Population Feasibility State Guided Autonomous Constrained Multi-Objective Evolutionary Optimization
Abstract
:1. Introduction
- (1)
- Proposing a self-guided optimization method that combines the objective and constraint separation method with the multi-operator method. Specifically, this method combines the constraint method with the multi-operator method, using the multi-operator approach to enhance the diversity of solutions with more search directions. Adjusting the population search direction helps reduce the dependence of the constraint method on human experience for setting values. Additionally, overcoming the limitations of the multi-operator method relying on human experience using reinforcement learning further enhances the algorithm’s performance. This holds potential to contribute to the field of combining various constraint handling methods.
- (2)
- Proposing a novel method for characterizing the population state. Existing research has not characterized the population state from the perspective of both feasibility and feasibility of population solutions, limiting the perception of the population’s feasibility state. The proposed method for characterizing population state in this paper contributes to the description of population state in the field of autonomous intelligent optimization.
2. Related Works
2.1. Constrained Multi-Objective Evolutionary Optimization
2.2. The Objective and Constraint Separation Method
2.2.1. Constraint Dominance Principle
- (1)
- Both and are feasible solutions, but has a better optimization objective value.
- (2)
- is a feasible solution, whereas is an infeasible solution.
- (3)
- Both and are infeasible solutions, but , where represents the degree of constraint violation for the optimization solution.
2.2.2. ε Constrained Method
- (1)
- Both and are feasible solutions, but has a better optimization objective value.
- (2)
- is an feasible solution, whereas is not an feasible solution.
- (3)
- Both and are not feasible solutions, but , where represents the degree of constraint violation for the optimization solution.
2.2.3. Stochastic Ranking
2.3. The Multi-Operator Method
2.4. Discussion
3. The Proposed Method
3.1. Overall Framework
- (1)
- Learning the optimal strategy: Q-learning can learn the optimal strategy, that is, taking actions that can maximize cumulative returns in each state. By continuously updating the Q-value table, the agent can gradually converge to the optimal strategy.
- (2)
- No model required: Q-learning does not require modeling of the environment; that is, it does not require prior knowledge of state transition probabilities and reward functions. It learns the Q-value function through interaction with the environment, thus achieving model free learning.
- (3)
- Adaptability: Q-learning can adapt to different environments and tasks. It can handle continuous state space and action space.
- (4)
- Convergence: Under certain conditions, the Q-learning algorithm can converge to the optimal Q-value function. This means that the agent can learn the optimal strategy within a limited time.
Algorithm 1 Overall framework |
Input: Maximum generation , Population size , Problem dimension , Scaling factor in DE operator , Crossover reward update parameter , reward predictability parameter , greedy strategy parameter . Output |
1. ; |
2. with equation (4) |
3. ; |
4. ; |
5. 0.005, 0.2, 0.8; |
6. 0 0; |
7. , ; |
8. ; |
9. do |
10. ; |
11. Select operator with Algorithm 2; |
12. ; |
13. ; |
14. For = 1: |
15. ; |
16. ; |
17. ; |
18. ; |
19. end |
20. ; |
21. ; |
22. with Algorithm 3 to update Q-table; |
23. ; |
24. , ; |
25. ; |
26. End while |
27. ; |
3.2. The Selection of Reproduction Operator
3.2.1. Reproduction Operator for Regulating Population State
Algorithm 2 Two-stage reproduction operator selection |
Input: current generation , Q-table, Count-table; ; |
1. then |
2. ; |
3. Else |
4. ; |
5. }; |
6. End if |
7. then |
8. ; |
9. End if |
3.2.2. Two-Stage Reproduction Operator Selection
3.3. Update the Q-Table
3.3.1. Characterizing the Population State Based on the Feasibility of Population
Algorithm 3 Update the Q-table |
Input parameters of Q-learning, Count-table, Q-table; |
Output: Q-table, Count-table; |
1. ; |
2. then |
3. ; |
4. Else |
5. ; |
6. End if |
7. ; |
8. Count-tableCount-table(; |
3.3.2. Stage-Wise Evaluation of Population Solutions
3.3.3. Q-Value Update
3.4. Further Explanation
- (1)
- The problem type aimed at is different. Our research is devoted to solving constrained multi-objective problems, while the mentioned research is proposed for the unconstrained optimization problems.
- (2)
- The definition of population state is different. Our research defines the population state with the population feasibility information, which is useful to reduce the complexity of the mapping model between state, action, and reward. This option is important for proposing an efficient online adaptive operator selection algorithm, without the offline training phase. Oppositely, the population state in the abovementioned paper is population individual. Hence, their mapping model will be more accurate once well trained. However, it is not convenient to transfer the mapping model between the problems with different dimensions.
- (3)
- The definition of reward function is different. Our research defines the reward function using the indicator-based evaluation method, while the abovementioned paper employs the population fitness method to evaluate the performed action. Oppositely, their method has low cost on the reward function, while our method is more stable on evaluating the effectiveness of the performed operator.
3.5. Complexity Analysis
- (1)
- The time complexity of extracting population states is .
- (2)
- The time complexity of executing evolutionary strategies in the worst-case scenario is .
- (3)
- The time complexity of evaluating evolutionary strategies in the worst-case scenario is .
- (4)
- The time complexity of pre-warming and applying Q-learning in the worst-case scenario is ).
4. Experiment
4.1. Benchmark Test Suits
4.2. Comparison Algorithms and Parameters Setting
4.3. Performance Indicator
4.4. Effectiveness of the Proposed Method
- (1)
- Collaboration of multiple operators can increase the population diversity. When solving constrained multi-objective optimization problems, the complexity of the problem often leads to the algorithm falling into local optima. By using multiple operators to work together, the population diversity is increased, thereby enlarging the coverage on search space, and helping to avoid getting stuck in local optima.
- (2)
- Multiple operators’ collaboration can improve the convergence speed of algorithms. When solving constrained multi-objective optimization problems, the algorithm needs to find the optimal solution within a limited number of iterations. By using multiple operators to work together, it is possible to search in different directions, thereby accelerating the convergence speed of the algorithm.
- (3)
- Collaboration of multiple operators can increase the search ability of algorithms. When solving constrained multi-objective optimization problems, the algorithm needs to find the optimal solution in the search space. By using multiple operators to work together, it is possible to search in different directions simultaneously, thereby increasing the search capability of the algorithm and helping to find better solutions.
