冰雪天气下基于MFOA-KELM残差修正的跑道温度混合预测

陈斌; 刘悦; 李庆真; 丁宇; 王立文

doi:10.13700/j.bh.1001-5965.2021.0646

冰雪天气下基于MFOA-KELM残差修正的跑道温度混合预测

doi: 10.13700/j.bh.1001-5965.2021.0646

陈斌^{1, 2, ,},
刘悦^{1, 2},
李庆真^{1, 2},
丁宇^{1, 2},
王立文²

1.
中国民航大学电子信息与自动化学院, 天津 300300
2.
中国民航大学航空地面特种设备研究基地, 天津 300300

基金项目:

国家自然科学基金委员会-中国民航局民航联合研究基金 U1933107

详细信息

通讯作者:
陈斌, E-mail: chenbindavid@163.com

中图分类号: V219;TP311
计量
- 文章访问数: 352
- HTML全文浏览量: 100
- PDF下载量: 28
- 被引次数: 0
出版历程
- 收稿日期: 2021-10-29
- 录用日期: 2022-02-28
- 网络出版日期: 2022-04-26
- 整期出版日期: 2022-11-20

Runway temperature hybrid prediction based on MFOA-KELM residual correction under ice and snow

CHEN Bin^{1, 2
, ,},
LIU Yue^{1, 2},
LI Qingzhen^{1, 2},
DING Yu^{1, 2},
WANG Liwen²

1.
School of Electronic Information and Automation, Civil Aviation University of China, Tianjin 300300, China
2.
Department of Aviation Ground Special Equipment, Civil Aviation University of China, Tianjin 300300, China

Funds:

Joint Fund of Civil Aviation Research of National Natural Science Foundation of China and Civil Aviation Administration of China U1933107

More Information

Corresponding author: CHEN Bin, E-mail: chenbindavid@163.com

摘要

摘要:
道面温度短时精准预测是跑道积冰预警的关键因素之一, 为了解决单一机理预测模型随预测时间延长而造成误差累积的问题, 提出了一种冰雪天气下跑道温度混合预测方法。将跑道温度机理预测模型与核极限学习机(KELM)相结合, 建立一种数据驱动修正残差的跑道温度机理预测模型。针对果蝇优化算法(FOA)收敛速度慢、易陷入局部最小值的问题, 引入权值更新函数和距离扩充因子, 调整果蝇的全局寻优效果, 避免陷入局部极小值。利用改进的果蝇优化算法(MFOA)对KELM的正则化参数与核参数联合优化, 以冰雪天气下跑道温度实际数据为例, 建立基于改进果蝇优化核极限学习机(MFOA-KELM)的跑道温度混合预测模型, 并在不同时间尺度下对该混合预测模型进行仿真测试。实验结果表明：与单一机理预测模型相比, 当预测时长为120 min时, MFOA-KELM混合预测模型的平均绝对误差至少减小了61.43%, 在残差阈值为±0.5℃时, 平均预测准确率为91.25%。可见, MFOA-KELM混合预测模型具有更高的预测准确性, 研究结论显示该混合预测方法能够为机场跑道温度短时精准预测提供新思路。
- 混合建模 /
- 核极限学习机 /
- 改进果蝇算法 /
- 跑道温度 /
- 预测
Abstract:
The runway surface temperature short-term accurate prediction is one of the key factors for runway icing warning.In order to solve the problem of error accumulation caused by a single mechanistic model with increasing prediction time, a hybrid runway temperature prediction method under ice and snow is proposed.The runway temperature mechanism model is combined with the kernel extreme learning machine (KELM) to develop a data-driven model for correcting the mechanism residuals. To address the problem that the fruit fly optimization algorithm (FOA) is slow to converge (converges slowly) and easily falls into local minima. By introducing a distance expansion factor and a weight update function, it is possible to modify the effect of the FOA's search for the global optimal solution and prevent falling into local minima. The modified fruit fly optimization algorithm (MFOA) is used to jointly optimize the KELM regularization parameter and the kernel parameter. A hybrid runway temperature prediction model is developed based on the modified fruit fly optimized kernel extreme learning machine (MFOA-KELM) with the actual data of runway temperature under ice and snow. The hybrid model is simulated and tested under different time lengths. The experimental results show that compared with the single mechanism prediction model, the mean absolute error of the MFOA-KELM hybrid model is reducedby at least 61.43% when the prediction length is 120 minutes, and the average prediction accuracy is 91.25% when the residual threshold is ±0.5℃. It can be seen that the MFOA-KELM hybrid model has higher prediction accuracy. The research findings show that this hybrid prediction method can provide a new idea for short time accurate prediction of airport runway temperatures.
- hybrid modeling /
- kernel extreme learning machine /
- modified fruit fly optimization algorithm /
- runway temperature /
- predict

