含间歇性能源的分布式电网优化配置理论与方法研究

发布时间：2019-01-11 11:07

【摘要】：在能源危机和环境污染日益严重的当今社会,具有清洁无污染、可循环利用等特点的风能和太阳能已引起广泛关注。但由于风能和太阳能受到季节气候的影响,具有间歇性和随机性的特点,则为了充分利用可再生能源,需要对电网中接入的可再生能源发电单元的容量和位置进行合理配置。本论文针对可再生能源发电的随机特性,对可再生能源电源的供电可靠性进行评估,深入研究含间歇性能源的电网优化配置问题。主要研究内容及工作包含以下方面：一、研究含储能的可再生能源发电系统的供电可靠性评估方法,首先分析风机和光伏电池发电功率的随机特性,提出利用离散状态马尔科夫链对储能装置充放电过程的建模方法,基于马尔科夫链的平稳分布概率分析系统供电可靠性。最后,以蒙特卡罗仿真方法的计算结果作为检验标准,验证算法的准确性。二、研究分布式风光互补发电系统的优化配置,首先,以经济性为目标应用简单迭代法对系统容量的配置进行优化,提出了储能单元剩余容量周期不变性原则,基于此原则使得满足负荷需求的系统风机个数、光伏个数和蓄电池容量三个优化变量呈近似线性化关系,可大大减少迭代次数。其次,采用改进的Pareto多目标粒子群优化算法,同时考虑供电可靠性和经济性两个优化指标,对系统中光伏单元、风机和蓄电池的容量,及光伏阵列的安装倾角和风机的安装高度进行优化配置。三、基于概率最优潮流算法实现风电机组接入配电网系统的接入点和注入容量的优化配置。以配电网的有功损耗为优化目标,并计及新能源发电单元分散式接入到配电网中所引起的不确定因素,得到的最佳接入点和接入容量能够减轻整个配电网的馈线压力、提高节点电压质量且获得环境和经济效益。提出了基于Cornish-Fisher级数展开的交流概率潮流计算方法,该算法避免了传统的蒙特卡罗模拟算法带来的计算量大的问题,提高了算法的效率。四、基于凸二阶锥优化理论研究新能源发电单元在辐射状配电网中的优化配置问题,首先,利用支路潮流方程对配电网系统进行潮流计算,分析新能源接入后对配电网电压分布及网络损耗的影响；其次,针对支路潮流方程提出两步松弛,即相角松弛和凸二阶锥松弛,把含有非线性潮流方程约束的优化问题,转换成凸二阶锥优化问题进行求解,避免了非线性优化问题求解复杂性随电力系统规模增大呈指数上升的问题；最后给出了松弛后问题的最优解具有保真性的充分条件,对于满足条件的配电网,可以基于凸二阶锥优化算法实现新能源电源在配电网中的优化配置。
[Abstract]:In today's society where the energy crisis and environmental pollution are increasingly serious, wind and solar energy, which have the characteristics of clean and pollution-free and recyclable, have attracted wide attention. However, wind and solar energy are influenced by seasonal climate and have the characteristics of intermittence and randomness. In order to make full use of renewable energy, it is necessary to allocate the capacity and position of renewable energy generation units connected to power grid reasonably. According to the stochastic characteristics of renewable energy generation, this paper evaluates the reliability of renewable energy supply, and deeply studies the optimal configuration of power grid with intermittent energy. The main research contents and work include the following aspects: first, the reliability evaluation method of renewable energy generation system with energy storage is studied. Firstly, the stochastic characteristics of power generation of fan and photovoltaic cell are analyzed. A method of modeling the charging and discharging process of energy storage device by discrete Markov chain is proposed. The reliability of the power supply system is analyzed based on the stationary distribution probability of Markov chain. Finally, the accuracy of the algorithm is verified by the results of Monte Carlo simulation. Secondly, the optimal configuration of distributed wind-wind complementary generation system is studied. Firstly, the simple iterative method is applied to optimize the configuration of system capacity, and the principle of cycle invariance of residual capacity of energy storage unit is proposed. Based on this principle, the number of system fans, the number of photovoltaic cells and the capacity of battery are approximately linearized, which can greatly reduce the number of iterations. Secondly, using the improved Pareto multi-objective particle swarm optimization algorithm, considering the reliability and economy of power supply, the capacity of PV unit, fan and battery in the system is studied. And the installation angle of photovoltaic array and the installation height of fan are optimized. Thirdly, based on the probabilistic optimal power flow algorithm, the optimal configuration of the access point and injection capacity of wind turbine to the distribution network system is realized. Taking the active power loss of the distribution network as the optimization objective, and taking into account the uncertainty caused by decentralized access to the distribution network by new energy generation units, the optimal access point and access capacity can reduce the feeder pressure of the whole distribution network. The voltage quality of the node is improved and the environmental and economic benefits are obtained. An AC probabilistic power flow calculation method based on Cornish-Fisher series expansion is proposed. The algorithm avoids the problem of large amount of computation brought by the traditional Monte Carlo simulation algorithm and improves the efficiency of the algorithm. Fourthly, based on convex second-order cone optimization theory, the optimal allocation of new energy generation units in radial distribution network is studied. Firstly, the power flow calculation of distribution network system is carried out by using branch power flow equation. The influence of new energy on distribution network voltage distribution and network loss is analyzed. Secondly, for the branch power flow equation, a two-step relaxation, namely phase angle relaxation and convex second-order cone relaxation, is proposed. The optimization problem with nonlinear power flow equation constraints is transformed into convex second-order cone optimization problem. The problem that the complexity of nonlinear optimization problem increases exponentially with the increase of power system scale is avoided. Finally, the sufficient condition that the optimal solution of the relaxation problem has fidelity is given. For the distribution network that satisfies the condition, the optimal configuration of new energy power supply in distribution network can be realized based on convex second-order cone optimization algorithm.
【学位授予单位】：浙江大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：TM61

【参考文献】