Design And Optimization On Microgrids With Multi-source In Multiple Scenarios

Paced with the development of society and economy, the contradiction between economic growth and environmental issues becomes increasingly prominent. The power supply, mostly driven by fossil fuels makes the rapid consumption of fossil fuels and produces a large number of pollutants and greenhouse gases. Meanwhile, due to the higher requirement on power quality and power reliability from folks, it is a piece of practical challenge to improve energy structure and promote reliability of power supply. Development of distributed generation technology, effectively bettering energy consumption structure and addressing environment protection issue, is unsuitable for large-scale accessing to power grid. However, it is the proposed microgrid that is capable to tackle this challenge. Microgrid, a micro power system constituted by distributed generation, load, energy storage devices and centralized controllers, is considered as a platform for distributed generation and a means to spur flexibility and economic efficiency of grid. As a fresh member of the smart grid system, microgrid has become a hot topic as well as an important direction in power system.Optimization and design on microgrids, a complex multi-dimensional problem, are crucial to developing microgrid technology. Therefore, this dissertation focuses on optimization and design on microgrids with a variety of microgrid models and typical application scenarios. Single-objective optimization algorithm and multi-objective optimization algorithm are presented to solve the optimization model. The research is carried out in four aspects which include:optimal design of distributed generation in a grid-connected microgrid, optimal sizing of distributed generation in a typical island microgrid with time-shifting load, optimal configuration of a combined cooling heating and power (CCHP) microgrid under uncertainty, multi-objective planning of a microgrid containing distributed generation and electric vehicle charging stations. The specific topics of this dissertation are as follows:(1) Focusing on the optimal location and capacity of distributed generation accessing to a grid-connected microgrid, a three-objective optimization model with objectives of minimizing total cost, minimizing net losses and maximizing voltage stability is proposed. In addition, the free-search algorithm has been backboned by improvement and extension including fitness selection strategy and adaptive archiving algorithm. And then, the progressiveness of multi-objective free-search algorithm is verified through algorithm performance tests. Finally, the research on the optimal location and capacity of distributed generation is carried out based on the IEEE-33 bus system that functions as an example for grid-connected microgrid. Moreover, besides validating the model and algorithm, this example optimized the three targets mentioned above and paved the way for further study.(2) Aimed at optimal sizing of distributed generation in a typical island microgrid with time-shifting load, the steady-state output model of the distributed generation is established at first. Apart from a novel discharge strategy is proposed with taking the influence of cycle count on storage life in account, the mathematical modeling and operation strategy of time-shifting load are developed. Furthermore, the objective function and constraints of optimal problem are established. Combined chaos optimization with free-search algorithm, an improved chaotic free-search algorithm is proposed with elite competitive strategy added. Then, the feasibility and superiority of chaotic free search algorithm are verified through algorithm performance tests. Finally, four cases are set for optimal sizing in a typical island microgrid as an example whose illustrates that optimization by using the proposed model can reduce redundant investment and improve system efficiency.(3) Corresponding to design CCHP microgrid under uncertainty, taking the probability of wind, radiation and load into account, the model is established by using chance constrained programming. Firstly, the heating device characteristic model and probabilistic model of wind power, solar power generation and load have been built. Secondly, combined with probability characteristics, the annualized cost of microgrid is developed based on chance constrained programming. At last, the chance constrained programming problem is solved by Monte Carlo simulation. Moreover, the example analysis has indicated that the economy and reliability of the system could be promoted by sound consideration and coordination of the uncertainty of the power supply and the load. Thus proposed stochastic model is able to handle the associated risks.(4) Catering to electric vehicle charging station and distributed generation simultaneously accessing to microgrid, the optimal planning is carried out. In the first place, the shortest path model of electric vehicles is established considering the travel time. Based on the model mentioned above, the traffic satisfaction model of electric vehicle charging station has been built according to the tradeoff between p-medium model and maximal covering location model. Furthermore, the output models of the distributed generation are studied including the timing characteristics of different types. Last but not least, a three-objective optimization model with objectives of minimizing total cost, minimizing net losses and maximizing traffic satisfaction is developed and solved by multi-objective free-search algorithm. The case study based on improved IEEE-33 bus system demonstrates that a reasonable allocation of charging stations and distributed power are capable to reduce investment costs and net losses, and improve traffic satisfaction.