HTML全文

图 1 跑道热量传导过程示意图

Figure 1. Schematic diagram of runway heat transfer process

下载: 全尺寸图片幻灯片

图 2 函数优化曲线

Figure 2. Function optimization curves

下载: 全尺寸图片幻灯片

图 3 混合预测模型结构

Figure 3. Structure of hybrid prediction model

下载: 全尺寸图片幻灯片

图 4 MFOA-KELM模型流程

Figure 4. Flow chart of MFOA-KELM model

下载: 全尺寸图片幻灯片

图 5 模拟跑道温度实验系统采集平台

Figure 5. Simulation of runway temperature experimental system collecting platform

下载: 全尺寸图片幻灯片

图 6 样本1~样本4的预测结果

Figure 6. Prediction results of samples 1, 2, 3 and 4

下载: 全尺寸图片幻灯片

表 1 波音737-800起降参数^[8]

Table 1. Taking-off and landing parameters of Boeing 737-800^[8]

参数	数值
l_a/m	1 431
w_a/m	25
h_a/m	10
m_f/kg	56
η_f	0.95
e_f/(J·kg^-1)	43×10⁶

下载: 导出CSV

表 2 标准测试函数^[25]

Table 2. Standard test functions^[25]

函数	表达式	搜索区间	理论最优值
F₁		[-100, 100]	0
F₂		[-1.28, 1.28]	0
F₃		[-600, 600]	0
F₄		[-10, 10]	0

下载: 导出CSV

表 3 优化结果比较

Table 3. Optimization result comparison

函数	MFOA		FOA		PSO		DE
函数	平均值	标准差	平均值	标准差	平均值	标准差	平均值	标准差
F₁	8.195 2×10^-26	7.752 3×10^-28	0.000 190 81	7.532 3×10^-7	3.609 5	2.145 4	23.354 1	5.936 5
F₂	5.769 5×10^-5	4.638 2×10^-5	0.001 278 8	0.000 277 66	0.185 83	0.065 835	0.329 56	0.085 208
F₃	0	0	1.275×10^-5	5.276 7×10^-8	0.016 197	0.020 383	1.211 3	0.084 048
F₄	0	0	81.204 6	95.036 1	53.307 8	12.502 6	212.245 5	9.562 8

下载: 导出CSV

表 4 实验数据样本

Table 4. Experimental data samples

样本	日期	时间	采样间隔/min
1	2020-01-05	20:00—08:00	1
2	2019-12-23	10:00—22:00	1
3	2019-12-28	19:00—07:00	1
4	2020-01-07	08:00—20:00	1

下载: 导出CSV

表 5 机理预测模型参数

Table 5. Mechanistic prediction model parameters

参数	数值
I₀/(W·m^-2)	1 367
P₀/kPa	101.325
c_L/(J·kg^-1·K^-1)	4 200
ρ_a/(kg·m^-3)	1.29
c_snow/(J·kg^-1·K^-1)	2 100
c_a/(J·kg^-1·K^-1)	1 005.46
σ/(W·m^-2·K^-4)	5.67×10^-8
c_r/(J·kg^-1·K^-1)	970
ρ/(kg·m^-3)	1 930
λ/(W·m^-1·K^-1)	1.7
W₀/(kg·m^-2)	0.5
ε_r	0.9
r	0.7
ε_a	0.929
τ₁/s	900
τ₂/s	1 800
τ₃/s	3 600
τ₄/s	7 200

下载: 导出CSV

表 6 样本1~样本4的平均绝对误差

Table 6. Mean absolute error of samples 1, 2, 3 and 4

样本	预测时长/min	平均绝对误差
样本	预测时长/min	机理预测模型	MFOA-KELM混合	FOA-KELM混合	KELM混合
1	15	0.285	0.111	0.129	0.232
	30	0.575	0.130	0.147	0.222
	60	1.144	0.149	0.207	0.369
	120	2.336	0.194	0.243	0.265
2	15	0.219	0.087	0.170	0.191
	30	0.268	0.124	0.128	0.218
	60	0.479	0.191	0.269	0.296
	120	0.879	0.339	0.459	0.516
3	15	0.184	0.007	0.014	0.017
	30	0.368	0.014	0.031	0.034
	60	0.736	0.028	0.067	0.069
	120	1.472	0.055	0.106	0.138
4	15	0.144	0.093	0.098	0.099
	30	0.230	0.095	0.107	0.111
	60	0.428	0.120	0.177	0.182
	120	0.849	0.186	0.415	0.503

