Direct Position Updating-based Trying-mutation Particle Swarm Optimization Algorithm and its Application on Reliability Optimization
-
摘要: 为了提升粒子群优化(Particle swarm optimization, PSO)算法的全局寻优能力,增强PSO算法在处理复杂的、高维的、多模态优化问题的寻优性能,提升PSO算法在可靠性优化应用中的优化效果,提出了一种基于直接位置更新策略的试变异(Direct position updating-based trying-mutation PSO, DTPSO)算法。设计了直接位置更新策略和试变异策略,有效维持了种群的多样性,维持了探索和开发的平衡,提升了获得全局最优解的概率。通过9种复杂测试函数的验证和比较,证明了DTSPO算法设计的合理性,以及算法拥有的优异的全局寻优能力。对可靠性冗余分配和可靠度分配问题进行了优化,并和其它先进的改进算法进行比较,结果证明了DTPSO算法的稳定性和寻优性能。Abstract: In order to improve the global optimization ability of particle swarm optimization (PSO), enhance the performance of PSO in dealing with those complex, high-dimensional, multimodal optimization problems, and furthermore, promote the optimization effect in reliability optimization applications, a direct position updating-based trying-mutation PSO (DTPSO) is proposed in this paper. In this algorithm, the direct position updating strategy and trying mutation strategy are designed, which can effectively maintain the diversity of population, balance the exploitation and exploration, and increase the probability of obtaining the global optimal solution. After being verified and compared by 9 complex test functions, the rationality of DTPSO algorithm design and the excellent global optimization performances are proved. In the reliability optimization, the reliability redundancy allocation and reliability allocation are optimized by different advanced improved algorithms, and the comparison results prove the stability and optimization performance of DTPSO.
-
表 1 典型测试函数特征
测试函数 搜索范围 维数 最优解 全局极值 [-100, 100] 100 (0, 0, …, 0)100 0 [-100, 100] 100 (0, 0, …, 0)100 0 [-100, 100] 100 (0, 0, …, 0)100 0 [-100, 100] 100 (0, 0, …, 0)100 0 [-100, 100] 100 (0, 0, …, 0)100 0 [-100, 100] 100 (1, 1, …, 1)100 0 [-100, 100] 100 (0, 0, …, 0)100 0 [-100, 100] 100 (0, 0, …, 0)100 0 [-100, 100] 100 (1, 1, …, 1)100 0 表 2 不同惯性权重对寻优性能的影响
函数 ω1-DTPSO ω2-DTPSO ω3-DTPSO Mean StD Mean StD Mean StD f1 0 0 7.094 9×10-309 0 0 0 f2 0 0 0 0 0 0 f3 0 0 0 0 0 0 f4 0 0 0 0 0 0 f5 0 0 0 0 0 0 f6 0.001 6 0.002 5 1.701 7×10-5 5.480 5×10-5 2.051 6×10-7 5.108 8×10-7 f7 0 0 7.736 1×10-313 0 0 0 f8 0 0 2.608 7×10-309 0 0 0 f9 5.503 4×10-5 8.930 3×10-5 1.590 9×10-7 1.912 9×10-7 5.691 7×10-10 1.183 24×10-9 表 3 试变异策略对寻优性能的影响
函数 TPSO DPSO DTPSO Mean StD Mean StD Mean StD f1 2.029 0×103 432.309 7 0 0 0 0 f2 0.485 4 0.004 4 0 0 0 0 f3 1.483 5 0.119 9 0 0 0 0 f4 20.214 2 1.149 5 0 0 0 0 f5 3.284 3×103 435.924 0 0 0 0 0 f6 6.416 5×106 2.809 1×106 98.744 2 0.0047 2.051 6×10-7 5.108 8×10-7 f7 1.383 4×1088 7.576 3×1088 0 0 0 0 f8 5.407 2×104 2.286 8×104 0 0 0 0 f9 855.031 9 191.828 5 8.722 6 0.369 8 5.691 7×10-10 1.183 24×10-9 表 4 不同PSO改进型算法的寻优性能比较
