A data fusion algorithm will be developed that can determine current and predict future availability and demand for soil-water and nitrogen and fruit load of cotton plants has been developed. This algorithm ingests information on season progression, weather data and visual plant response captured using remote, vehicle mounted and fixed sensing techniques. A neural-fuzzy model will be developed and integrated with statistic model for irrigation and fertigation optimisation. This model uses the mathematical theory of fuzzy set and neural network training to simulate the process of human reasoning by allowing the computer to behave less precisely and logically than conventional computer method require. The thinking behind this approach is that decision making is not always a matter of black and white or true or false: it often involves gray areas, or “maybes”. Through integration with biophysical crop models these models have potential to be used to enhance autonomous and semi-autonomous decision making within agricultural production systems. The procedure used to develop the model could be applied to any crop/production system.

Professor Keith Pembleton

• Professor Keith Pembleton

• Computational Engineering and Science Research Centre
• Institute for Agriculture and the Environment
• National Centre for Engineering in Agriculture

• Artificial Intelligence and Image Processing
• Crop and Pasture Production

• Doctor of Philosophy (DPHD)