下载: 导出CSV

表 7 不同残差阈值内平均预测准确率

Table 7. Average prediction accuracy under different residual threshold values

预测时长/min	预测模型	不同允许误差范围内平均预测准确率/%
预测时长/min	预测模型	±0.5℃	±1℃	±2℃
15	机理预测模型	97.29	100.00	100.00
15	MFOA-KELM混合	100.00	100.00	100.00
30	机理预测模型	58.75	81.04	100.00
30	MFOA-KELM混合	100.00	100.00	100.00
60	机理预测模型	39.58	73.96	99.58
60	MFOA-KELM混合	100.00	100.00	100.00
120	机理预测模型	18.75	40.42	74.38
120	MFOA-KELM混合	91.25	99.38	100.00

下载: 导出CSV

参考文献(31)

[1]	KRŠMANC R, SLAK A Š, DEMŠAR J. Statistical approach for forecasting road surface temperature[J]. Meteorological Applications, 2013, 20(4): 439-446. doi: 10.1002/met.1305
[2]	汤筠筠, 郭忠印. 基于自回归求和移动平均的冬季路温短临预测[J]. 同济大学学报(自然科学版), 2017, 45(12): 1824-1829. doi: 10.11908/j.issn.0253-374x.2017.12.012 TANG J J, GUO Z Y. Winter road temperature short prognosis based on autoregressive summation moving average[J]. Journal of Tongji University (Natural Science Edition), 2017, 45(12): 1824-1829(in Chinese). doi: 10.11908/j.issn.0253-374x.2017.12.012
[3]	秦健, 孙立军. 国外沥青路面温度预估方法综述[J]. 中外公路, 2005, 25(6): 19-23. doi: 10.3969/j.issn.1671-2579.2005.06.006 QIN J, SUN L J. A review of foreign asphalt pavement temperature prediction methods[J]. Chinese and Foreign Highways, 2005, 25(6): 19-23(in Chinese). doi: 10.3969/j.issn.1671-2579.2005.06.006
[4]	BARBERE S. Calculation of maximum pavement temperatures from weather reports[R]. Washington, D.C. : National Research Council, 1957: 1-8.
[5]	CHRISTISON J T, ANDERSON K O. The response of asphalt pavements to low temperature climatic environments[C]//International Conference on the Structural Design of Asphalt Pavements, 1972: 44-56.
[6]	严作人. 层状路面体系的温度场分析[J]. 同济大学学报, 1984(3): 76-85. https://www.cnki.com.cn/Article/CJFDTOTAL-TJDZ198403006.htm YAN Z R. Temperature field analysis of laminated pavement systems[J]. Journal of Tongji University, 1984(3): 76-85(in Chinese). https://www.cnki.com.cn/Article/CJFDTOTAL-TJDZ198403006.htm
[7]	HU Y, ALMKVIST E, LINDBERG F, et al. The use of screening effects in modelling route-based daytime road surface temperature[J]. Theoretical and Applied Climatology, 2016, 125(1): 303-319.
[8]	NUIJTEN A D W. Runway temperature prediction, a case study for Oslo Airport, Norway[J]. Cold Regions Science and Technology, 2016, 125: 72-84. doi: 10.1016/j.coldregions.2016.02.004
[9]	邱金晶, 陈锋, 张珏, 等. 浙江省高速公路路面温度预报模型性能检验及应用[J]. 气象科技, 2020, 48(4): 518-528. https://www.cnki.com.cn/Article/CJFDTOTAL-QXKJ202004011.htm QIU J J, CHEN F, ZHANG J, et al. Performance test and application of highway pavement temperature forecasting model in Zhejiang province[J]. Meteorological Science and Technology, 2020, 48(4): 518-528(in Chinese). https://www.cnki.com.cn/Article/CJFDTOTAL-QXKJ202004011.htm
[10]	赵鸿铎, 姚祖康, 张长安, 等. 飞机除冰液对停机坪水泥混凝土的影响[J]. 交通运输工程学报, 2004, 4(2): 1-5. doi: 10.3321/j.issn:1671-1637.2004.02.001 ZHAO H D, YAO Z K, ZHANG C A, et al. Effects of aircraft de-icing fluid on apron cement concrete[J]. Journal of Transportation Engineering, 2004, 4(2): 1-5(in Chinese). doi: 10.3321/j.issn:1671-1637.2004.02.001
[11]	孙立军, 秦健. 沥青路面温度场的预估模型[J]. 同济大学学报(自然科学版), 2006, 34(4): 480-483. doi: 10.3321/j.issn:0253-374X.2006.04.011 SUN L J, QIN J. Prediction model for temperature field of asphalt pavement[J]. Journal of Tongji University (Natural Science Edition), 2006, 34(4): 480-483(in Chinese). doi: 10.3321/j.issn:0253-374X.2006.04.011
