Open-loop Estimation Method Of Resonant Mode Of Power Systems With Wind Power Penetration

With the gradual increase of power demand on load-side,as well as the insufficient of electric power generated from traditional power plant,more and more wind farms are integrated into power systems.As for wind power plants,the energy produced by them is usually intermittent and fluctuating,and needs to be transmitted via long transmission lines or cables.Additionally,there are lots of power-electronic devices in a typical wind farm.Such issues can induce interactions between open-loop oscillation modes of different subsystems of a wind farm penetrated power system,and lead to open-loop modal resonance.Up to now,however,researches on the mechanism of open-loop modal resonance are imperfect.In addition,the influences of openloop modal resonance on oscillation in power system are not well-understood.Furthermore,existing methods of analyzing open-loop modal resonance are ineffective.Especially,they are not able to detect asymmetric open-loop modal resonance.As a result,this thesis focuses on the mechanism and detection of asymmetric open-loop modal resonance,and has achieved the following accomplishments:(1)A rigorous mathematical definition of asymmetric open-loop modal resonance is provided,and a proof of its formation caused by the relative motion of a pair of open-loop oscillation modes is presented.In the first place,the process of partitioning the original system into two multi-input-multi-output subsystems,constructing the transfer function matrices of the subsystems and establishing a closed-loop transfer function model for the original system is introduced.It is the foundation of studying the mechanism of open-loop modal resonance.Based on the transfer function model,the thesis analyzes the open-loop modal resonance between the two subsystems resulted from variation of system parameters.The analysis includes calculating the distance between open-loop oscillation modes and closed-loop oscillation modes,as well as finding patterns in the location of closed-loop oscillation modes.The analysis shows that,as system parameters vary that forces a pair of open-loop oscillation modes to move closer but not coincide,either asymmetric open-loop modal repulsion or asymmetric open-loop modal attraction will be encountered.If asymmetric open-loop modal repulsion happens,the closed-loop oscillation mode,corresponding to the open-loop oscillation mode having smaller stability margin,will move towards the right-half complex plane and leads to small-signal instability of the original system.In contrast,if asymmetric open-loop modal attraction occurs,that closed-loop oscillation mode will move in the opposite direction and the small-signal stability of the original system may be enhanced.(2)A residue-based method is proposed to detect asymmetric open-loop modal resonance in power systems,via estimating the location of closed-loop oscillation mode corresponding to the pair of open-loop oscillation mode participating in the resonance.The method consists of two steps.The first step is to design appropriate transfer function models for the two subsystems,connected in feedback,and having interacting open-loop oscillation modes.The second step is to calculate the residue of open-loop oscillation modes participating in the resonance,under different system parameters,so as to predict the trajectory of the corresponding closedloop oscillation modes.The predicted trajectory will be used to determine whether asymmetric open-loop modal attraction or asymmetric open-loop modal repulsion occurs.Compared with existing open-loop methods for analyzing open-loop modal resonance,the proposed method can provide more accurate estimate of the trajectory of closed-loop oscillation modes,especially under the condition of open-loop modal attraction.Besides,the proposed method can be applied to optimize the parameters of controllers to enhance the small-signal stability of power systems.(3)A framework of parameter identification based open-loop modal resonance analysis is put forward,in order to deal with the situation that several important parameters,such as the stiffness,inertia and damping coefficient of drive train,are difficult to be obtained.As for parameter identification,the thesis proposes the use of Offline Reinforcement Learning to get the identified value of Spring Constant,as well as the optimal estimate of Shaft Torque among the two masses simultaneously.After that,the Least-square Method can be adopted to estimate the remaining parameters of wind power plant.The identified parameters can be directly used not only for the analysis of open-loop modal resonance but also by existing methods for the calculation of small-signal stability margin.The proposed framework can enhance the adaptability of resonance analysis to parameter uncertainty.(4)An implicit function based method is proposed to study asymmetric open-loop modal resonance in complex power system.Compared to the residue-based method stated above,the implicit function based method also calculates the residue of open-loop oscillation modes participating in the resonance,under different system parameters.But,it then performs another iteration,whose idea is originated from the Fixed-point Theory,in order to further improve the accuracy of estimation of the trajectories of closed-loop oscillation modes.The iteration requires information of partial derivatives of the Characteristic Equation,which can be calculated based on Implicit Function Theorem.Besides,Parameter identification can be incorporated into the proposed implicit function based method.In summary,the above-mentioned contributions are beneficial for developing the theory of modal resonance.In particular,it is the first time that the formation,detection and influences of asymmetric open-loop modal attraction are thoroughly studied.Apart from that,the two proposed open-loop modal resonance detection methods improve the accuracy of estimates of closed-loop oscillation modes and enhance the adaptability of resonance analysis to parameter uncertainty.Last but not least,the contributions can bring new insights on small-signal stability analysis of power systems and controller design.