4.5. The Comprehensive Performance of the Proposed Method
4.5.1. Compared with State-of-the-Art Constrained Optimization Algorithms
4.5.2. Compared with -Feasibility-Based Constrained Optimization Algorithms
4.5.3. Sensitivity Analysis of Key Parameters
5. Conclusions
Author Contributions
Funding
Data Availability Statement
Conflicts of Interest
References
- Hu, H.; Sun, X.; Zeng, B.; Gong, D.; Zhang, Y. Enhanced evolutionary multi-objective optimization-based dispatch of coal mine integrated energy system with flexible load. Appl. Energy 2022, 307, 118130. [Google Scholar] [CrossRef]
- Gong, D.; Xu, B.; Zhang, Y.; Guo, Y.; Yang, S. A Similarity-Based Cooperative Co-Evolutionary Algorithm for Dynamic Interval Multiobjective Optimization Problems. IEEE Trans. Evol. Comput. 2020, 24, 142–156. [Google Scholar] [CrossRef]
- Zuo, M.; Dai, G.; Peng, L.; Wang, M.; Liu, Z.; Chen, C. A case learning-based differential evolution algorithm for global optimization of interplanetary trajectory design. Appl. Soft Comput. 2020, 94, 106451. [Google Scholar] [CrossRef]
- Kumar, A.; Wu, G.H.; Ali, M.Z.; Mallipeddi, R.; Suganthan, P.N.; Das, S. A test-suite of non-convex constrained optimization problems from the real-world and some baseline results. Swarm Evol. Comput. 2020, 56, 100693. [Google Scholar] [CrossRef]
- Liang, J.; Ban, X.; Yu, K.; Qu, B.; Qiao, K.; Yue, C.; Chen, K.; Tan, K.C. A Survey on Evolutionary Constrained Multiobjective Optimization. IEEE Trans. Evol. Comput. 2023, 27, 201–221. [Google Scholar] [CrossRef]
- Zuo, M.C.; Dai, G.M.; Peng, L.; Tang, Z.; Gong, D.W.; Wang, Q.X. A differential evolution algorithm with the guided movement for population and its application to interplanetary transfer trajectory design. Eng. Appl. Artif. Intell. 2022, 110, 104727. [Google Scholar] [CrossRef]
- Zuo, M.C.; Dai, G.M.; Peng, L. A new mutation operator for differential evolution algorithm. Soft Comput. 2021, 25, 13595–13615. [Google Scholar] [CrossRef]
- Zuo, M.C.; Dai, G.M.; Peng, L. Multi-agent genetic algorithm with controllable mutation probability utilizing back propagation neural network for global optimization of trajectory design. Eng. Optim. 2019, 51, 120–139. [Google Scholar] [CrossRef]
- Jiao, L.C.; Luo, J.J.; Shang, R.H.; Liu, F. A modified objective function method with feasible-guiding strategy to solve constrained multi-objective optimization problems. Appl. Soft Comput. 2014, 14, 363–380. [Google Scholar] [CrossRef]
- Maldonado, H.M.; Zapotecas-Martínez, S. A Dynamic Penalty Function within MOEA/D for Constrained Multi-objective Optimization Problems. In Proceedings of the 2021 IEEE Congress on Evolutionary Computation (CEC), Kraków, Poland, 28 June–1 July 2021; pp. 1470–1477. [Google Scholar]
- Fan, Z.; Ruan, J.; Li, W.J.; You, Y.G.; Cai, X.Y.; Xu, Z.L.; Yang, Z.; Sun, F.Z.; Wang, Z.J.; Yuan, Y.T.; et al. A Learning Guided Parameter Setting for Constrained Multi-Objective Optimization. In Proceedings of the 2019 1st International Conference on Industrial Artificial Intelligence, Shenyang, China, 22–26 July 2019. [Google Scholar]
- Morovati, V.; Pourkarimi, L. Extension of Zoutendijk method for solving constrained multiobjective optimization problems. Eur. J. Oper. Res. 2019, 273, 44–57. [Google Scholar] [CrossRef]
- Schütze, O.; Alvarado, S.; Segura, C.; Landa, R. Gradient subspace approximation: A direct search method for memetic computing. Soft Comput. 2017, 21, 6331–6350. [Google Scholar] [CrossRef]
- Long, Q. A constraint handling technique for constrained multi-objective genetic algorithm. Swarm Evol. Comput. 2014, 15, 66–79. [Google Scholar] [CrossRef]
- Vieira, D.A.G.; Adriano, R.L.S.; Vasconcelos, J.A.; Krähenbühl, L. Treating constraints as objectives in multiobjective optimization problems using niched pareto genetic algorithm. IEEE Trans. Magn. 2004, 40, 1188–1191. [Google Scholar] [CrossRef]
- Ming, F.; Gong, W.; Wang, L.; Gao, L. Constrained Multi-objective Optimization via Multitasking and Knowledge Transfer. In Proceedings of the IEEE Transactions on Evolutionary Computation, Online, 20 December 2022. [Google Scholar]
- Deb, K.; Pratap, A.; Agarwal, S.; Meyarivan, T. A fast and elitist multiobjective genetic algorithm: NSGA-II. IEEE Trans. Evol. Comput. 2002, 6, 182–197. [Google Scholar] [CrossRef]
- Takahama, T.; Sakai, S. Constrained optimization by the ε constrained differential evolution with gradient-based mutation and feasible elites. IEEE Congr. Evol. Comput. 2006, 3, 2322–2327. [Google Scholar]