函数 MELPSO SRPSO DNPSO MAPSO DTPSO Mean StD Mean StD Mean StD Mean StD Mean StD f1 2.7021×10-88 1.4797×10-87 1.3597 0.7925 5.3732×10-32 2.6408×10-31 2.6081 1.0807 0 0 f2 0 0 0.4900 0.0026 0.4999 2.6756×10-5 0.4842 0.0073 0 0 f3 0 0 0.0266 0.0141 0.1620 0.0928 0.0465 0.0217 0 0 f4 0 0 21.3397 0.0538 20.1824 0.0549 19.7378 1.2965 0 0 f5 0 0 742.2342 97.0136 1.6505×103 285.1539 803.6139 141.3295 0 0 f6 93.8743 1.3185 1.3941×103 470.5492 3.8874×103 2.2049×103 2.0078 ×103 1.3292×103 2.0516×10-7 5.1088×10-7 f7 3.9447×10-188 0 2.0038×1053 1.0181×1054 3.6985×1033 2.0051×1034 2.2577×1048 1.0191×1049 0 0 f8 1.0269×10-76 3.4135×10-76 1.1436×104 4.5527×103 8.9545×103 7.8814×103 9.0510×103 2.9310×103 0 0 f9 4.1419 0.5041 480.5421 88.6011 6.2407×103 2.2854×103 450.1516 90.6516 5.6917×10-10 1.1834×10-9 表 5 不同子系统中各单元的可靠度和成本关系
子系统xi 1 2 3 4 5 单元可靠度ri 0.70 0.85 0.75 0.80 0.90 单元成本ci 2 3 2 3 1 表 6 在可靠性冗余分配优化中各算法性能比较
算法 最优解 系统最优总成本 系统最优可靠度 最优解比例/% 最优解时平均迭代步数 时间开销/s DTPSO [1, 1, 3, 1, 2] 16 0.994 5 100 22.4 33.413 QPSO [1, 1, 3, 1, 2] 16 0.994 5 88 60.3 27.732 ACO [2, 1, 2, 1, 2] 16 0.993 9 84 45.7 25.983 GA [2, 1, 2, 1, 2] 16 0.993 9 78 56.3 32.871 SAA [2, 1, 3, 1, 1] 17 0.993 9 75 79.4 21.325 表 7 在可靠度分配优化中各算法性能比较
算法 系统总成本比较 最优输出解比较(各单元分配可靠度) 系统可靠度 时间开销/s DTPSO Mean 260.1879 Min 259.9945 0.8309 0.9375 0.5000 0.6070 0.99 35.332 StD 0.2287 Max 261.3611 0.8575 0.9348 0.5020 0.5447 0.99 H-PSO Mean 261.3318 Min 259.9923 0.8299 0.9377 0.5000 0.6086 0.99 30.757 StD 4.0935 Max 275.9935 0.7586 0.9900 0.5000 0.6049 0.99 ALCPSO Mean 260.2271 Min 259.9917 0.8305 0.9373 0.5000 0.6087 0.99 24.374 StD 0.5970 Max 263.1432 0.8695 0.9335 0.5053 0.5078 0.99 MELPSO Mean 261.9232 Min 259.9916 0.8302 0.9374 0.5000 0.6089 0.99 59.137 StD 3.8856 Max 276.3871 0.7358 0.9897 0.5000 0.6395 0.99 -
[1] ZHENG B, HUANG H Z, LI Y F. Intelligent fault recognition strategy based on adaptive optimized multiple centers[J]. Mechanical Systems & Signal Processing, 2018, 106:526-536 [2] 李昕艺, 刘三阳, 谢维.基于共轭梯度法的感知矩阵优化方法[J].浙江大学学报, 2019, 46(1):15-21 https://www.cnki.com.cn/Article/CJFDTOTAL-HZDX201901003.htmLI X Y, LIU S Y, XIE W. A novel conjugate gradient method for sensing matrix optimization for compressed sensing systems[J]. Journal of Zhejiang University, 2019, 46(1):15-21 (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-HZDX201901003.htm [3] 雒卫廷.基于应变模态的数控机床滚动轴承可靠性检测方法[J].科学技术与工程, 2019, 19(17):173-178 doi: 10.3969/j.issn.1671-1815.2019.17.024LUO W T. Reliability detection method for rolling bearing of numerical control machine tool based on strain modal[J]. Science Technology and Engineering, 2019, 19(17):173-178 (in Chinese) doi: 10.3969/j.issn.1671-1815.2019.17.024 [4] 王林军, 王锬, 杜义贤, 等.一种基于外罚函数法的结构可靠性分析方法[J].三峡大学学报, 2019, 41(1):92-96 https://www.cnki.com.cn/Article/CJFDTOTAL-WHYC201901021.htmWANG L J, WANG T, DU Y X, et al. Structural reliability analysis method based on exterior penalty function method[J]. Journal of China Three Gorges University, 2019, 41(1):92-96 (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-WHYC201901021.htm [5] MENG D B, ZHANG X L, YANG Y J, et al. Interaction balance optimization in multidisciplinary design optimization problems[J]. Concurrent Engineering, 2016, 24(1): 48--57 doi: 10.1177/1063293X15600894 [6] MAHATO S K, BHATTACHARYEE N, PRAMANIK R. Fuzzy reliability redundancy optimisation with signed distance method for defuzzification using genetic algorithm[J]. International Journal of Operational Research, 2020, 37(3):307-323 doi: 10.1504/IJOR.2020.105441 [7] 于建芳, 刘升.