【作者】 曾涛；

【导师】 葛少云；

【作者基本信息】 天津大学 ， 电力系统及其自动化， 2008， 硕士

【摘要】 随着我国经济的快速发展和人民生活水平的提高,全社会对电能的需求量越来越大,对电能质量要求越来越高。而我国当前电网普遍存在配电能力偏小、变电站容载比偏小、高压网架脆弱、供电可靠性低等缺点,这在一定程度上制约着经济等各方面的发展。近年来人们普遍意识到城市电网规划已经成为一项迫切任务。其中确定变电站容量与位置是城市电网规划中介于电力负荷预测和网络规划之间的关键环节,其结果直接影响未来电力系统的线路走径、网络结构、电网投资、运行经济性及供电可靠性等关键问题。针对变电站优化规划的特点,本文根据文化算法的结构,分析若干智能算法作为文化算法框架中底层算法的可能性,最终选定粒子群算法,并设计文化算法上层空间的知识源约束条件。基于上面的分析,本文提出了一种用于变电站规划的新算法——粒子群文化算法,可以完成对新建变电站数量、容量、站址及供电范围等方面的优化。该方法首先根据容载比等方面的约束条件得到新建变电站的个数范围,然后采用0-1型整数规划方法得到容量组合的最优解和几组次优解。进而,再根据负荷密度函数,得到一组满足新建变电站个数范围的初始位置坐标。在此基础上,再将初始粒子群的相关信息导入到粒子群文化算法中进化迭代,最后完成待规划区的变电站位置、容量、供电范围等规划要求。本文将粒子群文化算法运用到实际工程项目的算例中,并在年费用、收敛速度、运算时间三方面与其他算法进行了对比分析。规划结果表明本文算法在上述的三个指标中具有优势,特别是在年费用方面。本文算法继承粒子群算法对初始解要求不高的优点,这样加强了其在变电站优化规划工作的可操作性和实用性,使得其能运用于实际的工程项目中,为待规划区提供合适的变电站规划方案。更多还原

【Abstract】 With the rapid development of our country’s economic and the improvement of people’s living standards, we demand much more electricity and better quality. But our country’s current power grid has some shortcomings, such as lack of distribution capacity, low capacity-load ratio of substation, vulnerable of high-voltage grid, low reliability. To some certain extents, it restricts economic and other aspects of development. In recent years people are generally aware that urban distribution network planning is becoming an urgent task. Substation Locating and Sizing is a core step of the urban distribution network planning between Load Forecasting and Network Planning, whose result will affect many aspects directly such as power line routing, network structure, power network investment, operation economy level and power supply reliability.According to the structure of cultural algorithm based on characteristic of substation optimal planning, this paper analyses the possibility of a number of intelligent algorithms as an underlying algorithm of cultural algorithm’s framework. Eventually it selects PSO, and designs the cultural source knowledge of the upper space constraints. Based on the above analysis, this paper presents a novel method for substation planning--Particle Swarm of Culture Algorithm, which can optimize the quantities, locations, sizes and power supply areas of substation. Firstly, this method obtains the number of new substations in accordance with capacity-load constraints. Secondly, it uses 0-1 type integer programming approach to obtain the optimal combination of capacity and several groups of sub-optimal solution. Thirdly, under the load density function, this algorithm gets a group of substations’location coordinates, which satisfy the number of new-to-built substations. Fourthly, it puts the relevant information of initial particle swarm into the Particle Swarm of Culture Algorithm to compute and evolve. Finally, it completes the location, capacity, supply area of substation, and other requirements of the planning area.This paper applies Particle Swarm of Culture Algorithm to the actual project examples, it compares and analyses in the aspects of the annual costs, convergency, and the computing time with other algorithm. The results show that the planning algorithm has advantages in the above-mentioned three indicators, especially in the aspect of annual cost. This method inherits the advantages from PSO, which is not sensitive to the initial solution. This strengthens its operability and practicality in Substation optimal planning, which makes it adapt to the actual projects, provides a suitable method to planning the substation.更多还原