Education, Science, Technology, Innovation and Life
Open Access
Sign In

Data analysis of campus water supply system based on random forest algorithm and time series model

Download as PDF

DOI: 10.23977/acss.2022.060302 | Downloads: 16 | Views: 682

Author(s)

Changsheng Li 1, Shuo Liu 1, Sujun Liu 1

Affiliation(s)

1 College of Information Processing and Control Engineering, Lanzhou Petrochemical University of Vocational Technology, Lanzhou, China, 730060

Corresponding Author

Changsheng Li

ABSTRACT

Firstly, this paper makes statistics and Analysis on the water use characteristics of different functional areas in the campus, excavates the change law of each water meter data with time dimension and the proportion of total water use, so as to excavate the water use characteristics of different functional areas in the campus (such as office area, living area, logistics service area, etc.). Then, this paper constructs a random forest (DRF) model based on the decision tree algorithm, analyzes the "caliber" and "water consumption" of the water meter through the python language design algorithm and the hierarchical relationship of the campus water meter, then excavates the relationship model between the water meter data, and analyzes the classification error of the model through the existing data provided in the annex; At the same time, according to the collection time of each water meter reading, this paper analyzes the leakage of the water supply pipe network in the campus based on the time series model (ARMA), models and excavates the change law of water consumption from the first level table to the fourth level table in four quarters, and then monitors the leakage of the campus water supply pipe network system in real time through the error threshold of the model. Finally, combined with the VC dimension theory in statistical theory, this paper analyzes the relationship between the leakage degree of water transmission pipe network and the cost (labor cost and material cost) and water price of pipe network maintenance, and gives a feasible decision-making scheme for pipe network maintenance of campus water supply system.

KEYWORDS

Water supply system, Decision tree, Random forest (RF), Time series model (ARMA), Autocorrelation coefficient

CITE THIS PAPER

Changsheng Li, Shuo Liu, Sujun Liu, Data analysis of campus water supply system based on random forest algorithm and time series model. Advances in Computer, Signals and Systems (2022) Vol. 6: 10-18. DOI: http://dx.doi.org/10.23977/acss.2022.060302.

REFERENCES

[1] Theodoris et al., Li Jingjiao, Wang Aixia, Wang Jiao, et al., pattern recognition (Fourth Edition) Beijing: Electronic Industry Press, February 2010
[2] Li Ying, Qian Jianguo, Wang Xiao, Mo Jianguo, Shi zhengchai. Research on Intelligent Fault Diagnosis System of power grid system based on data mining [J]. Automation and instrumentation, 2020 (07): 205-208
[3] Shao Qi, Chen Yunhao, Yang Shuting, Zhao Yifei, Li Jing. Hyperspectral image identification of Maize Varieties Based on random forest algorithm [J]. Geography and geographic information science, 2019,35 (05): 34-39
[4] Wu Qinghua. Random forest algorithm and its application in metabolic fingerprint [D]. Central South University, 2013
[5] Lan Hua, Liao Zhimin, Zhao Yang. Output prediction of photovoltaic power station based on ARMA model [J]. Electrical measurement and instrumentation, 2011,48 (02): 31-35
[6] Yang Mao, Xiong Hao, Yan Gangui, Mu gang. Research on real-time prediction of wind power based on data mining and fuzzy clustering [J]. Power system protection and control, 2013,41 (01): 1-6
[7] Feng pan, Cao Xianbing. Empirical research on stock price analysis and prediction based on ARMA model [J]. Practice and understanding of mathematics, 2011,41 (22): 84-90

Downloads: 15903
Visits: 270739

Sponsors, Associates, and Links


All published work is licensed under a Creative Commons Attribution 4.0 International License.

Copyright © 2016 - 2031 Clausius Scientific Press Inc. All Rights Reserved.