A Study On Security Constrained Dispatching Based On Pattern Recognition And Successive Linear Programming Techniques In The Power Market Environment

Nowadays many countries are restructuring their electrical industry around the world, which intensively challenges the existing technologies of Security Constrained Dispatching (SCD) of the power systems. Based on a review of SCD in power markets at home and abroad, the point of view is put forward of introducing the technique of pattern recognition into the traditional algorithm of successive linear programming, in order to obtain a better performance during generation scheduling.Power Market Technical Support System (PMTSS) is an automatic system serving the power market operation, whose primary function includes transaction management and security analysis. The importance of application of SCD to the power markets is discussed in the thesis.An algorithm of bounded linear programming is formulized in detail. The linearization of various constraints is developed, in addition to control variables and objective function of the SCD problem.To improve the speed and precision of optimization, the technique of pattern recognition is introduced into the successive linear programming. In Chapter 4, some basic concepts on pattern recognition are discussed, e.g. the composition of a pattern recognition system, feature selection, and others. Moreover, a general flowchart is provided for the application of pattern recognition to SCD problem.A newly developed program, based on the proposed methods, has a capability to meet the requirements of the realtime environment, whose successful application to Liaoning Power Market shows the feasibility of the proposed algorithm, flexibility to various types of constraints, and the practicability of pattern recognition technique.