无人机多光谱参数与Shapley分析融合的青贮玉米生物量估算

doi:10.3864/j.issn.0578-1752.2025.18.005

摘要/Abstract

摘要：

【目的】地上部生物量是衡量作物生长状况的重要指标。探究单一光谱参数模型与融合不同生长阶段模型在青贮玉米地上部生物量（above ground biomass，AGB）估测中的精度差异，旨在比较无人机多光谱特征参数及模型融合方法对青贮玉米AGB估测模型精度的影响，以提高宁夏青贮玉米生物量监测的准确性，为青贮玉米生物量的动态监测提供一种可行的技术方案。【方法】采用大疆无人机M300 RTK搭载M600 Pro型多光谱相机，获取6个不同氮素水平下青贮玉米各个生长阶段的多光谱影像数据，分析不同处理下光谱反射率和植被指数与青贮玉米地上部生物量变化关系。将青贮玉米全生育期数据集划分为营养生长阶段数据集和生殖生长阶段数据集，并对2个不同生长阶段数据集进行相关性分析，选择关联程度高的多光谱植被指数作为建模数据的输入量，利用随机森林（RF）、卷积神经网络（BP）等机器学习方法构建青贮玉米不同生长阶段的AGB估测模型，并使用灰狼优化算法进行模型优化。最后利用Shapley分析法，将优化的青贮玉米不同生长阶段AGB估算模型进行组合，得到具有多光谱变化特征的青贮玉米全生育期AGB估算模型。【结果】划分2个不同生长阶段可以提高青贮玉米生物量与多光谱植被指数的关联性，其中绿色叶绿素植被指数（green chlorophyll vegetation index，GCVI）提升最高，相关性绝对值达到了0.61和0.64；利用Shapley组合后的RF模型精度相对较高，R²为0.89、均方根误差（root mean square error，RMSE）为1.31 kg·m^-2；灰狼算法优化后的RF模型经Shapley组合后精度最高，R²为0.92、RMSE为1.11 kg·m^-2。【结论】在青贮玉米各生长阶段筛选最优光谱参数，并利用Shapley分析集成多阶段建模，能够有效提升青贮玉米AGB预测模型的精度。

关键词: 青贮玉米, 地上部生物量, 无人机, 多光谱, Shapley分析, 机器学习

Abstract:

【Objective】Aboveground biomass is an important indicator of crop growth, in order to explore the accuracy difference between single spectral parameter model and fusion of different growth stage models in silage maize above ground biomass (AGB) estimation, this study aimed to compare the effects of unmanned aerial vehicle (UAV) multispectral feature parameters and model fusion methods on silage maize AGB estimation modeling accuracy, so as to improve the accuracy of silage maize biomass monitoring in Ningxia, and to provide a feasible technological solution for silage maize biomass dynamic monitoring. 【Method】DJI UAV M300 RTK equipped with M600 Pro multispectral camera was used to acquire multispectral image data of silage maize at each growth stage under six different nitrogen levels, and the relationship between the spectral reflectance and vegetation index and the change of biomass of silage maize in the upper part of the ground under different treatments were analyzed. The data set of silage maize in the whole life cycle were classified into the nutrient growth stage data set and reproductive growth stage data set, and the correlation analysis of the two different growth stage data sets were carried out. The multispectral vegetation index with high degree of correlation was selected as the input of modeling data, and the AGB estimation model of silage maize at different growth stages were constructed by using machine learning methods, such as Random Forest (RF) and Convolutional Neural Network (BP). The model was optimized by using the Gray Wolf Optimization Algorithm, and finally optimizing the model by using the Shapley Analysis. The optimized AGB estimation model of different growth stages of silage maize was combined to obtain the AGB estimation model of silage maize with multi-spectral change characteristics for the whole reproductive period. 【Result】The division of two different growth stages could improve the correlation between silage maize biomass and multispectral vegetation index, in which the green chlorophyll vegetation index (GCVI) improved with the highest value, and the absolute values of the correlation reached 0.61 and 0.64; the accuracy of the RF model after the combination using Shapley analysis was relatively high, with R² of 0.89 and root mean square error (RMSE) of 1.31 kg·m^-2; The RF model optimized by Gray Wolf algorithm with the Shapley combination had the highest accuracy with R² of 0.92 and RMSE of 1.11 kg·m^-2. 【Conclusion】 In this study, screening the optimal spectral parameters at each growth stage of silage maize and integrating multi-stage modeling using Shapley analysis could effectively improve the accuracy of silage maize AGB prediction model.

