

















收稿日期:2022-02-26;修回日期:2022-04-24" 基金項目:教育部人文社會科學研究青年基金項目(21YJC630087);上海市哲學社會科學規劃科課題資助項目(2019BGL014);上海理工大學科技發展資助項目(2020KJFZ040)
作者簡介:張雨婷(1997-),女,安徽宿州人,碩士研究生,主要研究方向為智能優化和系統工程;劉勇(1982-),男(通信作者),江蘇金湖人,副教授,碩導,博士(后),主要研究方向為智能優化、服務網絡設計與優化和系統工程(liuyong.seu@163.com).
摘 要:
針對SBO(school based optimization)算法搜索性能差、易陷入局部最優等缺陷,提出融入教育心理學的SBO算法(SBO based on educational psychology,SBO-EP)。在教階段,引入最近發展區理論,對學生進行分組動態教學,提高算法的探索能力;引用成就動機理論加入自學階段,針對每組學生的成就動機設計動態自學方式,提高算法的開發能力;在每輪學習過程結束后參考同伴效應設置班級重組操作,增加解的多樣性。采用40個CEC2021測試函數和20個其他類型測試函數進行數值實驗,并將SBO-EP算法與蟻群優化算法、基于球形矢量的粒子群優化算法、阿基米德優化算法、灰狼優化算法、教與學優化算法、融合認知心理學的教與學優化算法、學生心理學優化算法進行對比分析。結果表明,SBO-EP算法在收斂速度、尋優精度及穩定性上優勢明顯。最后,對三種策略的組合進行對比實驗,驗證了改進策略的有效性。
關鍵詞:SBO算法; 最近發展區理論; 成就動機理論; 同伴效應
中圖分類號:TP18"" 文獻標志碼:A"" 文章編號:1001-3695(2022)09-011-2631-09
doi: 10.19734/j.issn.1001-3695.2022.02.0080
SBO algorithm integrated into educational psychology
Zhang Yuting, Liu Yong
(Business School, University of Shanghai for Science and Technology, Shanghai 200093, China)
Abstract:Aiming at the shortcomings of SBO algorithm,such as poor search performance and local optimization,this paper proposed SBO algorithm integrated with educational psychology. The teaching stage used the theory of zone of proximal deve-lopment to carry out dynamic teaching for students in groups to improve the exploration ability of algorithms. It introduced the theory of achievement motivation into the self-study stage,and designed dynamic self-study methods according to the achievement motivation of each group of students to improve the development ability of algorithms. After each round of learning process,refer to the peer effect theory,it set up the class reorganization operation to increase the diversity of solutions. This paper used 40 CEC2021 test functions and 20 other types of test functions for numerical experiments,and compared SBO-EP algorithm with ant colony optimization algorithm,spherical vector-based particle swarm optimization,Archimedes optimization algorithm,gray wolf optimization algorithm,teaching and learning algorithm,cognitive psychology teaching-learning-based optimization and student psychology based optimization algorithm. The results show that SBO-EP algorithm has obvious advantages in convergence speed,optimization accuracy and stability. Finally,it conducts a comparative experiment on the combination of the three strategies,verifies the effectiveness of the improved strategy.
Key words:SBO algorithm; zone of proximal development theory; achievement motivation theory; peer effect
0 引言
SBO算法是Farshchin等人[1]于2018年提出的一種新型元啟發式算法。目前存在的元啟發式算法主要包括受生物群體社會性或自然現象規律啟發而開發的優化算法,如遺傳算法(genetic algorithm,GA)[2]、模擬退火算法(simulated annealing,SA)[3]等。……