An Intelligent Variable Spraying Decision-Making System Based on Fuzzy Neural Network for Greenhouse Mobile Robot - Intelligent Computing for Sustainable Energy and Environment - page 257

An Intelligent Variable Spraying
Decision-Making System Based on Fuzzy Neural
Network for Greenhouse Mobile Robot
Guoqin Gao, Haiyan Zhou, and Xuemei Niu
School of Electrical & Information Engineering, Jiangsu University,
Zhen Jiang 212013, P.R. China
{ gqgao,niuxm } @ujs.edu.cn, zhouhy [e-mail]
Abstract. To improve the effective utilization rate of pesticide and re-
duce the pesticide residues and chemical pollution during spraying pro-
cess, an intelligent decision-making method for variable spraying based
on fuzzy neural network is designed according to the feature of the mo-
bile robot spraying in greenhouse, combined with the spraying principle
of variable spraying system for row-walking mobile robot. The decision
system of oine training fuzzy neural network is built by integrating the
information of the level of plant diseases and insect pests, the distance
and area of spraying target. The simulation results show that the fuzzy
neural network intelligent decision-making method can realize real-time
and quick decisions by off-line training. It has the greater decision ac-
curacy than the fuzzy decision system on the samples not appearing in
training because of its strong adaptability and generalization ability and
has a good fit for the uncertain work environment in greenhouse.
Keywords: intelligence decision-making, fuzzy neural network, variable
spraying, mobile robot.
1 Introduction
In the 21 century, great pressure in supplying agricultural and sideline prod-
ucts, low utilization of water and land resources and increasingly competition
in international agriculture market require China to develop precise agriculture.
Greenhouse is an important part of precise agriculture, so it is significant to
study its effective variable spraying decision system.
Greenhouse environment is thick airtight, high temperature, high humidity,
and easy to cause the spraying operator poisoning [1], thus it is important to
study the variable spraying adapting to the greenhouse environment to improve
the usage rate of spraying pesticide, and guarantee good health of operators. At
present, greenhouse pesticides are sprayed in the extensive way. The spraying
scheme is mostly decided by operator's experience, which is heavy workload,
and influenced by subjective factors. For a long time, the technology of using
pesticide still stays on the level of large capacity and large droplets, and the