Optimization of control systems for near-ambient grain drying: Part 1, the optimization procedure

Abstract

A method was developed of finding the optimum values of up to 10 parameters defining the size of equipment for, and the method of operation of, a near-ambient drier. The optimization procedure does not need a linearization of the drying model. Initial estimates of the cost of energy and overdrying are obtained by the systematic and repeated running of the Agricultural and Food Research Council, Institute of Engineering Research (AFRC IER) near-ambient simulation program (STOREDRY) with a range of values of each of the parameters to be optimized and with weather data averaged over 20 years. After rejecting those combinations of parameters giving excessive spoilage or drying time, a selected number of the best combinations are tested and, if necessary, re-ordered by running for each individual year of the 20. New and narrower bounds within which to adjust the parameter combinations are calculated and the whole process is repeated until there is no further improvement in the best ‘score’ and an optimum combination is identified. The policy to be optimized is similar to that developed in earlier work by Ryniecki. The main features of this ‘fan-and-heater’ policy were that it allowed for the airflow to be altered between heating and non-heating periods and for the control set-point relative humidity to vary with the progress of drying.