果园变量喷雾技术研究现状与前景分析

果园变量喷雾是提高农药有效利用率、提升果品品质的重要手段之一,已经成为国内外学者研究的热点课题。为明确果园变量喷雾技术与装备已处的研究阶段、所面临的挑战和未来发展的方向,该文从果园变量喷雾技术中冠层结构探测与重构、施药智能决策和变量喷雾执行系统3个主要环节,重点概述了冠层结构探测的主要技术手段及其优缺点,认为机器视觉技术、超声波传感技术、LIDAR(light detection and ranging)探测技术及其相互之间的组合传感技术是未来最主要发展的冠层结构探测技术;综述了当前所采用的基于果园面积GA(ground area)模型、基于冠层高度的LWH(leaf wall height)模型、基于树体面积的LWA(leaf wall area)模型和基于冠层体积的TRV(tree row volum)模型,在此基础上阐述了这4种模型之间的内在联系;在分析了对靶开关决策、离散型决策和连续型决策模型的现状和特点的基础上,提出基于模糊算法的施药量智能连续决策是未来重要的发展方向;从果园变量喷雾机所采用的传感技术、决策模型和所取得的技术指标方面论述了当前世界最典型的装备现状,进一步分析了施药量调控系统、风量调控系统和喷雾位置调控系统的研究现状,提出了风量快速调控系统和喷雾位置快速响应系统的发展方向,以期为果园变量喷雾技术与装备研究提供参考。

Research situation and progress analysis on orchard variable rate spraying technology

Abstract: Orchard variable rate spraying technology is beneficial to enhance the pesticide utilization efficiency and fruit quality, which has been a hot research topic for scholars all over the world. In order to clear the research situation, facing challenge and future progress, first of all, a summarization of the variable rate spraying system was introduced in this paper. There are 3 technique links including geometrical characterization of tree canopy, spray volume calculation and decision- making model, variable rate spraying executive system. For the typical variable-rate sprayers from USA, China, Spain, Iran and Slovenia, the sensors mounted, the property index, the decision-making model used and the qualification acquired were listed in the paper. The first part of this paper presented the foundations of the main systems and their applications in agriculture for the geometrical characterization of tree canopy, including systems based on ultrasound, digital photographic techniques, light sensors, high-resolution radar images, computer vision and LIDAR (light detection and ranging). Amongst these methods, LIDAR laser scanners, computer stereo vision systems, ultrasound and inter-combination are probably the most promising and complementary techniques for achieving 3D (three-dimensional) pictures and maps of plants and canopies. In the second part, a range of empirical models for calculating appropriate plant protection products (PPP) dose in the orchard all over the world were presented. Empirical models contain the equations with variables that influence the PPP application in the orchard. Among apple producers, there are 4 different empirical models in common use, tree-row-volume (TRV) model, ground area (GA) model, leaf wall area (LWA) model and leaf wall height (LWH) model. A potential direction for future development of decision-making model could be focused on using a fuzzy logic system, which is able to continuously control the PPP dosage in a range from 0 to 100%. The advantage of a fuzzy logic system over other systems lies in its capacity to dynamically connect various types of input variables, such as the amount of the laser beams reflected from the tree canopy, the leaf area size, the LAI (leaf area index) and the sprayer travel speed, which enables the fuzzy system to quickly output the response values. The situations of the spray volume regulator system, air volume regulator system and nozzle position regulator system were introduced at the third part. After several years, the technology has been improved much. However, attention must be paid to the fact that discrepancies in the original measurement systems were too large, owing to imprecise operation of various sensor components. Nevertheless it is necessary to underline that LIDAR technology is a strong tool for developing variable rate spraying techniques, and with it, tree canopy properties are sensed in real time. So it is no longer a major problem in the precise application to establish the detection system of tree canopy properties, but to control artificial intelligence actuation, which in the future will properly direct the air flow and the dosage rate of PPP per tree canopy in the orchard. And the quick response system of air volume regulator system and nozzle position regulator system should be paid more attention in the future.