Stiffness Distributing Characteristic And Flexible Simulation Method Of Power System Dynamic Course

The unceasing strengthen of the power system will led to the coutinuous increase of power system operation complexity when the enormous benefits is brought. The possibility of the occurrence, affecting scope, and dangers of cascading accidents, will be enlarged in different extent with the continual increasing system scale. How to correctly and efficiently simulate power system dynamic response after the disturbance is a very important and basic work for reliable mastering system operation acts and setting the correct protection and control system against the safety. As a result of possible disturbances being subject to the link and system which constitutes the power system may exist diverse differences, therefore, the dynamic process of power system is often not only the existence of the coexistence of slow-fast phenomenon, but also the existence of slow and fast variables constitute a continual changing situation, which make the power system dynamic process show the obvious stiff characteristics, has seriously hampered the efficiency of power system dynamic simulation process. This thesis research on how to adaptively divide variables into different groups based on the changes of dynamic process so as to obtain flexible stiff process measures through the analysis for dynamic process of spatial and temporal distribution of localized stiff analysis. On this basis, a flexible algorithm for power system dynamic simulation process is built and applied to power system long dynamic process simulation including wind power.As the theoretical foundation of the implementation flexibility to stiff problem for power system dynamic process, this thesis firstly expounds the definition of the stiffness, defines the the distinction between local stiffness and the overall stiffness, and through the analysis for distribution characteristics of stiffness of system dynamic process, pointed out: stiffness often has the local distribution characteristics of time and variables space, that is, the degree of stiffness on the one hand, will change over time, the system can be conversed from more stiffness toward the weak-stiff or even flexible. On the other hand, if system shows sfiff some time, usually only in some variables sets, while the variables in internal sets often show the characteristics of a good flexiblity. During dymamic process for a large scale power system, especially a long term dynamic process, temporal and spatial characteristics of stiffness distribution show obvious, which provides precondition for flexibility of stiffness in power system dynamic process. Based temporal and spatial characteristics of stiffness distribution, this thesis analyze flexiblity strategies for stiff problem of power system dynamic process, which provides technical support for follow-up study.Variable grouping problem is a feasible measure to implement the flexiblity of stiff problem. In the past simulation methods, only a pre-fixed variables grouping form runs through the the whole process of system dynamic calculation, the grouping scale factor between different variables after the initial value is given, is no longer adjusted, this way is difficult to adapt the temporal and spatial characteristics of stiffness distribution for different dynamic process, which restricts efficiency of relevant computational methods. According to existing problems, this thesis provides a kind of flexiblity simulation method combining self-adaptive flexible grouping strategy and variables grouping coordination strategy with self-adjustment integral measures for power system dynamic process. Based on the local truncation error, this method can finish variables grouping adaptively, and adjust the relative step factor and the smallest integral step, which achieves the purpose of adaptively flexible grouping, and at the slow-varying step size spot, integral tuning is implemented in order to improve computational precision. Results of calculation example show that the proposed method improves computational efficiency in ensuring the accuracy.For many variables in power system dynamic process after influence of system disturbances attenuation to a certain extent, it will be very small to further variable range especially variable range of envelope value or envelope centre value. At that time, if these small changes are ignored, and make variables secede from integral computation, then the influence on the follow-up integral precision of system will not be out of range of engineering permission, which can improve the follow-up computational efficiency obviously. If these variables are retained in grouping and integral computation, it will restraint enlarge the integral step and influence the improvement of computational efficiency. Pointing to this problem, this thesis provides a variable seceding mechanism at proper time, which forms a flexible simulation method combining variables grouping and seceding strategies. Simulation examples show that this flexible simulation method including variables seceding strategy may further improve computational efficiency in ensuring the computational accuracy.At present, there are scarced relative ripe methods and computational programs used for computing efficiently long term dynamic behaviour of large scale power system, especially wind power integration on a large scale power system, which speeds up the urgent demand for simulation methods and computational programs. Pointing to this situation, flexible simulation method has been applied to large-scale power system simulation including wind electric power grid combined with asynchronous wind generators and doubly-fed wind turbine model, analyze dynamic behavior such as frequency, voltage for different wind speed change modes and system operating conditions after wind power integration on power system. The simulation results show that the proposed flexible simulation method can adapt to the need for rapid analysis for long term dynamic process of large power systems that contains wind power.This research work is supported by the National Nature Science Foundation of China "optimization theory research on green energy efficiency of large power systems including intermittent power supply" under Grant 50477008, and Heilongjiang Province research initiation funds for the postdoctoral researcher project "analysis method of the dynamic mode for numerical simulation of power system stability".