設計一個能夠有效且快速求解高目標最佳化問題的演算法，已成為當領域的 重要課題之一。我們以 Deb 與 Jain [8] 所提出的 NSGA-III 為基礎演算法，從 親代選擇機制、環境選擇機制以及平行化，試圖改良此優秀演算法中尚為欠缺的 部分，經由實驗佐證，不管在演算法的求解能力以及執行速度上，都收到相當不 錯的成效。
在親代選擇機制，我們希望藉由 NSGA-III 的參考點概念來加強選擇親代時 的辨別能力，以期能夠辨認出族群中哪些區域需要優先進行改進，並透過鄰居選 取機制，增強演算法整體的求解效能。鄰居親代選擇機制將目標空間中相近的個 體交配，能夠有效地加強演算法的求解能力，我們在實驗中測試了各種鄰居大小 的設定，當此值設為族群大小的 10% 效果最好。
關於強化劣勢參考點的親代選擇機制，由於其選取的取向集中性強，因此不 適合族群尚未收歛完整時使用，我們的實驗結果顯示，此機制大約在演化代數經 過 70%~90% 時使用較為適當，並能有效加強演算法的求解能力，此機制與鄰 居親代選擇機制配合下，與原版的 NSGA-III 演算法相比，於總計 15 個測試問 題中，能有著 12 勝 3 和的顯著改進，成效相當優異。
在將 NSGA-III 平行化的機制中，我們嘗試將權重向量用各種不同的方式切 割給予子族群去各自演化，其中效能最好的分割方法為保留角的依序整批分配法，
搭配遷徙以及我們所提出的親代選擇機制，演算法的效能不管在執行速度或是求
74
解能力上，都勝過原版的 NSGA-III，在總計 15 個測試問題中，有著 11 勝 3 和
1 負的顯著改善。
雖然我們所提出的親代選擇機制以及平行化，求解能力比 NSGA-III 演算法 優異，但這些機制都需要設定一些參數，在我們的實驗中，有提供建議的參數調 整方向。我們未來會致力於提出一套自動調整參數的版本，以加強這些機制的實 用性，並嘗試求解各種具限制以及現實世界中的高目標最佳化問題，驗證我們的 機制在各種情況下都能有著優異的表現。
75
參考文獻
[1] H. Ishibuchi, N. Tsukamoto & Y. Nojima, “Evolutionary many-objective optimization: A short review,” Proceedings of IEEE Congress on Evolutionary
Computation, pp. 2419–2426, June. 2008.
[2] T. W. Simpson, W. Chen, J. K. Allen & F. Mistree, “Conceptual design of a family of products through the use of the robust concept exploration method,” in
6
thAIAA/USAF/NASA/ISSMO Symposium on Multidisciplinary Analysis and Optimization, vol. 2, pp. 1535–1545, 1996.
[3] H. Jain & K. Deb, “An improved adaptive approach for elitist nondominated sorting genetic algorithm for many-objective optimization,” Lecture Notes in
Computer Science 7811: Evolutionary Multi-Criterion Optimization – EMO, pp.
307–321, 2013.
[4] K. Deb, A. Pratap, S. Agarwal & T. Meyarivan, “A fast and elitist multiobjective genetic algorithm: NSGA-II,” IEEE Transactions on Evolutionary Computation, vol. 6, issue 2, pp. 182–197, 2002.
[5] Q. Zhang & H. Li, “MOEA/D: A multiobjective evolutionary algorithm based on decomposition,” IEEE Transactions on Evolutionary Computation, vol. 11(6), pp. 712–731, 2007.
[6] K. Deb & R. B. Agrawal, “Simulated binary crossover for continuous search space,” Complex Systems, vol. 9, pp. 115–148, Apr. 1995.
[7] E. Cantú-Paz, “A survey of parallel genetic algorithms,” Calculateurs Paralleles, vol. 10, pp. 141–171, 1998.
[8] K. Deb & H. Jain, “Handling many-objective problems using an improved NSGA-II procedure,” Proceedings of IEEE World Congress on Computational
Intelligence, 2012
76
[9] H. Sato, H. E. Aguirre & K. Tanaka, “Controlling dominance area of solutions and its impact on the performance of MOEAs,” Lecture Notes in Computer
Science 4403: Evolutionary Multi-Criterion Optimization – EMO, pp. 55–20,
2007.[10] N. Drechsler, R. Drechsler & B. Becker, “Multi-objective optimization based on relation favour,” Lecture Notes in Computer Science 1993: Evolutionary
Multi-Criterion Optimization – EMO, pp. 154–166, 2001.