- Runarsson, T.P.; Xin, Y. Stochastic ranking for constrained evolutionary optimization. IEEE Trans. Evol. Comput. 2000, 4, 284–294. [Google Scholar] [CrossRef]
- Saha, A.; Ray, T. Equality Constrained Multi-Objective Optimization. In Proceedings of the IEEE Congress on Evolutionary Computation, Brisbane, QLD, Australia, 10–15 June 2012. [Google Scholar]
- Hamida, S.B.; Schoenauer, M. ASCHEA: New results using adaptive segregational constraint handling. In Proceedings of the 2002 Congress on Evolutionary Computation. CEC’02, Honolulu, HI, USA, 12–17 May 2002; pp. 884–889. [Google Scholar]
- Zapotecas-Martínez, S.; Ponsich, A. Constraint Handling within MOEA/D Through an Additional Scalarizing Function. Genet. Evol. Comput. Conf. 2020, 6, 595–602. [Google Scholar]
- Liu, Z.Z.; Wang, Y.; Wang, B.C.A. Indicator-Based Constrained Multiobjective Evolutionary Algorithms. IEEE Trans. Syst. Man Cybern.-Syst. 2021, 51, 5414–5426. [Google Scholar] [CrossRef]
- Jan, M.A.; Khanum, R.A. A study of two penalty-parameterless constraint handling techniques in the framework of MOEA/D. Appl. Soft Comput. 2013, 13, 128–148. [Google Scholar] [CrossRef]
- Ying, W.Q.; Peng, D.X.; Xie, Y.H.; Wu, Y. An Annealing Stochastic Ranking Mechanism for Constrained Evolutionary Optimization. In Proceedings of the 2016 International Conference on Information System and Artificial Intelligence, Hong Kong, China, 24–26 June 2016; pp. 576–580. [Google Scholar]
- Sabat, S.L.; Ali, L.; Udgata, S.K. Stochastic Ranking Particle Swarm Optimization for Constrained Engineering Design Problems; Lecture Notes in Computer Science; Springer: Berlin/Heidelberg, Germany, 2010; 672p. [Google Scholar]
- Li, L.; Li, G.P.; Chang, L. On self-adaptive stochastic ranking in decomposition many-objective evolutionary optimization. Neurocomputing 2022, 489, 547–557. [Google Scholar] [CrossRef]
- Ying, W.Q.; He, W.P.; Huang, Y.X.; Wu, Y. An Adaptive Stochastic Ranking Mechanism in MOEA/D for Constrained Multi-objective Optimization. In Proceedings of the 2016 International Conference on Information System and Artificial Intelligence, Hong Kong, China, 24–26 June 2016; pp. 514–518. [Google Scholar]
- Gu, Q.H.; Wang, Q.; Xiong, N.N.; Jiang, S.; Chen, L. Surrogate-assisted evolutionary algorithm for expensive constrained multi-objective discrete optimization problems. Complex Intell. Syst. 2022, 8, 2699–2718. [Google Scholar] [CrossRef]
- Wang, H.D.; Jin, Y.C. A Random Forest-Assisted Evolutionary Algorithm for Data-Driven Constrained Multiobjective Combinatorial Optimization of Trauma Systems. IEEE Trans. Cybern. 2020, 50, 536–549. [Google Scholar] [CrossRef]
- Li, B.D.; Tang, K.; Li, J.L.; Yao, X. Stochastic Ranking Algorithm for Many-Objective Optimization Based on Multiple Indicators. IEEE Trans. Evol. Comput. 2016, 20, 924–938. [Google Scholar] [CrossRef]
- Chen, Y.; Yuan, X.P.; Cang, X.H. A new gradient stochastic ranking-based multi-indicator algorithm for many-objective optimization. Soft Comput. 2019, 23, 10911–10929. [Google Scholar] [CrossRef]
- Yu, X.B.; Yu, X.R.; Lu, Y.Q.; Yen, G.G.; Cai, M. Differential evolution mutation operators for constrained multi-objective optimization. Appl. Soft Comput. 2018, 67, 452–466. [Google Scholar] [CrossRef]
- Xu, B.; Duan, W.; Zhang, H.F.; Li, Z.Q. Differential evolution with infeasible-guiding mutation operators for constrained multi-objective optimization. Appl. Intell. 2020, 50, 4459–4481. [Google Scholar] [CrossRef]
- Liu, Y.J.; Li, X.; Hao, Q.J. A New Constrained Multi-objective Optimization Problems Algorithm Based on Group-sorting. Proc. Genet. Evol. Comput. Conf. Companion 2019, 7, 221–222. [Google Scholar]
- He, C.; Cheng, R.; Tian, Y.; Zhang, X.Y.; Tan, K.C.; Jin, Y.C. Paired Offspring Generation for Constrained Large-Scale Multiobjective Optimization. IEEE Trans. Evol. Comput. 2021, 25, 448–462. [Google Scholar] [CrossRef]
- Tian, Y.; Li, X.P.; Ma, H.P.; Zhang, X.Y.; Tan, K.C.; Jin, Y.C. Deep Reinforcement Learning Based Adaptive Operator Selection for Evolutionary Multi-Objective Optimization. IEEE Trans. Emerg. Top. Comput. Intell. 2023, 7, 1051–1064. [Google Scholar] [CrossRef]
- Zuo, M.; Gong, D.; Wang, Y.; Ye, X.; Zeng, B.; Meng, F. Process Knowledge-guided Autonomous Evolutionary Optimization for Constrained Multiobjective Problems. IEEE Trans. Evol. Comput. 2023, 28, 193–207. [Google Scholar] [CrossRef]