黄金正弦模拟退火算法求解低碳有能力约束的车辆路径问题[J].科学技术与工程, 2020, 20(11):4202-4209 doi: 10.3969/j.issn.1671-1815.2020.11.002YU J F, LIU S. Golden sine simulated annealing algorithm for low carbon capacitated Vehicle Routing problem[J]. Science Technology and Engineering, 2020, 20(11):4202-4209 (in Chinese) doi: 10.3969/j.issn.1671-1815.2020.11.002 [8] 郑波.基于粒子群优化算法的航空发动机故障诊断与性能参数预测研究[D].成都: 电子科技大学, 2018ZHENG B. Research on aeroengine fault diagnosis and performance parameters prediction based on PSO Algorithm[D]. Chengdu: University of Electronic Science and Technology of China, 2018 (in Chinese) [9] 王春华, 郭月, 姜宗帅.基于改进粒子群算法的行星齿轮传动多目标可靠性优化设计[J].机械强度, 2018, 40(6):1364-1370 https://www.cnki.com.cn/Article/CJFDTOTAL-JXQD201806016.htmWANG C H, GUO Y, JIANG Z S. Multi-objective reliability optimization design of planetary gear transmission based on improved particle swarm algorithm[J]. Journal of Mechanical Strength, 2018, 40(6):1364-1370 (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-JXQD201806016.htm [10] 陈东宁, 于传宇, 姚成玉, 等.基于Lévy飞行微粒群算法的液压系统可靠性优化[J].液压与气动, 2017(3):17-21 https://www.cnki.com.cn/Article/CJFDTOTAL-YYYQ201703004.htmCHEN D N, YU C Y, YAO C Y, et al. Hydraulic system reliability optimization based on Lévy flight particle swarm algorithm[J]. Chinese Hydraulics & Pneumatics, 2017(3):17-21 (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-YYYQ201703004.htm [11] 鲁鹏, 章卫国, 李广文, 等.一种基于杂草克隆的多目标粒子群算法[J].西北工业大学学报, 2012, 30(2):286-290 doi: 10.3969/j.issn.1000-2758.2012.02.025LU P, ZHANG W G, LI G W, et al. A new and efficient multi-objective particle swarm optimization (MOPSO) algorithm based on invasive weed cloning[J]. Journal of Northwestern Polytechnical University, 2012, 30(2):286-290 (in Chinese) doi: 10.3969/j.issn.1000-2758.2012.02.025 [12] ANESCU G, ULMEANU A P. A no speeds and coefficients PSO approach to reliability optimization problems[J]. Computers & Industrial Engineering, 2018, 120:31-41 [13] SAHOO L, BANERJEE A, BHUNIA A K, et al. An efficient GA—PSO approach for solving mixed-integer nonlinear programming problem in reliability optimization[J]. Swarm and Evolutionary Computation, 2014, 19:43-51 doi: 10.1016/j.swevo.2014.07.002 [14] MELLAL M A, ZIO E. An adaptive particle swarm optimization method for multi-objective system reliability optimization[J]. Journal of Risk and Reliability, 2019, 233(6):990-1001 [15] MAHMOODABADI M J, SALAHSHOOR MOTTAGHI Z, BAGHERI A. HEPSO: high exploration particle swarm optimization[J]. Information Sciences, 2014, 273:101-111 doi: 10.1016/j.ins.2014.02.150 [16] ZHENG B, HUANG H Z, GUO W, et al. Fault diagnosis method based on supervised particle swarm optimization classification algorithm[J]. Intelligent Data Analysis, 2018, 22(1):191-210 doi: 10.3233/IDA-163392 [17] TANWEER M R, SURESH S, SUNDARARAJAN N. Self regulating particle swarm optimization algorithm[J]. Information Sciences: An International Journal, 2015, 294(C):182-202 [18] 郑波, 高峰.