《计算机应用》唯一官方网站 ›› 2022, Vol. 42 ›› Issue (8): 2511-2518.DOI: 10.11772/j.issn.1001-9081.2021061079
• 先进计算 • 上一篇
收稿日期:
2021-06-25
修回日期:
2022-02-25
接受日期:
2022-03-04
发布日期:
2022-03-17
出版日期:
2022-08-10
通讯作者:
李和成
作者简介:
沈瑜(1995—),女,重庆人,硕士研究生,主要研究方向:最优化;基金资助:
Yu SHEN, Hecheng LI(), Lijuan CHEN
Received:
2021-06-25
Revised:
2022-02-25
Accepted:
2022-03-04
Online:
2022-03-17
Published:
2022-08-10
Contact:
Hecheng LI
About author:
SHEN Yu, born in 1995, M. S. candidate. Her research interests include optimization.Supported by:
摘要:
双层规划涉及上层和下层两个最优化问题,上层规划问题的约束域由下层规划问题隐式确定,双层优化以上层目标为主,而下层目标在下层变量方面必须达到最优。双层规划问题的递阶结构使其具有很高的计算复杂度,特别是频繁计算下层问题会累计很大的计算量。为了有效求解这类问题,提出一种基于近似技术的进化算法。首先,采取多种群协同进化,分别利用交叉和变异算子平衡算法的开采和勘探能力;其次,基于灵敏度分析理论,设计了新个体的近似评价方式以减少算法的下层求解次数。一个算例的近似效果演示结果表明,由近似技术得到的近似后代个体与精确后代个体的位置大部分是重合的。除此之外,在10个常用算例上的结果显示,所提算法比多值映射算法获得了更好的最优解;并且根据CPU时间比较,说明近似技术有效地提高了找到最优解的速度,减少了运行时间,验证了所提算法采取的近似技术的有效性。
中图分类号:
沈瑜, 李和成, 陈黎娟. 基于近似技术的双层规划进化算法[J]. 计算机应用, 2022, 42(8): 2511-2518.
Yu SHEN, Hecheng LI, Lijuan CHEN. Evolutionary algorithm based on approximation technique for solving bilevel programming problems[J]. Journal of Computer Applications, 2022, 42(8): 2511-2518.
父代个体 | 近似后代个体 | 精确后代个体 |
---|---|---|
(0.00,2.00),(1.00,4.00) | (0.70,3.40) | (0.70,3.40) |
(0.10,2.20),(0.80,3.60) | (0.24,2.48) | (0.24,2.48) |
(0.40,2.80),(0.90,3.80) | (0.55,3.10) | (0.55,3.10) |
(1.20,3.90),(1.60,3.70) | (1.40,3.80) | (1.40,3.80) |
(1.80,3.60),(2.40,3.30) | (2.20,3.40) | (2.20,3.40) |
(2.00,3.50),(2.60,3.20) | (2.40,3.30) | (2.40,3.30) |
(3.00,3.00),(3.50,2.50) | (3.20,2.80) | (3.20,2.80) |
(3.60,2.40),(4.10,1.90) | (3.90,2.10) | (3.90,2.10) |
(4.50,1.50),(4.90,1.10) | (4.70,1.30) | (4.70,1.30) |
(1.75,3.63),(3.25,2.75) | (2.50,3.19) | (2.50,3.25) |
表1 近似下层解和精确下层解
Tab. 1 Approximate and precise follower solutions
父代个体 | 近似后代个体 | 精确后代个体 |
---|---|---|
(0.00,2.00),(1.00,4.00) | (0.70,3.40) | (0.70,3.40) |
(0.10,2.20),(0.80,3.60) | (0.24,2.48) | (0.24,2.48) |
(0.40,2.80),(0.90,3.80) | (0.55,3.10) | (0.55,3.10) |
(1.20,3.90),(1.60,3.70) | (1.40,3.80) | (1.40,3.80) |
(1.80,3.60),(2.40,3.30) | (2.20,3.40) | (2.20,3.40) |
(2.00,3.50),(2.60,3.20) | (2.40,3.30) | (2.40,3.30) |
(3.00,3.00),(3.50,2.50) | (3.20,2.80) | (3.20,2.80) |
(3.60,2.40),(4.10,1.90) | (3.90,2.10) | (3.90,2.10) |
(4.50,1.50),(4.90,1.10) | (4.70,1.30) | (4.70,1.30) |
(1.75,3.63),(3.25,2.75) | (2.50,3.19) | (2.50,3.25) |
4.80 | 4.80 | 0.00 | 0.048 |
2.96 | 2.96 | 0.00 | 0.030 |
4.20 | 4.20 | 0.00 | 0.042 |
6.60 | 6.60 | 0.00 | 0.066 |
7.80 | 7.80 | 0.00 | 0.078 |
8.10 | 8.10 | 0.00 | 0.081 |
