Crop Cultivation Control Systems

Simulation

Abstract

In perceding report, it was suggested that weed control system is able to be formulated as a stochastic multi stage decision problem by introducing prediction value to the system state variable.
Following the results of above formulation, computer models were built up and simulation was carried out for the purpose of examining characteristics and roles of informations about agricultural production systems.
The results obtained are as follows;
1. Taking rc a ratio of yield prices to current prices, investment efficiency is larger than 1.0 when γc ranges from 0.1 to 1.0, but smaller than 1.0 when γc is larger than 1.0.
2. In the case that prediction probability of the systems and yield prices are constant, amount of information available to decision makers has a maximum restriction.
3. A degree of use of information, which is available to decision makers in the process of agricultural production, is affected by yield prices C. In the case of this paper, the degree is nearly satulated and the rate of increase is small when C is larger than 1.0. Consequently, if yield prices raise up, it is not only difficult to obtain good results by acting on maximum expectation criterion but also is losing, without improvement of system prediction accuracy.
4. Values of information depreciate as stages go by and the rate is proportional to yield prices.