Study On Hierarchical Control And Optimization Operation Of Microgrid

With the influence of energy crisis, deteriorating environment and defect of centralized power supply, distributed generation technologies have been paid for attention gradually. By promoting utilization of renewable energy, distributed generation technologies become effective complements to the traditional power system. However, large-scale grid connected distributed generations can cause negative effects on the security and stability because of randomness and inertialess, so their application range is still limited. Under this circumstance, microgrid is proposed and becomes an approach for distributed generator (DG) to connect to utility grid friendly.Unlike traditional power system dominated by synchronous machines, there are many differences on characteristics and control modes among numerous types of DGs in microgrid, so the standardization of control is difficult to achieve. Under the architecture of hierarchical control, secondary control and tertiary control conduct global adjustment by the measurement in microgrid, and some control targets which cannot be met locally are achieved by communication. In addition, structure of local control can be simplified and control structure of microgrid could be standardized by hierarchical control. This dissertation which is supported by Special Fund of the National Basic Research Program of China (973Program) covers four aspects:local control of DGs, reactive power sharing enhancement based on secondary control, power oscillations suppression based on secondary control and economic operation of microgrid. The main content of this dissertation is as follows:(1)According to architecture of microgrid and models of DGs, the primary control strategy which coordinates DGs with different frequency regulation modes is proposed, and a microgrid containing energy storage, diesel generator, wind power generation and photovoltaic power generation can operate under the primary control. By average modeling technique, controllable power electronic devices could be transformed into combination models of controlled voltage source and controlled current source, simulations that contains several DG models are accelerated. Furthermore, the proposed control strategy is verified under circumstances such as power fluctuation of photovoltaic, load variation and switch between grid-connected mode and island mode.(2)In order to deal with uneven sharing of reactive power in classic droop control, mechanism of reactive power sharing is analyzed from the aspect of droop gain, load voltage and line impedance characteristics. By power information of DGs that is uploaded to the microgrid central controller(MGCC), a reactive control strategy based on secondary control is constructed. Potential functions which have the objective of reactive power sharing and deviation of output voltage at no-load are established in the proposed method. MGCC centrally regulates output voltage at no-load of DGs via potential functions, and reactive power sharing among distributed generation units is effectively improved. The classical structure of droop control need not be changed under secondary control, and communication between DGs and MGCC is needed. Furthermore, the magnitude of circulation can be quantified by the concept of circulating normalised reactive power, circulation characteristics and circulation suppression are tested under different load conditions.(3)Due to the presence of single-phase DGs and load in microgrid, unsymmetrical voltage in a three-phase system causes100-Hz power oscillations of DGs. By analyzing the mechanism of the active and reactive power oscillations, Analytic formula of power oscillations is firstly derived, and the impact of positive and negative sequence electrical quantities on power oscillations is revealed. Accordingly, a secondary control is proposed to control power oscillations. An optimization model is built following optimization theory. The model aims at minimizing the amplitude of oscillations, subject to constraints which include power oscillations rate and power reference of PQ control. By optimized results, the secondary control can simultaneously reduce active and reactive oscillations while reactive oscillations are large. Therefore, the indexes of oscillations amplitude and oscillations rate reach the depth optimization.(4) According to optimized results, the economic operation control adjusts power given values of each DG and can be considered as superior control strategy of microgrid. Due to large proportion of intermittent energy contained in microgrid, the power generation plan is difficult to draw up because of randomness fluctuation. An economic operation model of microgrid based on chance constrained programming is put forward, and charge-discharge constraint of energy storage, reserve capacity constraint and power balance constraint are considered in the model. In addition, a probabilistic model based on cloud amount and geographical factors is adopted for photovoltaic generation, randomness fluctuations is truly reflected. On the basis of sequence operation theory, probability sequences are calculated by probability distribution of wind power generation and photovoltaic power generation, the chance constraints of the proposed model can be transformed into equivalence class of deterministic constraints. Accordingly, the proposed model is compatible with solvers such as TOMLAB, the results of the scheme are stable and can be quickly calculated compared with stochastic simulations. Futhermore, in order to investigate the rationality of probability sequence, the impact of different discretized steps on results is discussed and the proposed method is verified.