scholarly journals Intelligent Performance Analysis with a Natural Language Interface

2017 ◽  
Vol 25 (3) ◽  
pp. 168-175 ◽  
Author(s):  
Esko K. Juuso
Keyword(s):  
Natural Language ◽  
Condition Monitoring ◽  
Asset Management ◽  
Performance Monitoring ◽  
Expert Knowledge ◽  
Temporal Analysis ◽  
Monitoring Data ◽  
Statistical Process ◽  
Domain Expertise ◽  
Condition Indices

Abstract Performance improvement is taken as the primary goal in the asset management. Advanced data analysis is needed to efficiently integrate condition monitoring data into the operation and maintenance. Intelligent stress and condition indices have been developed for control and condition monitoring by combining generalized norms with efficient nonlinear scaling. These nonlinear scaling methodologies can also be used to handle performance measures used for management since management oriented indicators can be presented in the same scale as intelligent condition and stress indices. Performance indicators are responses of the process, machine or system to the stress contributions analyzed from process and condition monitoring data. Scaled values are directly used in intelligent temporal analysis to calculate fluctuations and trends. All these methodologies can be used in prognostics and fatigue prediction. The meanings of the variables are beneficial in extracting expert knowledge and representing information in natural language. The idea of dividing the problems into the variable specific meanings and the directions of interactions provides various improvements for performance monitoring and decision making. The integrated temporal analysis and uncertainty processing facilitates the efficient use of domain expertise. Measurements can be monitored with generalized statistical process control (GSPC) based on the same scaling functions.

The Application of Smart, Connected Power Plant Assets for Enhanced Condition Monitoring and Improving Equipment Reliability

2016 ◽  
Author(s):  
Michael Reid ◽  
Bernie Cook
Keyword(s):  
Power Plant ◽  
Data Storage ◽  
Condition Monitoring ◽  
Asset Management ◽  
Performance Monitoring ◽  
Significant Proportion ◽  
Electric Utility ◽  
Sensor Technology ◽  
Steam Turbines ◽  
Equipment Reliability

The U.S. electric utility industry continues to undergo dramatic change due to a number of key trends and also prolonged uncertainty. These trends include: • Increasing environmental regulations uncertainty • Natural gas supply uncertainty and price • Economic / decoupling of electricity demand growth from GDP • Aging coal and nuclear generation fleet / coal retirements • Aging workforce • Increasing distributed energy resources • Increasing customer expectations The transformation ultimately demands significant increases in power plant generation operating capabilities (e.g. flexibility, operating envelop, ramp rates, turn-down etc.) and higher levels of equipment reliability, while reducing O&M and capital budgets. Achieving higher levels of equipment reliability and flexibility, with such tightening budget and resource constraints, requires a very disciplined approach to maintenance and an optimized mix of the following maintenance practices: • Reactive (run-to-failure) • Preventive (time-based) • Predictive (condition-based) • Proactive (combination of 1, 2 and 3 + root cause failure analysis) Many U.S. electric utilities with fossil generation have adopted and implemented elements of an equipment reliability process consistent with Institute of Nuclear Power Operations (INPO) AP-913. The Electric Power Research Institute has created a guideline modeled from the learnings of AP-913, that consists of six key sub-processes [1]: 1. Scoping and identification of critical components (identifying system and component criticality) 2. Continuing equipment reliability improvement (establishing and continuously improving system and component maintenance bases) 3. Preventive Maintenance (PM) implementation (implementing the PM program effectively) 4. Performance monitoring (monitoring system and component performance) 5. Corrective action 6. Life cycle management (long-term asset management) A significant proportion of Duke Energy’s coal fleet is of an age where individual components have reached their design intent end-of-life thereby creating an increased need for performance monitoring. Until recent times this was largely performed by maintenance technicians with handheld devices. This approach does not allow regular data collection for trending and optimization of maintenance practices across the fleet. Significant and recent advances in sensor technology, microprocessors, data acquisition, data storage, communication technology, and software have enabled the transformation of critical power plant assets such as steam turbines, combustion turbines, generators, transformers, and large balance-of-plant equipment into smart, connected power plant assets. These enhanced assets, in conjunction with visualization software, provide a comprehensive conditioning monitoring solution that continuously acquires sensory data and performs real time analysis to provide information and insight. This advanced condition monitoring capability has been successfully applied to obtain earlier detection of equipment issues and failures and is key to improving overall equipment reliability. This paper describes an approach by Duke Energy to create and apply smart, connected power plant assets to greatly enhance its fossil generation continuous condition monitoring capabilities. It will discuss the value that is currently being realized and also look at future possibilities to apply big data and analytics to enhance information, insight, and actionable intelligence.

Brisa's model for life cycle performance monitoring and decision- making process

2019 ◽  
Author(s):  
Carlos Biscaia de Oliveira
Keyword(s):  
Decision Making ◽  
Asset Management ◽  
Performance Indicators ◽  
Performance Monitoring ◽  
Decision Support Techniques ◽  
Cycle Performance ◽  
Decision Making Process ◽  
Risk Levels ◽  
Asset Portfolio ◽  
Asset Monitoring

<p>The asset management model must allow for visibility across the asset portfolio, enabling a more coherent and informed decision-making process. This topic addresses how the need to improve analytic capabilities and decision support techniques leads to the guidelines of Brisa’s Information System Dashboard, covering asset’s availability and condition indexes (Asset Monitoring), risk levels and relevant costs key performance indicators.</p>