- Jiao, R.W.; Zeng, S.Y.; Li, C.H.; Yang, S.X.; Ong, Y.S. Handling Constrained Many-Objective Optimization Problems via Problem Transformation. IEEE Trans. Cybern. 2021, 51, 4834–4847. [Google Scholar] [CrossRef]
- Tian, Y.; Cheng, R.; Zhang, X.; Jin, Y. PlatEMO: A MATLAB Platform for Evolutionary Multi-Objective Optimization [Educational Forum]. IEEE Comput. Intell. Mag. 2017, 12, 73–87. [Google Scholar] [CrossRef]
- Kumar, A.; Wu, G.H.; Ali, M.Z.; Luo, Q.Z.; Mallipeddi, R.; Suganthan, P.N.; Das, S. A Benchmark-Suite of real-World constrained multi-objective optimization problems and some baseline results. Swarm Evol. Comput. 2021, 67, 100961. [Google Scholar] [CrossRef]
- Liang, J.; Qiao, K.J.; Yu, K.J.; Qu, B.Y.; Yue, C.T.; Guo, W.F.; Wang, L. Utilizing the Relationship Between Unconstrained and Constrained Pareto Fronts for Constrained Multiobjective Optimization. IEEE Trans. Cybern. 2023, 53, 3873–3886. [Google Scholar] [CrossRef]
- Tian, Y.; Zhang, Y.; Su, Y.; Zhang, X.; Tan, K.C.; Jin, Y. Balancing Objective Optimization and Constraint Satisfaction in Constrained Evolutionary Multiobjective Optimization. IEEE Trans. Cybern. 2022, 52, 9559–9572. [Google Scholar] [CrossRef]
- Liu, Z.Z.; Wang, B.C.; Tang, K. Handling Constrained Multiobjective Optimization Problems via Bidirectional Coevolution. IEEE Trans. Cybern. 2022, 52, 10163–10176. [Google Scholar] [CrossRef]
- Panichella, A. An Improved Pareto Front Modeling Algorithm for Large-scale Many-Objective Optimization. Proc. Genet. Evol. Comput. Conf. 2022, 7, 565–573. [Google Scholar]
- Yu, K.J.; Liang, J.; Qu, B.Y.; Luo, Y.; Yue, C.T. Dynamic Selection Preference-Assisted Constrained Multiobjective Differential Evolution. IEEE Trans. Syst. Man Cybern.-Syst. 2022, 52, 2954–2965. [Google Scholar] [CrossRef]
- Ming, F.; Gong, W.Y.; Wang, L.; Lu, C. A tri-population based co-evolutionary framework for constrained multi-objective optimization problems. Swarm Evol. Comput. 2022, 70, 101055. [Google Scholar] [CrossRef]
- Zou, J.; Sun, R.Q.; Yang, S.X.; Zheng, J.H. A dual-population algorithm based on alternative evolution and degeneration for solving constrained multi-objective optimization problems. Inf. Sci. 2021, 579, 89–102. [Google Scholar] [CrossRef]
Problem | M | D | IDCNSGA-III | DCNSGA-III |
---|---|---|---|---|
TREE1 | 2 | 300 | 7.2952 × 10−1 (1.64 × 10−2) + | 7.0269 × 10−1 (7.73 × 10−3) |
TREE2 | 2 | 300 | 7.6076 × 10−1 (1.01 × 10−2) + | 7.4026 × 10−1 (6.86 × 10−3) |
TREE3 | 2 | 120 | 7.9388 × 10−1 (9.08 × 10−3) + | 7.6790 × 10−1 (1.20 × 10−2) |
TREE4 | 2 | 120 | 7.0300 × 10−1 (3.38 × 10−2) + | 6.3061 × 10−1 (3.22 × 10−2) |
TREE5 | 2 | 120 | 7.8099 × 10−1 (2.67 × 10−2) + | 7.4329 × 10−1 (2.03 × 10−2) |
LIRCMOP1 | 2 | 30 | 1.2444 × 10−1 (2.56 × 10−2) + | 1.0638 × 10−1 (2.28 × 10−2) |
LIRCMOP2 | 2 | 30 | 2.4912 × 10−1 (2.00 × 10−2) + | 2.2259 × 10−1 (2.44 × 10−2) |
LIRCMOP3 | 2 | 30 | 1.0955 × 10−1 (2.05 × 10−2) + | 9.0574 × 10−2 (1.32 × 10−2) |
LIRCMOP4 | 2 | 30 | 2.0854 × 10−1 (1.79 × 10−2) + | 1.8894 × 10−1 (1.89 × 10−2) |
LIRCMOP5 | 2 | 10 | 2.8561 × 10−1 (1.65 × 10−3) + | 2.1053 × 10−1 (2.88 × 10−2) |
LIRCMOP6 | 2 | 10 | 1.9183 × 10−1 (1.32 × 10−3) + | 1.1070 × 10−1 (4.64 × 10−2) |
LIRCMOP7 | 2 | 30 | 2.1168 × 10−1 (5.83 × 10−2) = | 1.9214 × 10−1 (7.85 × 10−2) |
LIRCMOP8 | 2 | 30 | 1.3642 × 10−1 (9.86 × 10−2) = | 1.4752 × 10−1 (9.08 × 10−2) |
LIRCMOP9 | 2 | 30 | 2.5944 × 10−1 (7.32 × 10−2) + | 1.0893 × 10−1 (3.78 × 10−2) |
LIRCMOP10 | 2 | 30 | 1.9340 × 10−1 (1.40 × 10−1) + | 5.9978 × 10−2 (3.49 × 10−2) |
LIRCMOP11 | 2 | 30 | 2.7786 × 10−1 (8.43 × 10−2) + | 1.6002 × 10−1 (4.40 × 10−2) |
LIRCMOP12 | 2 | 30 | 4.0423 × 10−1 (7.72 × 10−2) + | 2.8276 × 10−1 (1.00 × 10−1) |
LIRCMOP13 | 3 | 30 | 0.0000 × 10+0 (0.00 × 10+0) = | 3.7687 × 10−6 (2.06 × 10−5) |
LIRCMOP14 | 3 | 30 | 1.7358 × 10−5 (5.94 × 10−5) = | 1.6225 × 10−5 (6.98 × 10−5) |
MW1 | 2 | 30 | 1.4722 × 10−1 (1.30 × 10−1) | Infeasible |
MW2 | 2 | 30 | 3.8124 × 10−1 (1.29 × 10−1) = | 4.3084 × 10−1 (8.17 × 10−2) |
MW3 | 2 | 30 | 5.0566 × 10−1 (1.43 × 10−2) = | 4.3151 × 10−1 (1.60 × 10−1) |
MW4 | 3 | 30 | 4.3348 × 10−1 (5.15 × 10−2) | Infeasible |
MW5 | 2 | 30 | 1.1962 × 10−1 (7.62 × 10−2) | Infeasible |
MW6 | 2 | 10 | 2.9226 × 10−1 (1.89 × 10−2) = | 2.9375 × 10−1 (3.12 × 10−2) |
MW7 | 2 | 10 | 4.0304 × 10−1 (1.59 × 10−3) = | 4.0119 × 10−1 (3.99 × 10−3) |
MW8 | 3 | 10 | 4.8872 × 10−1 (2.38 × 10−2) - | 5.0081 × 10−1 (1.98 × 10−2) |
MW9 | 2 | 30 | 1.3763 × 10−1 (1.70 × 10−1) = | 1.4546 × 10−1 (0.00 × 10+0) |
