A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
Dynamic multiobjective optimization and multivariate analysis for power generation scheduling of the diesel generators in dynamically positioned vessels
Highlights A two-phase analysis framework is proposed for dynamic multiobjective optimization of power generation scheduling. Multi-objective ant lion optimizer (MOALO) is used to find the Pareto set. Employing Self-organizing maps (SOM) and cluster analysis methods to study the relationships among samples. Multidimensional scaling (MDS) is applied to investigate the relationships among performances and decision variables. An oil rig platform example is demonstrated.
Abstract Economic and environmental topics together with the higher performance of marine vessels have been becoming more demanding in recent years. This work aims to investigate the optimal operation of diesel generators in the dynamically positioned vessels. Due to the continuous external power demand, the dynamic multiobjective optimization model for optimal scheduling of the diesel generators is generated, taking the fuel consumption and greenhouse gas emission into consideration simultaneously. The present study proposes a 2-phase analysis framework for each time step. A newly developed meta-heuristic optimization algorithm, multi-objective ant lion optimizer (MOALO) is adopted to find the Pareto set in Phase-I. A necessary decision-making procedure based on a simple additive weighting (SAW) approach is utilized to find the final compromise solution in Phase-II. Furthermore, multivariate analysis (MVA) methods are employed to mine the historical features of economic and environmental performances as well as the loadings of all generators. Self-organizing mapping (SOM) together with cluster analysis is utilized to study the relationships among data samples. Multidimensional scaling (MDS) is adopted to examine the relationships among these attributes. An oil rig platform equipped with 8 diesel generators is selected as an illustrative example. The dynamic Pareto fronts and corresponding final compromise solutions obtained from SAW are provided. The effects of various weights upon the history of fuel consumption and emission are examined. The hidden information about the performances in history is studied through SOM and MDS. Results from the numerical example demonstrate that dynamic multiobjective optimization and multivariate analysis extend the application of optimization and data mining in the field of power scheduling for diesel generators in dynamically positioned vessels. The findings of this study add to the understanding of relationships among loads of the generators and corresponding economic and environmental features.
Dynamic multiobjective optimization and multivariate analysis for power generation scheduling of the diesel generators in dynamically positioned vessels
Highlights A two-phase analysis framework is proposed for dynamic multiobjective optimization of power generation scheduling. Multi-objective ant lion optimizer (MOALO) is used to find the Pareto set. Employing Self-organizing maps (SOM) and cluster analysis methods to study the relationships among samples. Multidimensional scaling (MDS) is applied to investigate the relationships among performances and decision variables. An oil rig platform example is demonstrated.
Abstract Economic and environmental topics together with the higher performance of marine vessels have been becoming more demanding in recent years. This work aims to investigate the optimal operation of diesel generators in the dynamically positioned vessels. Due to the continuous external power demand, the dynamic multiobjective optimization model for optimal scheduling of the diesel generators is generated, taking the fuel consumption and greenhouse gas emission into consideration simultaneously. The present study proposes a 2-phase analysis framework for each time step. A newly developed meta-heuristic optimization algorithm, multi-objective ant lion optimizer (MOALO) is adopted to find the Pareto set in Phase-I. A necessary decision-making procedure based on a simple additive weighting (SAW) approach is utilized to find the final compromise solution in Phase-II. Furthermore, multivariate analysis (MVA) methods are employed to mine the historical features of economic and environmental performances as well as the loadings of all generators. Self-organizing mapping (SOM) together with cluster analysis is utilized to study the relationships among data samples. Multidimensional scaling (MDS) is adopted to examine the relationships among these attributes. An oil rig platform equipped with 8 diesel generators is selected as an illustrative example. The dynamic Pareto fronts and corresponding final compromise solutions obtained from SAW are provided. The effects of various weights upon the history of fuel consumption and emission are examined. The hidden information about the performances in history is studied through SOM and MDS. Results from the numerical example demonstrate that dynamic multiobjective optimization and multivariate analysis extend the application of optimization and data mining in the field of power scheduling for diesel generators in dynamically positioned vessels. The findings of this study add to the understanding of relationships among loads of the generators and corresponding economic and environmental features.
Dynamic multiobjective optimization and multivariate analysis for power generation scheduling of the diesel generators in dynamically positioned vessels
Xuebin, Li (author) / Luchun, Yang (author) / Lihua, Huang (author) / Changjie, Wang (author)
Applied Ocean Research ; 122
2022-03-07
Article (Journal)
Electronic Resource
English
Dynamically Positioned Vessels — Their Suitability and Safe Operations
| Springer Verlag | 1991
Stochastic generation scheduling considering wind power generators
| American Institute of Physics | 2012
NORWAY - Dynamically positioned FPSO broadens options for ultra deep production
Online Contents | 2007