Research On A Multi-Timescale Voltage/Reactive Power Control Method For Microgrids Based On MAS

In the past years, researches on reliability of power supply and enhancing of power quality in microgrid mainly focus on adjusting DERs’(Distributed Energy Sources) active power to maintain frequency stabilization, meanwhile the voltage stabilization is becoming one of the hot-topics in microgrid for that the voltage is another important index to evaluate power quality. Two objects in microgrid voltage control are required: one is that the timely and effective voltage recover should be realized when voltage fluctuation happens in some nodes, another one is that global reactive power optimization should be performed to achieve economic operation of the microgrid. Focus on previously mentioned objectives, this paper studies the hierarchical control method for the voltage stabilization and the main innovations are presented as follows:1. A three-level hierarchical structure is established and a multi-timescale voltage/reactive power control method(VRPCM) for microgrid is proposed. First, optimal global DER reactive power scheduling strategy is drew up taking the system economy and the voltage stabilization of partial nodes into account. Then voltage/reactive power control zones should be selected and divided based on the structure and operating characteristic of the microgrid, the VRPC(voltage/reactive power controller) in every control zone calculates the reactive power adjustment value in its control zone real-timely to make the node voltage have the ability to acclimatize itself to the random fluctuation of the renewable energy and loads in microgrid. At last, local control of the DERs should be implemented to ensure the control instructions issued by the VRPC could be carried out successfully.2. MAS(Multi-Agent System) is adopted in this thesis to achieve the VRPCM explained above and a three-level MAS model for different voltage control objectives in microgrid is established. Moreover, two indexes assessing the reactive power, namely VQSF(voltage quality satisfaction factor) and RPPF(reactive power price factor), are introduced as the cooperation criterions of DER Agents.3. Under the background of the fluctuation of DERs and loads, the uncertainty of the line impedance parameters and the flexible structure of the microgrid, in order to enhance the VRPC’s adaptability regarding on the dynamic change in microgrid structure and parameters, the intelligent Q-learning algorithm is introduced in the secondary voltage/reactive power control to achieve the self-adapting voltage control in each control zone.4. A hybrid energy management system for the microgrid (HEMS-MG) platform which contains local DER Agents, ZVC(Zone Voltage Control) Agent and Global Reactive Power Optimization(GRPO) Agent is established in C++ Builder based on the proposed multi-timescale MAS model for voltage/reactive power control. The results verify the correctness and effectiveness of the proposed VRPCM when fluctuation of DERs and load happen in the microgrid.