MW10 | 2 | 30 | 2.2475 × 10−1 (9.60 × 10−2) = | 2.6132 × 10−1 (1.02 × 10−1) |
MW11 | 2 | 30 | 4.3830 × 10−1 (1.25 × 10−3) + | 4.3107 × 10−1 (4.68 × 10−3) |
MW12 | 2 | 30 | 1.4185 × 10−1 (2.45 × 10−1) = | 0.0000 × 10+0 (0.00 × 10+0) |
MW13 | 2 | 30 | 2.9592 × 10−1 (8.84 × 10−2) − | 3.6448 × 10−1 (5.60 × 10−2) |
MW14 | 3 | 30 | 1.1746 × 10−1 (4.12 × 10−2) + | 6.9888 × 10−2 (4.25 × 10−2) |
+/−/= | 17/2/11 |
Problem | M | D | CMOEA-MS | BiCo | IDCNSGA-III |
---|---|---|---|---|---|
TREE1 | 2 | 300 | 6.9857 × 10−1 (8.47 × 10−3) − | 6.6672 × 10−1 (1.07 × 10−2) − | 7.2952 × 10−1 (1.64 × 10−2) |
TREE2 | 2 | 300 | 7.3411 × 10−1 (6.04 × 10−3) − | 7.0974 × 10−1 (6.64 × 10−3) − | 7.6076 × 10−1 (1.01 × 10−2) |
TREE3 | 2 | 120 | 7.2314 × 10−1 (3.43 × 10−2) − | 6.7538 × 10−1 (2.71 × 10−2) − | 7.9388 × 10−1 (9.08 × 10−3) |
TREE4 | 2 | 120 | 5.8701 × 10−1 (7.96 × 10−2) − | 3.6920 × 10−1 (5.26 × 10−2) − | 7.0300 × 10−1 (3.38 × 10−2) |
TREE5 | 2 | 120 | 7.1495 × 10−1 (5.06 × 10−2) − | 5.8449 × 10−1 (3.32 × 10−2) − | 7.8099 × 10−1 (2.67 × 10−2) |
LIRCMOP1 | 2 | 30 | 9.7178 × 10−2 (1.24 × 10−2) − | 1.1009 × 10−1 (8.09 × 10−3) − | 1.2444 × 10−1 (2.56 × 10−2) |
LIRCMOP2 | 2 | 30 | 2.1006 × 10−1 (1.81 × 10−2) − | 2.2736 × 10−1 (1.08 × 10−2) − | 2.4912 × 10−1 (2.00 × 10−2) |
LIRCMOP3 | 2 | 30 | 9.0817 × 10−2 (9.06 × 10−3) − | 1.0056 × 10−1 (1.06 × 10−2) − | 1.0955 × 10−1 (2.05 × 10−2) |
LIRCMOP4 | 2 | 30 | 1.8370 × 10−1 (1.16 × 10−2) − | 1.9512 × 10−1 (1.11 × 10−2) − | 2.0854 × 10−1 (1.79 × 10−2) |
LIRCMOP5 | 2 | 10 | 1.1219 × 10−1 (1.01 × 10−1) − | 1.8104 × 10−2 (5.57 × 10−2) − | 2.8561 × 10−1 (1.65 × 10−3) |
LIRCMOP6 | 2 | 10 | 8.2188 × 10−2 (5.35 × 10−2) − | 1.6302 × 10−2 (3.54 × 10−2) − | 1.9183 × 10−1 (1.32 × 10−3) |
LIRCMOP7 | 2 | 30 | 2.3937 × 10−2 (6.37 × 10−2) − | 0.0000 × 10+0 (0.00 × 10+0) − | 2.1168 × 10−1 (5.83 × 10−2) |
LIRCMOP8 | 2 | 30 | 0.0000 × 10+0 (0.00 × 10+0)− | 0.0000 × 10+0 (0.00 × 10+0) − | 1.3642 × 10−1 (9.86 × 10−2) |
LIRCMOP9 | 2 | 30 | 1.0132 × 10−1 (2.56 × 10−2) − | 8.6803 × 10−2 (1.68 × 10−2) − | 2.5944 × 10−1 (7.32 × 10−2) |
LIRCMOP10 | 2 | 30 | 5.5279 × 10−2 (2.33 × 10−2) − | 5.3817 × 10−2 (1.60 × 10−2) − | 1.9340 × 10−1 (1.40 × 10−1) |
LIRCMOP11 | 2 | 30 | 1.6227 × 10−1 (3.12 × 10−2) − | 1.4853 × 10−1 (2.92 × 10−2) − | 2.7786 × 10−1 (8.43 × 10−2) |
LIRCMOP12 | 2 | 30 | 1.9414 × 10−1 (6.64 × 10−2) − | 2.1506 × 10−1 (8.50 × 10−2) − | 4.0423 × 10−1 (7.72 × 10−2) |
LIRCMOP13 | 3 | 30 | 7.8946 × 10−5 (9.93 × 10−5) + | 3.2686 × 10−5 (7.50 × 10−5) + | 0.0000 × 10+0 (0.00 × 10+0) |
LIRCMOP14 | 3 | 30 | 3.3711 × 10−4 (3.44 × 10−4) + | 1.7965 × 10−4 (2.07 × 10−4) + | 1.7358 × 10−5 (5.94 × 10−5) |
MW1 | 2 | 30 | Infeasible | Infeasible | 1.4722 × 10−1 (1.30 × 10−1) |
MW2 | 2 | 30 | 4.0040 × 10−1 (1.16 × 10−1) = | 4.7769 × 10−1 (8.73 × 10−2) + | 3.8124 × 10−1 (1.29 × 10−1) |
MW3 | 2 | 30 | 3.8630 × 10−1 (1.67 × 10−1) − | 4.3716 × 10−1 (6.28 × 10−2) − | 5.0566 × 10−1 (1.43 × 10−2) |
MW4 | 3 | 30 | Infeasible | Infeasible | 4.3348 × 10−1 (5.15 × 10−2) |
MW5 | 2 | 30 | Infeasible | 0.0000 × 10+0 (0.00 × 10+0) = | 1.1962 × 10−1 (7.62 × 10−2) |
MW6 | 2 | 10 | 2.9209 × 10−1 (2.72 × 10−2) = | 3.0992 × 10−1 (1.04 × 10−2) + | 2.9226 × 10−1 (1.89 × 10−2) |
MW7 | 2 | 10 | 3.9338 × 10−1 (2.98 × 10−2) − | 4.0454 × 10−1 (2.21 × 10−3) + | 4.0304 × 10−1 (1.59 × 10−3) |
MW8 | 3 | 10 | 4.8467 × 10−1 (7.40 × 10−2) = | 5.1693 × 10−1 (1.55 × 10−2) + | 4.8872 × 10−1 (2.38 × 10−2) |
MW9 | 2 | 30 | Infeasible | Infeasible | 1.3763 × 10−1 (1.70 × 10−1) |
MW10 | 2 | 30 | 2.5155 × 10−1 (8.48 × 10−2) = | 2.8290 × 10−1 (8.87 × 10−2) + | 2.2475 × 10−1 (9.60 × 10−2) |
MW11 | 2 | 30 | 3.4707 × 10−1 (6.92 × 10−2) − | 3.2792 × 10−1 (8.55 × 10−2) − | 4.3830 × 10−1 (1.25 × 10−3) |
MW12 | 2 | 30 | Infeasible | Infeasible | 1.4185 × 10−1 (2.45 × 10−1) |
MW13 | 2 | 30 | 5.5586 × 10−2 (1.14 × 10−1) − | 2.4713 × 10−1 (7.46 × 10−2) − | 2.9592 × 10−1 (8.84 × 10−2) |
MW14 | 3 | 30 | 7.1004 × 10−2 (4.95 × 10−2) − | 3.5529 × 10−2 (2.09 × 10−2) − | 1.1746 × 10−1 (4.12 × 10−2) |
+/−/= | 2/22/4 | 7/21/1 |
Problem | AGEMOEA-II | DSPCMDE | NSGA-II | IDCNSGA-III |
---|---|---|---|---|
TREE1 | 6.6896 × 10−1 (7.51 × 10−3) − | 7.1916 × 10−1 (1.06 × 10−2) − | 6.6897 × 10−1 (8.27 × 10−3) − | 7.2952 × 10−1 (1.64 × 10−2) |
TREE2 | 7.0870 × 10−1 (8.40 × 10−3) − | 7.5347 × 10−1 (7.02 × 10−3) − | 7.0990 × 10−1 (7.99 × 10−3) − | 7.6076 × 10−1 (1.01 × 10−2) |
