Intelligent drone plant protection solution——a new farming experience in the context of artificial intelligence- 【鑫 芯 物 联】
Aviation technology, as one of the most rapidly developing science and technology in the past century, has had a rapid development in the agricultural field. Agricultural aviation application is an important part of modern agriculture , and also an important symbol reflecting the level of agricultural modernization . The combination of agriculture and aviation technology has brought about a comprehensive reform of agricultural production methods, which will definitely play a more important role in the future.
The plant protection drone uses the drone platform equipped with a pesticide application device to obtain remote sensing information for crops and quantify and precisely apply pesticides. It has the advantages of strong adaptability to complex terrain and high operating efficiency. UAVs can integrate a variety of new mission loads such as intelligent flight control systems, composite optoelectronic pods, precision variable spray equipment, etc. to provide comprehensive support for agricultural plant protection. In 2014, the "MIT Technology Review" released by the Massachusetts Institute of Technology put agricultural drone technology as the first place, and listed brain map technology, gene editing technology, neuromorphic chip, etc. as the top ten breakthrough scientific and technological innovations of the year. technology. Especially in China, with the rapid decline of the rural labor force and the maturity of drone technology, drone plant protection has developed more rapidly in China than any other country in the world.
Currently, benefiting from breakthroughs in artificial intelligence technology, parallel computing technology, and intelligent hardware, UAV plant protection technology is developing in an intelligent, systematic, and precise direction. In the future, intelligent plant protection drones with automatic identification and positioning of the operating area, autonomous planning of the operating path, fully automatic unmanned flight, and visible and traceable operation throughout the process will greatly improve the efficiency of agricultural operations and the working environment of agricultural workers. Intelligent drone plant protection will also become the core practice technology of smart plant protection, green plant protection, and scientific plant protection in China, which will play a core role in promoting agricultural modernization and intelligent technological upgrading and transformation in China, and for agricultural practitioners in the new era Bring a new farming experience.
Core technology composition of intelligent drone plant protection
1.Recognition and extraction of remote sensing crop information based on machine learning
In recent years, remote sensing information acquisition technology based on micro-drones has been widely used in the agricultural field. The micro-drone is equipped with a variety of photoelectric loads, which can quickly obtain farmland crop information. Compared with satellite remote sensing technology, it has the characteristics of low operating cost, flexibility and convenience, and high local data accuracy. It has unparalleled advantages in crop growth estimation, field facility investigation, and crop phenotypic information acquisition. As deep learning-based pattern recognition technology is gradually applied to the field of agricultural intelligent remote sensing information processing, drones are used as remote sensing platforms equipped with multi- sensors , which can quickly extract the surface characteristics of jobs and classify and identify job targets, thereby providing intelligent plant protection operations. basis.
2. Field work environment intelligent perception technology
Based on the field work prescription map, intelligent and dynamic sensing is performed on working objects and other obstacles in the environment, and pattern recognition methods are used to obtain key information such as basic scale characteristics and type attributes of obstacles, thereby performing dynamic operation control. With the rapid development of artificial intelligence technology and sensor technology, unmanned system control technology is moving from autonomous driving under established conditions in the past to real-time sensing of the environment and autonomous decision-making in real-time under non-predetermined conditions, thereby completing tasks autonomously.
3. Intelligent path planning technology for complex field conditions
Due to the limited load capacity and endurance time of drones, and at the same time, the operating area is mostly scattered and small areas with irregular areas, so reasonable operation path planning can reduce the non-operational flight time of agricultural drones. At present, simulation and demonstration operations have proved the effectiveness of such methods, but compared to other application fields, the research on intelligent plant protection operation path planning methods that introduce obstacles and multi-machine cooperative collision avoidance is still insufficient, and currently cannot meet the booming development. Demand for drone plant protection applications.
4.Droplet movement modeling and precise spraying control technology during application
Unlike ground machinery operations, the effects of UAV plant protection are affected by a variety of factors such as the environmental wind field, flight altitude, and flight speed. According to the current environmental wind field and operating parameters, predicting the droplet drift deposition area, so as to adjust the operating parameters in real time, is a precise means and goal of UAV plant protection. At present, traditional research methods mainly focus on the quantitative statistics of experimental data. Through aerial spray droplet deposition and drift tests, the effects of various types of drone field operations on droplet deposition were evaluated, and the effects of droplet height on flight height, droplet size spectrum, nozzle configuration, and weather conditions were studied.
5. Comprehensive intelligent evaluation system and evaluation technology for plant protection effects
The important evaluation indicators of the droplet distribution characteristics of crop liquids on crops include uniformity and sedimentation. Among them, the droplets' sedimentation is a direct indicator of the effectiveness of the chemical liquid control; the coverage and uniformity of the sedimentary liquid is optimized for aerial application. Important guidance for pharmaceutical equipment and technology solutions.
6. Traceable, controllable and visual technology based on cloud data management technology
UAV plant protection operations are still difficult to supervise in China. It is urgent to use informationized and intelligent means to effectively monitor the entire process of drone operations. We expect that based on data mining, a single management and control model will be transformed into a big data -based intelligent operation supervision model based on information services as the future development direction. UAV intelligent supervision systems currently in use all adopt the application mode of "onboard terminal + network cloud server". The airborne data acquisition terminal includes a variety of key operation status information sensors and mobile data communication modules. The collected operation height, flow, and flight speed information can be uploaded directly to the data server through the public data network. Using intelligent data processing methods, we can obtain macro information such as total regional work volume, types of pests and diseases, basic operation effects, and use of pesticides. At the same time, we can monitor crop growth, pest and disease outbreak trends, comprehensive effects of pesticide use, planting area changes, and regional yields. The information is dynamically predicted to provide information support for agricultural producers and competent authorities in responding to changes in the market of agricultural-related products.
Artificial intelligence technology is leading the era with unprecedented development speed, and it will definitely have a profound impact on the production, research and development, and application of agricultural drones.
(1) The introduction of artificial intelligence technology has pushed the development of drone remote sensing technology from the initial stage of information acquisition to a new stage of information mining, fusion, and regeneration. With the support of deep learning technology and massive data, various heterogeneous information The inversion between them will show us the details and essential laws of the crop growth process.
(2) The UAV field operation mode will change from the human-in-the-loop control mode to an autonomous control method based on field environment perception. The control of the operation process no longer depends on human experience and on-site judgment, but based on multi-source information fusion Intelligent decision-making to ensure reliable data support for the entire operation process control.
(3) With the accumulation of operating experience and the introduction of intelligent planning methods, UAV operations will first be optimized and quickly digitally deduced in the form of simulations. Most process state variables have been accurately estimated before the actual operation. The introduction of an intelligent plant protection operation path planning method that simultaneously introduces obstacles and multi-machine cooperative collision avoidance will become a hotspot for further research.
(4) The drone spray droplet movement model based on computational fluid dynamics simulation will be integrated into the drone flight control system, and with the support of variable airborne spraying equipment, the dynamic adjustment of the work path will be realized. Precision variable spraying of real-time environmental operating parameters.
(5) The key status information of the operation process will be accurately and quantitatively recorded. With the support of big data technology, the entire operation process becomes transparent, all links become controllable and traceable, and the modelling and data of the operation effect A comprehensive evaluation system will be finally established.
It can be expected that in the future, smart plant protection solutions with agricultural drones as the application core will bring people a new kind of agricultural experience.
May 25, 2018
Chengdu Xinxin Electronic Technology Co., Ltd. Editorial Department