Key words: silage maize, above ground biomass, UAV, multispectral, Shapley analysis, machine learning

韩林蒲, 马纪龙, 齐勇杰, 高嘉琪, 谢铁娜, 贾彪. 无人机多光谱参数与Shapley分析融合的青贮玉米生物量估算[J]. 中国农业科学, 2025, 58(18): 3632-3647.

HAN LinPu, MA JiLong, QI YongJie, GAO JiaQi, XIE TieNa, JIA Biao. Multispectral Unmanned Aerial Vehicle Parameters Combined with Machine Learning to Predict Silage Maize Biomass[J]. Scientia Agricultura Sinica, 2025, 58(18): 3632-3647.

0
/ / 推荐

导出引用管理器 EndNote|Reference Manager|ProCite|BibTeX|RefWorks

链接本文: https://www.chinaagrisci.com/CN/10.3864/j.issn.0578-1752.2025.18.005

https://www.chinaagrisci.com/CN/Y2025/V58/I18/3632

图/表 17

表1

图1

表2

图2

表3

表4

图3

表5

图4

图5

图6

图7

表6

表7

表8

表9

图8

参考文献 44

[1]	王晗, 向友珍, 李汪洋, 史鸿棹, 王辛, 赵笑. 基于无人机多光谱遥感的冬油菜地上部生物量估算. 农业机械学报, 2023, 54(8): 218-229.
	WANG H, XIANG Y Z, LI W Y, SHI H Z, WANG X, ZHAO X. Estimation of winter rapeseed above-ground biomass based on UAV multi-spectral remote sensing. Transactions of the Chinese Society for Agricultural Machinery, 2023, 54(8): 218-229. (in Chinese)
[2]	刘明星, 李长春, 李振海, 冯海宽, 杨贵军, 陶惠林. 基于高光谱遥感与SAFY模型的冬小麦地上生物量估算. 农业机械学报, 2020, 51(2): 192-202, 220.
	LIU M X, LI C C, LI Z H, FENG H K, YANG G J, TAO H L. Estimation of dry aerial mass of winter wheat based on coupled hyperspectral remote sensing and SAFY model. Transactions of the Chinese Society for Agricultural Machinery, 2020, 51(2): 192-202, 220. (in Chinese)
[3]	DU M M, NOGUCHI N. Monitoring of wheat growth status and mapping of wheat yield’s within-field spatial variations using color images acquired from UAV-camera system. Remote Sensing, 2017, 9(3): 289.
[4]	沈思聪, 张靖雪, 陈鸣晖, 李志威, 孙盛楠, 严学兵. 基于无人机多光谱估测不同品种紫花苜蓿的地上生物量和叶绿素含量. 光谱学与光谱分析, 2023, 43(12): 3847-3852.
	SHEN S C, ZHANG J X, CHEN M H, LI Z W, SUN S N, YAN X B. Estimation of above-ground biomass and chlorophyll content of different alfalfa varieties based on UAV multi-spectrum. Spectroscopy and Spectral Analysis, 2023, 43(12): 3847-3852. (in Chinese)
[5]	韩颖, 裴亮, 杜嘉. 洪河湿地植被地上生物量遥感反演研究. 遥感技术与应用, 2014, 29(2): 224-231. doi: 10.11873j.issn.1004-0323.2014.2.0224
	HAN Y, PEI L, DU J. Remote sensing inversion of aboveground biomass over the Honghe wetland. Remote Sensing Technology and Application, 2014, 29(2): 224-231. (in Chinese)
[6]	肖武, 陈佳乐, 笪宏志, 任河, 张建勇, 张雷. 基于无人机影像的采煤沉陷区玉米生物量反演与分析. 农业机械学报, 2018, 49(8): 169-180.
	XIAO W, CHEN J L, DA H Z, REN H, ZHANG J Y, ZHANG L. Inversion and analysis of maize biomass in coal mining subsidence area based on UAV images. Transactions of the Chinese Society for Agricultural Machinery, 2018, 49(8): 169-180. (in Chinese)
[7]	LIU T, YANG T L, ZHU S L, MOU N N, ZHANG W J, WU W, ZHAO Y Y, YAO Z S, SUN J J, CHEN C, SUN C M, ZHANG Z J. Estimation of wheat biomass based on phenological identification and spectral response. Computers and Electronics in Agriculture, 2024, 222: 109076.
[8]	DUBE T, MUTANGA O. The impact of integrating WorldView-2 sensor and environmental variables in estimating plantation forest species aboveground biomass and carbon stocks in uMgeni Catchment, South Africa. ISPRS Journal of Photogrammetry and Remote Sensing, 2016, 119: 415-425.
