Study On Microgrid System Control Technology

The stand-alone microgrid integrates distributed generations, energy storage devices, the regional load and control devices effectively,to form a single control unit, which has broad application prospects in the development of islands, deserts and other remote areas.In terms of abundant access of intermittent renewable energy generation and the limited system capacity, it brings random consequences of stand-alone micro power grid to spotlight. There would be considerable significances and requirements on practical development and construction to gain stable and efficient operation of stand-alone microgrid by settling reasonable system control strategies.According to the fluctuation shown from the practical operation on loads and renewable energy power generation, which affects the scheduling plan made in advance to be effectively achieved, this project focuses on a coordinating controlling strategy of controllable generations active output for stand-alone microgrid, which is based on ultra-short-term forecasting of loads and renewable energy power generation.Firstly, this paper summed up the development and construction process of current system of stand-alone microgrid, described the typical stand-alone microgrid structure, established of mathematical models of the common distributed generations; studied the control theory of controllable generations under the centralized control mode, taking into account to maximize use of available renewable energy, studied its maximum power tracking mode power principle and method. Secondly, discussed the necessity of renewable power generation and load ultra-short-term forecasting in a stand-alone microgrid, proposed a direct power prediction method based on least squares support vector machine method of real-time data sequence by analyzing conventional ultra-short-term power system load forecasting method. Considering the actual situation of renewable energy intermittent, adopted the same method of load forecasting to achieve the ultra-short-term forecast of wind generation; added a similar day selecting to the method to achieve the direct forecast of photovoltaic power generation. Third, proposed a multi-time scale microgrid control strategy of stand-alone microgrid that includes day-ahead scheduling plan and on-line control strategy, established the network architecture of the control system. On the basis of meeting the load demand and supply constraints, established the day-ahead scheduling model of a centralized controlled stand-alone microgrid, with the consideration of multiple objectives such as fuel costs, maintenance costs and emissions costs, and then introduced the particle swarm optimization to solve the scheduling model. The concept of microgrid net load was introduced to this paper, and on the basis of ultra-short-term net load forecasting, an on-line control strategy was proposed to control the power output of the system controllable generations, response net load changes ahead of time, and ensure the validity of the dispatching plan. Eventually, in terms of an island microgrid example, the day-ahead scheduling has been made, the reliability of ultra-short-term forecasting of loads and renewable energy power generation has been confirmed. And then results of online control strategy is also obtained. Comparing with the system operation result simply with the day-ahead scheduling, the effects of the multi-time scales control strategy proposed by this paper has been verified.