Coordination between central and local control of photovoltaic generation system for controlling voltage violation and unbalance in distribution systems

Abstract

Solar energy is a clean and pollution-free energy. Economic growth has led to increased demand for electricity in many countries, which has encouraged the installation of solar power generation systems, especially in LV distribution system. Households have the potential to install solar power generation systems because the costs of installation and maintenance have been continuously decreased. However, when there are many connected solar power generation systems in LV distribution system, the following problems may occur: (1) Loss of real power generation due to the operation of overvoltage protection of solar power generation systems, especially the ones on downstream nodes and (2) Voltage unbalance due to the connection of single-phase solar power generation system. In this dissertation, solar power generation system is determined as only Photovoltaic (PV) system. To cope with the above mentioned problem in LV distribution systems, this dissertation applies a control strategy to support high PV penetration which is the coordination between central and local controls. Central control is used to assess parameter setting for local control by considering the following principles: (1) the total connected PV system must inject real power output at the maximum value and (2) the parameter setting result must not cause voltage violation under uncertainty condition in LV distribution system. 2-stage Particle Swarm Optimization (PSO) is applied in central control for the parameter assessment of local control. The parameter setting will be sent to local control which consists of P(U) and Q(U) functions. P(U) and Q(U) functions are used for real and reactive power output adjustment from connected PV system respectively. The operation of P(U) and Q(U) functions will be regulated when the voltage at the connection point of PV system changes. In this dissertation, the power flow algorithm with using local control P(U) and Q(U) functions applies Newton-Raphson Method with step-length adjustment. The test systems are modified 19 and 29 node distribution systems. The power flow results show that the parameter setting obtained from central control enables the local control operate effectively under uncertain conditions in LV distribution systems.