基于改进樽海鞘群寻优SVM的土壤含水量预测算法

doi:10.11959/j.issn.2096-3750.2021.00192

物联网学报 ›› 2021, Vol. 5 ›› Issue (1): 99-107.doi: 10.11959/j.issn.2096-3750.2021.00192

基于改进樽海鞘群寻优SVM的土壤含水量预测算法

赵小强, 杨帆, 晏珠峰

西安邮电大学，陕西西安 710121

修回日期:2021-01-15 出版日期:2021-03-30 发布日期:2021-03-01
作者简介:赵小强（1977- ），男，博士，西安邮电大学教授，主要研究方向为物联网技术及应用
杨帆（1994- ），男，西安邮电大学硕士生，主要研究方向为物联网技术及应用
晏珠峰（1995- ），男，西安邮电大学硕士生，主要研究方向为物联网技术及应用
基金资助:
国家自然科学基金委员会—雅砻江联合基金资助项目(U1965102);陕西省创新人才推进计划—科技创新团队(2019TD-28);西安市科技计划项目(201806117YF05NC13-2);陕西省教育厅产业化培育项目(18JF029);陕西省国际合作计划项目(2018KW-025)

Prediction method of soil water content based on SVM optimized by improved salp swarm algorithm

Xiaoqiang ZHAO, Fan YANG, Zhufeng YAN

Xi’an University of Posts and Telecommunications, Xi’an 710121, China

Revised:2021-01-15 Online:2021-03-30 Published:2021-03-01
Supported by:
The Yalong River Joint Funds of the National Natural Science Foundation of China(U1965102);The Shaanxi Innovative Talent Promotion Plan-Science and Technology Innovation Team(2019TD-28);The Science and Technology Projects of Xi’an(201806117YF05NC13-2);The Industrialization Cultivation Project of Shaanxi Provincial Department of Education(18JF029);The Shaanxi International Cooperation Project(2018KW-025)

摘要/Abstract

摘要：

针对传统土壤含水量预测算法存在的精度和效率较低等问题，采用支持向量机（SVM, support vector machine）建立预测模型，提出一种改进樽海鞘群算法（SSA, salp swarm algorithm）优化SVM的土壤含水量预测算法。首先，引入反向学习和混沌优化对标准樽海鞘群算法进行改进，解决算法易陷入局部最优解和收敛速度慢的问题；其次，利用改进的樽海鞘群算法对影响 SVM 性能的参数进行优化，构建对应的预测模型；最后，将所提模型与粒子群优化SVM预测模型、鲸鱼算法优化SVM预测模型进行对比。实验结果表明，所提模型的均方误差和决定系数分别为0.42和0.901，与其他两种模型相比性能更优，验证了所提算法的有效性。

关键词: 土壤含水量预测, 支持向量机, 樽海鞘群算法, 反向学习, 混沌优化

Abstract:

Aiming at the problems of low accuracy and low efficiency of traditional soil water content prediction methods, support vector machine (SVM) was used to establish a prediction model, and the soil water content prediction method based on SVM optimized was proposed by the improved salp swarm algorithm.Firstly, the opposition-based learning and chaotic optimization were introduced to improve the standard salp swarm algorithm to solve the problem that the algorithm was easy to fall into the local optimal solution and its convergence speed was slow.Secondly, the improved salp swarm algorithm was used to optimize the parameters that affect the performance of SVM and the corresponding prediction model was built.Finally, the proposed model was compared with the particle swarm optimization SVM and the whale algorithm optimized SVM prediction model.The experimental results show that the mean square error and decision coefficient of the proposed model are 0.42 and 0.901, which are better than the other two models which verified the effectiveness of the proposed method.

Key words: soil water content prediction, support vector machine, salp swarm algorithm, opposition-based learning, chaotic optimization

中图分类号:

TP181

赵小强, 杨帆, 晏珠峰. 基于改进樽海鞘群寻优SVM的土壤含水量预测算法[J]. 物联网学报, 2021, 5(1): 99-107.

Xiaoqiang ZHAO, Fan YANG, Zhufeng YAN. Prediction method of soil water content based on SVM optimized by improved salp swarm algorithm[J]. Chinese Journal on Internet of Things, 2021, 5(1): 99-107.