[11] E. Zitzler & S. Künzli, “Indicator-based selection in multiobjective search,”
Lecture Notes in Computer Science 3242: Parallel Problem Solving from Nature – PPSN VIII, pp. 832–842, 2004.
[12] N. Beume, B. Naujoks & M. Emmerich, “SMS-EMOA: multiobjective selection based on dominated hypervolume,” European Journal of Operational Research, pp. 1653–1669, 2007.
[13] K. Deb & J. Sundar, “Reference point based multi-objective optimization using evolutionary algorithms,” Proceedings of 2006 Genetic and Evolutionary
Computation Conference, pp. 635–642, 2006.
[14] H. Ishibuchi, Y. Sakane, N. Tsukamoto & Y. Nojima, “Evolutionary many-objective optimization by NSGA-II and MOEA/D with large populations,”
Proceedings of 2009 IEEE International Conference on Systems, Man, and Cybernetics, pp. 1820–1825, 2009
[15] K. Deb & K. Saxena, “Searching for Pareto-optimal solutions through dimensionality reduction for certain large-dimensional multi-objective optimization problems,” Proceedings of IEEE Congress on Evolutionary
Computation, pp. 3353–3360, 2006.
[16] D. Brockhoff & E. Zitzler, “Are all objectives necessary? On dimensionality reduction in evolutionary multiobjective optimization,” Lecture Notes in
Computer Science 4193: Parallel Problem Solving from Nature – PPSN IX, pp.
533
–
542, 2006.77
[17] A. Sülflow, N. Drechsler & R. Drechsler, “Robust multi-objective optimization in high dimensional spaces,” Lecture Notes in Computer Science 4403:
Evolutionary Multi-Criterion Optimization – EMO, pp. 715–726, 2007.
[18] D. Corne & J. Knowles, “Techniques for highly multiobjective optimization:
Some non-dominated points are better than others,” Proceedings of 2007 Genetic
and Evolutionary Computation Conference, pp. 773 –
780, 2007.[19] M. Garza-Fabre, G. T. Pulido & C. A. C. Coello, “Ranking methods for many-objective optimization,” MICAI 2009: Advances in Artificial Intelligence, pp. 633–645, 2009.
[20] K. Deb & S. Tiwari, “Omni-optimizer: A generic evolutionary algorithm for single and multi-objective optimization,” European Journal of Operational
Research, vol. 185, issue 3, pp. 1062–1087, 2006.
[21] K. Köppen & K. Yoshida, “Substitute distance assignments in NSGA-II for handling many-objective optimization problems,” Lecture Notes in Computer
Science 4403: Evolutionary Multi-Criterion Optimization – EMO, pp. 727–741,
2007.[22] T. Wagner, N. Beume & B. Naujoks, “Pareto-, aggregation-, and indicator-based methods in many-objective optimization,” Lecture Notes in Computer Science
4403: Evolutionary Multi-Criterion Optimization – EMO, pp. 742–756, 2007.
[23] E. J. Hughes, “MSOPS-II: A general-purpose many-objective optimizer,”
Proceedings of IEEE Congress on Evolutionary Computation, pp. 3944 –3951, 2007.
[24] L. Thiele, K. Miettinen, P. J. Korhonen & J. Molina, “A preference-based interactive evolutionary algorithm for multiobjective optimization,” Helsinki
School of Economics, Working Paper, 2007.
[25] H. Ishibuchi, N. Akedo & Y. Nojima, “Recombination of similar parents in SMS-EMOA on many-objective 0/1 knapsack problems,” PPSN’ 12 Proceedings
of the 12
thinternational conference on Parallel Problem Solving from Nature,
vol. Part II, pp. 132–142, 2012.78
[26] M. Garza-Fabre, G. T. Pulido & C. A. Coello Coello, “Two novel approaches for many-objective optimization,” Proceedings of IEEE Congress on Evolutionary
Computation, 2010.
[27] K. Deb, L. Thiele, M. Laumanns & E. Zitzler, “Scalable test problems for evolutionary multi-objective optimization,” A. Abraham, L. Jain & R. Goldberg