Decision support system to schedule and monitor crop production operations using fuzzy logic

Conventional analytical procedures based on classical set theory or first-order logic have led to only partial success in the tactical scheduling of agricultural operations, an important farm management problem. The partial success can be attributed to the fact that while information available in the agricultural environment is fuzzy in nature, the classical set theory and first-order logic manipulate information based on the assumption that it is not fuzzy. The theory of fuzzy subsets which allows representation of inexact information has been suggested as a possible tool that can be used successfully in such circumstances;Trafficability of the field has a great bearing in deciding whether or not to carry out a given agricultural operation. Classical set theory, which divides days into workable and not workable sets is inadequate for farm management decision making. An alternative approach using fuzzy set theory to assess the extent to which a field may be dry enough to be trafficable was suggested. This approach was compared with a classical set theory based approach and was found to yield results more informative in deciding whether or not to carry out an operation on a chosen day. This was particularly true at intermediate levels of soil moisture content;Agricultural operations are carried out to satisfy certain objectives like preparing the field for planting, planting the seeds, and applying fertilizer and herbicides at the right time. Deciding whether or not to carry out an agricultural operation on a chosen day requires evaluation of the extent to which each of the competing alternatives satisfy the manager's objectives. The multi-objective decision making approach using linguistic approximation was modified to define a methodology for decision making. Operations on two farms with different urgency levels were used to illustrate the behaviour of the model under different situations;A fuzzy logic based program in the C computer language was developed to assist in deciding whether or not to carry out tillage operations on a given day. The decision procedure accounts for factors like soil moisture content, urgency of the operation, weather forecast, and importance of the operation. The program was evaluated using weather forecast data collected daily for a cropping season and the decisions made by the managers of two farms on whether or not to carry out tillage operations. The proportion of the total number of days compared on which the results were in agreement with the two managers were 0.75 for one with a relatively small and compact farm and 0.45 for the other, whose operation was distributed and relatively large.