MANAGING PRODUCT INHERENT VARIANCE DURING TREATMENT

The natural variance of agricultural product parameters complicates re cipe planning for product treatment, i.e. the process of transforming a product batch from its initial state to a prespecified final state. For a specific product P, recipes are currently composed by human expe rts on the basis of heuristic matches between product state and recipe features. This approach makes use of standard recipes, that do not su fficiently reflect inherent differences between batches. Improvement o f the recipe design process requires three problems to be solved: (1) assessment of the initial product state, (2) fixation of the recipe re quirements and (3) design of a treatment recipe. To objectively assess the initial product state, additional measurement of a specific param eter is required. This parameter varies substantially between batches, requiring large measurement samples. Without objective assessment, ho wever, automated determination of the recipe requirements and recipe d esign is not possible. This paper describes a procedure to get an obje ctive initial state assessment, and presents a Product Treatment Suppo rt System that takes an initial state assessment, and performs the pro cess of recipe design. Artificial intelligence (AI) techniques are app lied at three points in the process. Induction of decision trees is us ed to determine rules that an understandable to experts and that selec t products most suitable for state assessment. Neural networks are app lied to transform the assessment of the initial state into the overall requirements of the recipe. Finally, the actual recipe is derived by means of constraint satisfaction.