为了预防运载火箭总装过程中因动态事件引发的工期延误问题，保证火箭总装任务的按期交付，提出一种工期延误预警方法.该方法包括3个关键步骤：警情监测、警兆识别与警度预报.通过分析火箭总装工期的各种扰动因素及其作用机理，设计定量模型衡量各扰动因素的预警指标，实现火箭总装任务的进度监测.通过充分考虑各警度等级样本数量的不平衡性，应用不平衡分类算法实现警兆识别.通过综合考虑预警样本拖期程度与预警时间节点调整的难易度，设计相应的警度等级以实现警度预报.将该预警方法应用于上海某航天总装厂的实际总装过程数据，以验证该方法的有效性与优越性.
To prevent the overdue risks caused by randomness and dynamic events during rocket final assembly process, an early warning method for tardiness precaution is proposed to ensure in-time delivery. The method includes three steps: indicator monitoring, warning sign recognition and warning level prognostication. The inputs are quantized indicators which are elaborately designed by analyzing key factors and their mechanisms influencing the cycle time. Its main task is to monitor the progress of the rocket final assembly. An imbalanced learning algorithm considering the unbalanced nature of the data under different warning levels is utilized to achieve warning sign recognition. By considering both the tardiness of the job and the difficulty in adjustment during sample collecting, the warning level of sample is designed to realize warning level prognostication. The effectiveness and superiority of the proposed model are proven by applying the model to practical production data collected from some aerospace equipment company in Shanghai.
［1］ZHANG R, LIU Q, ZHANG Y, et al. Study on lean management of carrier rocket final assembly based on the process［J］. Industrial Engineering and Management, 2016, 21(2): 108-111.
［2］APOSTU M V, MEYER M, BENDUL J. Early indication system for critical situations in job-shop production［J］. Procedia CIRP, 2014, 19: 33-38.
［3］KRZHIZHANOVSKAYA V V, SHIRSHOV G S, MELNIKOVA N B, et al. Flood early warning system: Design, implementation and computational modules［J］. Procedia Computer Science, 2011, 4: 106-115.
［4］BEHRENS J, ANDROSOV A, BABEYKO A Y, et al. A new multi-sensor approach to simulation assisted tsunami early warning［J］. Natural Hazards and Earth System Science, 2010, 10(6): 1085-1100.
［5］张云波. 工程项目工期延误原因及预警模型研究［D］. 天津: 天津大学, 2004.
ZHANG Yunbo. The study on the cause of delay and the model of early warning in construction project［D］. Tianjin: Tianjin University, 2004.
［6］KIM B C. Forecasting project progress and early warning of project overruns with probabilistic methods［D］. Texas, USA: Texas A&M University, 2007.
［7］YUAN Z M, WANG Y W, SUN C S. Construction schedule early warning from the perspective of probability and visualization［J］. Journal of Intelligent & Fuzzy Systems, 2017, 32(1): 877-888.
［8］DE BARCELOS TRONTO I F, DA SILVA J D S, SANT’ANNA N. An investigation of artificial neural networks based prediction systems in software project management［J］. Journal of Systems and Software, 2008, 81(3): 356-367.
［9］ZHENG G Z, ZHU N, TIAN Z, et al. Application of a trapezoidal fuzzy AHP method for work safety evaluation and early warning rating of hot and humid environments［J］. Safety Science, 2012, 50(2): 228-239.
［10］HOTZ I, HANISCH A, SCHULZE T. Simulation-based early warning systems as a practical approach for the automotive industry［C］//Proceedings of the 2006 Winter Simulation Conference. Monterey, CA, USA: IEEE, 2006: 1962-1970.
［11］HOPP W J, SPEARMAN M L. Factory physics［M］.3rd ed. New York, USA: McGraw-Hill Companies, 2007.
［12］KLUG F. The internal bullwhip effect in car manufacturing［J］. International Journal of Production Research, 2013, 51(1): 303-322.
［13］刘明周, 单晖, 蒋增强, 等.不确定条件下车间动态重调度优化方法［J］. 机械工程学报, 2009, 45(10): 137-142.
LIU Mingzhou, SHAN Hui, JIANG Zengqiang, et al. Dynamic rescheduling optimization of job-shop under uncertain conditions［J］. Journal of Mechanical Engineering, 2009, 45(10): 137-142.
［14］邵晓亮. 面向机械产品制造的项目进度管理技术研究［D］. 武汉: 武汉理工大学, 2015.
SHAO Xiaoliang. The research of project schedule management technology oriented on machinery manufacturing［D］. Wuhan: Wuhan University of Technology, 2015.
［15］CHAWLA N V, BOWYER K W, HALL L O, et al. SMOTE: Synthetic minority over-sampling technique［J］. Journal of Artificial Intelligence Research, 2002, 16: 321-357.
［16］LIAW A, WIENER M. Classification and regression by random forest［J］. R News, 2002, 2(3): 18-22.
［17］TAI Y T, PEARN W L, LEE J H. Cycle time estimation for semiconductor final testing processes with Weibull-distributed waiting time［J］. International Journal of Production Research, 2012, 50(2): 581-592.
［18］GOODFELLOW I, BENGIO Y, COURVILLE A, et al. Deep learning［M］. Cambridge, MA, USA: MIT Press, 2016.
［19］HE H B, GARCIA E A. Learning from imbalanced data［J］. IEEE Transactions on Knowledge and Data Engineering, 2009, 21(9): 1263-1284.
［20］周志华, 杨强. 机器学习及其应用2011［M］. 北京: 清华大学出版社, 2011: 171-191.
ZHOU Zhihua, YANG Qiang. Machine learning and applications 2011［M］. Beijing: Tsinghua University Press, 2011: 171-191.
［21］SUN Y M, KAMEL M S, WANG Y. Boosting for learning multiple classes with imbalanced class distribution［C］//Sixth International Conference on Data Mining. Hong Kong, China: IEEE, 2006: 1-11.
［22］MASSEY JR F J. The Kolmogorov-Smirnov test for goodness of fit［J］. Journal of the American Statistical Association, 1951, 46(253): 68-78.
［23］FAWCETT T. An introduction to ROC analysis［J］. Pattern Recognition Letters, 2006, 27(8): 861-874.