TREE3 | 7.0231 × 10−1 (2.60 × 10−2) − | 7.7027 × 10−1 (1.83 × 10−2) − | 7.0375 × 10−1 (2.25 × 10−2) − | 7.9388 × 10−1 (9.08 × 10−3) |
TREE4 | 4.9408 × 10−1 (5.11 × 10−2) − | 6.3215 × 10−1 (3.56 × 10−2) − | 4.6792 × 10−1 (6.48 × 10−2) − | 7.0300 × 10−1 (3.38 × 10−2) |
TREE5 | 6.5838 × 10−1 (3.60 × 10−2) − | 7.2691 × 10−1 (2.81 × 10−2) − | 6.5426 × 10−1 (3.35 × 10−2) − | 7.8099 × 10−1 (2.67 × 10−2) |
LIRCMOP1 | 1.0273 × 10−1 (7.69 × 10−3) − | 1.2727 × 10−1 (3.16 × 10−2) = | 1.0180 × 10−1 (8.88 × 10−3) − | 1.2444 × 10−1 (2.56 × 10−2) |
LIRCMOP2 | 2.0970 × 10−1 (1.44 × 10−2) − | 2.4769 × 10−1 (4.96 × 10−2) = | 2.1462 × 10−1 (1.32 × 10−2) − | 2.4912 × 10−1 (2.00 × 10−2) |
LIRCMOP3 | 9.4261 × 10−2 (8.89 × 10−3) − | 1.0803 × 10−1 (2.32 × 10−2) = | 9.1757 × 10−2 (6.70 × 10−3) − | 1.0955 × 10−1 (2.05 × 10−2) |
LIRCMOP4 | 1.8121 × 10−1 (1.18 × 10−2) − | 2.0798 × 10−1 (2.78 × 10−2) = | 1.8629 × 10−1 (1.39 × 10−2) − | 2.0854 × 10−1 (1.79 × 10−2) |
LIRCMOP5 | 2.7174 × 10−2 (6.09 × 10−2) − | 2.8432 × 10−1 (5.36 × 10−3) = | 2.0243 × 10−2 (6.18 × 10−2) − | 2.8561 × 10−1 (1.65 × 10−3) |
LIRCMOP6 | 2.1490 × 10−2 (4.01 × 10−2) − | 1.9216 × 10−1 (7.19 × 10−4) = | 1.5007 × 10−2 (3.17 × 10−2) − | 1.9183 × 10−1 (1.32 × 10−3) |
LIRCMOP7 | 1.9382 × 10−2 (6.00 × 10−2) − | 1.9844 × 10−1 (6.95 × 10−2) = | 7.4225 × 10−3 (4.07 × 10−2) − | 2.1168 × 10−1 (5.83 × 10−2) |
LIRCMOP8 | 0.0000 × 10+0 (0.00 × 10+0) − | 1.0965 × 10−1 (1.04 × 10−1) = | 6.0278 × 10−3 (3.30 × 10−2) − | 1.3642 × 10−1 (9.86 × 10−2) |
LIRCMOP9 | 9.6729 × 10−2 (2.87 × 10−2) − | 2.8800 × 10−1 (5.17 × 10−2) = | 9.7909 × 10−2 (2.76 × 10−2) − | 2.5944 × 10−1 (7.32 × 10−2) |
LIRCMOP10 | 6.5082 × 10−2 (2.83 × 10−2) − | 2.4906 × 10−1 (1.24 × 10−1) = | 5.9470 × 10−2 (1.97 × 10−2) − | 1.9340 × 10−1 (1.40 × 10−1) |
LIRCMOP11 | 1.3892 × 10−1 (4.07 × 10−2) − | 3.7183 × 10−1 (1.11 × 10−1) + | 1.6531 × 10−1 (3.18 × 10−2) − | 2.7786 × 10−1 (8.43 × 10−2) |
LIRCMOP12 | 2.3159 × 10−1 (7.98 × 10−2) − | 3.7287 × 10−1 (7.52 × 10−2) = | 1.9288 × 10−1 (7.46 × 10−2) − | 4.0423 × 10−1 (7.72 × 10−2) |
LIRCMOP13 | 3.3380 × 10−4 (1.46 × 10−4) + | 1.0577 × 10−2 (5.75 × 10−2) + | 5.3026 × 10−5 (9.73 × 10−5) + | 0.0000 × 10+0 (0.00 × 10+0) |
LIRCMOP14 | 7.9656 × 10−4 (2.55 × 10−4) + | 2.6131 × 10−2 (7.93 × 10−2) + | 1.8352 × 10−4 (2.95 × 10−4) + | 1.7358 × 10−5 (5.94 × 10−5) |
MW1 | Infeasible | Infeasible | Infeasible | 1.4722 × 10−1 (1.30 × 10−1) |
MW2 | 3.0609 × 10−1 (1.50 × 10−1) − | 3.0403 × 10−1 (7.47 × 10−2) − | 3.2396 × 10−1 (1.49 × 10−1) = | 3.8124 × 10−1 (1.29 × 10−1) |
MW3 | 2.0367 × 10−1 (2.06 × 10−1) − | 4.8768 × 10−1 (1.93 × 10−2) − | 2.2971 × 10−1 (1.71 × 10−1) − | 5.0566 × 10−1 (1.43 × 10−2) |
MW4 | Infeasible | 3.1094 × 10−1(0.00 × 10+0) = | Infeasible | 4.3348 × 10−1 (5.15 × 10−2) |
MW5 | Infeasible | 9.8648 × 10−3 (1.56 × 10−2) − | Infeasible | 1.1962 × 10−1 (7.62 × 10−2) |
MW6 | 2.2796 × 10−1 (5.47 × 10−2) − | 2.2878 × 10−1 (4.78 × 10−2) − | 2.0492 × 10−1 (6.42 × 10−2) − | 2.9226 × 10−1 (1.89 × 10−2) |
MW7 | 3.7160 × 10−1 (6.69 × 10−2) = | 4.0684 × 10−1 (1.45 × 10−3)+ | 3.8315 × 10−1 (5.65 × 10−2) − | 4.0304 × 10−1 (1.59 × 10−3) |
MW8 | 4.6429 × 10−1 (7.49 × 10−2) = | 4.0464 × 10−1 (4.27 × 10−2) − | 4.3323 × 10−1 (6.61 × 10−2) − | 4.8872 × 10−1 (2.38 × 10−2) |
MW9 | Infeasible | 2.8487 × 10−1 (1.60 × 10−2) = | Infeasible | 1.3763 × 10−1 (1.70 × 10−1) |
MW10 | 2.0985 × 10−1 (1.03 × 10−1) = | 8.2714 × 10−2 (1.02 × 10−2) − | 1.9919 × 10−1 (1.03 × 10−1) = | 2.2475 × 10−1 (9.60 × 10−2) |
MW11 | 2.4123 × 10−1 (3.84 × 10−2) − | 4.4117 × 10−1 (1.10 × 10−3)+ | 2.3732 × 10−1 (3.80 × 10−2) − | 4.3830 × 10−1 (1.25 × 10−3) |
MW12 | Infeasible | 1.3176 × 10−1 (1.04 × 10−1) = | Infeasible | 1.4185 × 10−1 (2.45 × 10−1) |
MW13 | 1.9544 × 10−1 (7.54 × 10−2) − | 7.7809 × 10−2 (9.91 × 10−2) − | 1.9767 × 10−1 (7.52 × 10−2) − | 2.9592 × 10−1 (8.84 × 10−2) |
MW14 | 4.7499 × 10−2 (2.15 × 10−2) − | 2.3013 × 10−2 (1.94 × 10−3) − | 3.3691 × 10−2 (1.39 × 10−2) − | 1.1746 × 10−1 (4.12 × 10−2) |
+/−/= | 2/23/3 | 5/13/14 | 2/24/2 |
Algorithm | Ranking |
---|---|
IDCNSGA-III | 3.8088 |
DSPCMDE | 4.4412 |
CMOEA-MS | 5.7794 |
BiCo | 6.0588 |
AGEMOEA-II | 6.3676 |
NSGA-II | 6.5441 |
Problem | TriP | DPPPS | CAEAD | IDCNSGA-III |
---|---|---|---|---|
RWMOP1 | 5.9308 × 10−1 (2.84 × 10−3) − | 5.9372 × 10−1 (2.23 × 10−3) − | 6.0515 × 10−1 (1.38 × 10−3) = | 6.0529 × 10−1 (8.02 × 10−4) |
RWMOP2 | 2.5903 × 10−1 (1.18 × 10−1) − | 2.3218 × 10−1 (9.43 × 10−2) − | 3.6619 × 10−1 (1.41 × 10−2) − | 3.9251 × 10−1 (7.04 × 10−4) |