9.20 | 9.20 | 0.00 | 0.092 |
9.90 | 9.90 | 0.00 | 0.099 |
9.40 | 9.40 | 0.00 | 0.094 |
8.19 | 8.25 | 0.06 | 0.082 |
表2 目标函数值的误差
Tab. 2 Errors of objective function values
4.80 | 4.80 | 0.00 | 0.048 |
2.96 | 2.96 | 0.00 | 0.030 |
4.20 | 4.20 | 0.00 | 0.042 |
6.60 | 6.60 | 0.00 | 0.066 |
7.80 | 7.80 | 0.00 | 0.078 |
8.10 | 8.10 | 0.00 | 0.081 |
9.20 | 9.20 | 0.00 | 0.092 |
9.90 | 9.90 | 0.00 | 0.099 |
9.40 | 9.40 | 0.00 | 0.094 |
8.19 | 8.25 | 0.06 | 0.082 |
参数 | 含义 | 取值 |
---|---|---|
种群规模 种群聚类组数 | 50 5 | |
好个体数 | 5 | |
交叉概率 | 0.8 | |
变异概率 | 0.2 | |
拟合精度系数 | 1/100 | |
Max_gen | 运行的最大代数 | 20 |
表3 相关参数及取值
Tab. 3 Related parameter and their values
参数 | 含义 | 取值 |
---|---|---|
种群规模 种群聚类组数 | 50 5 | |
好个体数 | 5 | |
交叉概率 | 0.8 | |
变异概率 | 0.2 | |
拟合精度系数 | 1/100 | |
Max_gen | 运行的最大代数 | 20 |
算法 | 上层目标函数 | 算例1 | 算例2 | 算例3 | 算例4 | 算例5 | 算例6 | 算例7 | 算例8 | 算例9 | 算例10 | 算例11 |
---|---|---|---|---|---|---|---|---|---|---|---|---|
本文算法 | 49 | -1 011.666 7 | -22 | -20 | -243.5 | 225 | 0 | -18.679 9 | -29.2 | 0 | 748.89 | |
49 | -1 011.666 7 | -22 | -20 | -243.5 | 225 | 0 | -18.679 9 | -29.2 | 0 | 748.89 | ||
49 | -1 011.666 7 | -22 | -20 | -243.5 | 225 | 0 | -18.679 9 | -29.2 | 0 | 748.89 | ||
文献[31-32,29,33]算法 | 49 | -1 011.666 7 | -22 | -20 | -243.5 | 225 | 0 | -18.678 7 | -29.2 | 0 | 741.56 |
表4 上层目标函数值的最好值、最差值和均值
Tab. 4 Best values, worst values and mean values of leader objective functions
算法 | 上层目标函数 | 算例1 | 算例2 | 算例3 | 算例4 | 算例5 | 算例6 | 算例7 | 算例8 | 算例9 | 算例10 | 算例11 |
---|---|---|---|---|---|---|---|---|---|---|---|---|
本文算法 | 49 | -1 011.666 7 | -22 | -20 | -243.5 | 225 | 0 | -18.679 9 | -29.2 | 0 | 748.89 | |
49 | -1 011.666 7 | -22 | -20 | -243.5 | 225 | 0 | -18.679 9 | -29.2 | 0 | 748.89 | ||
49 | -1 011.666 7 | -22 | -20 | -243.5 | 225 | 0 | -18.679 9 | -29.2 | 0 | 748.89 | ||
文献[31-32,29,33]算法 | 49 | -1 011.666 7 | -22 | -20 | -243.5 | 225 | 0 | -18.678 7 | -29.2 | 0 | 741.56 |
算法 | 算例1 | 算例2 | 算例3 | 算例4 | 算例5 | 算例6 | 算例7 | 算例8 | 算例9 | 算例10 |
---|---|---|---|---|---|---|---|---|---|---|
本文算法 | (3,4,2,0,3,0) | (0,1,0,1,0.00,75.00,21.67) | (2,2) | (8,6) | (3,8, 0.5,0) | (20,5, 10,5) | (0,30, -10,10) | (0,2,1.877 0,0.910 0) | (0.000 1,0.899 9,0,0.599 9,0.400 1) | (0,30, -10,10) |
文献[31-32,29] 算法 | (3,4,2,0,3,0) | (0,1,0,1,0.00,75.00,21.67) | (2,2) | (8,6) | (3,8, 0.5,0) | (20,5, 10,5) | (0,30, -10,10) | (0,2,1.876 8,0.907 6) | (0.000 1,0.899 9,0,0.599 9,0.400 1) | (0,30, -10,10) |
表5 本文算法和比较算法的最优解对比
Tab. 5 Best solutions comparison of the proposed algorithm and compared algorithms
