the mirror approach. In the second step of the integrated approach, the machined surface is inspected to identify and remove any residual errors on the surface. Since comparison between manufactured surface and design surface is conducted by aligning the two surfaces in different coordinate systems, an optimization method is introduced to identify the best alignment with minimum area of the error regions to reduce the re-machining efforts. When error regions on the manufactured surface are identified, CNC machining tool paths need to be generated to remove these errors. Since various tool path generation methods may be used for the identified error islands on the surface, mapping between the error islands and the tool path generation methods needs to be studied. In this research, the mapping is conducted by analyzing the geometric characteristics of each error island and the capabilities of the tool path generation algorithms. The results obtained through different evaluation methods, such as simulations, measurement of machine tool's volumetric accuracy, and machining experiments, reveal significant improvement in the accuracy and efficiency of freeform surface machining process.
Open Access
An intelligent optimal product distribution scheduling approach
(2001) Wang, Huicheng; Xue, Deyi
Open Access
Analysis of Radiation Therapy in Cancer Treatment using Machine Learning
(2021-09-24) Yarschenko, Adam H; Sun, Qiao; Smith, Wendy L.; Kirkby, Charles; Xue, Deyi; Smith, Wendy; Sun, Qiao
Advancements in machine learning and data science have allowed researchers and clinicians to generate key insights from the vast amount of data generated in healthcare. This is currently a topic of research with great interest. With the advancement in algorithm design, and computing power, machine learning has proven to be a capable tool to augment or partially automate decision making. In this thesis, patient reported outcome surveys (PROs) for head and neck radiotherapy, and the relationship between the radiation dose distribution and breast size for whole-breast radiotherapy were investigated using statistical and machine learning methods. Two PRO measures; the M.D. Anderson symptom inventory for head and neck cancer, and the M.D. Anderson dysphagia inventory, were examined for a cohort of patients post radiotherapy for head and neck cancer. A strategy for administering a single PRO instrument is proposed which would reduce the questionnaire burden on patients, and allow physicians to identify patients who require specialized treatment for dealing with radiotherapy side-effects, such as referral to a dietician or speech language pathologist. Dosiomic features were extracted from the 3D radiation dose cloud in whole-breast radiotherapy plans. Feature reduction was achieved through hierarchical clustering and random forests were trained to stratify treated volume based on the distribution in the dose. Permutation feature importance was used to rank features’ classification utility in this task. The top 3 features were used to achieve superior performance when compared to the entire feature set. Dosiomics gives new insight into 3D dose distribution, and these features can be used in future studies to relate to treatment outcomes associated with whole breast radiotherapy for large volumes.
Open Access
Application of Fuzzy Numbers for Environmental Assessments: A Preliminary Study
(2016) Cheng, Xin; Li, Simon; Du, Ke; Xue, Deyi; Zareipour, Hamidreza
This thesis preliminarily explores the use of fuzzy numbers in two applications: life-cycle assessment (LCA) and building energy analysis. In LCA, this thesis compares the results in the context of concept selection between the fuzzy number and Monte Carlo approaches and finds that the numerical outcomes of both approaches are comparable. In building energy analysis, the traditional degree-day method is adapted by allowing the data inputs as fuzzy numbers. Instead of providing the “average point” results, the fuzzy number approach can yield an “interval” of energy estimations. The proposed fuzzy number approach has been applied to a building located in Calgary, and the results are compared to historical data and building energy simulation (i.e., eQUEST). As the ranking of fuzzy numbers is fundamental to support decision making in both applications, this thesis also investigates the axiomatic properties of one well-known ranking method, namely, the centroid index method. This thesis has explored using a numerical approach to identify counter-examples in the proof process.
