Click on the article title below “High Technology and Industrialization”
Or search for the public account: hitechcas
Track the development trends of high-tech industries
Smart Applications of Agricultural IoT
With the development of smart agriculture, the application of IoT technologies such as intelligent sensing chips and mobile embedded systems in modern agriculture is gradually expanding. By using wireless sensor networks, labor consumption and the impact on farmland environments can be effectively reduced, obtaining precise crop environmental and crop information, thus allowing for extensive use of various automated, intelligent, and remote-controlled production equipment, enabling monitoring of farmland information without leaving home. The application of agricultural IoT technology can better control the growth environment of crops, making them better suited for growth, improving crop yield and quality, and contributing to achieving high and stable yields of crops, enhancing land productivity, and improving agriculture’s ability to withstand natural disasters.
The technical system of agricultural IoT is illustrated in Figure 2, which roughly includes four parts: information collection, transmission storage, analysis decision-making, and intelligent control. Information collection mainly applies agricultural information collectors to manage agricultural environments, crops, products, etc., forming a series of economical and practical sensing devices, including sensors for temperature, humidity, light, CO2, and dew point. Transmission storage is developed for different agricultural production environments, integrating various types of information transmission devices such as wired and wireless, focusing on research into agricultural IoT information fusion, knowledge discovery, heterogeneous network access, etc., ensuring timely, reliable, and accurate information transmission. Analysis decision-making builds various crop growth models based on different crops (animals), varieties, growth periods, and diurnal growth patterns, determining the suitable environment required for crops based on these models, serving as the basis for environmental regulation. Intelligent control uses agricultural intelligent control devices for real-time measurement, display, storage of indoor and outdoor environmental data, and controlling the conditions of on-site environmental control devices, while also expanding additional features such as multilingual support, multiple control conditions, multiple modes, and historical data queries.
The analysis and decision-making of agricultural information is key to smart agriculture, being the most important link in the entire smart agriculture chain. The agricultural expert system is based on agricultural expert knowledge, mimicking the reasoning and decision-making of agricultural experts, integrating multiple agricultural technologies and knowledge to provide users with production technology consulting and decision-making services for variety selection, fertilizer and water management, pest and weed control, and agronomic management.
Agricultural Expert System Development Platform
The agricultural expert system development platform is a tool designed for the rapid development of agricultural expert systems. Based on this platform, secondary development can significantly reduce the workload and technical difficulty of developing agricultural expert systems. The expert system development platform helps researchers acquire, represent, and apply knowledge; assists system designers in structuring expert systems; and provides an internal software environment to enhance internal communication capabilities.
The agricultural expert system integrates scattered, localized individual agricultural technologies through intelligent information processing, providing systematic and adaptable solutions to various agricultural problems based on different soil and climate conditions.
From a business perspective, the agricultural expert system can simulate agricultural experts in addressing all issues during agricultural production. As shown in Figure 3, it includes over ten directions: breeding of plants and animals, selection of quality breeds, pest detection, pest forecasting and early warning, agricultural teaching (videos, animations, charts, text), testing of new varieties for plant protection, soil quality safety testing, soil testing and formula fertilization, suitability evaluation for production areas, and crop yield estimation.
The expert system can solve various problems for agricultural producers like an agricultural expert, providing various suggestions and guidance at any time, enabling agricultural producers to engage in more effective agricultural production. At the same time, the agricultural expert system can perform routine daily tasks and reduce the time taken to address issues in the agricultural production field, thus freeing agricultural experts from heavy workloads to focus on more critical research.
The agricultural expert system emphasizes the role of specialized agricultural knowledge and reasoning judgment more than ordinary computer information systems, possessing a stronger decision-making and consulting ability tailored to specific needs, compared to agricultural experts who have comprehensive knowledge and fast knowledge processing capabilities, unaffected by time, space, or human emotions. The agricultural expert system consolidates high-level agricultural knowledge, overcoming barriers posed by inconvenient transportation and information in rural areas, guiding agricultural production on-site, alleviating the shortage of agricultural technicians and the disparity in their skill levels. The agricultural expert system gathers local agricultural knowledge, solidifying the valuable experiences of agricultural experts and skilled farmers, forming a local agricultural production technology database and an agricultural environmental database. The timeliness of the agricultural expert system in agricultural production is significant, addressing problems encountered by farmers in real-time, greatly reducing the delay in problem resolution.
Core Principles of the Agricultural Expert System
Knowledge acquisition mechanism: The problem-solving ability of the agricultural expert system mainly depends on the quantity and quality of agricultural knowledge it uses. Knowledge engineers input agricultural knowledge summarized from agricultural field experts into the knowledge base, running the agricultural expert system for verification, and repeatedly modifying it with agricultural field experts to achieve the best results.
