The Reactive Power Detection Method And The Control Strategy Of Static Var Generator(SVG)

With the development of global economy, the effect of shock load on the grid is moreand more serious. Because traditional SVG can’t compensate rapidly dynamic andcompensation range of small, it can’t meet the need of users. As a new type of static varcompensation device, static var generator (SVG) overcomes the defects, shows greatsuperiority. The technology has become one of the important subjects to study at home andabroad. So, the key technologies of SVG is studied in this paper.First of all, the development process and the latest direction of development of SVGhave been given. The equivalent circuit and working principle of SVG without consideringthe loss and considering the loss have been given, too. The mathematical model is quoted,according to the three-phase instantaneous reactive power theory (Based on the rotationvector product, Based on the0coordinate transformation, Based on the dq0coordinatetransformation). Thus, the indirect current control and the direct current control of SVG wereelicited. The advantages and disadvantages were analyzed. The direct current control methodwhich based on the dq0coordinate transformation be chosen in this paper. Secondly,different reactive current detection methods were analyzed, their applicability and limitationswere researched in this paper. Wattless current detection methods which based on the theoryof neural network were introduced mainly. And, the principle of reactive current detectionmethod which based on neural network was research in this paper. Then, the overall systemstructure and the overall process of the SVG was designed. The main hardware circuit (Themain circuit part, The main controller, Signal conditioning circuit) and the main softwaremodule (Zero crossing detection module, PI control module) were designed. Finally, theneural network for wattles detection was built based on MATLAB. And, the training wascarried out, the training process diagram, the training state diagram, the training error curveand the simulation curve fitting were given. Then, the model of SVG system be set up in theSimulink simulation environment. The inductive compensation and the capacitivecompensation were studied.