基于S-PSO分类算法的故障诊断方法[J].航空学报, 2015, 36(11):3640-3651 https://www.cnki.com.cn/Article/CJFDTOTAL-HKXB201511015.htmZHENG B, GAO F. Fault diagnosis method based on S-PSO classification algorithm[J]. Acta Aeronautica et Astronautica Sinica, 2015, 36(11):3640-3651 (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-HKXB201511015.htm [19] YANG B, WANG C, HUANG H, et al. A multi-agent and PSO based simulation for human behavior in emergency evacuation[C]//Proceedings of 2007 International Conference on Computational Intelligence and Security. Harbin: IEEE, 2007: 296-300 [20] 王正初, 赵燕伟.复杂系统可靠性冗余优化的量子粒子群算法研究[J].中国制造业信息化, 2007, 36(11):72-75 doi: 10.3969/j.issn.1672-1616.2007.11.020WANG Z C, ZHAO Y W. QPSO-based reliability redundancy optimization for complex system[J]. Manufacture Information Engineering of China, 2007, 36(11):72-75 (in Chinese) doi: 10.3969/j.issn.1672-1616.2007.11.020 [21] REHMAN O U, YANG S Y, KHAN S. An improved quantum based particle swarm optimizer applied to electromagnetic optimization problems[J]. Compel: International Journal for Computation and Mathematics in Electrical and Electronic Engineering, 2018, 37(1):319-332 doi: 10.1108/COMPEL-04-2017-0160 [22] 胡启国, 胡小华, 吴泳龙.改进蚁群算法在系统可靠度最优冗余分配的应用[J].重庆交通大学学报, 2013, 32(3):543-546 doi: 10.3969/j.issn.1674-0696.2013.03.40HU Q G, HU X H, WU Y L. Application of improved ant colony algorithm in system reliability optimization of redundancy allocation[J]. Journal of Chongqing Jiaotong University, 2013, 32(3):543-546 (in Chinese) doi: 10.3969/j.issn.1674-0696.2013.03.40 [23] 张民悦, 李丹丹.基于遗传算法的桥式网络可靠度和冗余度分配优化[J].湖北大学学报, 2014, 36(6):505-510 doi: 10.3969/j.issn.1000-2375.2014.06.005ZHANG M Y, LI D D. Reliability and redundancy distribution optimization of bridge network based on genetic algorithm (GA)[J]. Journal of Hubei University, 2014, 36(6):505-510 (in Chinese) doi: 10.3969/j.issn.1000-2375.2014.06.005 [24] 沈博, 潘忠文, 林宏, 等.基于模拟退火算法的元器件布局优化设计[J].导弹与航天运载技术, 2019(6):17-21 https://www.cnki.com.cn/Article/CJFDTOTAL-DDYH201906005.htmSHEN B, PAN Z W, LIN H, et al. Component layout optimization design based on simulated annealing algorithm[J]. Missiles and Space Vehicles, 2019(6):17-21 (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-DDYH201906005.htm [25] 郑波, 马昕.改进的Kohonen网络在航空发动机分类故障诊断中的应用[J].航空发动机, 2020, 46(2):23-29 https://www.cnki.com.cn/Article/CJFDTOTAL-HKFJ202002005.htmZHENG B, MA X. Application of improved Kohonen network in aeroengine classification fault diagnosis[J]. Aeroengine, 2020, 46(2):23-29 (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-HKFJ202002005.htm [26] SINGH R P, MUKHERJEE V, GHOSHAL S P. Particle swarm optimization with an aging leader and challengers algorithm for optimal power flow problem with FACTS devices[J]. International Journal of Electrical Power & Energy Systems, 2015, 64:1185-1196