19 April 2019 in IEEE Transactions on Neural Networks and Learning Systems
This paper provides necessary and sufficient conditions for the existence of the static output-feedback (OPFB) solution to the H∞ control problem for linear discrete-time systems. It is shown that the solution of the static OPFB H∞ control is a Nash equilibrium point. Furthermore, a Q-learning algorithm is developed to find the H∞ OPFB solution online using data measured along the system trajectories and without knowing the system matrices. This is achieved by solving a game algebraic Riccati equation online and using the measured data. A simulation example shows the effectiveness of the proposed method.
1 September 2013 in IET Control Theory & Applications
Model predictive control (MPC) is an effective control strategy in the presence of system constraints. The successful implementation of MPC in practical applications requires appropriate tuning of the controller parameters. An analytical tuning strategy for MPC of first‐order plus dead time (FOPDT) systems is presented when the constraints are inactive. The available tuning methods are generally based on the user's experience and experimental results. Some tuning methods lead to a complex optimisation problem that provides numerical results for the controller parameters. On the other hand, many industrial plants can be effectively described by FOPDT models, and this model is therefore used to derive analytical results for the MPC tuning in a pole placement framework. Then, the issues of closed‐loop stability and possible achievable performance are addressed. In the case of no active constraints, it is shown that for the FOPDT models, control horizons subsequent to two do not improve the achievable performance and control horizon of two provides the maximum achievable performance. Then, MPC tuning for higher order plants approximated by FOPDT models is considered. Finally, simulation results are employed to show the effectiveness of the proposed tuning formulas.
1 January 2016 in IFAC-PapersOnLine
2 October 2019 in IEEE Transactions on Power Delivery
Loss of Excitation (LOE) is an important fault in synchronous generators which may cause the generator outage due to out-of-step(OS) conditions. Since the LOE protective relays act based on the impedance or admittance measurement at the generator terminals, they suffer from mal-operations caused by non-LOE phenomenon such as stable power swings. To preserve the security and dependability of the LOE protection, in this paper, some unique characteristics of LOE event are determined using the transient model of synchronous generators. In this paper, using the transient model of synchronous generators, some unique characteristics of LOE event are determined. Frequency decomposition and time-domain analysis of active and reactive powers as well as the terminal voltages are used to define a suitable rule for LOE protection. A comprehensive list of full and partial LOE scenarios, power swing and other non-LOE scenarios at different loading conditions and topological configurations are considered. The input-output data of the simulated scenarios are then utilized by a support vector machine(SVM) to construct a relay logic for partial and total LOE protection. The efficiency of proposed LOE protection algorithm is verified using a dynamic test system. The results are analyzed using performance criteria such as dependability, security, and accuracy.
21 September 2011 in International Journal of Robust and Nonlinear Control
31 July 2008 in Journal of Process Control