Studies On Wide-Area Damping Control In Power System Based On Output Prediction And LMI Method

The inter-area low-frequency oscillation in large-scale power system has been a bottle neck limiting system's transfer capacity and a key factor influencing system's safe and stable operation. Although it plays an important role in suppressing power oscillation, limiting to the local feedback signals, the current power system stabilizer (PSS) cannot utilize the relative angular velocity or relative power-angle to form a closed loop and cannot effectively suppress the inter-area low-frequency oscillation in the interconnected power system. The wide area measurement system (WAMS) offers a new facility for the monitoring, analysis and control of power system and brings new opportunities to the damping control of the interconnected power system.In the damping control strategy that based on local signal, the influence of time delay is usually ignored as the delay is very small. However, in wide-area geographical distributed and interconnected power system, the time delay in collection, treatment and transfer of wide-area information reaches tens to hundreds of millisecond, which cannot be ignored. The signal transfer time delay usually deteriorates system performance, degrades system stability and make the controller unable to reach its ideal effect.This dissertation investigates the inter-area low-frequency oscillation phenomenon caused by inadequate damping in modern interconnected power system, focusing on the wide-area supplementary inter-area damping control considering time delay. To overcome the adverse effect on stability of the closed-loop system caused by the wide-area signal time delay, relative methods with regards to the modern control theory, such as model based prediction method and linear matrix inequality (LMI) method, which enrich and expand the methods of suppressing the inter-area low-frequency oscillation.The main research work and innovative achievements are as follows:(1) An inter-area supplementary damping control strategy based on output prediction is proposed to overcome the poor reliability of the prediction method that does not use system model. The open-loop state observer is established based on the reduced-order system state space model. Then the output prediction model is built in which prediction error is used as input and a correction module is also contained. The module is used to correct the prediction error, and then the current predicted output can be obtained by the model from the delayed output. The feedback control law can be designed via the control theory for the system without time delay. In the correction module, both the fixed parameter and the dynamic correction method can be applied. The latter proportionally corrects the prediction error. Then two different wide-area supplementary inter-area damping control strategies based on the output prediction model are proposed, i.e., linear quadratic optimal part output feedback control and sliding-mode control.The simulation result of the two-area and four-machine test system proves the effectiveness of the output prediction model and the adaptability to various control strategies. The damping controllers based on this model are insensitive to time delay to some extent, and has better effect on suppressing the inter-area oscillation. Compared with the case that only local PSS is installed, the controller can enhance the inter-area damping significantly. Simulation results show that the dynamic correction is more effective in the aspect of suppressing the system oscillation than the fixed parameter correction method.(2) A cone complementarity linearization iterative algorithm which can directly solve the state feedback control law is proposed to overcome the shortcoming of the damping control strategy based on a class of stability criterion of time delayed closed-loop system which essentially is a tentative method and has large amount of computation. Two matrix inequality theorems based on state feedback as an asymptotic stability criterion on time-delay system are deduced from two known linear matrix inequality theorems. The damping controller design is converted to a cone complementarity problem subjecting to linear matrix inequalities (LMIs) and solved by cone complementarity linearization iterative algorithm. In the applied two known LMI theorems, the second one contains less unknown matrix variable, less algorithm calculation amount and is easier to converge.Based on the cone complementarity linearization iterative algorithm, firstly, the wide area inter-area supplementary state feedback damping control strategy adopting the delayed state observer is proposed. Considering the time delay of the wide-area output, the delayed state observer is designed to obtain the delayed state estimation which will be used to construct the state feedback. And the delayed state feedback matrix is solved by the first matrix inequality theorem. Secondly, the state direct feedback wide-area supplementary inter-area damping control strategy is proposed. The state direct feedback method is applied to deduct the delayed state vector directly to be used to construct the state feedback. And the delayed state feedback matrix is solved by the second matrix inequality theorem.The damping control strategy based on cone complementarity linearization iterative algorithm is an effective method to implement this idea that can directly use the delayed wide-area signals to actualize the supplementary inter-area damping control. The simulation result of the two-area and four-machine test system proves the damping controller based on cone complementarity linearization iterative algorithm is insensitive to time delay to some extent, and has better effect on suppressing the inter-area oscillation. Compared with the case that only local PSS is installed, the controller can enhance the inter-area damping significantly.(3) A time delay system modeling method is proposed, and then a two-layer control strategy of wide-area inter-area supplementary damping control strategy is put forward to overcome the limitation of current wide-area supplementary inter-area damping control method in which power system is modeled without time delay and cannot directly apply the time delay system control theory. The first layer is the inner control, which applies the output feedback control considering time delay to the power system without time delay to constitute the time delayed closed-loop system. The second layer is the outer layer, which carries out feedback control aiming at the time delayed system model by using the control theory of time delayed system. For the outer control, firstly, the dynamic output feedback control strategy is proposed. The state feedback matrix and observer gain matrix in the strategy are solved by the LMI method based on a stability criterion of time delay system with quadratic matrix inequality type. Secondly, the delayed state direct feedback control strategy is proposed. The delayed state vector is obtained from the delayed output directly, and then the delayed state feedback matrix is solved by LMI method based on a stability criterion of time delay system with quadratic matrix inequality type.The two-layer control strategy can directly apply the time delayed system control theory to overcome the adverse effect on the closed-loop system stability caused by the time delay of wide-area feedback signals. Compared with the single-layer control strategy in which the controlled power system is established as a system model without time delay, and the time delay influence is treated in the feedback phase, the proposed two-layer control strategy is more direct and effective.The simulation results of the two-area and four-machine test system and 10-machine 39-bus New England test system prove the damping controllers based on the two-layer control strategy are insensitive to time delay to some extent and has better effect on suppressing the inter-area oscillation. Compared with the case that only local PSS is installed, the controllers can enhance the inter-area damping significantly.The simulation results of the two-area and four-machine test system show that the time delay stability margins provided by the two-layer damping controller and the damping controller based on cone complementarity algorithm are greater than that provided by the damping controller based on the output prediction.