Lexington, KY, United States

Adaptive Intelligent Systems

Lexington, KY, United States
Time filter
Source Type

Liu Y.K.,University of Kentucky | Shao Z.,Adaptive Intelligent Systems | Zhang Y.M.,University of Kentucky
Welding Journal | Year: 2014

Robotic welding is often preferred for its outperformance over human welders who are subject to physical limitations to maintain the needed consistency. Unfortunately, industrial welding robots are basically articulated arms with a preprogrammed set of movements, lacking the intelligence skilled human welders possess. This paper aims to present a virtualized welding system that enables learning from human welder intelligence for transferring into a welding robot. In particular, a 6-DOF UR-5 industrial robot arm equipped with sensors observed the welding process and performed actual welding. A human welder operated a virtualized welding torch to adjust the welding speed based on the visual feedback from the sensors, and the motion of the virtualized torch was recorded and tracked by the robot arm. Nine such teleoperated welding experiments were conducted on pipe using gas tungsten arc welding (GTAW) under different welding currents to correlate the welding speed to the welding current. Robotic welding experiments, with the robot travel speed determined per the given welding current from the resultant correlation, verified that for top part of the pipe between 11 and 1 o'clock, adjusting the welding speed per the current used is adequate to generate acceptable welds. The obtained correlation between the welding speed and welding current could be used in human-machine cooperative control. It may also provide a constraint for automated welding process control. A foundation is thus established to utilize human intelligence and transfer it to welding robots.

Li X.R.,Adaptive Intelligent Systems | Shao Z.,Adaptive Intelligent Systems | Zhang Y.M.,University of Kentucky | Kvidahl L.,Huntington Ingalls Industries
Welding Journal | Year: 2013

Gas tungsten arc welding (GTAW) is the most widely used arc welding process for joining of metals in critical applications such as welding of pipes that are subject to high temperature and pressure. For such applications, assurance of weld joint penetration is the most critical requirement. Unfortunately, industry lacks a simple yet effective approach to monitor and control weld joint penetration. Existing automated orbital welding systems use welding parameters preprogrammed per welding position and heavily rely on accurate joint preparation and specific groove shape, both of which are costly and time consuming, to produce the required weld joint penetration. Highly skilled pipe welders often have to be employed to manually weld pipes to ensure the weld joint penetration. In this paper, a novel method is proposed to determine the weld joint penetration from arc voltage measurements. This method is based on an observation of the authors on the dynamic development of the weld pool surface during the penetrating process in GTAW; the weld pool surface tends to first expand toward the electrode and then be pushed away from the electrode after complete penetration is established. The development of the arc voltage that reflects the change in the arc length and weld pool surface can thus be analyzed to determine the weld joint penetration. A control algorithm has been accordingly designed and implemented to monitor and control weld joint penetration in pipe welding using a square butt joint at the 5G fixed position. Smooth and consistent weld beads were made with guaranteed/feedback-controlled complete joint penetration for both automated orbital welding and manual welding by an electrical engineer as a novice welder.

Li X.R.,Adaptive Intelligent Systems | Shao Z.,Adaptive Intelligent Systems | Zhang Y.M.,University of Kentucky
Welding Journal | Year: 2012

Plasma arc welding (PAW) keyhole mode uses highly penetrating plasma jets to produce narrow inner diameter (ID) beads but causes bead irregularity, undesirable wetting angles, and spatter. For the melt-in mode, plasma jets are less penetrating and greater heat inputs are needed to achieve the needed penetration. As a result, wider weld beads are produced similarly as gas tungsten arc welding (GTAW). For orbital pipe welding, reduced heat inputs and weld pools are needed to reduce the ID convexity around 12 o'clock and achieve the needed convexity around 6 o'clock. To this end, a novel PAW mode, namely the double-stage PAW, is proposed that uses a keyhole stage to build a partially penetrated keyhole and then a melt-in stage to finally reach complete joint penetration. While the keyhole stage helps reduce the heat inputs and weld pools, the melt-in stage finishes the complete penetration at reduced impacts from the plasma jets producing the desired weld bead geometry and regularity. A closed-loop control system has been developed to adaptively determine the duration of the melt-in stage using arc signals to assure the complete joint penetration. Extensive experiments have been conducted on stainless steel pipes around the full circumferance producing smooth weld beads of moderate width with no spatter. In comparison with keyhole PAW, bead geometry and regularity are significantly improved with slightly increased net heat inputs. In comparison with melt-in PAW and GTAW, the net heat input is reduced approximately 40% resulting in 40% reduction in ID weld bead width.

Li X.R.,Adaptive Intelligent Systems | Zhang Y.M.,Adaptive Intelligent Systems | Kvidahl L.,Huntington Ingalls Industries
Welding Journal | Year: 2013

Submerged arc welding (SAW) is known for its high productivity. However, there is a lack of capability to monitor and control weld penetration. Because penetration is believed to be primarily determined by base metal current, a gas metal arc welding (GMAW) gun is added into the SAW process to bypass part of the total current. The base metal current that controls weld penetration is directly reduced, and the ability to adjust the base metal current to control weld penetration without reducing deposition rate is introduced into SAW. To conveniently monitor weld penetration and acquire the needed feedback for weld penetration control, welding parameters and conditions affecting weld penetration were analyzed and specific variables subject to variation and fluctuation were identified. Experiments were conducted to see what parameters affect the weld penetration and what their significances are. It was found that the base metal current is the dominant parameter that determines weld penetration with a sufficient accuracy when other major parameters are in their stated ranges. A control system has been established to monitor and control weld penetration using a proportional integral derivative (PID) control algorithm. This algorithm is based on penetration feedback provided by the penetration model that correlates weld penetration depth to base metal current. Experiments on DH36 square butt joints verified the effectiveness of the proposed method.