Knowledge base and its management system: The knowledge base includes general knowledge and domain knowledge required during reasoning.
Inference engine: The inference engine is the module or program based on knowledge reasoning, serving as the “thinking” mechanism (core) of the agricultural expert system. The inference engine mainly consists of reasoning and control aspects.
Explanation mechanism: The explanation mechanism provides a clear, complete, and easy-to-understand answer to questions posed by users or system designers, thus rationally explaining its behavior and results. The explanation mechanism increases the system’s transparency, thereby enhancing its acceptability. Similarly, explanations are an important means for knowledge engineers to discover and correct defects and errors in the knowledge base.
Another crucial role of explanations is to educate novice users, allowing them to possess knowledge similar to that of agricultural field experts.
Database and its management system: The database, also known as the “blackboard,” records control information, intermediate hypotheses, intermediate results, and conclusions used during the system’s reasoning process. When the agricultural expert system starts running, the initial facts obtained are first placed on the blackboard, and the inference engine reasons based on these facts and the knowledge base, continuously placing intermediate results and hypotheses on the blackboard until a final conclusion is reached or reasoning fails, resulting in an exit.
Human-computer interface: The interface is the medium for information exchange between humans and the system, providing users with a relatively intuitive and convenient means. The interface system receives user questions, known data, and other information, processes them, and forwards them to the expert system. At the same time, the interface system must output the expert system’s responses to user inquiries, reasoning results, and explanations of reasoning results in a user-understandable format.
Technical Integration of Agricultural Expert Systems
Due to the characteristics of agricultural production, the foundational data in expert systems must not only be massive but also dynamic. The knowledge base, database, and model library must continuously incorporate new knowledge, data, and technologies to support and quickly resolve practical agricultural production issues.
The Global Positioning System (GPS) can accurately determine the spatial location of objects, providing foundational data. Geographic Information Systems (GIS) can achieve qualitative and quantitative descriptions of geographical entities, supplying essential foundational data and data analysis and visualization for expert systems. Remote sensing technology (RS) can also provide data for expert systems. These large and comprehensive data sets alleviate the shortage of knowledge and data sources in agricultural expert systems, providing data support for the establishment of foundational databases, knowledge databases, and model libraries; additionally, the data is dynamic, updatable, and offers a wide range of information.
The integration of agricultural expert systems with multimedia technology enhances the system’s visualization. Visuals often play a crucial role in decision-making. By combining expert experience with multimedia technology in pest images, weed morphology, disease morphology, etc., the system’s diagnostic performance is improved.
The IoT-based agricultural expert system utilizes wireless sensor nodes deployed in target areas to collect environmental information essential for plant growth, such as light intensity, air temperature, soil temperature, air humidity, soil moisture, and CO2 concentration in real-time. It employs ZigBee short-range wireless communication technology to transmit data collected by various sensors, aggregating the transmitted data to a computer, connecting with the computer through a wireless gateway, and using various devices to control the growth environment information in the target area in real-time. The software in the expert system can assess plant growth conditions based on environmental parameters and adjust various growth conditions accordingly.
For the field of smart agriculture, the application of the agricultural expert system platform frees agricultural production from the reliance on natural environments and experiential production methods, enabling scaled, refined, and scientific production, allowing users to make rational real-time guidance on agricultural production based on high-tech technologies, significantly enhancing the technological level of agricultural production.
Outlook
Today, smart agriculture is already present in our lives, and the agricultural expert system platform has broad application prospects in smart agriculture. The agricultural expert system platform fully utilizes modern agricultural information technology to facilitate communication between farmers and experts, enabling farmers to quickly solve agricultural production problems. It can be said that every link in agricultural production can utilize the agricultural expert system.
Authors: Yang Baozhu, Wang Xiuhui, Lu Weikang
Author Affiliation: National Center for Agricultural Information Engineering Research
Copyright Statement:
The text, images, etc., published in this public account are owned by the editorial department of “High Technology and Industrialization.” For authorized cooperation, please contact [email protected]
For more content, please subscribe to the magazine “High Technology and Industrialization”
Address: 33 West Fourth Ring North Road, Zhongguancun, Haidian District, Beijing (100190) Phone: 010-82627674
Fax: 010-86267674 Email: [email protected]
Website:
http://www.hitech.ac.cn
12 issues per year, 58 yuan per issue, annual subscription price 696 yuan
Postal distribution code: 82-741
Publication number: ISSN1006-222X CN11-3556/N