Visualization of the Non-Dominated Solutions in Many-Objective Optimization

Xiong, Minghui; Xiong, Wei; Jian, Ping

doi:10.1109/dsc.2019.00036

Cited by 4 publications

(3 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Therefore, in this section, the experimental results of different algorithms would be analyzed from the aspects of the effectiveness, convergence and distribution. The distributions of final solutions the three algorithms obtained solving case 1,2,3 are visualized based on the visual clustering method [15], as shown in figure 4. The red and blue lines represent the preferred and unpreferred solutions, respectively.…”

Section: Resultsmentioning

confidence: 99%

A Region Preference-based Optimization Algorithm for Satellite Mission Planning Problem

Xiong

2022

J. Phys.: Conf. Ser.

Self Cite

View full text Add to dashboard Cite

For the earth observation satellite mission planning problem, objectives such as observed target quantity, observation profit, energy consumption, image quality should be considered simultaneously, which is a many-objective optimization problem. Classical optimization-based mission planning algorithms obtain a set of non-dominated solutions in the entire search space, while only a single satisfy final plan is desired by decision maker. In this paper, a five-objective optimization model for satellite mission planning problem is constructed, then a region preference-based evolutionary algorithm, HMOEA-T, is applied to obtain the desired solutions. The decision makers describe the preference on each objective in target region form, then the algorithm guides a more detailed search within the preference region rather than the entire Pareto front. Comparative studies with preference-based methods (T-NSGA-III) and classical methods (NSGA-III) are conducted. We have exemplified the proposed method manage to obtain the solutions satisfying the mission planning preference and achieve better performance in convergence and diversity.

show abstract

Section: Resultsmentioning

confidence: 99%

A Region Preference-based Optimization Algorithm for Satellite Mission Planning Problem

Xiong

2022

J. Phys.: Conf. Ser.

Self Cite

View full text Add to dashboard Cite

show abstract

“…Therefore, this section will compare and analyze the experimental results from three aspects: the effectiveness, convergence and diversity, time cost. [15] is utilized to intuitively examine and compare the performance of different algorithms. In the plot, the red and blue lines denotes to the solutions located inside and outside the preference region, respectively.…”

Section: B Results and Discussionmentioning

confidence: 99%

“…Step 4: Interaction and determine. Once the termination criterion is satisfied, a non-dominated solution visualization method [15] is utilized to help the DM intuitively analysis the current population and determine. If satisfied, the adjusted preference region and current population Pc serve as the new input, back to Step 2 and start the new optimization;…”

Section: Proposed Methodsmentioning

confidence: 99%

Preference-based Evolutionary Many-objective Optimization for Regional Coverage Satellite Constellation Design

Xiong¹,

Xiong²

2019

Proceedings of the 2019 International Conference on Mathematics, Big Data Analysis and Simulation and Modelling (MBDASM 2019)

Self Cite

View full text Add to dashboard Cite

For the satellite constellation design problem, the computational complexity of regional coverage performance evaluation has posed great difficulty to classical Pareto dominance-based algorithms discovering the entire Pareto front of the problem, while the decision makers are often interested in a limited part of it. In this paper, a preference-based manyobjective evolutionary algorithm, HMOEA-T, is utilized to solve this problem. The optimization includes two steps. The first step describes the preference information of decision maker by target region, while the second step focuses the search process on the preferred region and maintaining well convergence and diversity within the region. A visualization method is applied to intuitively analysis the performance of different methods handling the problem. Experimental results have shown the advantage of integrating preference information into the optimization process, and the comparative study with other state-of-the-art preferencebased methods (T-MOEA/D and T-NSGA-III) indicate that the proposed method can achieve competitive and better performance.

show abstract

Visualisation for Decision Support in Many-Objective Optimisation: State-of-the-art, Guidance and Future Directions

Hakanen

Gold

Miettinen

et al. 2023

Natural Computing Series

View full text Add to dashboard Cite

Visualization of the Non-Dominated Solutions in Many-Objective Optimization

Cited by 4 publications

References 23 publications

A Region Preference-based Optimization Algorithm for Satellite Mission Planning Problem

A Region Preference-based Optimization Algorithm for Satellite Mission Planning Problem

Preference-based Evolutionary Many-objective Optimization for Regional Coverage Satellite Constellation Design

Visualisation for Decision Support in Many-Objective Optimisation: State-of-the-art, Guidance and Future Directions

Contact Info

Product

Resources

About