基于自适应变异PSO-SVM的APU性能参数预测模型

doi:10.12305/j.issn.1001-506X.2021.02.27

摘要/Abstract

摘要：

为提高辅助动力装置(auxiliary power unit, APU)性能参数预测的精度,针对支持向量机(support vector machine, SVM)模型在实际使用中遇到的参数选择问题,采用自适应变异粒子群优化(particle swarm optimization, PSO)算法实现对SVM惩罚参数和核参数的优化选择,提出一种基于自适应变异PSO算法优化SVM的APU性能参数预测模型。进一步分析了预测模型不同预测步长对短期预测精度的影响。利用某型APU性能参数数据进行了验证,并与多种预测模型进行了对比实验。实验结果表明,对于排气温度(exhaust gas temperature, EGT)的预测,自适应变异PSO-SVM模型的平均绝对百分比误差(mean absolute percentage error, MAPE)比标准PSO-SVM模型低47%;对于滑油温度(oil temperrature, OT)的预测,自适应变异PSO-SVM模型的MAPE比标准PSO-SVM低29%,为短期APU性能变化趋势预测提供了一定的参考。

关键词: 辅助动力装置, 粒子群优化算法, 性能参数预测, 支持向量机

Abstract:

In order to improve the accuracy of performance parameter prediction of auxiliary power unit (APU), and solve the problem of parameter selection of support vector machine (SVM) model in practical use, adaptive mutation particle swarm optimization (PSO) is adopted to optimize the selection of penalty parameters and kernel parameters of SVM. And a prediction model of APU performance parameters based on adaptive mutation PSO algorithm is proposed. Furthermore, the influence of different prediction steps on short-term prediction accuracy is analyzed. The performance parameters of an APU are verified and compared with other prediction models. The experimental results show that for the prediction of exhaust gas temperature (EGT), the mean absolute percentage error (MAPE) of adaptive mutation PSO-SVM model is 47% lower than that of standard PSO-SVM model; for the prediction of oil temperature (OT), the MAPE of adaptive mutation PSO-SVM model is 29% lower than that of standard PSO-SVM, which provides some reference for short-term prediction of APU performance trend.

Key words: auxiliary power unit (APU), particle swarm optimization (PSO) algorithm, performance parameter prediction, support vector machine (SVM)

中图分类号:

V241.07

王坤, 侯树贤, 王力. 基于自适应变异PSO-SVM的APU性能参数预测模型[J]. 系统工程与电子技术, 2021, 43(2): 526-536.

Kun WANG, Shuxian HOU, Li WANG. APU performance parameter prediction model based on adaptive variation PSO-SVM[J]. Systems Engineering and Electronics, 2021, 43(2): 526-536.