Using Natural Language Processing to Assess the Psychological Effect of COVID-19 Pandemic on Insomnia via Tweets: A Pre-Post Retrospective Pilot Study (Preprint)

2021 ◽  
Author(s):  
Arash Maghsoudi ◽  
Sara Nowakowski ◽  
Ritwick Agrawal ◽  
Amir Sharafkhaneh ◽  
Sadaf Aram ◽  
...  
Keyword(s):  
Mental Health ◽  
Social Media ◽  
Natural Language Processing ◽  
Natural Language ◽  
Language Processing ◽  
Temporal Analysis ◽  
Health Condition ◽  
Additional Stress ◽  
Mental Health Condition ◽  
Twitter Users

BACKGROUND The COVID-19 pandemic has imposed additional stress on population health that may result in a higher incidence of insomnia. In this study, we hypothesized that using natural language processing (NLP) to explore social media would help to identify the mental health condition of the population experiencing insomnia after the outbreak of COVID-19. OBJECTIVE In this study, we hypothesized that using natural language processing (NLP) to explore social media would help to identify the mental health condition of the population experiencing insomnia after the outbreak of COVID-19. METHODS We designed a pre-post retrospective study using public social media content from Twitter. We categorized tweets based on time into two intervals: prepandemic (01/01/2019 to 01/01/2020) and pandemic (01/01/2020 to 01/01/2021). We used NLP to analyze polarity (positive/negative) and intensity of emotions and also users’ tweets psychological states in terms of sadness, anxiety and anger by counting the words related to these categories in each tweet. Additionally, we performed temporal analysis to examine the effect of time on the users’ insomnia experience. RESULTS We extracted 268,803 tweets containing the word insomnia (prepandemic, 123,293 and pandemic, 145,510). The odds of negative tweets (OR, 1.31; 95% CI, 1.29-1.33), anger (OR, 1.19; 95% CI, 1.16-1.21), and anxiety (OR, 1.24; 95% CI: 1.21-1.26) were higher during the pandemic compared to prepandemic. The likelihood of negative tweets after midnight was higher than for other daily intevals, comprising approximately 60% of all negative insomnia-related tweets in 2020 and 2021 collectively. CONCLUSIONS Twitter users shared more negative tweets about insomnia during the pandemic than during the year before. Also, more anger and anxiety-related content were disseminated during the pandemic on the social media platform. Future studies using an NLP framework could assess tweets about other psychological distress, habit changes, weight gain due to inactivity, and the effect of viral infection on sleep.

Multi-Zone Proportional Hazard Model for a Multi-Stage Degradation Process

2013 ◽  
Author(s):  
Lin Li ◽  
Zeyi Sun ◽  
Xinwei Xu ◽  
Kaifu Zhang
Keyword(s):  
Decision Making ◽  
Condition Monitoring ◽  
Proportional Hazards ◽  
Proportional Hazards Model ◽  
Degradation Process ◽  
Remaining Useful Life ◽  
Monitoring Data ◽  
Status Change ◽  
Multi Stage ◽  
And Performance

Conditional-based maintenance (CBM) decision-making is of high interests in recent years due to its better performance on cost efficiency compared to other traditional policies. One of the most respected methods based on condition-monitoring data for maintenance decision-making is Proportional Hazards Model (PHM). It utilizes condition-monitoring data as covariates and identifies their effects on the lifetime of a component. Conventional modeling process of PHM only treats the degradation process as a whole lifecycle. In this paper, the PHM is advanced to describe a multi-zone degradation system considering the fact that the lifecycle of a machine can be divided into several different degradation stages. The methods to estimate reliability and performance prognostics are developed based on the proposed multi-zone PHM to predict the remaining time that the machine stays at the current stage before transferring into the next stage and the remaining useful life (RUL). The results illustrate that the multi-zone PHM effectively monitors the equipment status change and leads to a more accurate RUL prediction compared with traditional PHM.

scholarly journals Automated processing of railway track deflection signals obtained from velocity and acceleration measurements

2018 ◽  
Vol 232 (8) ◽  
pp. 2097-2110 ◽  
Author(s):  
David Milne ◽  
Louis L Pen ◽  
David Thompson ◽  
William Powrie
Keyword(s):  
Performance Monitoring ◽  
Low Frequency ◽  
Cumulative Distribution ◽  
Railway Track ◽  
Statistical Process ◽  
Vibration Data ◽  
Rest Position ◽  
Low Frequency Vibration ◽  
Automated Processing ◽  
And Performance

Measurements of low-frequency vibration are increasingly being used to assess the condition and performance of railway tracks. Displacements used to characterise the track movement under train loads are commonly obtained from velocity or acceleration signals. Artefacts from signal processing, which lead to a shift in the datum associated with the at-rest position, as well as variability between successive wheels, mean that interpreting measurements is non-trivial. As a result, deflections are often interpreted by inspection rather than following an algorithmic or statistical process. This can limit the amount of data that can be usefully analysed in practice, militating against widespread or long-term use of track vibration measurements for condition or performance monitoring purposes. This paper shows how the cumulative distribution function of the track deflection can be used to identify the at-rest position and to interpret the typical range of track movement from displacement data. This process can be used to correct the shift in the at-rest position in velocity or acceleration data, to determine the proportion of upward and downward movement and to align data from multiple transducers to a common datum for visualising deflection as a function of distance along the track. The technique provides a means of characterising track displacement automatically, which can be used as a measure of system performance. This enables large volumes of track vibration data to be used for condition monitoring.

Sign in / Sign up

Export Citation Format

Share Document