New agricultural monitoring technologies based on satellite remote sensing, unmanned aerial vehicle remote sensing, intelligent sensing terminals and IoT network have made significant progress in their respective fields. However, it is difficult for single monitoring ways to meet the comprehensive sensing needs of modern agriculture, so it is urgent to develop a multi-source, multi-scale of space-air-ground cooperative monitoring and intelligent sensing system. The study first introduces some problems in China's agricultural development and points out the necessity of using space-air-ground information fusion technology; then analyzes the traditional sensing information technology of space, air and ground respectively. The applications of satellite remote sensing technology in different fields and the data acquisition and processing methods of UAV remote sensing technology in the subdivision fields of pest and disease detection, phenotype analysis and drought stress detection are summarized. Finally, the key technologies and networking methods of intelligent sensing terminal and IoT networks in ground networks are analyzed, and advantages and disadvantages, key technologies, and future development trends of space, air, ground network are analyzed respectively in detail. Combined with the deficiencies of the existing monitoring technology, this study summarizes the latest research results and applications of space, air and ground sensing technology in the integrated agricultural situation monitoring at home and abroad. Based on above analysis, the key technical problems that have not been solved in the research and application of space-air-ground integrated agricultural information sensing and fusion technology are pointed out. In addition, this study proposes that China's space-air-ground information sensing and fusion technology should develop in the direction of stability, refinement, and systematization, which provides a new perspective for the development and application of multi-source information fusion monitoring technology in the future. This study provides a reference for analyzing the hotspots of space, air and ground information sensing, breaking through the bottleneck of their applications, and grasping the development trend of integrated sensing and fusion technology, so as to enhance development of China's agricultural information perception in the way of three'dimension, preciseness and intelligence.

The livestock and poultry breeding industry is an important part of China's agriculture. The continuous growth of the scale of breeding has constantly increased the requirements for intelligent breeding technology. With the lack of breeding talents, the demand for intelligent and unmanned breeding technology is increasing. At present, China's aquaculture industry is at the initial stage of intelligent animal husbandry, and intelligent technology is gradually being applied to the livestock and poultry breeding industry, which is in its infancy. By summarizing the machine learning, image recognition, expert system, neural network, natural language processing and other technologies used by experts and scholars at home and abroad in the field of artificial intelligence in livestock and poultry breeding industry, this paper analyzes the application and development process of current artificial intelligence technology in the main application scenarios of livestock and poultry breeding industry, such as environmental control, precision feeding, health management, egg and milk management, and manure treatment. Based on the application of the most mainstream artificial intelligence technology at home and abroad in the field of livestock and poultry breeding, and the development of robot technology in replacing artificial breeding，through the continuous efforts of researchers and production personnel, artificial intelligence technology has been gradually applied in conventional feeding, environmental control and fecal sewage treatment, and as a result, intelligent robots for livestock breeding production management have being including disinfection and sterilization robots, egg picking robots, milking robots, intelligent patrol robots, and fecal cleaning robots. However, due to the late start of China's intelligent breeding industry, independent research and development, integrated intelligent breeding equipment and intelligent breeding technology have not been formed and there is also a lack of attention to the welfare of breeding animals, and the pollution generated in the breeding process is also relatively serious. On this basis, a proposal to research and develop multifunctional breeding robots based on domestic chips is put forward. By drawing on the experience of developed countries in animal husbandry, looking forward to the industrial demand and development prospect of the application of artificial intelligence technology in China's intelligent livestock and poultry farms, and putting forward technical research and industrial application suggestions, this study aims to provide a certain idea for the development of artificial intelligence technology in the field of livestock and poultry breeding in the future.

Plant protection UAVs (unmanned aerial vehicles) are usually operated in the form of flying defense teams. At present, the scheduling organization of the flying defense teams is relatively extensive, resulting in the low efficiency of the flying defense teams. Reasonable scheduling of UAVs can improve the operational efficiency of the flying defense teams. By taking the rice smut prevention and control scenarios as the research object, according to the operation specifications of “single-spray triple-prevention” against rice smut, referring to the current situation of the order modes of flying defense operation, an order management and plant protection UAVs scheduling model is proposed in this study, to ensure the operation quality and operation efficiency of plant protection UAVs. The model has two parts: (1) Order management, which is an order sorting method that comprehensively considers order work area, time window, and order urgency; (2) Scheduling model, which is a UAV scheduling model based on simulated annealing algorithm. Taking 16 plots in Nanjing area in China and rice smut control tasks of 4 flying defense teams for case study, the simulated annealing algorithm and the greedy algorithm were used to make a comparative study on the time window lengths of 3-6 days. Research results showed that, when the operation time window length is 3-5 days, the longer the time window, the shorter the scheduling distance and waiting time, the longer the total operation time, and the higher the total revenue. When the time window length is 6 days, the total operation time and operation income will not change. This research can provide a scientific basis for the deployment and decision-making analysis of the UAVs flying defense teams, and provide a reference for the development of the intelligent scheduling system for agricultural machinery.

