Study Of Stochastic Control Based On Power System Self-Organized Criticality

How to prevent large-scale power system accident is the common concern issue of world's power engineers. In the past, the physical process of power grid blackout is researched on more and more, but the study of the blackout inherent macroscopical characteristics are not sufficient. In recent years, there has been a series of achievements in power system complexity research. By analyzing the system of Chinese and American power blackout accident, it has been founded that the power law existed in the scale of the blackouts and frequency, and this relationship was considered to be the mathematical characteristics of complex systems.Main purpose of simulation of the blackouts is to abstract the overall power system blackout rule, which is used to guide the power system operation and planning. But the application of these rules to prevent the blackout has not been solved.Various random factors exist in power system, and the failure blackout sequence is a random process. The data of large-scale accident is analyzed, that the failure sequence of normalization is a stationary stochastic process. The random factors are added to power system state equation, that reducing the mathematical expectation of random process by adding control rules is proved. Based on this theory, control strategy is proposed for power law random process of blackouts, and control rules are designed by applying the expert system.In order to verify the effectiveness of control measures, the sand model was established. Through simulation on this model, the power law relationship between cascading scale and the corresponding frequency is verified. Three control rules are added into sand pile models, such as small-stick mechanism, decrease of fault transfer and active splitting control, and new power law functions are got respectively satisfying the self-organized criticality. Compared to the power law function from sand pile model without imposed rules, slopes of the new power-law functions increased in log-log coordinates, which means that the probability of large-scale accidents is reduced under the three rules. Therefore, correct control rule is an important factor to prevent large power outage.Reference to the OPA model, DC power flow model is established. Specific control measure named by progressive reduction control is put forward after the power law relation is verified in blackouts rate of the scale and frequency:New blackouts sequence expectation decrease after data statistics with progressive reduction control control added to DC current model.To reduce power system fault sequence expectation, regional balance indicator is proposed to monitor overall risk of cascading failure. Indicators of regional balance can distingish the global risk of power outages, and also be used as the target balance of control. Balance control method is optimimal adjustment of the power law. When the regional balance index out-of-limit, find the load with greatest contribution to the regional power flow, and adjust the load and generation in minimized object function of amount of removed load.