DEVELOPMENT OF A HIGH-SPEED ALGORITM OF NEUROLOGICAL CONCLUSION

One of the ways to increase the efficiency of the process of managing continuous dynamic objects is to develop new or improve existing control systems based on modern methods involving the achievements of information technology. The article deals with the creation of highly efficient control algorithms for technological objects, operating in conditions of uncertainty, designed to manage real-life objects. An algorithm is proposed for the structural-parametric adaptation of the PID parameters (proportional-integral-differential) -regulator, which allows to reduce the number of iterations in the learning process of the fuzzy-logical inference algorithm by reducing empty solutions. To determine the empty solutions, hybrid algorithms are used, which include modernized genetic and immune algorithms, which in turn allow you to configure the adaptation parameters of artificial neural network models. A block diagram of an automated control system for executive mechanisms is proposed, which includes a block for adapting the correction of not only parameters, but also the structure of the control system, which allows to reduce the error in the results of training a neuro-fuzzy network from 8 to 1%. The proposed algorithm is simple to implement on microcontrollers, which allows it to be implemented in the tasks of process control in the conditions of information uncertainty in real conditions at the operation stage.