Based on the patent analysis platform of patents, this article uses the bibliometric analysis method to measure, summarize and visually analyze the patent application amounts, patentees, inventors, technical topics and other classification indicators in the patent literature, so as to find out the development context, main innovation institutions, major innovation teams, core patents, key technologies concerned by the main applicants and others in Chinese tea and vegetables intelligent harvesting technology. The result and the conclusion can provide information support for promoting the innovation and development of tea and vegetables industry.

According to the characteristics of forage short season cotton continuous cropping mode in coastal saline alkali land, a new intelligent cotton planter based on Beidou navigation was designed. and the seed is made, and the layering is fertilized. The machine can complete the unmanned Beidou navigation, sowing belt cleaning, layered fertilization under seed, seed and fertilizer isolation, intelligent precision sowing, seed and fertilizer isolation, soil covering and pressing, surface shaping and other processes at one time. Field experiments are carried out with the machine. The results show that the accuracy of the link line and the straight end of the broadcast line are accurate, which improves the accuracy of the link line and the straightness of the broadcast line, and the average depth of the broadcast is 2.8 cm; The qualified rate of sowing depth was 92.5%; The accuracy of connecting line spacing is 1.2 cm; The straightness of sowing line was 0.8 cm, and the land use efficiency was increased by 5%. It lays the foundation for the follow-up standardized operation in the field, realizes the accurate control of sowing and fertilization, and saves seed and fertilizer.

With the rapid development of agriculture and animal husbandry in China, the demand for agricultural and animal husbandry machinery is increasing. Based on the design idea of agricultural machinery sharing and on-line cloud management, the cloud intelligent agricultural machinery sharing system based on Beidou platform was developed and applied to agricultural and animal husbandry machinery. It was described the overall structure and structure of the sharing system for agricultural machinery, developed the sharing platform APP for agricultural machinery, and used the Internet of things technology to realize the online visualization of the position of agricultural machinery, the Beidou satellite application provided a navigation service for agricultural machine operators to farm the fields of farmers, and can interact with information through an APP downloaded from a smartphone. Through the cloud intelligent system, farmers can choose the nearest farm machinery order, farm machinery can also login APP to receive orders or rush orders, so as to achieve rapid information exchange, improve production efficiency.

To realize the outdoor autonomous navigation of tracked robots, this paper proposes a localization method based on 3D lidar. First, lidar-IMU tightly coupled odometry and pose-graph optimization algorithms are used to establish a priori maps of the navigation area feature points. Second, to ensure that the robot can locate any position on the prior map, the satellite positioning information after latitude and longitude conversion is introduced into the point-cloud map information. Finally, the design uses the lightweight lidar odometry method to match the prior map to calculate the current pose of the robot. At the same time, to reduce the number of calculations, the nonlinear optimization area is limited to a rasterized small map. We use the GNSS carried by the robot to evaluate the algorithm in this paper: the maximum positioning error less than 0.26 m when the normal speed of the robot is 1.0 m/s, and the average positioning error less than 0.125 m. No localization failures occurred during the experiment, and the test results are good enough to meet the localization requirements of the tracked robot for outdoor navigation.

According to the demand of water and fertilizer integration in orchard, combined with the actual situation of experimental pear orchard, a set of water and fertilizer integration management and control system was designed. It is composed of water and fertilizer integrated system and management and control system, and adopts various irrigation forms, such as micro sprinkler irrigation, drip irrigation, alternate drip irrigation. The management and control system takes the water fertilizer integrated machine as the core, including data acquisition, water and fertilizer management and control and remote control unit, which can monitor the environment, equipment status and parameters, and make decisions and control. Users can view and operate through a variety of ways to achieve information collection, transmission, management, decision-making and control in one. A variety of water and fertilizer application schemes were designed and tested. The results show that the system runs well, and can accurately control different amounts of water and fertilizer. Under the condition of 19% fertilizer reduction, the growth, yield and quality of micro sprinkler irrigation remain unchanged, and the labor saving reaches 90%. Under the condition of 40% drip irrigation water reduction and 35% fertilizer reduction, the growth and quality remain unchanged and the yield increases. Combined with the water and fertilizer application rules, the system realized the integrated management and control of water and fertilizer, and achieved good results, which can provide reference for the design and application of orchard water and fertilizer integration.

