Induction Motor VSS Control Using Neural Networks

Avoiding singularities in recurrent neural control for induction motors

“…It is worth mentioning, that the selection of the constraint set W * i remains an open problem. In this paper, such a selection was made based on the observation of experimental results, which were published in [16].…”

“…The second goal can be reached by designing a control law based on the RHONN model. To design such a controller, the NBC methodology [16] can be used.…”

“…In order to illustrate the application of the proposed approach let us consider the induction motors control. The − coordinate system model [16] is given bẏ…”

“…On the other hand, indirect adaptive neural control [10,15], which consists of two tasks: first, designing a master-slave synchronization based on an adaptive parameter estimation, where a recurrent neural network identifier (slave) is synchronized with the real plant (master), in order to minimize a synchronization error; second, design a nonlinear controller based on the neural model. The basic idea is the identifier adaptive capture information about the disturbances in the plant model, such information allow the nonlinear controller to overcome the internal variations in the real plant.Neural Block Control is an adaptive indirect neural control technique that has been successfully applied for controlling induction motors [16], stepper motors [17], synchronous electrical generators [18] and electro-hydraulic systems [19]. In this neural control scheme, a high-order recurrent neural network (RHONN) [20] is proposed for the identifier model and based on such neural identifier, a control law is derived combining the Block Control [21] and Sliding Modes techniques [22].Comparing with other neural control techniques [10,13,14] that require full-state full-connected neural identifiers, the NBC strategy has the advantage that only a partial-state partially connected recurrent neural network [17] is required, reducing significantly the mathematical analysis and the computational burden.…”

“…Neural Block Control is an adaptive indirect neural control technique that has been successfully applied for controlling induction motors [16], stepper motors [17], synchronous electrical generators [18] and electro-hydraulic systems [19]. In this neural control scheme, a high-order recurrent neural network (RHONN) [20] is proposed for the identifier model and based on such neural identifier, a control law is derived combining the Block Control [21] and Sliding Modes techniques [22].…”