[12]	姚源朝, 仇鹏, 许建良, 等. 基于混合模型的气流床气化炉建模[J]. 化工学报, 2021, 72(5): 2727-2734. https://www.cnki.com.cn/Article/CJFDTOTAL-HGSZ202105034.htm YAO Y Z, QIU P, XU J L, et al. Modeling of a gasifier in a fluid bed based on a hybrid model[J]. Journal of Chemical Engineering, 2021, 72(5): 2727-2734(in Chinese). https://www.cnki.com.cn/Article/CJFDTOTAL-HGSZ202105034.htm
[13]	成庶, 刘嘉文, 伍珣, 等. 基于特征提取与误差补偿的金属化薄膜电容器剩余寿命预测[J]. 中国电机工程学报, 2022, 42(7): 2672-2680. https://www.cnki.com.cn/Article/CJFDTOTAL-ZGDC202207021.htm CHENG S, LIU J W, WU X, et al. Residual life prediction of metalized film capacitors based on feature extraction and error compensation[J]. Proceedings of the CSEE, 2022, 42(7): 2672-2680(in Chinese). https://www.cnki.com.cn/Article/CJFDTOTAL-ZGDC202207021.htm
[14]	HUANG G B. An insight into extreme learning machines: Random neurons, random features and kernels[J]. Cognitive Computation, 2014, 6(3): 376-390. doi: 10.1007/s12559-014-9255-2
[15]	亓孝武, 李可军, 于小晏, 等. 基于核极限学习机和Bootstrap方法的变压器顶层油温区间预测[J]. 中国电机工程学报, 2017, 37(19): 5821-5828. https://www.cnki.com.cn/Article/CJFDTOTAL-ZGDC201719033.htm QI X W, LI K J, YU X Y, et al. Transformer top oil temperature interval prediction based on nuclear limit learning machine and Bootstrap method[J]. Proceedings of the CSEE, 2017, 37(19): 5821-5828(in Chinese). https://www.cnki.com.cn/Article/CJFDTOTAL-ZGDC201719033.htm
[16]	PAN W T. A new fruit fly optimization algorithm: Taking the financial distress model as an example[J]. Knowledge-Based Systems, 2012, 26: 69-74.
[17]	MYTHILI S, KUMAR A V. Hybrid feature selection and heterogeneous clustering ensemble framework for detection of circulating tumor cells[J]. Journal of Medical Imaging and Health Informatics, 2016, 6(5): 1160-1166.
[18]	王东. 在线核极限学习机的改进与应用研究[D]. 蚌埠: 安徽财经大学, 2015. WANG D. Research on the improvement and application of online kernel limit learning machine[D]. Bengbu: Anhui University of Finance and Economics, 2015(in Chinese).
[19]	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. https://www.cnki.com.cn/Article/CJFDTOTAL-HKDI201901002.htm PI J, MA S, ZHANG Q Q, et al. Optimized GRNN aero-engine exhaust temperature prediction model based on improved fruit fly algorithm[J]. Journal of Aerodynamics, 2019, 34(1): 8-17(in Chinese). https://www.cnki.com.cn/Article/CJFDTOTAL-HKDI201901002.htm
[20]	周淑贞. 气象学与气候学[M]. 北京: 高等教育出版社, 2004. ZHOU S Z. Meteorology and climatology[M]. Beijing: Higher Education Press, 2004(in Chinese).
[21]	翁笃鸣, 李炬, 高歌. 晴天太阳总辐射的参数化及气候计算[J]. 气象科学, 1997, 17(1): 1-9. https://www.cnki.com.cn/Article/CJFDTOTAL-QXKX199701000.htm WENG D M, LI J, GAO G. Parameterization of total solar radiation on sunny days and climate calculation[J]. Meteorological Science, 1997, 17(1): 1-9(in Chinese). https://www.cnki.com.cn/Article/CJFDTOTAL-QXKX199701000.htm
[22]	戚春香, 李瑶, 杨简, 等. 青藏高原机场跑道多年冻土地基温度场特征[J]. 交通运输工程学报, 2019, 19(1): 33-47. https://www.cnki.com.cn/Article/CJFDTOTAL-JYGC201901007.htm QI C X, LI Y, YANG J, et al. Characteristics of multi-year permafrost foundation temperature field of airport runway on Qinghai-Tibet Plateau[J]. Journal of Transportation Engineering, 2019, 19(1): 33-47(in Chinese). https://www.cnki.com.cn/Article/CJFDTOTAL-JYGC201901007.htm