算法 | 算例1 | 算例2 | 算例3 | 算例4 | 算例5 | 算例6 | 算例7 | 算例8 | 算例9 | 算例10 |
---|---|---|---|---|---|---|---|---|---|---|
本文算法 | (3,4,2,0,3,0) | (0,1,0,1,0.00,75.00,21.67) | (2,2) | (8,6) | (3,8, 0.5,0) | (20,5, 10,5) | (0,30, -10,10) | (0,2,1.877 0,0.910 0) | (0.000 1,0.899 9,0,0.599 9,0.400 1) | (0,30, -10,10) |
文献[31-32,29] 算法 | (3,4,2,0,3,0) | (0,1,0,1,0.00,75.00,21.67) | (2,2) | (8,6) | (3,8, 0.5,0) | (20,5, 10,5) | (0,30, -10,10) | (0,2,1.876 8,0.907 6) | (0.000 1,0.899 9,0,0.599 9,0.400 1) | (0,30, -10,10) |
算法 | 算例1 | 算例2 | 算例3 | 算例4 | 算例5 | 算例6 | 算例7 | 算例8 | 算例9 | 算例10 |
---|---|---|---|---|---|---|---|---|---|---|
本文算法 | -27 | -4 673.36 | 2 | -6 | -19.5 | 100 | 100 | -1.026 8 | 3.2 | 100 |
文献[31-32,29]算法 | -27 | -4 673.36 | 2 | -6 | -19.5 | 100 | 100 | -1.015 6 | 3.2 | 100 |
表6 最优解对应的下层目标函数值
Tab. 6 Follower objective function values corresponding to best solutions
算法 | 算例1 | 算例2 | 算例3 | 算例4 | 算例5 | 算例6 | 算例7 | 算例8 | 算例9 | 算例10 |
---|---|---|---|---|---|---|---|---|---|---|
本文算法 | -27 | -4 673.36 | 2 | -6 | -19.5 | 100 | 100 | -1.026 8 | 3.2 | 100 |
文献[31-32,29]算法 | -27 | -4 673.36 | 2 | -6 | -19.5 | 100 | 100 | -1.015 6 | 3.2 | 100 |
算例 | CPU/s | UACPU/s | 算例 | CPU/s | UACPU/s |
---|---|---|---|---|---|
1 | 0.4 | 0.7 | 7 | 0.5 | 0.9 |
2 | 0.3 | 0.8 | 8 | 0.5 | 0.9 |
3 | 0.5 | 0.9 | 9 | 0.4 | 0.9 |
4 | 0.3 | 0.7 | 10 | 0.6 | 1.0 |
5 | 0.5 | 0.9 | 11 | 200.0 | 310.0 |
6 | 0.4 | 0.8 |
表7 本文算法的CPU时间和UACPU时间比较
Tab. 7 Comparison of CPU time and UACPU time of the proposed algorithm
算例 | CPU/s | UACPU/s | 算例 | CPU/s | UACPU/s |
---|---|---|---|---|---|
1 | 0.4 | 0.7 | 7 | 0.5 | 0.9 |
2 | 0.3 | 0.8 | 8 | 0.5 | 0.9 |
3 | 0.5 | 0.9 | 9 | 0.4 | 0.9 |
4 | 0.3 | 0.7 | 10 | 0.6 | 1.0 |
5 | 0.5 | 0.9 | 11 | 200.0 | 310.0 |
6 | 0.4 | 0.8 |
1 | DEMPE S. Foundations of Bilevel Programming[M]. New York: Springer, 2002 :7-10. |
2 | DEMPE S. Annotated bibliography on bilevel programming and mathematical programs with equilibrium constraints[J]. Optimization, 2003, 52(3): 333-359. 10.1080/0233193031000149894 |
3 | BARD J F. Practical Bilevel Optimization: Algorithms and Applications[M]. New York: Springer, 1998 :8-20. 10.1007/978-1-4757-2836-1_7 |
4 | WHITTAKER G, FÄRE R, GROSSKOPF S, et al. Spatial targeting of agri-environmental policy using bilevel evolutionary optimization[J]. Omega, 2017, 66(Pt A): 15-27. 10.1016/j.omega.2016.01.007 |
5 | AN B, ORDÓÑEZ F, TAMBE M, et al. A deployed quantal response-based patrol planning system for the U.S. Coast Guard[J]. Interfaces, 2013, 43(5): 400-420. 10.1287/inte.2013.0700 |