Open Access
B-Spline Based 3D Model Reconstruction and Finite Element Analysis of Human Knee Joint
(2017) Zhu, Di; Li, Leping; Xue, Deyi; Johansen, Craig; Lu, Qingye
Knee joint is the largest diarthrodial joint in the human body, and the normal joint mechanics is essential for our daily life. Finite element analysis provides an efficient tool for studying knee joint mechanical behaviour under different conditions. Due to the complex shapes of the knee joint, it is important to obtain accurate models with realistic geometries prior to FE simulation. A semi-automatic 3D point cloud fitting procedure in MATLAB based on B-Splines that accounts for contact geometries and CAD compatibility was developed in this thesis. The reconstructed model was then used for nonlinear stress-relaxation simulations under ramp compressions, where pore pressure, contact pressure and reaction forces were investigated. The reconstruction procedure has successfully reduced overclosures at contact surfaces, promoted faster convergence and enhanced simulation performances. This study helps further build the bridge between 2D medical images and FE simulations.
Open Access
Chatter Suppression in Boring Operations through Altering Tool Dynamics
(2016) Alammari, Youssef; Park, Simon; Freiheit, Theodor; Dalton, Colin; Tu, Paul; Xue, Deyi
In machining, a phenomenon known as chatter, which is a self-excited excessive vibration, is considered one of the most limiting factors for productivity. Particularly in boring process, the long cantilevered structure of boring bar makes it the most flexible part of machine tool system. As a result, there is a high possibility for chatter to occur compared with other machine tool components. In this research, two methods were investigated to attenuate chatter and increase stability in boring operations by manipulating the boring bar structure dynamics. First, the effect of altering the boring bar’s natural frequency on chatter is investigated through varying the boring bar’s overall mass. Second, a tuned liquid column damper (TLCD) is developed to improve the dynamic stiffness for boring bars application for improving the overall stability. Several models and experiments are carried out for evaluating the performance of the proposed methods.
Open Access
Concept and Methods for the Development of Blockchain based Cloud Manufacturing
(2023-04-30) Wang, Binni; Tu, Paul; Li, Simon; Xue, Deyi; Thekinen, Joseph; Patterson, Raymond; Zeng, Yong
Cloud manufacturing is seen as a promising improvement for networked manufacturing, with potential for achieving higher service quality at a lower cost, although the progress of its application has not met expectations. The main obstacle hindering its adoption is the reluctance of potential users to share data with cloud manufacturing due to concerns about data safety issues caused by central management. To address this drawback, the integration of a decentralized blockchain system has been proposed as a promising solution. However, the traceable data and provable system features of blockchain technology cannot guarantee data security, particularly for small or moving objects that lack the ability for data verification. Therefore, blockchain-based cloud manufacturing systems must have the ability to ensure data reliability. Moreover, the scalability problem must be addressed for the adoption of a blockchain-based system. This PhD thesis proposes a feasible architecture for a blockchain-based cloud manufacturing system and evaluates the reliability of perceived data through data similarity measurement. Firstly, the architecture of the system is outlined, with the use of rollup technology to address scalability issues. Second, the proper fee setting for the system is analyzed. For data similarity measurement, comparable data is selected using a modified clustering algorithm, and the data is described using polygon-based descriptions, including existing and proposed polygon descriptions. Finally, data similarity measurement is transformed into a similarity comparison of polygon descriptions, with methods such as multi-objective programming-based similarity comparison and overlap area-based similarity comparison applied for this purpose. The feasibility of the proposed methods is verified through a case study.
Open Access
Control of a Highly Maneuverable Autonomous Underwater Vehicle
(2020-01-29) Garcia Rodriguez, Jaime Arnoldo; Ramírez-Serrano, Alejandro; Dalton, Colin; Xue, Deyi
A Highly Maneuverable Autonomous Underwater Vehicle (HM-AUV) is a novel type of underwater vehicle designed to navigate in constrained and difficult to reach environments. The vehicle used in this thesis has a unique 3-thruster configuration that allows it to produce motion among the different axis of motion by adjusting its thruster tilt angles. The focus of this thesis is to develop a control scheme that can take advantage of this unique thruster configuration in order for the HM-AUV to maneuver among tight spaces. The developed controller achieves control over most of the HM-AUV's axis of motion directly while managing to control its lateral, or sway, movement via a novel direct sway control technique. With the implemented control scheme, the HM-AUV is able to achieve complex motions, albeit at a slow rate due to the underpowered thrusters that are installed in the vehicle.