[23]	顾尔祚. 流体力学有限差分法基础[M]. 上海: 上海交通大学出版社, 1988. GU E Z. Fundamentals of finite difference method for fluid mechanics[M]. Shanghai: Shanghai Jiaotong University Press, 1988(in Chinese).
[24]	杨锡运, 邢国通, 马雪, 等. 一种核极限学习机分位数回归模型及风电功率区间预测[J]. 太阳能学报, 2020, 41(11): 300-306. https://www.cnki.com.cn/Article/CJFDTOTAL-TYLX202011040.htm YANG X Y, XING G T, MA X, et al. A kernel limit learning machine quantile regression model and wind power interval prediction[J]. Journal of Solar Energy, 2020, 41(11): 300-306(in Chinese). https://www.cnki.com.cn/Article/CJFDTOTAL-TYLX202011040.htm
[25]	SURJANOVIC S, BINGHAM D. Virtual library of simulation experiments: Test functions and datasets[EB/OL]. [2021-10-01]. http://www.sfu.ca/~ssurjano.
[26]	蔡赛男, 宋卫星, 班利明, 等. 基于鲸鱼算法优化LSSVM的滚动轴承故障诊断[J]. 控制与决策, 2022, 37(1): 230-236. https://www.cnki.com.cn/Article/CJFDTOTAL-KZYC202201025.htm CAI S N, SONG W X, BAN L M, et al. Fault diagnosis of rolling bearings based on whale algorithm optimized LSSVM[J]. Control and Decision Making, 2022, 37(1): 230-236(in Chinese). https://www.cnki.com.cn/Article/CJFDTOTAL-KZYC202201025.htm
[27]	谷远利, 张源, 芮小平, 等. 基于免疫算法优化LSSVM的短时交通流预测[J]. 吉林大学学报(工学版), 2019, 49(6): 1852-1857. https://www.cnki.com.cn/Article/CJFDTOTAL-JLGY201906015.htm GU Y L, ZHANG Y, RUI X P, et al. Short-time traffic flow prediction based on immune algorithm optimized LSSVM[J]. Journal of Jilin University (Engineering Edition), 2019, 49(6): 1852-1857(in Chinese). https://www.cnki.com.cn/Article/CJFDTOTAL-JLGY201906015.htm
[28]	吴桂芳, 崔勇, 刘宏, 等. 基于差分进化算法的三维电场传感器解耦标定方法[J]. 电工技术学报, 2021, 36(19): 3993-4001. https://www.cnki.com.cn/Article/CJFDTOTAL-DGJS202119004.htm WU G F, CUI Y, LIU H, et al. A decoupled calibration method for three-dimensional electric field sensors based on differential evolutionary algorithm[J]. Journal of Electrical Engineering Technology, 2021, 36(19): 3993-4001(in Chinese). https://www.cnki.com.cn/Article/CJFDTOTAL-DGJS202119004.htm
[29]	中国民用航空局. 民用机场水泥混凝土道面设计规范: MH/T 5004—2010[S]. 北京: 中国民用航空局, 2010. Civil Aviation Administration of China. Specification for airportcement concrete pavement design: MH/T 5004—2010[S]. Beijing: Civil Aviation Administration of China, 2010(in Chinese).
[30]	周哲韬, 刘路, 宋晓, 等. 基于Transformer模型的滚动轴承剩余使用寿命预测方法[J/OL]. 北京航空航天大学学报, 2021(2021-07-12)[2021-10-26]. https://doi.org/10.13700/j.bh.1001-5965.2021.0247. ZHOU Z T, LIU L, SONG X, et al. A method for predicting the remaining service life of rolling bearings based on Transformer model[J/OL]. Journal of Beijing University of Aeronautics and Astronautics, 2021(2021-07-12)[2021-10-26]. https://doi.org/10.13700/j.bh.1001-5965.2021.0247(in Chinese).
[31]	陈景龙, 王日新, 李玉庆, 等. 一种基于SD-ICA的卫星电池健康状态估计方法[J]. 北京航空航天大学学报, 2021, 47(10): 2058-2067. doi: 10.13700/j.bh.1001-5965.2020.0376#viewType=Abstract CHEN J L, WANG R X, LI Y Q, et al. A SD-ICA-based method for satellite battery health state estimation[J]. Journal of Beijing University of Aeronautics and Astronautics, 2021, 47(10): 2058-2067(in Chinese). doi: 10.13700/j.bh.1001-5965.2020.0376#viewType=Abstract