This paper presents a new approach to analyzing measurement records from industrial processes. The proposed methodology is based on the model of contextual processing and uses big data from experimental process tomography datasets. Electrical capacitance tomography is used for monitoring noninvasive flow and for data acquisition. The measurement data are collected, stored, and processed to identify process regimes and process threats. A specific physical modification was introduced into the pneumatic conveying flow rig in order to study flow behavior under extreme conditions, extending the available knowledge base. A support vector machine was applied for data classification. This study illustrates how contextual processing can facilitate data interpretation and opens the way for the development of methods for detecting pre-emergency flow patterns.
In this article, we present a novel application domain for human computation, specifically for crowdsourcing, which can help in understanding particle-tracking problems. Through an interdisciplinary inquiry, we built a crowdsourcing system designed to detect tracer particles in industrial tomographic images, and applied it to the problem of bulk solid flow in silos. As images from silo-sensing systems cannot be adequately analyzed using the currently available computational methods, human intelligence is required. However, limited availability of experts, as well as their high cost, motivates employing additional nonexperts. We report on the results of a study that assesses the task completion time and accuracy of employing nonexpert workers to process large datasets of images in order to generate data for bulk flow research. We prove the feasibility of this approach by comparing results from a user study with data generated from a computational algorithm. The study shows that the crowd is more scalable and more economical than an automatic solution. The system can help analyze and understand the physics of flow phenomena to better inform the future design of silos, and is generalized enough to be applicable to other domains.
Abstract-This paper covers design, implementation and evaluation of a system that may be used to predict future stock prices basing on analysis of data from social media services. The authors took advantage of large datasets available from Twitter micro blogging platform and widely available stock market records. Data was collected during three months and processed for further analysis. Machine learning was employed to conduct sentiment classification of data coming from social networks in order to estimate future stock prices. Calculations were performed in distributed environment according to Map Reduce programming model. Evaluation and discussion of results of predictions for different time intervals and input datasets proved efficiency of chosen approach is discussed here.
An intricate problem associated with fixed bed operation is liquid maldistribution, which denotes the fact that the liquid does not homogeneously flow through the bed. In a comparative study we evaluated two capacitance imaging methodsscapacitance wire-mesh sensor and electrical capacitance tomography (ECT)swith respect to their capability of measuring static and dynamic liquid holdup in a fixed bed. The capacitance wire-mesh sensor as an invasive instrument is able to disclose flow structures at higher spatial resolution and was therefore considered the reference instrument for liquid holdup measurement. We found that both methods predict dynamic liquid holdup in the column in a similar way with only small systematic deviation. The results therefore prove that noninvasive electrical capacitance tomography can reliably measure cross-sectional dynamic liquid holdup in a fixed bed, even with a simple and fast linear back projection reconstruction algorithm.
In Mediterranean seas and coastal zones, rivers can be the main source of mercury (Hg). Catchment management therefore affects the load of Hg reaching the sea with surface runoff. The major freshwater inflows to the Baltic Sea consist of large rivers. However, their systems are complex and identification of factors affecting the outflow of Hg from its catchments is difficult. For this reason, a study into the impact of watershed land use and season on mercury biogeochemistry and transport in rivers was performed along two small rivers which may be considered typical of the southern Baltic region. Neither of these rivers are currently impacted by industrial effluents, thus allowing assessment of the influence of catchment terrain and season on Hg geochemistry. The study was performed between June 2008 and May 2009 at 13 sampling points situated at different terrain types within the catchments (forest, wetland, agriculture and urban). Hg analyses were conducted by CVAFS. Arable land erosion was found to be an important source of Hg to the aquatic system, similar to urban areas. Furthermore, inflows of untreated storm water discharge resulted in a fivefold increase of Hg concentration in the rivers. The highest Hg concentration in the urban runoff was observed with the greatest amount of precipitation during summer. Moderate rainfalls enhance the inflow of bioavailable dissolved mercury into water bodies. Despite the lack of industrial effluents entering the rivers directly, the sub-catchments with anthropogenic land use were important sources of Hg in the rivers. This was caused by elution of metal, deposited in soils over the past decades, into the rivers. The obtained results are especially important in the light of recent environmental conscience regulations, enforcing the decrease of pollution by Baltic countries.
This work covers deployment of contextual processing of measurement data in application to temporal modeling of pneumatic conveying industrial process. Electrical capacitance tomography (ECT) used as a non-invasive process monitoring tool is supported by data mining for regularization of nonlinear inverse problem solution. Processing of a larger number of archived experimental datasets enables extracting additional constraints for inference. Contextual data processing model (CDPM) extracts demanded information from the data in order to incorporate it as an expert knowledge about the process temporal behavior. Then it is incorporated into the Bayesian inference framework. Comparative analysis with previous work and domain expert prepared baseline to the proposed approach is demonstrated. Additionally, simplified parameterization is tested and verified by the quantitative experimental analysis.
This paper presents a new approach to the analysis of data on powder flow from electrical capacitance tomography (ECT) using probability modelling and Bayesian statistics. The methodology is illustrated for powder flow in a hopper. The purpose, and special features, of this approach is that ‘high‐level’ statistical Bayesian modelling combined with a Markov chain Monte Carlo (MCMC) sampling algorithm allows direct estimation of control parameters of industrial processes in contrast to usually applied ‘low‐level’, pixel‐based methods of data analysis. This enables reliable recognition of key process features in a quantitative manner. The main difficulty when investigating hopper flow with ECT is due to the need to measure small differences in particle packing density. The MCMC protocol enables more robust identification of the responses of such complex systems. This paper demonstrates the feasibility of the approach for a simple case of particulate material flow during discharging of a hopper. It is concluded that these approaches can offer significant advantages for the analysis and control of some industrial powder and other multi‐phase flow processes.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.