The Problem of Conditional Interval Forecasting and Its Solution by Using a Generalized Plant Model

The paper is concerned with algorithms forecasting the plant's response to a specified input signal. Because the plant's equation is not known in advance these algorithms are developed on the knowledge of experimental data obtained in observation of normal functioning. Algorithm versions insuring a compromise between the requirements to forecasting accuracy and simplicity of implementation are proposed and studied. An example of practical application in weather forecasting for agriculture is given.