Virtual multiphase flow metering using diverse neural network ensemble and adaptive simulated annealing

AL-Qutami, Tareq Aziz; Ibrahim, Rosdiazli; Ismail, Idris; Ishak, Mohd Azmin

doi:10.1016/j.eswa.2017.10.014

Cited by 72 publications

(32 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…This task requires a minimum of time to solve it and a maximum of reliability of the found solution to the identified problem [12,35]. The complexity of the problem also lies in the fact [26,38,43] that the state and dynamics of the processes of functioning of a cellular communication network cannot be unambiguously described using clear mathematical models.…”

Section: Research Frameworkmentioning

confidence: 99%

Problems Solving of Cell Subscribers based on Expert Systems Neural Networks

AlRababah¹

2019

IJACSA

View full text Add to dashboard Cite

With the growing demand for telecommunications services, the number of calls to telecommunications companies related to issues of using services, setting up and maintaining equipment, as well as resolving possible problems arising in the process of using services is also growing. From the point of view of system analysis, the problem is the mismatch between the existing and the required (target) state of the system for a given state of the environment at the moment in time. Based on this definition, we consider the problem of the subscriber of the cellular network a mismatch between the existing and the required state of the cellular network in this state of the environment at the moment in time. The state of the cellular network is characterized by the functioning of all devices, the proposed range of services. A short time for analyzing problem situations and making decisions, a large amount of information characterizing the current situation, the difficulty of solving poorly formalized and poorly structured tasks in the absence of complete and reliable information about the state of the cellular communication network and the functioning of its elements lead to a mismatch of human capabilities to effectively solve these problems. In this regard, the development and implementation of a precedent based neural network expert system for solving the problems of subscribers of a cellular communication network is an urgent scientific and technical task.

show abstract

Section: Research Frameworkmentioning

confidence: 99%

Problems Solving of Cell Subscribers based on Expert Systems Neural Networks

AlRababah¹

2019

IJACSA

View full text Add to dashboard Cite

show abstract

“…Soft sensors for continuous processes such as gas and liquid flow estimation also had recent developments in the industrial applications. AL-Qutami et al [12] introduce an ensemble of diverse neural networks to learn a soft sensor for multiphase flow metering on continuous process. In [13], a soft sensor based on State Dependent Parameter model, with the goal to estimate concentration of gas on a stirring tank is used for quality prediction of the continuous process.…”

Section: Related Workmentioning

confidence: 99%

Multivariate Time Series for Data-Driven Endpoint Prediction in the Basic Oxygen Furnace

Sala¹,

Jalalvand

Deyne³

et al. 2018

2018 17th IEEE International Conference on Machine Learning and Applications (ICMLA)

View full text Add to dashboard Cite

Industrial processes are heavily instrumented by employing a large number of sensors, generating huge amounts of data. One goal of the Industry 4.0 era is to apply datadriven approaches to optimize such processes. At the basic oxygen furnace (BOF), molten iron is transformed into steel by lowering its carbon content and achieving a certain chemical endpoint. In this work, we propose a data-driven approach to predict the endpoint temperature and chemical concentration of phosphorus, manganese, sulfur and carbon at the basic oxygen furnace. The prediction is based on two distinct datasets. First, a collection of static features is used which represent a more classic data-driven solution. The second approach includes time-series data that provide a better estimate of the final endpoint and enable further tuning of the process parameters, if necessary. For both approaches, model-based feature selection is used to filter the most relevant information. Results obtained by both models are compared in order to estimate the added value of including the time series data analysis on the performance of the BOF process. Results show that a simple feature extraction approach can enhance the prediction for phosphorus, manganese and temperature.

show abstract

“…The literature contains numerous applications of neural networks to model and locally optimize complex processes; a list all previous works would be quite comprehensive so we only point to a few examples here [39,40,41]. We firmly believe that ReLU networks, which are the de-facto standard in deep learning [23], is an important class of surrogate models that deserves being studied.…”

Section: Introductionmentioning

confidence: 99%

ReLU networks as surrogate models in mixed-integer linear programs

Grimstad

Andersson

2019

Computers & Chemical Engineering

117

View full text Add to dashboard Cite

We consider the embedding of piecewise-linear deep neural networks (ReLU networks) as surrogate models in mixed-integer linear programming (MILP) problems. A MILP formulation of ReLU networks has recently been applied by many authors to probe for various model properties subject to input bounds. The formulation is obtained by programming each ReLU operator with a binary variable and applying the big-M method. The efficiency of the formulation hinges on the tightness of the bounds defined by the big-M values. When ReLU networks are embedded in a larger optimization problem, the presence of output bounds can be exploited in bound tightening. To this end, we devise and study several bound tightening procedures that consider both input and output bounds. Our numerical results show that bound tightening may reduce solution times considerably, and that small-sized ReLU networks are suitable as surrogate models in mixed-integer linear programs.

show abstract

Virtual multiphase flow metering using diverse neural network ensemble and adaptive simulated annealing

Cited by 72 publications

References 19 publications

Problems Solving of Cell Subscribers based on Expert Systems Neural Networks

Problems Solving of Cell Subscribers based on Expert Systems Neural Networks

Multivariate Time Series for Data-Driven Endpoint Prediction in the Basic Oxygen Furnace

ReLU networks as surrogate models in mixed-integer linear programs

Contact Info

Product

Resources

About