[9]	HAN S Y, ZHAO Y, CHENG J P, ZHAO F, YANG H, FENG H K, LI Z H, MA X M, ZHAO C J, YANG G J. Monitoring key wheat growth variables by integrating phenology and UAV multispectral imagery data into random forest model. Remote Sensing, 2022, 14(15): 3723.
[10]	贾元, 张琳, 吴冬秀, 宋创业, 袁伟影, 李凌浩. 基于无人机多光谱实测数据的草地生物量反演模型比较. 生态学报, 2024, 44(15): 6854-6864.
	JIA Y, ZHANG L, WU D X, SONG C Y, YUAN W Y, LI L H. Comparative analysis of grassland biomass inversion models based on unmanned aerial vehicle multispectral data. Acta Ecologica Sinica, 2024, 44(15): 6854-6864. (in Chinese)
[11]	AKBARIAN S, XU C Y, WANG W J, GINNS S, LIM S. An investigation on the best-fit models for sugarcane biomass estimation by linear mixed-effect modelling on unmanned aerial vehicle-based multispectral images: A case study of Australia. Information Processing in Agriculture, 2023, 10(3): 361-376.
[12]	WEI L L, YANG H S, NIU Y X, ZHANG Y N, XU L Z, CHAI X Y. Wheat biomass, yield, and straw-grain ratio estimation from multi- temporal UAV-based RGB and multispectral images. Biosystems Engineering, 2023, 234: 187-205.
[13]	ZHANG P P, LU B, GE J Y, WANG X Y, YANG Y D, SHANG J L, LA Z, ZANG H D, ZENG Z H. Using UAV-based multispectral and RGB imagery to monitor above-ground biomass of oat-based diversified cropping. European Journal of Agronomy, 2025, 162: 127422.
[14]	TEN HARKEL J, BARTHOLOMEUS H, KOOISTRA L. Biomass and crop height estimation of different crops using UAV-based lidar. Remote Sensing, 2020, 12(1): 17.
[15]	张宏鸣, 侯贵河, 孙志同, 杨欢瑜, 韩柯城, 韩文霆. 基于XGBoost-Shapley的玉米不同生育期LAI遥感估算. 农业机械学报, 2022, 53(7): 208-216, 225.
	ZHANG H M, HOU G H, SUN Z T, YANG H Y, HAN K C, HAN W T. Remote sensing estimation of maize leaf area index at different growth periods based on XGBoost-Shapley algorithm. Transactions of the Chinese Society for Agricultural Machinery, 2022, 53(7): 208-216, 225. (in Chinese)
[16]	GONG P, PU R L, YU B. Conifer species recognition: an exploratory analysis of in situ hyperspectral data. Remote Sensing of Environment, 1997, 62(2): 189-200.
[17]	RONDEAUX G, STEVEN M, BARET F. Optimization of soil- adjusted vegetation indices. Remote Sensing of Environment, 1996, 55(2): 95-107.
[18]	KYRATZIS A C, SKARLATOS D P, MENEXES G C, VAMVAKOUSIS V F, KATSIOTIS A. Assessment of vegetation indices derived by UAV imagery for durum wheat phenotyping under a water limited and heat stressed Mediterranean environment. Frontiers in Plant Science, 2017, 8: 1114. doi: 10.3389/fpls.2017.01114 pmid: 28694819
[19]	DANG Y P, PRINGLE M J, SCHMIDT M, DALAL R C, APAN A. Identifying the spatial variability of soil constraints using multi-year remote sensing. Field Crops Research, 2011, 123(3): 248-258.
[20]	CAMPOS I, NEALE C M U, ARKEBAUER T J, SUYKER A E, GONÇALVES I Z. Water productivity and crop yield: A simplified remote sensing driven operational approach. Agricultural and Forest Meteorology, 2018, 249: 501-511.
[21]	井宇航, 郭燕, 张会芳, 戎亚思, 张少华, 冯伟, 王来刚, 贺佳, 刘海礁, 郑国清. 无人机飞行高度对冬小麦植株氮积累量预测模型的影响. 河南农业科学, 2022, 51(2): 147-158.
	JING Y H, GUO Y, ZHANG H F, RONG Y S, ZHANG S H, FENG W, WANG L G, HE J, LIU H J, ZHENG G Q. Effects of UAV flight height on prediction model of plant nitrogen accumulation in winter wheat. Journal of Henan Agricultural Sciences, 2022, 51(2): 147-158. (in Chinese) doi: 10.15933/j.cnki.1004?3268.2022.02.018
