The Optimization Algorithm of Aviation Equipment Maintenance Cost Forecast and its Applied Research

Abstract:

The maintenance cost forecast of aviation equipment is a multifactor influenced, non-linear and little samples problem. Aiming at the problem, genetic algorithm (GA) and support vector machine (SVM) were combined to build a GA-SVM forecast model for maintenance cost of aviation equipment. The model used GA to optimize the parameters of SVM, which can avoid the blindness choice of parameters and improve its forecast efficiency. Through the example analysis, the model has more accurate results and extensibility than PSO-SVM, SVM and multivariate linear regression in the forecast of maintenance cost of aviation equipment.

• Periodical: Advanced Materials Research (Volumes 760-762)
• Pages: 1851-1855
• DOI: https://doi.org/10.4028/www.scientific.net/AMR.760-762.1851
• Online since: September 2013
• Authors: Li Zhi Xiao, Dong Ping Yang, De Xiang Sun, Xiao Kun Wang, Zhi Liang Li
• Keywords: Genetic Algorithm (GA), Maintenance Cost, Support Vector Machine (SVM)

References

[1] Xiaodong Liu and Hengxi Zhang, Research on relationship of aircraft reliability and develop cost, " Journal of Engineering University of Air Force(Science Edition). Xi, an, 1st ed., vol. 1, 2000, pp.63-66.

[2] Hengxi Zhang, Xiaodong Liu and Jilian Guo., Research report on aircraft whole life cost forecast, " Engineering Academy of Engineering University of Air Force. Xi, an, (1994).

[3] Yuhui Yang, Hongfu Zuo and Guo Chen, Parameter effect analysis and adaptive optimization on the SVM time series forecast model, Journal of aviation power. 4st ed., vol. 21, 2006, pp.767-772.

[4] Shaoxiong Wu and Enzhou Huang, Control chart pattern recognition based on SVM, Computer application. 1st ed., vol. 27, 2007, pp.61-64.

[5] Keerth I, Lin C J, Asymptotic behaviors of support vector machines with Gaussian kerne, Neural Computation. 15st ed., 2003, pp.1667-1689.

DOI: 10.1162/089976603321891855

[6] Lin H T and Lin C J Keerth I, A study on sigmoid kernels for SVM and the training of non-PSD kernels by SMO-type methods, Taibei: National Taiwan University. (2003).

[7] Panchi Li and Shaohua Xu, Kernel function character analysis of SVM on pattern recognition, Computer engineer and design. 2nd ed., vol. 26, 2005, pp.302-304.

[8] Shiqing Wang and Yan Cao, Research of character selection based on GA-SVM, Computer engineer and design. 18st ed., vol. 31, 2010, pp.4088-4092.

[9] Feng Guo, Hengxi Zhang and Shou'an Li, Aircraft maintenance and support cost forecast based on PLS, Journal of Engineering University of Air Force(Science Edition). 3rd ed., vol. 6, 2010, pp.10-15.

[10] Gaobo Chen, Aircraft maintenance and support cost forecast model based on PSO-LSSVM, Computer application and software. 9st ed., vol. 27, 2010, p.183.