Aiming at the lack of on-line monitoring system of rice harvester's crushing rate, impurities rate and loss rate, this paper constructs an on-line monitoring system of rice harvester's operation quality. The GPS module in the system can realize the real-time monitoring of field operation speed and operation position. On-line detection devices are used to monitor the crushing rate, impurities rate and the loss rate during field operations. Each function module realizes data communication with a man-machine interaction system through CAN bus. A field experiment was carried out to verify the accuracy of the system. The results show that the accuracy of the on-line monitoring system for the rice harvester operation quality is 82.76% for crushing rate, 78.69% for impurities rate, and 73.53% for loss rate. In the field test, when the manual test results of operation quality increase, the system test results increase correspondingly, and when the manual test results decrease, the system test results decrease correspondingly. Therefore, the detection results of manual and system on the change trend of operation quality are consistent. Therefore, the on-line monitoring system of operation quality of rice harvester constructed in this paper can realize visual monitoring, give an alarm in time when the working quality of harvester becomes worse and provide powerful technical support for intelligent rice harvester and research on adaptive control strategy.

Traditional plant protection machinery has low operating efficiency and low pesticide utilization rate, resulting in environmental pollution and excessive pesticide residues in agricultural products. It is urgent to improve the intelligent level of plant protection machinery by using artificial intelligence technology. This paper introduces the basic definition and connotation of artificial intelligence, the basic theory and key technology, analyzed the artificial intelligence technology in the plant protection machinery operation NongQing information acquisition, agricultural machinery, navigation and aviation plant protection technology development direction, puts forward six Suggestions of artificial intelligence, speed up the development of agriculture in order to speed up the artificial intelligence and plant protection mechanization technology provides the basis.

Due to the special natural climate conditions of Xinjiang, on the basis of drawing lessons from foreign existing models and combining with the characteristics of local cotton planting, agricultural machinery designers have developed a series of models of cotton drip irrigation seeder and its supporting monitoring technology and equipment. This paper introduces the development process of drip irrigation seeder under film in Xinjiang and the evolution process of the mainstream model structure, typical representative technology and the core working parts seeder, and forecasts the development trend of the seeder in the future by reviewing and combing the results.

The grain loss during the operation of combine harvester is directly related to the economic benefits of farmers. Mastering the accurate real-time grain loss information not only provide intuitive basis for the manipulator to adjust the working parameters of combine harvester that ensures the operation quality and improves the operation efficiency but also facilitate farmers or relevant government departments to track, supervise and measure the harvest loss of each region. The grain loss mainly includes the header loss, the unthreshed grain, the separating loss, the cleaning loss and leakage loss. The cleaning loss and separating loss account for a large proportion of grain loss. Therefore, it is necessary to monitor them in real time. Chinese researches on the monitoring methods of the separating loss are still at the theoretical and experimental stage, which is different from the researches oversea. Methods to improve the accuracy, reliability and adaptability of the separating loss sensors are difficult and essential in the information developing process of the grain combine harvester in our country. This paper summarizes the research progress of the monitoring methods and devices of grain combine harvester's separating loss at home and abroad from the aspects of the monitoring principles of the separating loss, the sensor signal processing system, the sensor damping method, the probability distribution model of separating loss. In addition, the developing trend will be analyzed. This will contribute to the continuous improving and upgrading of separation loss sensor and lay a solid foundation for the completion of information harvest in China .

It is necessary to distribute functions reasonably between a human operator and an automation system in a teleoperated agricultural robotic tractor to accomplish a task cooperatively. This paper proposes a strategy of dynamic function allocation on the basis of a BP neural network, genetic algorithm and adaptive genetic algorithm. Here, the operator's state, workload, and task demand are chosen as trigger mechanism of dynamic function allocation. Then, a traditional BP neural network, genetic algorithm based BP neural network, and adaptive genetic algorithm based BP neural network are established by taking the operator's state, workload, and task demand as inputs of the network and automation level as output. The three network are compared to obtain more effective dynamic function allocation. Simulation tests show that the adaptive genetic algorithm based BP neural network has minimum training time and has highest prediction accuracy.