Instruments and Systems: Monitoring, Control, and Diagnostics Annotation << Back Comparative analysis of neural tuner efficiency as alternative to DC drive adaptive control using state observer Yu.I. EREMENKO, A.I. GLUSHCHENKO, V.A. PETROV A problem of DC electric drive adaptive control system development is the scope of this research. Two methods are considered to solve it – a neural tuner and an observer. The task is to compare their efficiency. Experiments for these methods are conducted using a rolling mill DC drive mathematical model. The drive inertia moment value is being changed during experiments. The aim of the adaptive control systems under consideration is to compensate such parameter drift adjusting speed controller parameter value. Obtained results analysis shows that both methods allow to achieve almost the same results under the condition when controller parameter initial value is optimal for the initial drive state. But this is feasible only if the accurate drive mathematical model is known. If this condition is not met, then the observer is not able to keep required overshoot for the drive speed loop. In its turn, the neural tuner does not require the plant model and can be applied in situations, when speed controller parameter initial value is not optimal. This allows to make a conclusion that the neural tuner is more applicable under the real industrial conditions than the observer. The main reason for this is that the plant nonlinear model identification is rather complicated task in production. Keywords: electric drive; P-controller; transients; adaptive control; neural tuner; observer. Contacts: E-mail: erem49@mail.ru Pp. 11-17.