[22]	邵国敏, 王亚杰, 韩文霆. 基于无人机多光谱遥感的夏玉米叶面积指数估算方法. 智慧农业, 2020, 2(3): 118.
	SHAO G M, WANG Y J, HAN W T. Estimation method of leaf area index for summer maize using UAV-based multispectral remote sensing. Smart Agriculture, 2020, 2(3): 118. (in Chinese) doi: 10.12133/j.smartag.2020.2.3.202006-SA001
[23]	任枫, 王琦, 杨佳, 任庆福. 基于高分六号卫星红边波段的森林蓄积量遥感反演: 以西宁市针叶林为例. 中国农业气象, 2022, 43(5): 408-420.
	REN F, WANG Q, YANG J, REN Q F. Retrieving of the forest stock volume based on the red edge bands of GF-6 remote sensing satellite: A case study of coniferous forest in Xining City. Chinese Journal of Agrometeorology, 2022, 43(5): 408-420. (in Chinese)
[24]	XING N C, HUANG W J, XIE Q Y, SHI Y, YE H C, DONG Y Y, WU M Q, SUN G, JIAO Q J. A transformed triangular vegetation index for estimating winter wheat leaf area index. Remote Sensing, 2020, 12(1): 16.
[25]	肖芳芳, 张洪艳, 贺威, 张良培. 自动化样本生成策略用于冬季作物制图——以兰陵县为例. 测绘通报, 2022 (12): 121. doi: 10.13474/j.cnki.11-2246.2022.0367
	XIAO F F, ZHANG H Y, HE W, ZHANG L P. Automated sample generation strategy for winter crop mapping: A case study in Lanling county. Bulletin of Surveying and Mapping, 2022 (12): 121. (in Chinese) doi: 10.13474/j.cnki.11-2246.2022.0367
[26]	姜友谊, 刘博伟, 张成健, 赵丹, 陈日强, 徐波, 龙慧灵, 杨贵军, 杨浩. 利用无人机多光谱影像的多品种玉米成熟度监测. 农业工程学报, 2023, 39(20): 84-91.
	JIANG Y Y, LIU B W, ZHANG C J, ZHAO D, CHEN R Q, XU B, LONG H L, YANG G J, YANG H. Multi-variety maize maturity monitoring based on UAV multi-spectral images. Transactions of the Chinese Society of Agricultural Engineering, 2023, 39(20): 84-91. (in Chinese)
[27]	苏子龙, 严文亮, 李慧敏, 高林燕, 寿文琪, 吴军. 基于改进麻雀搜索算法优化BP神经网络的农业碳排放预测. 环境科学, 2024, 45(12): 6818-6827.
	SU Z L, YAN W L, LI H M, GAO L Y, SHOU W Q, WU J. Prediction of agricultural carbon emission based on improved BP neural network with optimized sparrow search algorithm. Environmental Science, 2024, 45(12): 6818-6827. (in Chinese)
[28]	张晓凤, 王秀英. 灰狼优化算法研究综述. 计算机科学, 2019, 46(3): 30-38. doi: 10.11896/j.issn.1002-137X.2019.03.004
	ZHANG X F, WANG X Y. Comprehensive review of grey wolf optimization algorithm. Computer Science, 2019, 46(3): 30-38. (in Chinese)
[29]	王俊, 冯军, 张戈, 王建林, 王胜, 郑泰皓. 基于改进灰狼优化算法的医学数据特征选择应用研究. 河南大学学报(自然科学版), 2020, 50(5): 570-578.
	WANG J, FENG J, ZHANG G, WANG J L, WANG S, ZHENG T H. Application research of medical data feature selection based on improved grey wolf optimization algorithm. Journal of Henan University (Natural Science), 2020, 50(5): 570-578. (in Chinese)
[30]	AN Q X, WEN Y, DING T, LI Y L. Resource sharing and payoff allocation in a three-stage system: Integrating network DEA with the Shapley value method. Omega, 2019, 85: 16-25.
[31]	运彬媛, 张昊, 翟勇全, 马健祯, 姬丽, 李稼润, 金学兰, 贾彪. 基于土壤氮素水平的玉米冠层SPAD值估算方法. 生态学杂志, 2024, 43(5): 1488-1497.
	YUN B Y, ZHANG H, ZHAI Y Q, MA J Z, JI L, LI J R, JIN X L, JIA B. Estimation method for canopy SPAD values of maize based on soil nitrogen level. Chinese Journal of Ecology, 2024, 43(5): 1488-1497. (in Chinese) doi: 10.13292/j.1000-4890.202405.005
[32]	王帝, 孙榕, 苏勇, 杨博. 基于无人机多光谱影像的水稻生物量估测. 农业工程学报, 2024, 40(17): 161-170.
