-
摘要: 基于现代排序理论的建模思想, 分析了飞机对发动机的需求特性和发动机使用维修的周转特点, 引入工作节的概念, 建立了基于视情维修策略的航空发动机调度模型, 提出了求解模型的改进在线LPT算法, 并以CF6型发动机的使用数据验证了本方法的有效性。发现排序规则是调度计划的重要影响因素, LPT规则与传统的FCFS规则相比, 发动机在翼时间长, 送修次数少, 保障率高, 租赁损失成本低。分析结果表明: LPT规则优于传统的FCFS规则, 基于视情维修策略的航空发动机调度模型有效、可行。Abstract: Based on the modeling thought of modern scheduling theory, the requirement property for aeroengines and the turnover property of aeroengine maintenance were analyzed, the concept of work note was introduced, a scheduling model of aeroengines under condition-based maintenance strategy was established, an online longest processing time (LPT) rule algorithm was improved for solving the model, and the used data of CF6 aeroengines were studied. It is pointed that aimed at aeroengine, on-wing time is longer, maintenance time is lower, guarantee rate is higher, and rental loss cost is lower when LPT rule is compared with traditional first come first serve (FCFS) rule. Analysis result indicates that LPT rule is better than FCFS rule, and the model is feasible.
-
表 1 寿命矩阵
Table 1. Life matrix
发动机编号 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 第1次在翼寿命/d 218 1 069 271 366 152 713 592 151 420 487 117 681 300 118 49 151 216 674 321 205 537 411 205 186 98 第2次在翼寿命/d 1 069 139 530 900 243 216 718 90 860 491 373 437 705 188 684 281 664 553 698 214 741 378 858 971 352 
下载: 导出CSV
表 2 发动机调度计划
Table 2. Aeroengine scheduling
序列号 机位序号 发动机号 送修次数 送修返回时刻(1) /d 在翼时间/d 开始时刻/d 下发时刻/d 送修返回时刻(2) /d 1 1 2 0 0 1 069 0 1 069 1 129 ή ή ή ή ή ή ή ή ή 64 8 2 1 1 129 139 1 209 1 348 1 408 65 17 23 2 1 188 103 1 225 1 328 1 388
下载: 导出CSV
表 3 计算结果比较
Table 3. Comparison of computation results
排序次数 缺发次数 总缺发时间/d 租赁成本/元 备件保障率/% 实际情况 67 10 194 213 400 99.249 改进在线LPT算法 65 4 100 110 000 99.613 比较结果 2 6 94 103 400 0.364
下载: 导出CSV
-
[1] 梁剑. 基于成本优化的民用航空发动机视情维修决策研究[D]. 南京: 南京航空航天大学, 2005. [2] 徐俊刚, 戴国忠, 王宏安. 生产调度理论和方法研究综述[J]. 计算机研究与发展, 2004, 41 (2): 257-267. https://www.cnki.com.cn/Article/CJFDTOTAL-JFYZ200402000.htmXu Jun-gang, Dai Guo-zhong, Wang Hong-an. An overview of theories and methods of production scheduling[J]. Journal of Computer Research and Development, 2004, 41 (2): 257-267. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-JFYZ200402000.htm [3] 王小平, 齐欢, 肖恒辉, 等. 基于串联排队网络的三峡-葛洲坝水利枢纽联合调度模型[J]. 交通运输工程学报, 2006, 6 (3): 82-86. http://transport.chd.edu.cn/article/id/200603018Wang Xiao-ping, Qi Huan, Xiao Heng-hui, et al. Co-scheduling model of Three Gorges-Gezhou Dam based on seriesqueuing network[J]. Journal of Traffic and Transportation Engineering, 2006, 6 (3): 82-86. (in Chinese) http://transport.chd.edu.cn/article/id/200603018 [4] 唐恒永, 赵传立. 排序引论[M]. 北京: 科学出版社, 2002. [5] 唐国春, 张峰, 罗守成, 等. 现代排序论[M]. 上海: 上海科学普及出版社, 2003. [6] 左洪福, 张海军, 戎翔. 基于比例风险模型的航空发动机视情维修决策[J]. 航空动力学报, 2006, 21 (4): 716-721. doi: 10.3969/j.issn.1000-8055.2006.04.018Zuo Hong-fu, Zhang Hai-jun, Rong Xiang. Condition-based aeroengine maintenance decision method using proportional hazards model[J]. Journal of Aerospace Power, 2006, 21 (4): 716-721. (in Chinese) doi: 10.3969/j.issn.1000-8055.2006.04.018 [7] Graham R L. Bounds on multiprocessing ti ming anomalies[J]. SIAM Journal on Applied Mathematics, 1969, 17 (2): 416-429. doi: 10.1137/0117039 [8] Chen Bo. Areviewof on-line machine scheduling: algorithms and competitiveness[J]. Mathematical Theory and Application, 1999, 19 (3): 1-15. [9] Chen Bo, Vestjens AP A. Scheduling onidentical machines: howgoodis LPTin an on-line setting?[J]. Operations Research Letters, 1997, 21 (2): 165-169. [10] 陈仕平, 张国川. 两台平行机的实时到达在线排序[J]. 应用数学学报, 2000, 23 (1): 31-37. https://www.cnki.com.cn/Article/CJFDTOTAL-YYSU200001003.htmChen Shi-ping, Zhang Guo-chuan. On-line scheduling on two parallel machines where jobs arrive over ti me[J]. Acta Mathematicae Applicatae Sinica, 2000, 23 (1): 31-37. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-YYSU200001003.htm -
点击查看大图
图(2) / 表(3)
计量
- 文章访问数: 348
- HTML全文浏览量: 121
- PDF下载量: 235
- 被引次数: 0
