Reactive Power Optimization For Urban High Voltage Power Systems

The purpose of reactive power optimization is to reduce the active power loss and to improve the voltage quality in power networks by adjusting reactive power devices. Research of reactive power optimization has great significance in theory and practical application.Reactive power optimization problem is a multi-variable, multi-bound mixed integer programming problem, the operating variables include continuous and discrete variables, so the optimization problem becomes complex. In substations of an urban power system, there are shunt capacitor sets connected to the low voltage buses. The reactive power optimization for urban high voltage power systems can be carried out by adjusting shunt capacitor sets and transform taps. Transformer taps and capacitor sets may be represented as discrete variables, we adjust capacitor sets to reduce network loss and adjust transformer taps to meet voltage, transformers and capacitors can be adjusted separately. In traditional methods, the branch-bound approach of mixed integer programming can be used to solve the problem with discrete variables effectively, but a huge amount of calculation is required.Due to the discrete characteristics of transformer taps and compensation capacitor sets, this thesis adopts a simple and practical approach, a auxiliary function is structured to express the input or withdrawal of capacitor sets. This method transforms discrete variables into continuous variables. Setting the state of the initial value of capacitor is 0.5, when the calculated values deviate from 0.5, let it slide into the value 0 or 1 directly. The results output is in the form of an integer.The optimal objective is to minimize the compositive expenses of power loss and the cost of equipment regulation. The adjustment cost of equipment includes the cost for reactive power compensation and transformer tap. The equation constraint is composed of the nodal power balance equations, and the inequation constraint is composed of the constraints of breach power, nodal voltage, reactive power compensation capacity and the transformer taps. In order to improve the iterative characteristics, the weight coefficient is introduced in this thesis.Variable Metric Algorithm is adopted to solve the optimal problem. In the calculate process, this thesis adopts Lagrange multipliers and penalty function method to constitute the objective function augmented, it related to the one-dimensional search, the gradient calculation and so on.Proposed approach is examined on two systems, computed results provide the active power losses and capacitor states. Testing computation shows that the approach is simple and practical, and the economic index has been improved.