Robust nonlinear damping backstepping with augmented observer for position control of permanent magnet stepper motors

Abstract

We innovate an approach to position control, a robust nonlinear damping backstepping with augmented observer for position control of permanent magnet stepper motors (PMSMs). The proposed method requires only position feedback and nominal value of K m JL among PMSM parameters. We propose new 3 rd order single-input single-output PMSM dynamics that consist of position, velocity, and acceleration using a commutation scheme. Instead of separated phase A and B current dynamics, the actual torque dynamics are proposed. The external disturbance, acceleration dynamic, and uncertainty of K m JL are regarded as a disturbance. In order to estimate the position, the velocity, the accleration, and the disturbance, an augmented observer is designed. A nonlinear damping backstepping is developed to suppress a position tracking error increased as the estimation error of the disturbance is increased. This approach simplifies the design process such that the control algorithm is suitable for real time control. The performance of the proposed method was validated via experiments.

Active disturbance attenuation control for permanent magnet synchronous motor via feedback domination and disturbance observer

“…A sudden change of speed reference shall impact the tracking performance of the control system. When t = 0.05 s, ω r rise from 200 to 220 rpm, it's obviously awarded that the proposed DO-based domination feedback control (21)- (26) provides a faster trailing performance than any other control laws. When t = 0.13 s, ω r suddenly lowered 20 rpm, there arise no overshoot of speed appeared in the proposed control system, while the other three controllers takes about 3 ms to revive from the overshoot of speed.…”

“…Moreover, the design principle of the parameters involved in the simulations is illustrated in Remark 3. Hence the control parameters for (30), (31) and the proposed controller (21)- (26) are selected as k 1 = 4000, k 2 = 8000, k 3 = 4000; the observer parameters for (31) and the proposed controller (21)- (26) are selected as λ 1 = 2000, λ 2 = 1000, λ 3 = 500; the parameters of speed-loop PI controller chosen as k p = 3.5, k i = 0.01 and the parameters of current-loop PI controller chosen as k p = 35, k i = 0.1.…”

“…Actually, the results mentioned above yield rather complex controller forms result from a popular backstepping-based handling strategy is utilised. These designs undoubtedly increase the complexity and costly calculations have to be conquered when practically implemented, such as control methods mentioned in [2] and backstepping-based methods with disturbance estimator [25,26] and so on. In [27], an interesting feedback domination technique is proposed which can simplify the design and render very concise controllers.…”