	WANG D, SUN R, SU Y, YANG B. Rice biomass estimation based on multispectral imagery from unmanned aerial vehicles. Transactions of the Chinese Society of Agricultural Engineering, 2024, 40(17): 161-170. (in Chinese)
[33]	PRUDNIKOVA E, SAVIN I, VINDEKER G, GRUBINA P, SHISHKONAKOVA E, SHARYCHEV D. Influence of soil background on spectral reflectance of winter wheat crop canopy. Remote Sensing, 2019, 11(16): 1932.
[34]	GAMON J A, PEÑUELAS J, FIELD C B. A narrow-waveband spectral index that tracks diurnal changes in photosynthetic efficiency. Remote Sensing of Environment, 1992, 41(1): 35-44.
[35]	GITELSON A A, KAUFMAN Y J, STARK R, RUNDQUIST D. Novel algorithms for remote estimation of vegetation fraction. Remote Sensing of Environment, 2002, 80(1): 76-87.
[36]	MUTANGA O, SKIDMORE A K, VAN WIEREN S. Discriminating tropical grass (Cenchrus ciliaris) canopies grown under different nitrogen treatments using spectroradiometry. ISPRS Journal of Photogrammetry and Remote Sensing, 2003, 57(4): 263-272.
[37]	贾洁琼, 刘万青, 孟庆岩, 孙云晓, 孙震辉. 基于GF-1 WFV影像和机器学习算法的玉米叶面积指数估算. 中国图象图形学报, 2018, 23(5): 719-729.
	JIA J Q, LIU W Q, MENG Q Y, SUN Y X, SUN Z H. Estimation of maize leaf area index based on GF-1 WFV image and machine learning random algorithm. Journal of Image and Graphics, 2018, 23(5): 719-729. (in Chinese)
[38]	梁昔明, 李星, 龙文. 融合鲸鱼算法的混合灰狼优化算法. 数学的实践与认识, 2022, 52(6): 130-138.
	LIANG X M, LI X, LONG W. Hybrid grey wolf optimization algorithm fused with whale algorithm. Mathematics in Practice and Theory, 2022, 52(6): 130-138. (in Chinese)
[39]	曾妍, 王迪. 极化SAR旱地作物分类的后向消除特征选择研究. 中国农业资源与区划, 2023, 44(3): 222-232.
	ZENG Y, WANG D. Backward elimination feature selection for classification of dryland crops using polarimetric SAR data. Chinese Journal of Agricultural Resources and Regional Planning, 2023, 44(3): 222-232. (in Chinese)
[40]	LUNDBERG S M, LEE S I. A unified approach to interpreting model predictions. Advances in Neural Information Processing Systems, 2017, 30: 4768-4777.
[41]	LI K F, WAN Y T, MA A L, ZHONG Y F. A lightweight multiscale and multiattention hyperspectral image classification network based on multistage search. IEEE Transactions on Geoscience and Remote Sensing, 2025, 63: 5509418.
[42]	ZHANG P P, LU B, SHANG J L, WANG X Y, HOU Z W, JIN S J, YANG Y D, ZANG H D, GE J Y, ZENG Z H. Ensemble learning for oat yield prediction using multi-growth stage UAV images. Remote Sensing, 2024, 16(23): 4575.
[43]	周敏姑, 闫云才, 高文, 何景源, 李鑫帅, 牛子杰. 基于多光谱遥感和CNN的玉米地上生物量估算模型. 农业机械学报, 2024, 55(9): 238-248.
	ZHOU M G, YAN Y C, GAO W, HE J Y, LI X S, NIU Z J. Estimating aboveground biomass of maize based on multispectral remote sensing and convolutional neural network. Transactions of the Chinese Society for Agricultural Machinery, 2024, 55(9): 238-248. (in Chinese)
[44]	运彬媛, 谢铁娜, 李虹, 岳翔, 吕明玥, 王佳琦, 贾彪. 融合无人机光谱与纹理信息的玉米氮素营养估测. 中国农业科学, 2024, 57(16): 3154-3170. doi:10.3864/j.issn.0578-1752.2024.16.005.
	YUN B Y, XIE T N, LI H, YUE X, LÜ M Y, WANG J Q, JIA B. Nitrogen nutrition estimation of maize based on UAV spectrum and texture information. Scientia Agricultura Sinica, 2024, 57(16): 3154-3170. doi:10.3864/j.issn.0578-1752.2024.16.005. (in Chinese)

