A Decision Support System For Precision Farming

Precision farming theory and technology is a research frontier in the area of agricultural information technology, and is of significant importance for development of digital farming and modern agriculture. This research investigated the variability and the quantitative relationships among soil nutrients ( soil pH, total N, organic matter, available P), growth status ( SPAD, LAI and SPAD LAI) and yield of rice. All data were analyzed by both classical statistics and geostatistics based on GIS. By further integrating GIS, GPS and crop management knowledge model, a decision support system for precision farming (DSSPF) was developed, which laid a technical foundation for the practical application of precision farming system.In this research, geostatistics was adopted to analyse crop growth status, soil nutrient and yield, with emphasized on quantitative characterization of the spatial variation and investigating the relationship between crop growth indices and yield from spatial distribution. All parameters exhibited spatial correlation except for soil pH. Soil total N and organic matter, SPAD, LAI and SPAD LAI were all positively correlated to rice yield. Kriged interpolation maps provided good indication of the spatial variability in crop yield and growth status. Spatial interpolation and correlation analysis proved that SPAD LAI was more indicative of crop growth status than individual variable, and useful for implementing growth season regulation and topdressing as needed.An integration model called component combination was presented and applied to development of precision farming system ( DSSPF ) . Based on the component characteristics as independency of development environment, flexibility of soft upgrading, convenience of integration, the DSSPF was developed on the platform of Visual Basic and ComGIS. The programming processes of soft component and double project test with Visual Basic platform were introduced, the techniques of integrating GIS and model, GIS and GPS were presented. The DSSPF overcoming the shortcoming of running with GIS soft platform by using GIS re-development language, and can independently upgrade and maintain, thus providing a component-based framework for integrative development and independent running of precison farming system.A knowledge model was incorporated to establish GIS- and GPS- and model-based DSSPF. The models used in DSSPF included crop management knowledge modelestablished in our laboratory, new model based on spatial variability analysis. The knowledge model was developed based on the dynamic relationships between crop and environments, and quantitative representation of the temporal and spatial variation in crop growth and management indices. This provided the decision-making of DSSPF with interpretation and wide applicability, thus overcoming the shortcomings of the traditional expert system with strong empiricism and localization.