微电网中光伏发电预测及随机优化调度方法的研究
文献类型：学位论文
| 作者 | 李鹏梅 |
| 学位类别 | 硕士 |
| 答辩日期 | 2014-05-28 |
| 授予单位 | 中国科学院沈阳自动化研究所 |
| 导师 | 臧传治 |
| 关键词 | 光伏发电预测 BP神经网络 随机规划 粒子群算法 微电网能源优化调度 |
| 其他题名 | Research on methods about solar power forecasting and stochastic programming strategy for micro-grid |
| 学位专业 | 控制理论与控制工程 |
| 中文摘要 | 近年来，微电网技术迅速发展，以降低能源成本，减少碳排放量，提高服务可靠性为目标，协调微电网内的可再生能源，存储装置，可控负荷等设备，进而实现微电网的安全、稳定、经济运行。光伏发电预测和能源优化调度是微电网控制和调度系统的重要组成部分，其精确的光伏预测精度和合理的调度策略能够保证微电网系统安全稳定的运行。同时对电力系统调度、电力市场营销以及发电公司竞价上网都具有很大的影响。本文针对微电网中的光伏发电预测和能源优化调度这两个方面进行研究。 （1）光伏发电系统的输出功率受到太阳辐射强度、温度和湿度等环境因素和气象因素的影响，呈现出时变性、间歇性和随机性。对其进行精确预测具有很大的挑战，因此选择合适的预测算法进行建模，提高光伏发电系统输出功率的预测精度。本文提出了基于相似日原理和改进的BP神经网络预测算法，对晴天、阴天、雨天三个样本集分别进行了预测并与实际值进行分析比较，结果显示改进的BP算法比传统的BP算法预测精度提高很多，证明了该算法的正确性及有效性。 （2）可再生能源的间歇性和负荷的随机性对微电网能源管理系统产生了巨大的挑战。在随机环境下的能源优化调度问题在微电网的研究中具有重要意义。以微电网中光伏发电系统的功率预测为基础，基于光伏发电预测的不确定性，在随机环境下的能源优化调度的问题成为本文的另一个研究重点。本文将光伏预测误差作为随机变量，建立了一种基于期望模型的能源随机优化调度模型。用Monte Carlo模拟方法生成了光伏发电预测误差的情景集，应用粒子群优化算法来解决随机优化调度模型。通过与确定性模型产生的调度方案相对比，证明了基于随机规划的调度模型能考虑光伏发电预测的不确定性，更符合微电网系统运行中的实际情况，调度结果也可有效地减少不确定性带来的经济损失及各种风险。 |
| 索取号 | TM73/L33/2014 |
| 英文摘要 | In recent years, micro-grid technology is developing rapidly, aimed at reducing energy costs, reducing carbon emissions and improving service reliability. Equipments in micro-grid such as renewable energy, storage devices, controllable loads can coordinate with each other, to realize the security, stability and economic operation of the micro-grid. Photovoltaic power prediction and energy scheduling optimization are important parts of the micro-grid control and scheduling system, which are related to security, stability and operation of the power system. It has a great impact to power system scheduling, power marketing and power generation companies bidding. The methods about photovoltaic power generation forecasting and stochastic programming strategy for micro-grid have been studied in my study. (1) Output power of photovoltaic generation system has characteristics of time-varying, intermittence and randomness due to the various meteorological factors such as season, solar radiation, temperature, humidity, etc. Therefore, the establishment of appropriate forecasting model to improve the accuracy of photovoltaic output prediction is one of the main contents of this paper. In this paper, a forecasting method is proposed based on the principle of similar days and improved BP neural network. We make predictions on sunny, cloudy, rainy days respectively and analyze the results compared with the actual values. The results show that the improved BP algorithm is more accuracy than the traditional BP algorithm. It proves the correctness and effectiveness of the algorithm proposed. (2) Uncertainties in intermittent renewable and stochastic loads are creating new challenges for micro- grid energy management systems (EMS). This paper focuses on the energy optimal scheduling problems of micro-grid under stochastic environment. Energy optimization scheduling problem in random environment considering photovoltaic power prediction is another focus of this paper based on the uncertainties associated with photovoltaic forecasting error. In my study, photovoltaic prediction error is treated as a random variable, then an energy stochastic optimization scheduling model is builded based on the expectation model. A scenarios generation approach based on Monte Carlo model is employed to characterize the random nature of uncertainty according to probability density function of photovoltaic forecasting error. Then, the particle swarm optimization algorithm is proposed to solve the stochastic programming model. Through the comparison of the simulation results with deterministic method, it proves that stochastic scheduling optimization strategy model can consider the uncertainties associated with photovoltaic forecasting, which confirms to actual operation situation of micro-grid system. The scheduling result can be effective in reducing economic loss and various risks caused by uncertainty. |
| 语种 | 中文 |
| 产权排序 | 1 |
| 页码 | 69页 |
| 分类号 | TM73 |
| 源URL | [http://ir.sia.ac.cn/handle/173321/14828]  |
| 专题 | 沈阳自动化研究所_工业控制网络与系统研究室 |
| 推荐引用方式 GB/T 7714 | 李鹏梅. 微电网中光伏发电预测及随机优化调度方法的研究[D]. 中国科学院沈阳自动化研究所. 2014. |
入库方式： OAI收割
来源：沈阳自动化研究所
浏览0
下载0
收藏0
其他版本
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。