图/表 8

表1

测试函数的数学表达式及属性"

函数名称	计算式	范围	最优值
Sphere	$F_{1} (x) = \sum_{i = 1}^{n} x_{i}^{2}$	[-100,100]	0
Rastrigin	$F_{9} (x) = \sum_{i = 1}^{n} [x_{i}^{2} - 10 \cos (2 π x_{i}) + 10]$	[-5.12,5.12]	0
Griewank	$F_{11} (x) = \frac{1}{4 000} \sum_{i = 1}^{n} x_{i}^{2} - \prod_{i = 1}^{n} \cos (\frac{x_{i}}{\sqrt{i}}) + 1$	[-600,600]	0
Penalize2	$\begin{array}{l} F_{13} (x) = 0.1 {\sin^{2} (3 π x_{1}) + \sum_{i = 1}^{n} {(x_{i} - 1)}^{2} [1 + \sin^{2} (3 π x_{i} + 1)] + \\ {(x_{n} - 1)}^{2} [1 + \sin^{2} (2 π x_{n})]} + \sum_{i = 1}^{n} u (x_{i}, 5, 100, 4) \end{array}$	[-50,50]	0

表1

表2

图1

表3

表4

表5

图2

图3

参考文献 13

[1]	QIU J X , GAO Q Z , WANG S ,et al. Comparison of temporal trends from multiple soil moisture data sets and precipitation:the implication of irrigation on regional soil moisture trend[J]. International Journal of Applied Earth Observation and Geoinformation, 2016,48(5): 17-27.
[2]	王露娇 . 基于神经网络的茶园土壤墒情预测方法研究[D]. 合肥:安徽农业大学, 2018.
	WANG L J . Study on the prediction method of soil moisture in tea garden based on neural network[D]. Hefei:Anhui Agricultural University, 2018.
[3]	LEWIN J . Simple soil water simulation model for assessing the irrigation requirements of wheat[J]. Israel Journal of Agricultrical Research, 1972,22(4): 201-213.
[4]	王玮 . 黄河干流灌区土壤墒情动态预测与用水需求模型研究[D]. 西安:西安理工大学, 2019.
	WANG W . Study on dynamic prediction of soil moisture and water demand in the main reaches irrigation area of yellow river[D]. Xi’an:Xi’an University of Technology, 2019.
[5]	白冬妹, 郭满才, 郭忠升 . 时间序列自回归模型在土壤水分预测中的应用研究[J]. 中国水土保持, 2014(2): 42-45.
	BAI D M , GUO M C , GUO Z S . Application of time series autoregressive model in soil moisture prediction[J]. Chinese Society of Soil and Water Conservation, 2014(2): 42-45.
[6]	安小宇, 鲁奎豪, 崔光照 . 基于改进樽海鞘优化 BP 神经网络的土壤墒情预测[J]. 中国农机化学报, 2019,40(11): 124-130.
	AN X Y , LU K H , CUI G Z . Prediction of soil moisture based on BP neural network optimized by adaptive salp swarm algorithm[J]. Journal of Chinese Agricultural Mechanization, 2019,40(11): 124-130.
[7]	PRAKASH S , SHARMA A , SAHU S S . Soil moisture prediction using machine learning[C]// Proceedings of Second International Conference on Inventive Communication and Computational Technologies (ICICCT).[S.l.:s.n.], 2018: 1-6.
[8]	MIRJALILI S , GANDOMI A , MIRJALILI S Z ,et al. Salp swarm algorithm:a bio-inspired optimizer for engineering design problems[J]. Advances in Engineering Software, 2017,114(12): 163-191.
[9]	TIZHOOSH H R , . Opposition-based learning:a new scheme for machine intelligence[C]// Proceedings of IEEE International Conference on Computational Intelligence for Modelling,Control and Automation and International Conference on Intelligent Agents,Web Technologies and Internet Commerce (CIMCA-IAWTIC’06). Piscataway:IEEE Press, 2005: 695-701.
[10]	WANG H , WU Z , RAHNAMAYAN S ,et al. Enhancing particle swarm optimization using generalized opposition-based learning[J]. Information Sciences, 2011,181(20): 4699-4714.
[11]	KOHLI M , ARORA S . Chaotic grey wolf optimization algorithm for constrained optimization problems[J]. Journal of Computing and Information Science in Engineering, 2018,5(4): 458-472.
[12]	BONYADI M , MICHALEWICZ Z . Particle swarm optimization for single objective continuous space problems:a review[J]. Evolutionary Computation, 2017,25(1): 1-54.
[13]	MIRJALILI S , LEWIS A . The whale optimization algorithm[J]. Advances in Engineering Software, 2016,95(5): 51-67.

