| With the development of renewable energy and distributed power technology,the research of reliability of distributed power systems has become a new focus in recent years.It is widely used in aerospace and other equipment platforms.The characteristic of renewable energy and the ‘source-load uncertainty’ of distributed power systems make it largely required to be improved.Because of the difference between distributed power systems and traditional power systems,typical reliability research methods are not able to solve the existing evaluation and optimization problems.It is urgent to conduct relevant research on the reliability of distributed power systems.The application background of this dissertation is equipment platforms.The main content is to study the reliability evaluation and optimization methods of the distributed generation systems(DGS).There are three parts to the paper.First of all,this paper analyzes the research status of DGS’s reliability.It summarizes the advantages and disadvantages of the traditional indexes.The paper proposes the value of minimum expected load shedding(MELS)as the index of the DGS’s reliability.The main contribution is as follows: The paper uses N-K cascading fault model to imitate failure scenarios,and applies DC power flow optimization model to find the scheme of minimizing load shedding.This method has better improvement on calculation speed and the rationality of MELS.MELS index expresses the average loss degree of network function in various fault scenarios.It can analyze the fault details of each load,which can represent the network reliability more intuitively and effectively.The above views are proved by four experiments.Secondly,this paper constructs the reliability evaluation model.This paper proposes a topology optimization model with a greedy algorithm and a genetic algorithm.It bases on the MELS index.The main contribution is: According to the idea of coordinated programming,the paper uses the bi-level programming model to solve the problem.The upper model is used to make plans for distributed generations(DG).The lower model is used for DC power flow planning.Different constraints and optimization strategies can be added according to the actual application scenarios.The paper uses the algorithms previously mentioned to solve the problem.Thirdly,this paper proposes the reconstruction strategy.It is divided into static and dynamic reconfiguration.This paper presents a framework of scene and task oriented DGS management and reconstruction method.In this paper,the IEEE33 example and the data of a solar UAV are used as experimental examples.Simulation results show the effectiveness of the above method.The main contribution is as below: First,the management is divided into two parts: static reconfiguration and dynamic active management.Secondly,use scenario reduction method to construct typical fault and timing scenarios.Thirdly,divides the reconfiguration and management process into four types: DG location and capacity,load reduction,energy storage dispatching and distribution,and on-off of tie switch.And finally,uses the optimization model to solve and sort out the schemes in different scenarios.The reconstruction and management method proposed in this paper is suitable for active management in short time scale and system reconstruction in long time scale. |
