Improving Multiobjective Multidisciplinary Optimization With a Data Mining-Based Hybrid Method

2015 ◽  
Author(s):  
Hongyi Xu ◽  
Ching-Hung Chuang ◽  
Ren-Jye Yang
Keyword(s):  
Data Mining ◽  
Bias Correction ◽  
Multidisciplinary Design Optimization ◽  
Multidisciplinary Optimization ◽  
Side Impact ◽  
Computational Time ◽  
Hybrid Strategy ◽  
Clustering And Classification ◽  
Product Design Process ◽  
Optimization Search

Multiobjective, multidisciplinary design optimization (MDO) of complex system is challenging due to the long computational time needed for evaluating new designs’ performances. Heuristic optimization algorithms are widely employed to overcome the local optimums, but the inherent randomness of such algorithms leads to another disadvantage: the need for a large number of design evaluations. To accelerate the product design process, a data mining-based hybrid strategy is developed to improve the search efficiency. Based on the historical information of the optimization search, clustering and classification techniques are employed to detect low quality designs and repetitive designs, and which are then replaced by promising designs. In addition, the metamodel with bias correction is integrated into the proposed strategy to further increase the probability of finding promising design regions within a limited number of design evaluations. Two case studies, one mathematical benchmark problem and one vehicle side impact design problem, are conducted to demonstrate the effectiveness of the proposed method in improving the searching efficiency.

Methodology Development for Simulating Full Frontal and Offset Frontal Impacts Using Full Vehicle MADYMO Models

1999 ◽  
Author(s):  
Bala Deshpande ◽  
Gunasekar TJ ◽  
Russell Morris ◽  
Sudhanshu Parida ◽  
Mostafa Rashidy ◽  
...  
Keyword(s):  
Rigid Body ◽  
Joint Stiffness ◽  
Side Impact ◽  
Computational Time ◽  
Test Results ◽  
Vehicle Interior ◽  
Significant Saving ◽  
Methodology Development ◽  
Full Vehicle ◽  
Frontal Impacts

Abstract MADYMO articulated full vehicle models of the 1992 Ford Taurus, 1995 Chevrolet Lumina and the 1994 Dodge Intrepid for frontal and side impact modes have been developed and validated against test data. MADYMO (Mathematical Dynamic Model) is typically used to model occupants in the environment of the vehicle interior and thus finds application mainly in assessing occupant injuries. In this study however, MADYMO has been employed not only to model the occupants but also to represent the major load bearing structures in the vehicles. Input for the MADYMO models consisting of rigid body joint stiffness was obtained from corresponding full vehicle Finite Element (FE) models. Model validation was done by comparing the vehicle and dummy numbers with the New Car Assessment Program (NCAP) test results. Models correlated very well with both test and FE data. This modeling approach demonstrates the utility of rigid body based full car models for crashworthiness analysis. Such models result in significant saving in computational time and resources. In this paper, we describe the simulation of two different crash modes: full frontal and offset frontal impacts using the full vehicle MADYMO models. These simulations were validated with the corresponding test results in full frontal mode and IIHS offset mode. The models are useful for simulating a variety of impact situations, for example, with different occupant sizes, occupant positions, impact velocities, and in car to car impacts for performing compatibility studies.

scholarly journals Modeling of Hypervelocity Impact Experiments Using Gamma-SPH Technique

2017 ◽  
Author(s):  
Jérôme Limido ◽  
Mohamed Trabia ◽  
Shawoon Roy ◽  
Brendan O’Toole ◽  
Richard Jennings ◽  
...  
Keyword(s):  
Computation Time ◽  
3D Models ◽  
Hypervelocity Impact ◽  
Side Impact ◽  
Particle Methods ◽  
Computational Time ◽  
Back Surface ◽  
Tensile Instability ◽  
Particle Hydrodynamics ◽  
And Performance

A series of experiments were performed to study plastic deformation of metallic plates under hypervelocity impact at the University of Nevada, Las Vegas (UNLV) Center for Materials and Structures using a two-stage light gas gun. In these experiments, cylindrical Lexan projectiles were fired at A36 steel target plates with velocities range of 4.5–6.0 km/s. Experiments were designed to produce a front side impact crater and a permanent bulging deformation on the back surface of the target without inducing complete perforation of the plates. Free surface velocities from the back surface of target plate were measured using the newly developed Multiplexed Photonic Doppler Velocimetry (MPDV) system. To simulate the experiments, a Lagrangian-based smooth particle hydrodynamics (SPH) is typically used to avoid the problems associated with mesh instability. Despite their intrinsic capability for simulation of violent impacts, particle methods have a few drawbacks that may considerably affect their accuracy and performance including, lack of interpolation completeness, tensile instability, and existence of spurious pressure. Moreover, computational time is also a strong limitation that often necessitates the use of reduced 2D axisymmetric models. To address these shortcomings, IMPETUS Afea Solver® implemented a newly developed SPH formulation that can solve the problems regarding spurious pressures and tensile instability. The algorithm takes full advantage of GPU Technology for parallelization of the computation and opens the door for running large 3D models (20,000,000 particles). The combination of accurate algorithms and drastically reduced computation time now makes it possible to run a high fidelity hypervelocity impact model.