RWMOP3 | 9.0001 × 10−1 (5.74 × 10−4) + | 8.9997 × 10−1 (5.96 × 10−4) + | 8.9798 × 10−1 (1.18 × 10−3) + | 8.9312 × 10−1 (1.65 × 10−3) |
RWMOP4 | 8.5421 × 10−1 (3.76 × 10−3) − | 8.5443 × 10−1 (3.07 × 10−3) − | 8.5315 × 10−1 (3.39 × 10−3) − | 8.5873 × 10−1 (1.41 × 10−3) |
RWMOP5 | 4.3215 × 10−1 (8.51 × 10−4) + | 4.3186 × 10−1 (1.17 × 10−3) + | 4.3341 × 10−1 (3.27 × 10−4) + | 3.4774 × 10−1 (5.79 × 10−2) |
RWMOP6 | 2.7559 × 10−1 (3.70 × 10−4) − | 2.7536 × 10−1 (7.04 × 10−4) − | 2.7448 × 10−1 (1.06 × 10−3) − | 2.7585 × 10−1 (2.07 × 10−3) |
RWMOP7 | 4.8340 × 10−1 (1.53 × 10−4) = | 4.8319 × 10−1 (2.54 × 10−4) − | 4.8398 × 10−1 (2.39 × 10−4) + | 4.8340 × 10−1 (6.23 × 10−4) |
RWMOP8 | 2.5752 × 10−2 (1.95 × 10−4) + | 2.5758 × 10−2 (1.98 × 10−4) + | 2.5913 × 10−2 (8.24 × 10−5) + | 2.5611 × 10−2 (2.82 × 10−4) |
RWMOP9 | 3.9371 × 10−1 (9.07 × 10−3) − | 3.9425 × 10−1 (7.85 × 10−3) − | 4.0953 × 10−1 (1.10 × 10−4) − | 4.0970 × 10−1 (1.01 × 10−4) |
RWMOP10 | 8.4260 × 10−1 (2.08 × 10−3) + | 8.4157 × 10−1 (1.52 × 10−3) + | 8.4204 × 10−1 (1.91 × 10−3) + | 8.1216 × 10−1 (8.48 × 10−3) |
RWMOP11 | 9.3081 × 10−2 (1.38 × 10−3) − | 9.2283 × 10−2 (1.15 × 10−3) − | 9.1840 × 10−2 (2.02 × 10−3) − | 9.4492 × 10−2 (8.63 × 10−4) |
RWMOP12 | 5.5804 × 10−1 (1.29 × 10−3) = | 5.5752 × 10−1 (1.85 × 10−3) − | 5.5619 × 10−1 (1.81 × 10−3) − | 5.5854 × 10−1 (1.34 × 10−3) |
RWMOP13 | 8.7506 × 10−2 (2.66 × 10−4) − | 8.7487 × 10−2 (2.89 × 10−4) − | 8.6793 × 10−2 (4.26 × 10−4) − | 8.7645 × 10−2 (1.40 × 10−4) |
RWMOP14 | 6.1141 × 10−1 (4.00 × 10−3) − | 6.1217 × 10−1 (3.50 × 10−3) − | 6.1320 × 10−1 (1.23 × 10−3) − | 6.1683 × 10−1 (4.09 × 10−4) |
RWMOP15 | 5.3143 × 10−1 (3.35 × 10−3) − | 5.2874 × 10−1 (5.41 × 10−3) − | 5.4037 × 10−1 (6.61 × 10−4) = | 5.4029 × 10−1 (5.74 × 10−4) |
RWMOP16 | 7.6265 × 10−1 (2.48 × 10−4) − | 7.6259 × 10−1 (2.49 × 10−4) − | 7.6250 × 10−1 (2.70 × 10−4) − | 7.6316 × 10−1 (2.82 × 10−5) |
RWMOP17 | 2.2013 × 10−1 (4.95 × 10−2) = | 2.2862 × 10−1 (3.68 × 10−2) − | 2.4094 × 10−1 (2.48 × 10−2) = | 2.4837 × 10−1 (2.04 × 10−2) |
RWMOP18 | 4.0510 × 10−2 (6.33 × 10−6) + | 4.0509 × 10−2 (6.77 × 10−6) + | 4.0505 × 10−2 (4.03 × 10−6) + | 4.0481 × 10−2 (2.63 × 10−5) |
RWMOP19 | 2.5834 × 10−1 (2.00 × 10−2) − | 2.5561 × 10−1 (2.33 × 10−2) − | 3.4259 × 10−1 (7.91 × 10−3) − | 3.5698 × 10−1 (3.83 × 10−3) |
RWMOP21 | 3.1570 × 10−2 (8.25 × 10−5) + | 3.1621 × 10−2 (5.17 × 10−5) + | 3.1760 × 10−2 (8.09 × 10−7) + | 3.1508 × 10−2 (4.65 × 10−4) |
RWMOP22 | infeasible | infeasible | 8.0800 × 10−1 (2.15 × 10−1) | infeasible |
RWMOP23 | 9.9856 × 10−1 (4.53 × 10−16) = | 9.9856 × 10−1 (0.00 × 10+0) = | 1.0187 × 10+0 (1.15 × 10−1) = | 1.0510 × 10+0 (1.21 × 10−1) |
RWMOP25 | 2.4150 × 10−1 (2.47 × 10−5) + | 2.4149 × 10−1 (2.37 × 10−5) + | 2.4149 × 10−1 (1.95 × 10−5) + | 2.4136 × 10−1 (8.03 × 10−5) |
RWMOP26 | 1.2680 × 10−1 (1.78 × 10−2) − | 1.2454 × 10−1 (1.62 × 10−2) − | 1.4953 × 10−1 (9.48 × 10−3) = | 1.4490 × 10−1 (6.30 × 10−3) |
RWMOP27 | 1.3860 × 10+8 (4.40 × 10+8) − | 1.2481 × 10+8 (4.08 × 10+8) = | 1.2597 × 10+9 (5.96 × 10+9) = | 4.0381 × 10+8 (1.79 × 10+9) |
RWMOP28 | infeasible | infeasible | infeasible | 3.4356 × 10−2 (1.09 × 10−2) |
RWMOP29 | 7.5012 × 10−1 (1.71 × 10−2) + | 7.4954 × 10−1 (2.02 × 10−2) − | 7.7187 × 10−1 (7.56 × 10−3) = | 7.4998 × 10−1 (6.20 × 10−2) |
TREE1 | 7.1186 × 10−1 (9.69 × 10−3) − | 7.4386 × 10−1 (8.51 × 10−3) + | 7.3571 × 10−1 (6.27 × 10−3) = | 7.2952 × 10−1 (1.64 × 10−2) |
TREE2 | 7.5052 × 10−1 (7.85 × 10−3) − | 7.7012 × 10−1 (7.14 × 10−3) + | 7.6484 × 10−1 (4.35 × 10−3) = | 7.6076 × 10−1 (1.01 × 10−2) |
TREE3 | 7.2121 × 10−1 (4.69 × 10−2) − | 7.5576 × 10−1 (1.42 × 10−2) − | 7.1745 × 10−1 (1.63 × 10−2) − | 7.9388 × 10−1 (9.08 × 10−3) |
TREE4 | 5.8872 × 10−1 (6.74 × 10−2) − | 6.0943 × 10−1 (2.95 × 10−2) − | 4.7716 × 10−1 (5.14 × 10−2) − | 7.0300 × 10−1 (3.38 × 10−2) |
TREE5 | 7.2876 × 10−1 (5.40 × 10−2) − | 7.2453 × 10−1 (2.13 × 10−2) − | 6.3036 × 10−1 (3.01 × 10−2) − | 7.8099 × 10−1 (2.67 × 10−2) |
LIRCMOP1 | 1.0835 × 10−1 (1.46 × 10−2) − | 1.0634 × 10−1 (1.33 × 10−2) − | 1.0751 × 10−1 (2.72 × 10−2) − | 1.2444 × 10−1 (2.56 × 10−2) |
LIRCMOP2 | 2.2951 × 10−1 (1.86 × 10−2) − | 2.2789 × 10−1 (1.92 × 10−2) − | 2.3089 × 10−1 (3.32 × 10−2) − | 2.4912 × 10−1 (2.00 × 10−2) |
LIRCMOP3 | 9.9237 × 10−2 (1.21 × 10−2) − | 1.0128 × 10−1 (1.02 × 10−2) − | 8.9164 × 10−2 (1.50 × 10−2) − | 1.0955 × 10−1 (2.05 × 10−2) |
LIRCMOP4 | 1.8951 × 10−1 (1.70 × 10−2) − | 1.8867 × 10−1 (1.62 × 10−2) − | 1.8284 × 10−1 (3.20 × 10−2) − | 2.0854 × 10−1 (1.79 × 10−2) |