图/表 12

图1

图2

图3

图4

表1

图5

图6

表2

图7

图8

表3

图9

参考文献 31

1	赵运生, 胡骏, 吴铁鹰, 等. 大型民用飞机辅助动力装置性能仿真[J]. 航空动力学报, 2011, 26 (7): 1590- 1598.
	ZHAO Y S , HU J , WU T Y , et al. Performance simulation of large civil aircraft auxiliary power unit[J]. Journal of Aerospace Power, 2011, 26 (7): 1590- 1598.
2	FLORES J J , GRAFF M , RODRIGUES H . Evaluative design of ARMA and ANN models for time series forecasting[J]. Renewable Energy, 2012, 44, 225- 230. doi: 10.1016/j.renene.2012.01.084
3	王辉, 唐启东. 组合预测方法在APU状态监控中的应用[J]. 航空维修与工程, 2013, (6): 47- 49. doi: 10.3969/j.issn.1672-0989.2013.06.022
	WANG H , TANG Q D . Application for combination prediction method in apu condition monitoring[J]. Aviation Maintenance & Engineering, 2013, (6): 47- 49. doi: 10.3969/j.issn.1672-0989.2013.06.022
4	董平. APU系统的状态监测与维修决策研究[D].南京:南京航空航天大学, 2018.
	DONG P. Research on condition monitoring and maintenance decision of APU system[D]. Nanjing: Nanjing University of Aeronautics and Astronautics, 2018.
5	宋云雪, 彭朋, 史永胜. 民用发动机状态混沌预测算法[J]. 航空动力学报, 2011, 26 (3): 675- 679.
	SONG Y X , PENG P , SHI Y S . Chaos arithmetic for civil aeroengine forecasting[J]. Journal of Aerospace Power, 2011, 26 (3): 675- 679.
6	杜维仲, 王硕. 基于LSTM的辅助动力装置系统辨识与仿真[J]. 计算机测量与控制, 2020, 28 (2): 157- 161.
	DU W Z , WANG S . System identification and simulation of auxiliary power plant based on LSTM[J]. Computer Measurement & Control, 2020, 28 (2): 157- 161.
7	皮骏, 马圣, 张奇奇, 等. 基于改进果蝇算法优化的GRNN航空发动机排气温度预测模型[J]. 航空动力学报, 2019, 34 (1): 8- 17.
	PI J , MA S , ZHANG Q Q , et al. Aero-engine exhaust gas temperature prediction model based on IFOA-GRNN[J]. Journal of Aerospace Power, 2019, 34 (1): 8- 17.
8	丁刚, 徐敏强, 侯立国. 基于过程神经网络的航空发动机排气温度预测[J]. 航空动力学报, 2009, 24 (5): 1035- 1039.
	DING G , XU M Q , HOU L G . Prediction of aeroengine exhaust gas temperature using process neural network[J]. Journal of Aerospace Power, 2009, 24 (5): 1035- 1039.
9	YILMAZ I . Evaluation of the relationship between exhaust gas temperature and operational parameters in CFM56-7 Bengines[J]. Proceedings of the Institution of Mechanical Engineers, 2009, 223 (4): 433- 440.
10	陈庆贵, 李洪伟, 李明, 等. 基于径向基过程神经网络的航空发动机排气温度预测[J]. 兵器装备工程学报, 2019, 40 (6): 154- 157. doi: 10.11809/bqzbgcxb2019.06.032
	CHEN Q G , LI H W , LI M , et al. Prediction of aeroengine exhaust gas temperature based on RBF neural networks[J]. Journal of Ordnance Equipment Engineering, 2019, 40 (6): 154- 157. doi: 10.11809/bqzbgcxb2019.06.032
11	张一震, 钟诗胜, 付旭云, 等. 基于动态集成算法的航空发动机气路参数预测[J]. 航空动力学报, 2018, 33 (9): 2285- 2295.
	ZHANG Y Z , ZHONG S S , FU X Y , et al. Aeroengine gas path parameter prediction based on dynamic ensemble algorithm[J]. Journal of Aerospace Power, 2018, 33 (9): 2285- 2295.
12	刘小鱼.基于数据驱动的航空发动机气路参数预测研究[D].大连:大连理工大学, 2019.
	LIU X Y. Data driven prediction of the performance of aero-engine gas path[D]. Dalian: Dalian University of Technology, 2019.
13	唐敏杰. 基于状态维修的APU起动机监控模型的建模与实现[J]. 航空维修与工程, 2014, (3): 73- 75. doi: 10.3969/j.issn.1672-0989.2014.03.043
	TANG M J . Modeling and implementation of the APU starter monitoring model under CBM strategy[J]. Aviation Maintenance & Engineering, 2014, (3): 73- 75. doi: 10.3969/j.issn.1672-0989.2014.03.043
14	KUMAR A, SRIVASTAVA A, GOEL N, et al. Exhaust gas temperature data prediction by autoregressive models[C]//Proc.of the IEEE 28th Canadian Conference on Electrical and Computer Engineering, 2015: 976-981.
15	戴邵武, 陈强强, 丁宇. 基于改进EMD的排气温度裕度预测[J]. 兵器装备工程学报, 2020, 41 (1): 157- 162.
