Font Size: a A A

Study On Disease Detection And Precision Spray Method Of Fruit Trees And Key Technical Equipment

Posted on:2018-04-16Degree:DoctorType:Dissertation
Country:ChinaCandidate:F QuFull Text:PDF
GTID:1313330518497419Subject:Mechanical Manufacturing and Automation
Abstract/Summary:PDF Full Text Request
Precision spray for apple leaf diseases is the key section to improve the efficiency of pesticide used in apple orchard disease prevention. The premise of precision spray is the accurate acquisition of disease information, and then spray the inflected area according to the disease degree and distribution. This research put forward a grading evaluation method of apple leaf diseases under complex background using alternaira alternata apple pathotype and apple mosaic disease as examples, studied the profile modeling spray method for orchard air-assisted sprayer, and designed the key components for the tracked self-propelled orchard air-assisted sprayer. On the basis, the system integration, test and analysis were carried out. The main research contents and conclusions are as follows:(1) A method of grading evaluation for apple leaf diseases based on singular value decomposition(SVD) was studied. The color operator was used to preprocess the image of apple leaf disease,which removed the irrelevant region on the focal plane of the image and the background area with large chromatic aberration; an image segmentation method based on SVD for apple leaf diseases was proposed.This method removed the near color areas such as weeds, which were hard to be segmented by the conventional segmentation methods, by using the difference between the image clear foreground and the fuzzy background, then the complete disease leaf regions were obtained. The average segmentation accuracy was 96.28%. On the basis, the disease spots were segmented by the color feature, and then evaluated the disease grade according to the occupation standard.(2) A method of profile modeling spray for orchard air-assisted sprayer based on flow field simulation was studied. According to the design of response surface methodology (RSM) and the simulation of airflow velocity field based on Fluent, the influence of different factors on the distribution of airflow velocity field was analyzed. The second order response surface fitted model was established through comparison, and the multi-island genetic algorithm (MIGA) was used to optimize the parameters of the model. The optimum parameter combination was verified by the pole test, the results showed that the average relative error of the airflow velocity distribution at the approximate spindle profile was 8.82%,and the droplet coverage rate could also meet the needs of profile spray.(3) The key components for the tracked self-propelled orchard air-assisted sprayer were designed.The component of driving control unit was designed, which realized the remote control of gasoline engine drived sprayer. The component of spraying control based on the method of ultrasonic target and the method of flow control was designed, which realized the target variable spraying control of sprayer by using ultrasonic sensors and electric regulating valve. The monitoring component of air supply system of sprayer based on multi-sensor method was designed, the average accuracy of temperature measurement was 98.9%, and the speed measurement resolution was 30r/min. It could remind and upload fault information timely when the component monitored the abnormal of the air supply system.(4) An evaluation system for apple leaf diseases based on Android was developed. Combined the apple leaf diseases grading evaluation method with the mobile client, the evaluation software was developed and tested in an Android mobile phone. The software could identify the small disease spots and the accuracy was high. The average time of grading the disease leaf of artificial background and natural background was 0.91s and 8.74s, which showed the high efficiency of the software.(5) The performance of tracked self-propelled orchard air-assisted sprayer was tested. The effective remote control distance of the sprayer was about 50m, and the average traveling straightness error on the hardened road was 0.07m/10m. By using the target spraying mode and target variable spraying mode, the usage of pesticide could effectively be reduced. The parameters of air supply system of sprayer varied with the total working hours increased of the belt drive system, characterized the variations in belt drive stability.
Keywords/Search Tags:disease detection, image processing, singular value decomposition, computational fluid dynamics (CFD), precision spray
PDF Full Text Request
Related items