函数名称	统计值	OCSSA	PSO	SSA	WOA
Sphere	最优值	0	4.13×10^-6	4.15×10^-9	3.68×10^-88
	平均值	0	2.81×10^-5	5.69×10^-9	9.25×10^-78
	标准差	0	2.88×10^-5	1.74×10^-9	1.81×10^-77
Rastrigin	最优值	0	2.60×10¹	2.98×10¹	0
	平均值	0	3.23×10¹	4.76×10¹	5.68×10^-15
	标准差	0	5.90	9.96	1.04×10^-14
Griewank	最优值	0	3.94×10^-2	4.89×10^-7	0
	平均值	0	1.83×10^-1	6.41×10^-3	0
	标准差	0	1.18×10^-1	6.33×10^-3	0
Penalize2	最优值	6.00×10^-12	2.25×10^-7	1.10×10^-2	4.47×10^-2
	平均值	9.00×10^-12	2.10×10^-6	1.43×10^-2	1.81×10^-1
	标准差	1.69×10^-12	1.09×10^-6	6.62×10^-3	1.02×10^-1

时间	空气温度/℃	空气湿度	土壤温度/℃	土壤含水量	光照强度/lx
2019/9/9 12:00	40.4	47.6%	32.1	27.6%	23.9
2019/9/10 14:00	41.0	44.5%	34.3	26.4%	28.0
2019/9/11 16:00	41.4	38.9%	34.4	24.9%	31.8
2019/9/12 18:00	37.4	47.2%	32.2	24.4%	1.1
…	…	…	…	…	…
2019/9/23 13:00	40.4	17.3%	31.1	22.8%	23.3
2019/9/24 16:00	39.1	16.7%	31.7	22.0%	28.4

运行次数	模型参数C	模型参数γ	MSE	R²
1	14.843 0	7.356 1	0.42	0.903
2	20.999 9	7.047 6	0.39	0.909
3	17.957 1	6.583 7	0.44	0.897
4	9.728 3	5.233 8	0.44	0.897
5	17.326 9	6.878 5	0.43	0.901

模型名称	MSE	R²
OCSSA-SVM	0.42	0.901
SSA-SVM	0.59	0.864
PSO-SVM	0.67	0.845
WOA-SVM	0.62	0.859

基于改进樽海鞘群寻优SVM的土壤含水量预测算法

Prediction method of soil water content based on SVM optimized by improved salp swarm algorithm

在线阅读

PDF下载

可视化

摘要/Abstract

引用本文

使用本文

图/表 8

参考文献 13

相关文章 4

Metrics

推荐阅读 0

[1]	张彪, 汪西明, 徐逸凡, 李文, 韩昊, 刘松仪, 陈学强. 基于多智能体深度强化学习的多域协同抗干扰方法研究[J]. 物联网学报, 2022, 6(4): 104-116.
[2]	徐勇军, 杨浩克, 叶迎晖, 陈前斌, 卢光跃. 反向散射通信网络资源分配综述[J]. 物联网学报, 2021, 5(3): 56-69.
[3]	冯仕民,丁恩杰,刘忠育,俞啸,孟磊,赵志凯. 矿山人员状态信息描述方法研究[J]. 物联网学报, 2019, 3(4): 9-16.
[4]	冯仕民,刘忠育,俞啸,孟磊,赵志凯,丁恩杰. 矿山物联网人员情境描述与不安全行为识别[J]. 物联网学报, 2018, 2(4): 93-98.