The article describes the signal paths of modern control systems. The method of error accumulation to the signal is shown, an example of calculation of such an error is given. Some variants of solving the problem of reducing this error are offered.
The article describes the possibilities of application of neural network technologies in the control system of the melting process in the furnace. The variants of using neural networks both in local control loops and in the control system of the whole process are shown. Advantages and disadvantages of such systems are considered. The example of creation of the simplest neural network regulator for stabilization of parameters is given.
Development and deepening of the full economic calculation in the national economy of the country is on the way to expanding the independence of associations, enterprises and their subdivisions.
