Nonlinear Control Systems
Adaptive sliding-mode load-torque observer: its stability aspects
WSEAS TRANSACTIONS on SYSTEMS
| Hi-index | 0.00 |
As a contribution towards improving the environment, a new position controller for vector controlled electric drives employing permanent magnet synchronous motors (PMSM) is presented that achieves approximately 27% less frictional energy loss than a conventional controller adjusted to give the same manoeuvre time for the same initial and final positions. The frictional energy loss minimisation is carried out within the constraint of a velocity-time profile of fixed form to achieve prescribed manoeuvre times for relatively large step reference position changes. Optimal settings of the parameters of the velocity-time profile are made to minimise the energy loss. The control system robustness is achieved through the use of two control techniques, forced dynamic control (FDC) and sliding mode control (SMC), which maintain the velocity-time profile regardless of the mechanical load presented to the motor. The closed loop system automatically enters a linear operational mode with prescribed dynamics as the demanded position is approached, enabling derivative feed forward pre-compensation techniques to eliminate dynamic lag if needed. Vector control is achieved by an FDC direct axis current control loop with zero reference input, assuming a non-salient PMSM. The only information required from the user is the reference position and the required manoeuvre time. For commissioning, no time consuming adjustment of PI controllers is needed in contrast to conventional vector controlled drives. Only the following parametric entries are needed: a) the motor power and voltage ratings, the permanent magnet flux, the number of pole pairs and an estimate of the total moment of inertia presented to the rotor, to calculate the maximum rotor angular acceleration magnitude for which no motor control torque saturation can occur and b) the settling time for the FDC direct axis current control. Since a specified manoeuvre time may be realised, the method is especially suited to applications in which more than one position control loop has to be coordinated. The performance of the new controller and its advantages over a conventional controller adjusted to yield the same settling time are demonstrated by simulations.