6 | NASROLAHPOUR E, KAZEMPOUR J, ZAREIPOUR H, et al. A bilevel model for participation of a storage system in energy and reserve markets[J]. IEEE Transaction on Sustainable Energy, 2018, 9(2): 582-598. 10.1109/tste.2017.2749434 |
7 | 李丽滢,付寒梅. 考虑多种运输方式的整车物流服务供应链订单分配问题[J]. 计算机应用, 2019, 39(6): 1836-1841. 10.11772/j.issn.1001-9081.2018122461 |
LI L Y, FU H M. Order allocation problem of vehicle logistics service supply chain considering multiple modes of transportation[J]. Journal of Computer Applications, 2019, 39(6): 1836-1841. 10.11772/j.issn.1001-9081.2018122461 | |
8 | 尹胜男,李引珍,张长泽. 基于客流分配的高铁票价调整策略[J]. 计算机应用, 2020, 40(1): 278-283. |
YIN S N, LI Y Z, ZHANG C Z. High-speed railway fare adjustment strategy based on passenger flow assignment[J]. Journal of Computer Applications, 2020, 40(1): 278-283. | |
9 | SINHA A, MALO P, DED K. A review on bilevel optimization: from classical to evolutionary approaches and applications[J]. IEEE Transactions on Evolutionary Computation, 2018, 22(2): 276-295. 10.1109/tevc.2017.2712906 |
10 | JENSEN T V, KAZEMPOUR J, PINSON P. Cost-optimal ATCs in zonal electricity markets[J]. IEEE Transactions on Power Systems, 2018, 33(4): 3624-3633. 10.1109/tpwrs.2017.2786940 |
11 | DUPIN R, MICHIORRI A, KARINIOTAKIS G. Optimal dynamic line rating forecasts selection based on ampacity probabilistic forecasting and network operators’ risk aversion[J]. IEEE Transactions on Power Systems, 2019, 34(4): 2836-2845. 10.1109/tpwrs.2018.2889973 |
12 | SINHA A, SOUN T, DEB K. Using Karush-Kuhn-Tucker proximity measure for solving bilevel optimization problems[J]. Swarm and Evolutionary Computation, 2019, 44: 496-510. 10.1016/j.swevo.2018.06.004 |
13 | SINHA A, SOUN T, DEB K. Evolutionary bilevel optimisation using KKT proximity measure [C]// Proceedings of 2017 IEEE Congress on Evolutionary Computation. Piscataway: IEEE, 2017: 2412-2419. 10.1109/cec.2017.7969597 |
14 | LIU S N, WANG M Z, KONG N, et al. An enhanced branch-and-bound algorithm for bilevel integer linear programming[J]. European Journal of Operational Research, 2021, 291(1): 661-679. 10.1016/j.ejor.2020.10.002 |
15 | FRANK S, MEHLITZ P, PILECKA M. Optimality conditions for the simple convex bilevel programming problem in Banach spaces[J]. Optimization, 2017, 67(2): 237-268. 10.1080/02331934.2017.1394296 |
16 | LIU J E, HONG Y F, ZHENG Y. A new variant of penalty method for weak linear bilevel programming problems[J]. Wuhan University Journal of Natural Sciences, 2018, 23(4): 328-332. 10.1007/s11859-018-1330-1 |