年份 Year	pH	有机质 Organic matter (g·kg^-1)	全氮 Total nitrogen (g·kg^-1)	全磷 Total phosphorus (g·kg^-1)	碱解氮 Alkaline hydrolysis nitrogen (mg·kg^-1)	速效磷 Rapidly available phosphorus (mg·kg^-1)	速效钾 Rapidly available potassium (mg·kg^-1)
2023	7.85	12.75	0.39	0.66	36.75	14.96	98.54
2024	8.19	13.77	0.44	0.78	40.25	15.19	110.42

光谱波段 Spectral band	光谱名称 Spectral name	中心波长 Center wavelength (nm)	波宽 Wave length (nm)	灰板反射率 Gray board reflectance (%)
B1	蓝Blue (B)	450	30	0.60
B2	绿Green (G)	555	27	0.60
B3	红Red (R)	660	22	0.59
B4	红边Red edge (RE)	720	10	0.60
B5	红边750 Red edge 750 (750)	750	10	0.60
B6	近红外Near infrared reflectance (NIR)	840	30	0.60

植被指数 Vegetation index	公式 Formula	参考文献 Reference
绿色归一化差异植被指数 Green normalized difference vegetation index (GNDVI)	$G N D V I=\left(B_{6}-B_{2}\right) /\left(B_{6}+B_{2}\right) $	[18]
归一化差异植被指数 Normalized difference vegetation index (NDVI)	$N D V I=\left(B_{6}-B_{3}\right) /\left(B_{6}+B_{3}\right) $	[19]
修正土壤植被指数 Modified soil vegetation index (MSAVI)	$M S A V I=0.5\left\{2 B_{6}+1-\left[\left(2 B_{6}+1\right)^{2}-8\left(B_{6}-B_{3}\right)\right]^{0.5}\right\}$	[20]
重归一化植被指数 Re-normalized difference vegetation index (RDVI)	$R D V I=\sqrt{\left(B_{6}-B_{3}\right) /\left(B_{6}+B_{3}\right)} $	[21]
优化土壤调节植被指数 Optimized soil adjusted vegetation index (OSAVI)	$O S A V I=1.16\left(B_{6}-B_{3}\right) /\left(B_{6}+B_{3}+0.16\right) $	[22]
红边叶绿素指数 Red edge chlorophyll index (NDREI)	$N D R E I=\left(B_{5}-B_{4}\right) /\left(B_{5}+B_{4}\right) $	[23]
改进简单比值植被指数 Improved simple ratio vegetation index (MSR)	$M S R=\left(B_{5}-B_{4}\right) /\left(B_{4}+B_{3}\right) $	[24]
绿色叶绿素植被指数 Green chlorophyll vegetation index (GCVI)	$G C V I=\mathrm{B}_{6} /\left(\mathrm{B}_{2}-1\right) $	[25]

生长阶段 Growth period	样本数量 Sample number	最大值 Maximum values	最小值 Minimum value	平均值 Average value	标准差 Standard deviation	变异系数 Coefficient of variation
V6—V9	60	7.15	2.66	5.33	1.09	20.58%
V9—V12	60	8.88	4.18	7.11	1.16	16.35%
V12—R1	60	13.47	5.16	9.20	2.44	26.56%
R1—R5	60	18.38	7.43	13.50	3.36	24.93%

模型 Model	V6—V9	V9—V12	V12—R1	R1—R5
随机森林RF	0.31	/	/	0.49
灰狼优化的随机森林 GWO-RF	0.39	/	/	0.52
反向传播神经网络 BP	/	/	/	0.13
灰狼优化的反向传播神经网络 GWO-BP	0.05	/	/	0.24