Minev M.,Adaptive Intelligent Systems
Proceedings - IEEE 13th International Conference on Data Mining Workshops, ICDMW 2013 | Year: 2013

This study concerns financial news articles, which reflect the monetary policy during the US sub prime mortgage crisis. In particular we consider official announcements conducted by the Federal Reserve and its leading representatives. We aim to quantify such information using dependency parsing techniques and statistical measures. In addition, we examine the correlations between the monetary policy and the stock markets by modeling composite index volatilities as functions of key publications. A prototype for the classification of news is targeted, which should reveal the economical impact of events. An eminent aspect of our study is the identification, extraction, and representation of topic-related features and the corresponding instances. © 2013 IEEE.

Yau J.Y.K.,Adaptive Intelligent Systems | Joy M.,Adaptive Intelligent Systems
International Journal of Mobile Learning and Organisation | Year: 2011

The purpose of this paper is to present the data analysis obtained from our interview study, which showed that participants had different individual mobile learning (hereafter, abbreviated as m-learning) preferences. The understanding of these preferences for different m-learning requirements can be used as a foundation for building successful personalised m-learning applications catered to learners' individual m-learning needs. Participants' dynamic m-learning preferences (including location of study, noise/distraction level in a location and time of day) are described. We propose a context-aware personalised m-learning application based on these m-learning preferences. Six scenarios are given to illustrate the m-learning preferences of different learners. The system architecture consists of a learner profile, personalisation mechanism and learning object repository. An initial m-learning preference questionnaire is used to obtain learners' dynamic m-learning preferences. Current context values are retrieved from context-aware technologies. Appropriate learning objects are selected to learners based on their preferences and context values. Copyright © 2011 Inderscience Enterprises Ltd.

Minev M.,Adaptive Intelligent Systems | Schommer C.,Adaptive Intelligent Systems
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) | Year: 2014

Financial news carry information about economical figures and indicators. However, these texts are mostly unstructured and consequently hard to be processed in an automatic way. In this paper, we present a representation formalism that supports a linguistic composition for machine learning tasks. We show an innovative approach to structuring financial texts by extracting principal indicators. Considering announcements in the monetary policy domain, we distinguish between attributes and their values and argue that attributes are to be represented as an aggregated set of economic terms, keeping their values as corresponding conditional expressions. We close with a critical discussion and future perspectives. © 2014 Springer-Verlag.

Shi L.,Adaptive Intelligent Systems | Cristea A.I.,Adaptive Intelligent Systems
IEEE Multimedia | Year: 2016

This article explores open social learner modeling (OSLM) - a social extension of open learner modeling (OLM). A specific implementation of this approach is presented by which learners' self-direction and self-determination in a social e-learning context could be potentially promoted. Unlike previous work, the proposed approach, multifaceted OSLM, lets the system seamlessly and adaptively embed visualization of both a learner's own model and other learning peers' models into different parts of the learning content, for multiple axes of context, at any time during the learning process. It also demonstrates the advantages of visualizing both learners' performance and their contribution to a learning community. An experimental study shows that, contrary to previous research, the richness and complexity of this new approach positively affected the learning experience in terms of perceived effectiveness, efficiency, and satisfaction. This article is part of special issue on social media for learning. © 1994-2012 IEEE.

Adaptive Intelligent Systems | Date: 2010-03-17

A welding system and method includes a main torch including a main electrode configured to form a first arc with a base metal; a first bypass torch including a first bypass electrode configured to form a second arc with the main electrode; and a second bypass torch including a second bypass electrode configured to form a third arc with the main electrode.

Agency: Department of Defense | Branch: Navy | Program: SBIR | Phase: Phase I | Award Amount: 70.00K | Year: 2010

This SBIR project aims at the development of an innovative technology that can be attached to existing submerged arc welding (SAW) systems to monitor and feedback control the depth of weld penetration. Existing SAW systems will be operated using existing welding procedures without modifications. In many applications skilled welders may adjust the welding parameters based on their observation of dynamic weld pool surface to successfully control the weld penetration. Although for SAW the weld pool surface is not observable, it may/should still provide necessary information to estimate the weld penetration. A method is thus proposed to generate a dynamic change in submerged arc weld pool and then estimate this change from arc electrical signals that can be easily measured. The Phase I project aims at proving the feasibility of the proposed method through developing a laboratory monitoring/control system, using it to feedback control the weld penetration at different levels, and verifying the monitoring/control accuracy through comparison with actual welds. The Phase II project will develop a commercial prototype system, demonstrate benefits through increasing the thickness for no back-gouging two-sided butt welding, on-site demonstrate the system and benefits at selected shipyards, and prepare for the military qualification and transition.

Loading Adaptive Intelligent Systems collaborators
Loading Adaptive Intelligent Systems collaborators