17 | TUO Q, LAN H Y. New exact penalty function methods with ϵ-approximation and perturbation convergence for solving nonlinear bilevel programming problems[J]. Journal of Computational Analysis and Applications, 2019, 5(1): 449-458. |
18 | AGOR J, ÖZALTIN O Y. Feature selection for classification models via bilevel optimization[J]. Computers and Operations Research, 2019, 106: 156-168. 10.1016/j.cor.2018.05.005 |
19 | LI H C. A genetic algorithm using a finite search space for solving nonlinear/linear fractional bilevel programming problems[J]. Annals of Operations Research, 2015, 235(1): 543-558. 10.1007/s10479-015-1878-5 |
20 | LI H C, WANG Z C. An evolutionary algorithm using parameter space searching for interval linear fractional bilevel programming problems[J]. International Journal of Pattern Recognition and Artificial Intelligence, 2016, 30(4): No.1659011. 10.1142/s0218001416590114 |
21 | LI H C, FANG L. Co-evolutionary algorithm: an efficient approach for bilevel programming problems[J]. Engineering Optimization, 2014, 46(3): 361-376. 10.1080/0305215x.2013.772601 |
22 | ABO-ELNAGA Y, NASR S. Modified evolutionary algorithm and chaotic search for bilevel programming problems[J]. Symmetry, 2020, 12(5): No.767. 10.3390/sym12050767 |
23 | GOSHU N N, KASSA S M. A Systematic Sampling Evolutionary (SSE) method for stochastic bilevel programming problems[J]. Computers and Operations Research, 2020, 120: No.104942. 10.1016/j.cor.2020.104942 |
24 | 曾明华,全轲. 双层规划的改进混合布谷鸟搜索量子行为粒子群优化算法[J]. 计算机应用, 2020, 40(7): 1908-1912. |
ZENG M H, QUAN K. Improved hybrid cuckoo search-based quantum-behaved particle swarm optimization algorithm for bi-level programming[J]. Journal of Computer Applications, 2020, 40(7): 1908-1912. | |
25 | CASAS-RAMÍREZ M S, CAMACHO-VALLEJO J F. Solving the p-median bilevel problem with order through a hybrid Heuristic[J]. Applied Soft Computing, 2017, 60: 73-86. 10.1016/j.asoc.2017.06.026 |
26 | ISLAM M M, SINGH H K, RAY T, et al. An enhanced memetic algorithm for single-objective bilevel optimization problems[J]. Evolutionary Computation, 2017, 25(4): 607-642. 10.1162/evco_a_00198 |
27 | ABO-ELNAGA Y, EI-SHORBAGY M A. Multi-sine cosine algorithm for solving nonlinear bilevel programming problems[J]. International Journal of Computational Intelligence Systems, 2020, 13(1): 421-432. 10.2991/ijcis.d.200411.001 |
28 | WANG G M, MA L M. The estimation of particle swarm distribution algorithm with sensitivity analysis for solving nonlinear bilevel programming problems[J]. IEEE Access, 2020, 8: 137133-137149. 10.1109/access.2020.3011017 |