	DAI S W , CHEN Q Q , DING Y . Prediction of exhaust gas temperature margin based on improved EMD[J]. Journal of Ordnance Equipment Engineering, 2020, 41 (1): 157- 162.
16	李艳军, 张建, 曹愈远, 等. 基于模糊信息粒化和优化SVM的航空发动机性能趋势预测[J]. 航空动力学报, 2017, 32 (12): 3022- 3030.
	LI Y J , ZHANG J , CAO Y Y , et al. Forecasting of aero-engine performance trend based on fuzzy information granulation and optimized SVM[J]. Journal of Aerospace Power, 2017, 32 (12): 3022- 3030.
17	NIETO P J G , GARCÍA-GONZALO E , LASHERAS F S , et al. Hybrid PSO-SVM-based method for forecasting of the remaining useful life for aircraft engines and evaluation of its reliability[J]. Reliability Engineering & System Safety, 2015, 138 (6): 219- 231.
18	于广滨, 丁刚, 姚威, 等. 基于支持过程向量机的航空发动机排气温度预测[J]. 电机与控制学报, 2013, 17 (8): 30- 36. doi: 10.3969/j.issn.1007-449X.2013.08.005
	YU G B , DING G , YAO W , et al. Aeroengine exhaust gas temperature prediction using support process vector machine[J]. Electric Machines and Control, 2013, 17 (8): 30- 36. doi: 10.3969/j.issn.1007-449X.2013.08.005
19	皮骏, 黄江博, 黄磊, 等. 基于改进QPSO-SVR的航空发动机排气温度预测[J]. 振动测试与诊断, 2019, 39 (2): 267- 272, 440.
	PI J , HUANG J B , HUANG L , et al. Aeroengine exhaust gas temperature prediction based on IQPSO-SVR[J]. Journal of Vibration, Measurement & Diagnosis, 2019, 39 (2): 267- 272, 440.
20	ZHAN L W , LI C W . A hybrid PSO-SVM-based method for predicting the friction coefficient between aircraft tire and coating[J]. Measurement Science & Technology, 2017, 28 (2): 025004.
21	曹惠玲, 王冉. 基于滑动时窗策略自适应优化支持向量机的航空发动机性能参数在线预测[J]. 推进技术, 2020, 41 (8): 1887- 1894.
	CAO H L , WANG R . Adaptively optimized support vector machine online prediction of aeroengine performance parameters based on sliding time window strategy[J]. Journal of Propulsion Technology, 2020, 41 (8): 1887- 1894.
22	蔡坤烨, 蔡景, 周迪, 等. 基于SVM方法的APU故障预测方法[J]. 南京航空航天大学学报, 2019, 51 (4): 466- 473.
	CAI K Y , CAI J , ZHOU D , et al. APU fault prediction based on SVM method[J]. Journal of Nanjing University of Aeronautics & Astronautics, 2019, 51 (4): 466- 473.
23	ZHANG Y , VISSER C C D , CHU Q P . Aircraft damage identification and classification for database-driven online flight-envelope prediction[J]. Journal of Guidance, Control, and Dynamics, 2018, 41 (2): 449- 460. doi: 10.2514/1.G002866
24	CORTES C , VAPNIK V . Support-vector networks[J]. Machine Learning, 1995, 20 (3): 273- 297.
25	BURGES C J C . A tutorial on support vector machines for pattern recognition[J]. Data Mining and Knowledge Discovery, 1998, 2 (2): 121- 167. doi: 10.1023/A:1009715923555
26	ABOUHAWWASH M , SEADA H , DEB K . Towards faster convergence of evolutionary multi-criterion optimization algorithms using Karush Kuhn Tucker optimality based local search[J]. Computers and Operations Research, 2017, 79 (C): 331- 346.
27	SMITS G F, JORDAAN E M. Improved SVM regression using mixtures of kernels[C]//Proc.of the International Joint Conference on Neural Networks, 2002: 2785-2790.
28	魏瑾瑞. 对支持向量机混合核函数方法的再评估[J]. 统计研究, 2015, 32 (2): 90- 96.
	WEI J R . A reevaluation of mixed kernel function for support vector machine[J]. Statistical Research, 2015, 32 (2): 92- 98.
29	KENNEDY J, EBERHART R. Particle swarm optimization[C]//Proc.of the Intenational Conference on Neural Networks, 2011: 1942-1948.
30	SHI Y H, EBERHART R C. A modified particle swarm optimizer[C]//Proc.of the IEEE World Congress on Computational Intelligence, 1998: 69-73.
31	HOLLAND J . Genetic algorithms[J]. Scientific American, 1992, 267 (1): 66- 72. doi: 10.1038/scientificamerican0792-66