scholarly journals Data-mining Techniques for Image-based Plant Phenotypic Traits Identification and Classification

2019 ◽  
Vol 9 (1) ◽  
Author(s):  
Md. Matiur Rahaman ◽  
Md. Asif Ahsan ◽  
Ming Chen
Keyword(s):  
Data Mining ◽  
Agricultural Development ◽  
Computational Time ◽  
Support Vector ◽  
Phenotypic Traits ◽  
Linear Discriminant ◽  
Plant Phenomics ◽  
Classification Prediction ◽  
Linear Svm ◽  
Integrated Or

AbstractStatistical data-mining (DM) and machine learning (ML) are promising tools to assist in the analysis of complex dataset. In recent decades, in the precision of agricultural development, plant phenomics study is crucial for high-throughput phenotyping of local crop cultivars. Therefore, integrated or a new analytical approach is needed to deal with these phenomics data. We proposed a statistical framework for the analysis of phenomics data by integrating DM and ML methods. The most popular supervised ML methods; Linear Discriminant Analysis (LDA), Random Forest (RF), Support Vector Machine with linear (SVM-l) and radial basis (SVM-r) kernel are used for classification/prediction plant status (stress/non-stress) to validate our proposed approach. Several simulated and real plant phenotype datasets were analyzed. The results described the significant contribution of the features (selected by our proposed approach) throughout the analysis. In this study, we showed that the proposed approach removed phenotype data analysis complexity, reduced computational time of ML algorithms, and increased prediction accuracy.

scholarly journals Hybrid Method for Detecting Duplicate Image by Using Image Retrieval Technique in Data Mining

2021 ◽  
Vol 09 (04) ◽  
Author(s):  
Dr. S. Thavamani ◽  
Keyword(s):  
Data Mining ◽  
Image Retrieval ◽  
Hybrid Method ◽  
Video Retrieval ◽  
New Method ◽  
Computational Time ◽  
Storage Space ◽  
Copy Detection ◽  
Retrieval Technique ◽  
Improved Accuracy

Duplicated images cause several problems in online sites, so these demand special attention. To address the disadvantages of frames copy detection, the Hybrid Method of Detecting Duplicate Image by Using Image Retrieval Technique in Data Mining was proposed. We use the new method of eliminating duplicates in this example. To address the disadvantages of frames copy detection, the Hybrid Method of Detecting Duplicate Image by Using Image Retrieval Technique in Data Mining was proposed. The new method of eliminating duplicates in this example has proposed. Using this method, you can get rid of frames that aren't relevant to the video. This makes for more precise and faster video retrieval with fewer duplicates. As a back end, this technique is implemented in C# and SQL. The findings are put to the test and compared to the current SIFT process. The results showed that the output improved accuracy while reducing storage space, computational time, and memory use.

scholarly journals A Data Mining Technique for Tourist Destination Brand Image Building

2020 ◽  
Vol 8 (6) ◽  
pp. 4617-4622
Keyword(s):  
Data Mining ◽  
Clustering Algorithm ◽  
Tourism Industry ◽  
Destination Image ◽  
Data Mining Technique ◽  
Tourist Destination ◽  
Text Data ◽  
Clustering And Classification ◽  
Google Search

The destination image branding is the domain of tourism industry where the facts and information is collected and evaluated for finding the credibility of a target tourist destination. Manual collection and processing of collected information accurately is a complicated and time consuming task therefore a data mining model is suggested ,in this presented work that collect and evaluate the destination image accurately and based on evaluation can make the recommendations about visits of tourist. In order to perform this task data mining techniques are applied on text data source. In first the data is extracted from the Google search engine and it is preprocessed for make it impure. In further the data is labeled based on the positive and negative words available in the collected facts. Finally the clustering and classification of text is performed. For clustering of data FCM (fuzzy c means) clustering algorithm and for classification the Bayesian classifier is used. Based on final classification of text data the decision is made for the destination visits.