29 | SINHA A, MALO P, DEB K. Evolutionary algorithm for bilevel optimization using approximations of the lower level optimal solution mapping[J]. European Journal of Operational Research, 2017, 257(2): 395-411. 10.1016/j.ejor.2016.08.027 |
30 | LIU Y H, LI H C, CHEN H F. A genetic algorithm for solving linear bilevel programming problems [C]// Proceedings of the 14th International Conference on Computational Intelligence and Security. Piscataway: IEEE, 2018: 40-44. 10.1109/cis2018.2018.00017 |
31 | LI H, ZHANG L, JIAO Y C. Solution for integer linear bilevel programming problems using orthogonal genetic algorithm[J]. Journal of Systems Engineering and Electronics, 2014, 25(3): 443-451. 10.1109/jsee.2014.00051 |
32 | YUE D J, GAO J Y, ZENG B, al at. A projection-based reformulation and decomposition algorithm for global optimization of a class of mixed integer bilevel linear programs[J]. Journal of Global Optimization, 2019, 73(1): 27-57. 10.1007/s10898-018-0679-1 |
33 | KALASHNIKOV V, FLORES MUÑIZ J G, KALASHNYKOVA N. Bilevel optimal tolls problems with non linear costs: a heuristic solution method[M]// KOSHELEVA O, SHARY S P, XIANG G, et al. Beyond Traditional Probabilistic Data Processing Techniques: Interval, Fuzzy etc. Methods and Their Applications, SCI 835. Cham: Springer, 2020: 481-516. |
[1] | 聂青青, 万定生, 朱跃龙, 李致家, 姚成. 基于时域卷积网络的水文模型[J]. 《计算机应用》唯一官方网站, 2022, 42(6): 1756-1761. |
[2] | 史非凡, 史旭华. 基于参考向量的自适应约束多目标进化算法[J]. 《计算机应用》唯一官方网站, 2022, 42(2): 542-549. |
[3] | 李二超, 毛玉燕. 基于空间收缩技术的约束多目标进化算法[J]. 《计算机应用》唯一官方网站, 2021, 41(12): 3419-3425. |
[4] | 严华健, 张国富, 苏兆品, 刘扬. 救灾物资高维多目标自适应分配问题建模与求解[J]. 计算机应用, 2020, 40(8): 2410-2419. |
[5] | 曾明华, 全轲. 双层规划的改进混合布谷鸟搜索量子行为粒子群优化算法[J]. 计算机应用, 2020, 40(7): 1908-1912. |
[6] | 赵志学, 李夏苗, 周鲜成. 考虑拥堵区域的多车型绿色车辆路径问题优化[J]. 计算机应用, 2020, 40(3): 883-890. |
[7] | 尹胜男, 李引珍, 张长泽. 基于客流分配的高铁票价调整策略[J]. 计算机应用, 2020, 40(1): 278-283. |
[8] | 李丽滢, 付寒梅. 考虑多种运输方式的整车物流服务供应链订单分配问题[J]. 计算机应用, 2019, 39(6): 1836-1841. |
[9] | 张志强, 鲁晓锋, 孙钦东, 王侃. 增强开发能力的改进人工蜂群算法[J]. 计算机应用, 2019, 39(4): 949-955. |
[10] | 张鑫, 邹德旋, 沈鑫. 含交叉项的混合二范数粒子群优化算法[J]. 计算机应用, 2018, 38(8): 2148-2156. |
[11] | 刘宝, 董明刚, 敬超. 改进的排序变异多目标差分进化算法[J]. 计算机应用, 2018, 38(8): 2157-2163. |
[12] | 袁亦川, 杨洲, 罗廷兴, 秦进. 求解动态优化问题的多种群竞争差分进化算法[J]. 计算机应用, 2018, 38(5): 1254-1260. |
[13] | 郭后钱, 王微微, 尚颖, 赵瑞莲. 基于动态集合进化算法的弱变异测试用例集生成[J]. 计算机应用, 2017, 37(9): 2659-2664. |
[14] | 熊瑞琦, 马昌喜. 多配送中心危险货物配送路径鲁棒优化[J]. 计算机应用, 2017, 37(5): 1485-1490. |
[15] | 吕宏武, 谷雷, 王慧强, 邹世辰, 冯光升. 分层检查点的近似最优周期计算模型[J]. 计算机应用, 2017, 37(1): 103-107. |
阅读次数 | ||||||
全文 |
|
|||||
摘要 |
|
|||||