控制变量			EGT	OT	RS	BAF	BAP	GL
气压海拔外界温度	GL	相关性	0.619	0.579	0.395	0.260	-0.025	1.000
		显著性(双侧)	0	0	0	0	0
		DF	638	638	638	638	638	0
	BAF	相关性	0.709	0.389	0.279	1.000	-0.056	0.260
		显著性(双侧)	0	0	0		0	0
		DF	638	638	638	0	638	638

模型	MAE	MSE	MAPE/%	RMSE
小波神经网络	9.571	0.072	3.98	0.268
Elman神经网络	6.434	0.036	2.81	0.189
标准PSO-SVM	2.826	0.011	1.19	0.109
SVM	4.399	0.041	1.72	0.202
LSSVM	3.128	0.032	1.35	0.178
自适应变异PSO-SVM	1.603	0.009	0.63	0.095
BP神经网络	7.977	0.052	3.47	0.228

模型	MAE	MSE	MAPE/%	RMSE
小波神经网络	0.929	0.253	1.06	0.503
Elman神经网络	0.164	0.088	0.20	0.296
标准PSO-SVM	0.213	0.061	0.24	0.246
SVM	0.310	0.087	0.35	0.294
LSSVM	0.235	0.074	0.25	0.272
自适应变异PSO-SVM	0.152	0.045	0.17	0.212
BP神经网络	0.192	0.062	0.22	0.248

[1]	杨建峰, 肖和业, 李亮, 白俊强, 董维浩. 基于模糊聚类和专家评分机制的无人机多层次模块划分方法[J]. 系统工程与电子技术, 2022, 44(8): 2530-2539.
[2]	杜思予, 全英汇, 沙明辉, 方文, 邢孟道. 基于进化PSO算法的稀疏捷变频雷达波形优化[J]. 系统工程与电子技术, 2022, 44(3): 834-840.
[3]	薛春岭, 曹菲, 孙庆, 秦建强, 冯晓伟. 基于多特征信息融合的海面微弱目标检测[J]. 系统工程与电子技术, 2022, 44(11): 3338-3345.
[4]	朱峰, 蒋倩倩, 林川, 杨啸. 基于支持向量机的典型宽带电磁干扰源识别[J]. 系统工程与电子技术, 2021, 43(9): 2400-2406.
[5]	董庆, 李本威, 闫思齐, 钱仁军. 基于BSO-ELM的涡轴发动机加速过程性能参数预测[J]. 系统工程与电子技术, 2021, 43(8): 2181-2188.
[6]	李昂, 聂党民, 温祥西, 王泽坤, 杨诚修. 基于相依网络和SVM的管制系统运行态势评估[J]. 系统工程与电子技术, 2021, 43(5): 1287-1294.
[7]	赵帅, 刘松涛, 汪慧阳. 基于PSO-CNN的LPI雷达波形识别算法[J]. 系统工程与电子技术, 2021, 43(12): 3552-3563.
[8]	李玖阳, 胡敏, 王许煜, 徐家辉, 李菲菲. 基于ALPSO算法的低轨卫星小推力离轨最优控制方法[J]. 系统工程与电子技术, 2021, 43(1): 199-207.
[9]	何春蓉, 朱江. 基于注意力机制的GRU神经网络安全态势预测方法[J]. 系统工程与电子技术, 2021, 43(1): 258-266.
[10]	马愈昭, 王强强, 王瑞松, 熊兴隆. 基于SVD和MPSO-SVM的光纤周界[J]. 系统工程与电子技术, 2020, 42(8): 1652-1661.
[11]	苏志刚, 陈欣然, 郝敬堂. 基于粒子群优化的圆阵列波束形成方法[J]. 系统工程与电子技术, 2020, 42(7): 1449-1454.
[12]	李亚兰, 金炜东, 葛鹏. 基于VMD和特征融合的辐射源信号识别[J]. 系统工程与电子技术, 2020, 42(7): 1499-1503.
[13]	吴明功, 叶泽龙, 温祥西, 王宏军. 基于局部弹性路由层的关键航路段改航规划[J]. 系统工程与电子技术, 2020, 42(7): 1534-1542.
[14]	孙胜祥, 韩霜. 基于双层决策的装备订购多因素激励定价模型与算法[J]. 系统工程与电子技术, 2020, 42(6): 1338-1347.
[15]	李红光, 郭英, 眭萍, 齐子森, 苏令华. 基于高维特征选择的跳频电台细微特征识别[J]. 系统工程与电子技术, 2020, 42(2): 445-451.