اکبر رهیده

In this paper, a method is presented for the electrical rotor fault detection in wound rotor induction generators. The fault is a rotor electrical fault in induction generators employed for wind turbine applications. Due to the tower shadow effect, a stray fault signature is introduced in the generator current which may result in false fault detection. The method is based on the time-frequency analysis in which the frequency is transformed and the time averaging of the signal energy is obtained. The method is known as harmonics order tracking analysis. For reliable fault detection, statistic indices are used in post-processing that increase the ability to distinguish between the fault effect and tower shadow effect. The simulation results illustrate that the proposed method is effectively successful in discriminating these two effects. Finally, the harmonics order tracking analysis is evaluated using the experimental results.

In this paper, the event-triggered strategy in the case of finite-horizon model predictive control (MPC) is studied and its advantages over the input to state stability (ISS) Lyapunov based triggering rule is discussed. In the MPC triggering rule, all the state trajectories in the receding horizon are considered to obtain the triggering rule. Clearly, the finite horizon MPC is sub-optimal with respect to the infinite-horizon case. This sub-optimality is considered as a coefficient to implement the MPC triggering rule and to increase the inter-execution time. However, to design a proper processing board to control this coefficient suitably, it is necessary to specify lower and upper bounds for it. The contribution of this paper is to determine an effective upper bound for this coefficient. By incorporating the sub-optimal lower bound instead of being 1, the conservatism decreases. This coefficient can be used as a tunable parameter in integrated or network control systems to compromise between the performance and inter-execution time. Simulation results depict the applicability of the proposed method and validate the theoretical results.

Current sensorless control methods, which eliminate the current sensor by using a current observer, can bring cost effective and reliable solutions to digital control of DC-DC converters. Since any inaccuracy in the model of the converter can deviate the observed current from the actual inductor current and consequently result in the steady state error of the output voltage, an accurate model is derived, incorporating the parasitic resistances of the switch, inductors and capacitors and the forward voltage of the diode. According to this model, Cuk converter is a nonlinear system. Therefore an extended Kalman filter (EKF) is used for inductor current estimation and output voltage filtering. In this paper, an accurate current observer is presented for Cuk converters by considering the influential parasitic parameters in the converter model. The estimated current from the observer, is used for current control of the converter. Simulation results of the system show the accuracy and favorable efficiency of the observer and controller.

This paper presents the implementation of the event-triggered technique on sliding mode control (SMC) method in the case of linear systems with bounded uncertainty. Event-triggered technique is a scheduling strategy to reduce the number of control updates in a feedback loop. The Main contribution of this paper is to implement event-triggered SMC on the linear multi-input linear systems with parameter uncertainties. Although the chattering is inevitable in SMC, to tackle this phenomenon, hysteresis bounds are offered to reduce the frequency of control updates; nevertheless it induces a limit cycle. By updating control law at hysteresis boundaries, as the event-triggering rule, system finally ends to limit cycle. Another contribution of this paper is to reduce the amplitude of this limit cycle according to the variable control input, while guaranteeing no request for successive control updates too. In this paper by properly designing dynamic control law and hysteresis bands based on state trajectory, as the system moves towards the stable point, its amplitude decreases and the limit cycle amplitude fades accordingly. This method is illustrated by an example of linear system with uncertainty in the system matrix. Simulation results depict the applicability of SMC combination with the event-triggered technique.