A reference point selection and direction guidance-based algorithm for large-scale multi-objective optimization

Author(s):  
Xiangjuan Wu ◽  
Yuping Wang ◽  
Shuai Tian ◽  
Ziqing Wang
2009 ◽  
Vol 15 (2) ◽  
pp. 325-375 ◽  
Author(s):  
Willem K. Brauers ◽  
Edmundas K. Zavadskas

The definition of robustness in econometrics, the error term in a linear equation, was not only broadened, but in addition moved to the meaning of common language: from a cardinal to a qualitative one. These interpretations were tested by an application on the Facilities Sector in Lithuania. The application is multi‐objective: like costs, experience and effectiveness at the side of the contractors; quality, duration of the work and cost price at the side of the owners. These objectives having all different units the dimensionless ratios of the MOORA method avoids the difficulties of normalization. In a first part of MOORA these ratios were aggregated and in a second one they were used as distances to a reference point. The results of both parts control each other, a test on robustness. Additionally, MOORA shows a robust domination on all other methods of multi‐objective optimization. For the Facilities Sector in Lithuania, both parts of MOORA resulted in a comparable ranking. In this way a double check was made on the robustness of the results. Santrauka Patikimumo apibrėžimas ekonometrikoje, kaip neteisingas terminas tiesinėje lygtyje, buvo ne tik papildytas, bet ir išreikštas įprasta kalba: nuo kiekybinio prie kokybinio. Šios interpretacijos buvo patikrintos taikant jas Lietuvos paslaugų sektoriuje. Taikymas yra daugiatikslis: iš rangovo pusės kaip išlaidos, patirtis, efektyvumas; kokybė, darbo trukmė, kaina iš užsakovo pusės. Minėtieji tikslai turi skirtingus matavimo vienetus. O jų santykiniai dydžiai neturi mato vienetų, todėl taikant MOORA metodą yra išvengiama sunkumų juos normalizuojant. Pirmoje MOORA metodo taikymo dalyje šie santykiai yra sujungiami, o antroje dalyje ieškoma atstumo iki geriausio sprendinio. Abiejų metodo dalių rezultatai pagrindžia sprendinio teisingumą. Tai rodo aiškų MOORA metodo pranašumą, palyginti su kitais daugiatikslio optimizavimo metodais. Taikant abi MOORA metodo dalis Lietuvos paslaugų sektoriui buvo sudarytas lyginamasis rangavimas, buvo atliktas dvigubas rezultatų patikimumo patikrinimas.


2017 ◽  
Vol 58 ◽  
pp. 25-34 ◽  
Author(s):  
Rui Wang ◽  
Jian Xiong ◽  
Hisao Ishibuchi ◽  
Guohua Wu ◽  
Tao Zhang

2011 ◽  
Vol 90-93 ◽  
pp. 2734-2739
Author(s):  
Ruan Yun ◽  
Cui Song Yu

Non-dominated sorting genetic algorithms II (NSGAII) has been widely used for multi- objective optimizations. To overcome its premature shortcoming, an improved NSGAII with a new distribution was proposed in this paper. Comparative to NSGAII, improved NSGAII uses an elitist control strategy to protect its lateral diversity among current non-dominated fronts. To implement elitist control strategy, a new distribution (called dogmatic distribution) was proposed. For ordinary multi-objective optimization problem (MOP), an ordinary exploration ability of improved NSGAII should be maintained by using a larger shape parameter r; while for larger-scale complex MOP, a larger exploration ability of improved NSGAII should be maintained by using a less shape parameter r. The application of improved NSGAII in multi-objective operation of Wohu reservoir shows that improved NSGAII has advantages over NSGAII to get better Pareto front especially for large-scale complex multi-objective reservoir operation problems.


Sign in / Sign up

Export Citation Format

Share Document