Open Access
Customer and supplier involved one-of-a-kind product design and manufacture
(2010) Dong, Ying; Tu, Yiliu (Paul); Xue, Deyi
Open Access
Design for product adaptability
(2007) Li, Yi; Xue, Deyi; Gu, Peihua
Open Access
Design for project change management
(2004) Cheing, Siu Ying; Xue, Deyi; Gu, Peihua
In the pipeline design and construction industry, the construction projects are carried out based upon the design descriptions including design configuration and design attributes. The design configuration and attributes of the projects occasionally have to be modified after the construction has been scheduled or during the construction process, thus resulting in construction project changes. These project changes usually bring significant, negative impacts on cost, quality, and construction schedules of the project. In this research, a new systematic approach is proposed to identifying the optimal design to minimize the potential construction project changes. This research focuses on the aspects of pipeline system design. In the configuration design aspect, relations between design configurations and construction tasks are modeled by matrices. The design configuration that is most independent to the construction tasks is identified based upon Axiomatic Design approach. In addition, estimation of project change cost due to the design configuration changes is also discussed In attribute design aspect, potential changes of design attribute values are modeled by probability distribution functions. Attribute values of the design whose construction tasks are least sensitive to the changes of these attribute values are identified based upon Taguchi Method. In addition, estimation of project change cost due to the design attribute value changes is also discussed. Two case studies are conducted to show that the proposed approach can effectively manage the design and construction changes.
Open Access
Design Optimization of Truss Structures Using Artificial Neural Networks
(2023-09-22) Nourian, Navid; El-Badry, Mamdouh; Dann, Markus; Billah, Muntasir; Xue, Deyi
One of the primary objectives of structural design optimization is to achieve a design possessing the lowest possible weight, while it can safely withstand the effects of external loads. In the case of a truss of a specific topology, the role of an optimization algorithm is to determine the configuration and number of the truss elements as well as their cross-sectional areas. In this study, a novel model is proposed, by which the main optimization problem is decomposed into two more manageable problems: a size optimization within a shape optimization problem. A Deep Neural Network (DNN) is trained to approximate the optimal cross-sectional areas of the elements of a truss with a given shape and support positions. Furthermore, truss structures are characterized by pin joints connected by truss members, a concept that can be analogized to vertices and edges in a mathematical graph. Leveraging this analogy, a Graph Neural Network (GNN) is utilized to exploit the advantages of representing trusses as graphs. Specifically, a graph neural network-based surrogate model integrated with Particle Swarm Optimization (PSO) algorithm is developed to approximate nodal displacements of trusses during the design optimization process. Several truss examples are used to investigate the validity and effectiveness of the proposed optimization techniques in comparison with conventional FEM-based models.
Open Access
Design Project Scheduling with Probabilistic Iterations: Optimization & Investigation of Robustness
(2018-09-05) Ebufegha, Akposeiyifa Joseph; Li, Simon; Tu, Paul; Xue, Deyi; Sadeghpour, Farnaz
This research focuses on reliable prediction of product development (PD) project duration within reasonable solution times. It is based on the model of PD projects presented by Smith and Eppinger which suggests that PD projects are akin to Markov reward chains. The research was divided into two phases; the development of a hybrid algorithm to minimize the problem’s solve time without compromising solution quality and the identification of features of a robust PD schedule. The first phase of the research shows that the hybrid algorithm yields similar quality solutions whilst reducing solve times by 50% - 97.4%. The second phase highlights four features that affect robust PD schedule as well as patterns observed in the schedule. In conclusion, this research demonstrates the utility of the developed hybrid algorithm and also illustrates the importance of examining expected project duration and the standard deviation in the results when developing robust PD schedules.
Open Access
Development of a feature-based intelligent design system
(1998) Yadav, Swatantra; Xue, Deyi