Information Technology Reference
In-Depth Information
Input
Electromagnetic
Valve
PW M
Decision Maker
Controllor
Sprayer
Image
Collector
Ultrasonic
Sensor
Flow Sensor
Control
Decision
Fig. 1.
Structure of the variable spraying system
3 Decision-Making System of Variable Spraying
3.1 Design of Decision-Making System
To realize the automatic variable spraying, the decision is the key link and the
decision precision has a vital role in the implementation of variable spraying.
In this paper, an intelligent decision-making system based on the oine train-
ing fuzzy neural network is innovative designed. According to the bond crop in
greenhouse, it fuses not only the information of distance and area, but also the
level of plant diseases and insect pests in order to make the uncertain green-
house mobile robot spray eciently and accurately, reduce the use of pesticides,
improve the pesticides eciency, and cuts down the pollution of environment.
The decision-making system of variable spraying is shown in Fig. 2.
Offline Learning
Image Collector
Area
s
Fuzzy Neural Network
Pesticide Quantity
q
Ultrasonic Sensor
Distance
d
Damage
Level
n
Input
Decision Maker
Fig. 2.
Decision-making system of variable spraying
The camera sends the collected crop image to the computer through the USB
interface, and computer extracts and calculates the crop row from the back-
ground to decide the spraying area. In addition, the ultrasonic sensor is used to
measure the distance information of spraying target. Because of the complex-
ity working environment, especially the existences of uncertainty illumination
condition, and individual difference, it is dicult to distinguish and locate the
damage position quickly by collecting and processing the images real-timely. In
order to improve the eciency and practicality of spraying robots, and consid-
ering the features of plant diseases and insect pests in common glasshouse, the
Search WWH ::
Custom Search