摘要
针对现有的优化方法在优化支持向量机(SVM)参数时出现的搜索效率低、易陷入局部最优等问题,提出一种基于变异策略的改进型花朵授粉算法——MFPA算法,并将其应用于SVM的参数优化问题,通过UCI中的相关数据集进行测试,测试结果表明,基于MFPA算法参数优化的MFPA-SVM模型其分类性能要优于现有PSO-SVM模型和BA-SVM模型。最后将MFPA-SVM模型应用于风速预测问题,经测试证明,MFPA-SVM模型相较于PSO-SVM模型和BA-SVM模型可有效提高模型的预测精度。
In view of the existing optimization method in the optimization of support vector machine(SVM) parameters of low searching efficiency and easy falling into the master problem,a kind of modified pollinate flowers algorithm based on mutation strategy—MFPA algorithm is put forward,and applies to SVM parameter optimization problem,through the relevant test data sets from UCI,test results show that MFPA-SVM parameter optimization model of its classification performance is superior to the existing PSO-SVM model and the BA-SVM model. Finally MFPA-SVM model is applied to wind speed forecasting problem,proved by tests,MFPA-SVM model compared with PSO-SVM model and the BA–SVM model can effectively improve the prediction accuracy of the models.
引文
[1] KUN-CHIEH W. The grey-based artificial intelligence modeling of thermal error for machine tools[J]. Journal of Grey System,2010,22(4):353-366.
[2]GUO Q,XU R,MAO C,et al. Application of information fusion to volumetric error modeling of CNC machine tools[J].The International Journal of Advanced Manufacturing Technology,2015,78(1-4):439-447.
[3]CHEN Z,LIU X,PENG D,et al. Dynamic model of NC rotary table in angle measurements with time series[J].Transactions of the Institute of Measurement and Control,2012,23(5):153-159.
[4]HUANG Y,ZHANG J,LI X,et al. Thermal error modeling by integrating GA and BP algorithms for the high-speed spindle[J]. The International Journal of Advanced Manufacturing Technology,2014,71(9-12):1669-1675.
[5]SUGANYADEVI M V,BABULAL C K. Support vector regression model for the prediction of loadability margin of a power system[J]. Applied Soft Computing,2014,24:304-315.
[6]JIANG M,LUO J,JIANG D,et al. A cuckoo search-support vector machine model for predicting dynamic measurement errors of sensors[J]. IEEE Access,2013,5(4):452-459.
[7]XUE Z,DU P,SU H. Harmonic analysis for hyperspectral image classification integrated with PSO optimized SVM[J].IEEE Journal of Selected Topics in Ap plied Earth Observations and Remote Sensing, 2014, 7(6):2131-2146.
[8]ENACHE A C,SGARCIU V. Anomaly intrusions detection based on support vector machines with bat algorithm[C]//System Theory,Control and Computing(ICSTCC),201418th International Conference. IEEE,2014:856-861.
[9]YANG X S. Flower pollination algorithm for global optimization[C]//International Conference on Unconventional Computing and Natural Computation.Springer Berlin Heidelberg,2012:240-249.
[10] BENSOUYAD M,SAIDOUNI D E. A discrete flower pollination algorithm for graph coloring problem[C]//Cybernetics(CYBCONF),2015 IEEE 2nd International Conference on. IEEE,2015:151-155.
[11] PRAEHIBA R,MOSES M B,SAKTHIVEL S. Flower pollination algorithm applied for different economic load dispatch problems[J]. International Journal of Engineering and Technology,2014,6(2):1009-1016.
[12] RODRIGUES D,YANG X S,DE S A N,et al. Binary flower pollination algorithm and its application to feature selection[M]. Recent Advances in Swarm Intelligence and Evolutionary Computation. Springer International Publishing,2015:85-100.
[13] OCHOA A, GONZALEZ S, MARGAIN L, et al.Implementing flower multi-objective algorithm for selection of university academic credits[C]//Nature and Biologically Inspired Computing(Na BIC),2014 Sixth World Congress on. IEEE,2014:7-11.
[14]ZHENG W,FU H,YANG G. Targeted mutation:a novel mutation strategy for differential evolution[C]//Tools with Artificial Intelligence(ICTAI), 2015 IEEE 27th International Conference on. IEEE,2015:286-293.
[15]欧阳海滨,高立群,孔祥勇.随机变异差分进化算法[[J].东北大学学报(自然科学版),2013,34(3):330-334.
[16]WALTON S,HASSAN O,MORGAN K,et al. Modified cuckoo search:a new gradient free optimisation algorithm[J]. Chaos,Solitons&Fractals,2011,44(9):710-718.
[17]邵璠,孙育河,梁岚珍.基于时间序列法的风电场风速预测研究[J].电力环境保护,2008,24(4):1-5.
[18]潘迪夫,刘辉,李燕飞.基于时间序列分析和卡尔马滤波算法的风电场风速预测优化模型[J].电网技术,2008,32(7):82-86.
[19] LI X,LIU Y,XIN W. Wind speed prediction based on genetic neural network[C]//2009 4th IEEE Conference on Industrial Electronics and Applications. IEEE, 2009:2448-2451.
[20]YANG X,CUI Y,ZHANG H,et al. Research on modeling of wind turbine based on LS-SVM[C]//2009 International Conference on Sustainable Power Generation and Supply.IEEE,2009:1-6.