基于人工神经网络的短期风功率预测研究

发布时间：2018-05-15 03:32

本文选题：风电功率 + 短期预测　；参考：《华北电力大学》2015年硕士论文

【摘要】：随着世界能源短缺、国家能源供应安全形势的日趋严峻,风能等可再生能源产业发展迅猛,风电在电网中所占比重越来越大。然而相较于一些传统的能源(如水电、火电),风力发电具有波动性、间歇性、随机性的特点,风电大规模并网后会严重影响到电力系统的电能质量和稳定运行。风电功率预测技术对风电场生产安排和指导系统调度运行意义非常重大,因此,亟需对风电场输出功率预测技术开展深入研究,对风功率进行较为准确的预测。本文主要进行风电场的短期风功率的预测工作,以某200MW风电场的现场测量数据和运行数据为基础,进行数据预处理、分析和短期风功率预测。首先分析了风电机组的功率输出曲线,以及风速、风向等因素对风功率的影响。而后应用功率直接预测方法,利用采集到的风速、风向、温度、气压等气象历史数据和风机运行数量作为预测模型的输入,对风电场的短期功率进行了预测。分别建立了BP神经网络预测模型、RBF神经网络预测模型和RBF-BP组合神经网络预测模型进行短期风功率预测,并进行预测结果的比较。经结果对比分析,证明RBF-BP组合神经网络预测模型预测精度更高,具有适应时变特性的能力,以及很好的非线性映射能力,可以在风电功率预测及其它相似的预测中应用。
[Abstract]:With the shortage of energy in the world, the security situation of national energy supply is becoming more and more severe, the renewable energy industry of wind energy and other renewable energy industries is developing rapidly, and the proportion of wind power in the power grid is increasing. However, compared to some traditional energy (such as hydropower, thermal power), wind power has the characteristics of volatility, intermittence and randomness, and the wind power is strict with the grid. The power quality and the stable operation of the power system are seriously affected. The prediction technology of wind power is of great significance to the scheduling and operation of the wind farm production arrangement and guidance system. Therefore, it is urgent to carry out an in-depth study on the prediction technology of the output power of the wind farm and to make a more accurate prediction of the wind power. This paper mainly carries out the short-term wind power of the wind farm. On the basis of the field measurement data and operating data of a 200MW wind farm, data preprocessing, analysis and short-term wind power prediction are carried out. First, the power output curve of the wind turbine and the influence of wind speed, wind direction and other factors on wind power are analyzed. Then the direct prediction method of power is used to use the wind speed collected. The wind direction, temperature, air pressure and other meteorological historical data are used as the input of the prediction model, and the short-term power of the wind farm is predicted. The BP neural network prediction model, the RBF neural network prediction model and the RBF-BP combination neural network prediction model are used to predict the short-term wind power, and the prediction results are compared. Through the comparison and analysis of the results, it is proved that the prediction model of RBF-BP combined neural network prediction model is more accurate, has the ability to adapt to time-varying characteristics, and has good nonlinear mapping ability, and can be applied to wind power prediction and other similar prediction.

【学位授予单位】：华北电力大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TM614;TP183

【参考文献】