System of automatic control of temperature of non-stationary object

The method of dynamic correction of signal of management of automatic regulator is considered. A method is based on measuring of size of thermal stream of nonlinear dissipative object. The thermal stream is interpreted as ?internal? managing influence of adjusting object. A management signal is corrected on the change of size of thermal stream of object. An adjusting algorithm will be realized by the artificial neural network of the simplest structure. The number of input signals of the neural network is equal to four: task, dynamic error, thermal stream, external disturbances. Function of activating of exit of neural network ? liner with limitations. Such structure of the automatic control system provides high quality of transients in her at the large changes of parameters of object. External indignations and changes of task also poorly influence on the dynamics of the adjusting system. The borders of area of the normal modes of operations of the automatic control system are certain. The researches results on the simulation model of the automatic control system are presented.