Taguchi-EM-AI Design Optimization Environment for SynRM Drives in Traction Applications

2021 ◽  
Vol 35 (11) ◽  
pp. 1372-1373
Author(s):  
A.A. Arkadan ◽  
N. Al Aawar
Keyword(s):  
Design Optimization ◽  
Search Algorithm ◽  
Motor Drives ◽  
Computational Time ◽  
Synchronous Reluctance Motor ◽  
Reluctance Motor ◽  
Motor Drive System ◽  
Taguchi Orthogonal Arrays ◽  
Optimization Search

Multi-objective design optimization environments are used for electric vehicles and other traction applications to arrive at efficient motor drives. Typically, the environment includes characterization modules that involve the use of Electromagnetic Finite Element and State-Space models that require large number of iterations and computational time. This work proposes the utilization of a Taguchi orthogonal arrays method in conjunction with a Particle Swarm Optimization search algorithm to reduce computational time needed in the design optimization of electric motors for traction applications. The effectiveness of the Taguchi method in conjunction with the optimization environment is demonstrated in a case study involving a prototype of a Synchronous Reluctance Motor drive system.

Institutional Research Using Data Mining

2011 ◽  
pp. 66-102 ◽  
Author(s):  
Constanta-Nicoleta Bodea ◽  
Vasile Bodea ◽  
Radu Mogos
Keyword(s):  
Data Mining ◽  
Academic Performance ◽  
Virtual Learning Environment ◽  
Student Activities ◽  
Participatory Behavior ◽  
Student Profiles ◽  
Clustering And Classification ◽  
E Learning ◽  
Using Data ◽  
And Behavior

The aim of this chapter is to explore the application of data mining for analyzing academic performance in connection with the participatory behavior of the students enrolled in an online two-year Master degree program in project management. The main data sources were the operational database with the students’ records and the log files and statistics provided by the e-learning platform. One hundred eighty-one enrolled students, and more than 150 distinct characteristics/ variables per student were used. Due to the large number of variables, an exploratory data analysis through data mining was chosen, and a model-based discovery approach was designed and executed in Weka environment. The association rules, clustering, and classification were applied in order to identify the factors explaining the students’ performance and the relationship between academic performance and behavior in the virtual learning environment. Data mining has revealed interesting patterns in data. These patterns indicate that academic performance is related to the intensity of the student activities in virtual environment. If the student understands how to work and she/he is motivated to communicate with others, then he might have a good academic performance. Based on clustering analysis, different student profiles were discovered, explaining the academic performance. The results are very encouraging and suggest several future developments.

Overview of Predictive Modeling Approaches in Health Care Data Mining

2016 ◽  
pp. 73-95 ◽  
Author(s):  
Sunita Soni
Keyword(s):  
Data Mining ◽  
Predictive Modeling ◽  
Clinical Information ◽  
Ease Of Use ◽  
Classification Model ◽  
Computational Time ◽  
Data Sets ◽  
Advantages And Disadvantages ◽  
Mining Methods ◽  
Predictive Data Mining

Medical data mining has great potential for exploring the hidden pattern in the data sets of the medical domain. A predictive modeling approach of Data Mining has been systematically applied for the prognosis, diagnosis, and planning for treatment of chronic disease. For example, a classification system can assist the physician to predict if the patient is likely to have a certain disease, or by considering the output of the classification model, the physician can make a better decision on the treatment to be applied to the patient. Once the model is evaluated and verified, it may be embedded within clinical information systems. The objective of this chapter is to extensively study the various predictive data mining methods to evaluate their usage in terms of accuracy, computational time, comprehensibility of the results, ease of use of the algorithm, and advantages and disadvantages to relatively naive medical users. The research has shown that there is not a single best prediction tool, but instead, the best performing algorithm will depend on the features of the dataset to be analyzed.

A Model for Extracting Most Desired Web Pages

2020 ◽  
pp. 119-145
Author(s):  
Jayanti Mehra ◽  
Ramjeevan Singh Thakur
Keyword(s):  
Data Mining ◽  
Path Length ◽  
Web Mining ◽  
Statistical Information ◽  
Web Pages ◽  
Systems Practice ◽  
Clustering And Classification ◽  
Using Data ◽  
Access Logs ◽  
Manipulation Process

Weblog analysis takes raw data from access logs and performs study on this data for extracting statistical information. This info incorporates a variety of data for the website activity such as average no. of hits, total no. of user visits, failed and successful cached hits, average time of view, average path length over a website; analytical information such as page was not found errors and server errors; server information, which includes exit and entry pages, single access pages, and top visited pages; requester information like which type of search engines is used, keywords and top referring sites, and so on. In general, the website administrator uses this kind of knowledge to make the system act better, helping in the manipulation process of site, then also forgiving marketing decisions support. Most of the advanced web mining systems practice this kind of information to take out more difficult or complex interpretations using data mining procedures like association rules, clustering, and classification.

Sign in / Sign up

Export Citation Format

Share Document