Constant Control Input Research Articles

Current control of three-phase PWM converter with model reference adaptive identifier.

The DC source of an inverter is, in many cases, composed of a diode bridge rectifier circuit with a smoothing capacitor in the DC link. Due to the existence of a large amount of harmonics in the AC line current, the inductive interference and low effective power factor have been serious problems.In this paper, the authors propose a new current control scheme of the three-phase PWM converter, which can suppress the harmonics of the input current and achieve the unity power factor. The current control is performed with the instantaneous voltage equation of the converter and the system parameters in the equation are identified by the model reference adaptive system. Therefore, the high performance current control is achieved even if the converter parameters are unknown. In the parameter identification, the number of the identified parameters can be halved to reduce the calculation time by introducing the complex plane to describe the state equation. Furthermore, the identified parameters can converge to the real parameters even if constant control input is given. The stability of the identification algorithm is proved with the hyperstability approach.The effectiveness of the proposed current control has been demonstrated by simulations and experiments. In the prototype, a DSP (TMS 320C25) can perform all necessary controls by software, and, as a result, a simple controller has been constructed. The sampling period was 255 μs and the switching frequency of the converter was 1.5 kHz. The time required to converge the identified parameters was within 5 ms and the current was settled with one sampling (255 μs) when reference current was changed from 2 A to 10 A.

Characterization of state controllable time-delay systems with piecewise constant inputs. Part I. Derivation of general conditions†

The possibility of sampled-data feedback control for linear time-invariant multiple delay systems is studied using the constraint of piecewise constant control inputs. A split system representation is utilized to derive both necessary and sufficient conditions for control of the system state to the null function in the state space.