VARIABLE RATE PRESCRIPTION GENERATION USING HETEROGENOUS PRESCRIPTION SOURCES WITH LEARNED WEIGHTING FACTORS

摘要

A method for prescribing a field operation (4) by generating an optimized prescription (36) with a weighted prescription subprocess (20), executing the field operation prescribed (22), and then updating the weighted prescription subprocess (20) using a learning subprocess (26). The weighted prescription subprocess (20) calculates and sums weighted output (32) from two or more site-specific models to generate the optimized prescription (36). The learning subprocess (26) determines new model weights (34) as a function of relative model error (44) calculated by comparing model output (30) against actual results (42) and desired results (48) of the executed field operation.