LIRCMOP5 | 2.8560 × 10−1 (6.95 × 10−3) − | 2.7503 × 10−1 (1.98 × 10−2) − | 2.6212 × 10−1 (5.20 × 10−2) − | 2.8561 × 10−1 (1.65 × 10−3) |
LIRCMOP6 | 1.8882 × 10−1 (2.17 × 10−2) − | 1.8259 × 10−1 (1.95 × 10−2) = | 1.6545 × 10−1 (2.05 × 10−2) − | 1.9183 × 10−1 (1.32 × 10−3) |
LIRCMOP7 | 2.1642 × 10−1 (1.13 × 10−2) + | 1.4293 × 10−1 (7.74 × 10−2) − | 4.2776 × 10−3 (2.34 × 10−2) − | 2.1168 × 10−1 (5.83 × 10−2) |
LIRCMOP8 | 2.0367 × 10−1 (1.07 × 10−2) = | 4.6456 × 10−2 (7.33 × 10−2) − | 0.0000 × 10+0 (0.00 × 10+0) − | 1.3642 × 10−1 (9.86 × 10−2) |
LIRCMOP9 | 1.6996 × 10−1 (7.00 × 10−2) − | 1.7749 × 10−1 (1.06 × 10−1) − | 2.5966 × 10−1 (6.40 × 10−2) = | 2.5944 × 10−1 (7.32 × 10−2) |
LIRCMOP10 | 8.4000 × 10−2 (2.52 × 10−2) − | 6.2384 × 10−2 (2.41 × 10−2) − | 1.3664 × 10−1 (7.23 × 10−2) = | 1.9340 × 10−1 (1.40 × 10−1) |
LIRCMOP11 | 1.9344 × 10−1 (3.97 × 10−2) − | 1.5943 × 10−1 (6.25 × 10−2) − | 2.0644 × 10−1 (1.02 × 10−1) − | 2.7786 × 10−1 (8.43 × 10−2) |
LIRCMOP12 | 3.4834 × 10−1 (5.79 × 10−2) − | 3.2052 × 10−1 (6.36 × 10−2) − | 2.7427 × 10−1 (6.43 × 10−2) − | 4.0423 × 10−1 (7.72 × 10−2) |
LIRCMOP13 | 5.4827 × 10−2 (7.41 × 10−2) + | 2.6023 × 10−2 (5.53 × 10−2) + | 3.5855 × 10−5 (8.81 × 10−5) + | 0.0000 × 10+0 (0.00 × 10+0) |
LIRCMOP14 | 5.1190 × 10−2 (7.33 × 10−2) + | 2.2962 × 10−2 (4.48 × 10−2) + | 2.2361 × 10−4 (2.69 × 10−4) + | 1.7358 × 10−5 (5.94 × 10−5) |
MW1 | infeasible | infeasible | infeasible | 1.4722 × 10−1 (1.30 × 10−1) |
MW2 | 4.2862 × 10−1 (5.85 × 10−2) = | 4.3252 × 10−1 (5.50 × 10−2) = | infeasible | 3.8124 × 10−1 (1.29 × 10−1) |
MW3 | 4.2942 × 10−1 (1.84 × 10−2) − | 4.2029 × 10−1 (2.10 × 10−2) − | 1.3968 × 10−1 (0.00 × 10+0) = | 5.0566 × 10−1 (1.43 × 10−2) |
MW4 | infeasible | infeasible | infeasible | 4.3348 × 10−1 (5.15 × 10−2) |
MW5 | infeasible | infeasible | infeasible | 1.1962 × 10−1 (7.62 × 10−2) |
MW6 | 3.1386 × 10−1 (1.20 × 10−2) + | 3.1343 × 10−1 (1.26 × 10−2) + | 8.1863 × 10−2 (4.78 × 10−2) − | 2.9226 × 10−1 (1.89 × 10−2) |
MW7 | 4.0831 × 10−1 (9.17 × 10−4) + | 4.0742 × 10−1 (1.21 × 10−3) + | 4.0216 × 10−1 (4.37 × 10−3) = | 4.0304 × 10−1 (1.59 × 10−3) |
MW8 | 5.2629 × 10−1 (9.09 × 10−3) + | 5.2888 × 10−1 (9.52 × 10−3) + | 3.1607 × 10−1 (7.02 × 10−2) − | 4.8872 × 10−1 (2.38 × 10−2) |
MW9 | infeasible | infeasible | infeasible | 1.3763 × 10−1 (1.70 × 10−1) |
MW10 | 1.6516 × 10−1 (6.00 × 10−2) − | 1.4508 × 10−1 (5.25 × 10−2) − | infeasible | 2.2475 × 10−1 (9.60 × 10−2) |
MW11 | 4.3417 × 10−1 (4.16 × 10−3) − | 4.0024 × 10−1 (4.32 × 10−2) − | infeasible | 4.3830 × 10−1 (1.25 × 10−3) |
MW12 | infeasible | infeasible | infeasible | 1.4185 × 10−1 (2.45 × 10−1) |
MW13 | 2.0227 × 10−1 (7.81 × 10−2) − | 2.1554 × 10−1 (7.95 × 10−2) − | 0.0000 × 10+0 (0.00 × 10+0) − | 2.9592 × 10−1 (8.84 × 10−2) |
MW14 | 1.7216 × 10−2 (2.06 × 10−3) − | 1.6526 × 10−2 (1.52 × 10−3) − | 1.3666 × 10−2 (9.90 × 10−4) − | 1.1746 × 10−1 (4.12 × 10−2) |
+/−/= | 14/36/7 | 14/37/5 | 13/33/5 |
Algorithm | Ranking |
---|---|
IDCNSGA-III ( = 0.005) | 2.6029 |
IDCNSGA-III ( = 0.02) | 3.0294 |
IDCNSGA-III ( = 0.01) | 3.0735 |
IDCNSGA-III ( = 0.015) | 3.1471 |
IDCNSGA-III ( = 0.001) | 3.1471 |
Algorithm | Ranking |
---|---|
IDCNSGA-III ( = 0.2) | 2.8382 |
IDCNSGA-III ( = 0.3) | 2.9118 |
IDCNSGA-III ( = 0.4) | 3.0147 |
IDCNSGA-III ( = 0.5) | 3.1471 |
IDCNSGA-III ( = 0.1) | 3.2059 |
Algorithm | Ranking |
---|---|
IDCNSGA-III ( = 0.8) | 2.6618 |
IDCNSGA-III ( = 0.5) | 2.8382 |
IDCNSGA-III ( = 0.9) | 3.0294 |
IDCNSGA-III ( = 0.6) | 3.1618 |
IDCNSGA-III ( = 0.7) | 3.3088 |
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
© 2024 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
Share and Cite
Zuo, M.; Xue, Y. Population Feasibility State Guided Autonomous Constrained Multi-Objective Evolutionary Optimization. Mathematics 2024, 12, 913. https://doi.org/10.3390/math12060913
Zuo M, Xue Y. Population Feasibility State Guided Autonomous Constrained Multi-Objective Evolutionary Optimization. Mathematics. 2024; 12(6):913. https://doi.org/10.3390/math12060913
Chicago/Turabian StyleZuo, Mingcheng, and Yuan Xue. 2024. "Population Feasibility State Guided Autonomous Constrained Multi-Objective Evolutionary Optimization" Mathematics 12, no. 6: 913. https://doi.org/10.3390/math12060913
APA StyleZuo, M., & Xue, Y. (2024). Population Feasibility State Guided Autonomous Constrained Multi-Objective Evolutionary Optimization. Mathematics, 12(6), 913. https://doi.org/10.3390/math12060913