基于混沌与遗传规划的电力系统短期负荷预测-电力系统及其自动化专业论文.docxVIP

摘要电力系统负荷是电力系统运行计划与规划编制的重要依据。精确的负荷预 摘要 电力系统负荷是电力系统运行计划与规划编制的重要依据。精确的负荷预 测能够提高系统运行的可靠性和经济性。 ／电力系统是一个非线性大系统，表现出一定的混沌特性。本文对电力系统 的负诗时间序列进行分析和计算，得出负荷时间序列属于混沌时间序列的结论， 并计算出了表征混沌程度的特征量——维数和最大Lyapunov指数。由混沌理论 指出其可预测性，并进而求出最大可预测时问。 基于混沌理论的负荷预测大都采用局域预测法，通过找到预测点邻近状态 点与其后续时间序列的函数关系，近似代替预测点与其后续时间序列之间的函 数关系，从而实现对未来的预测，就成为局域预测法的基本思想。在拟合预测 函数时，多采用线性拟合函数，虽然整体上是非线性，但具体到每个点来说仍 然存在着预测模型的偏差。}一厂一 ， ，v矿 本文钲此基础且提出了一种新的电力负荷短期预测方法一基于混沌理论的 局域遗传规划法。利用混沌理论为遗传规划法提供自变量以及自变量个数的选 取，同时利用遗传规划法能够自动寻找相关变量形成函数表达式的优点，结合 基于混沌理论的局域预测法，自动形成局域预测模型，从而有效地克服了局域 预测模型选取单一、固定的缺点。实例证明该方法是～种有效的短期负荷预测 方法。 (由于局域遗传规划法是每一个点建一个模型，因此所需要的计算时间较 长。y4为了减少预测所需时间，本文采用改进局域遗传规划法的初始群体形成的 方法来加快收敛速度，减少寻优时间。从仿真结果来看，该方法能够有效地减 少预测所需时间，加快收敛速度。 关键词：负荷预测混沌相空间重构局域法局域遗传规划法 AbstractThe Abstract The electrical power load is the important foundation for electrical departments to do operating and programming．Precise load forecasting will improve power system reliability and make it operate more economically． Power system is a very strong nonlinear system，and appears as chaos behavior． In this paper，it can be drawn that the load time series are chaotic ones by calculating and anal)zing the load time series The major chaotic character exponents can also be calculated and the predictable time cail also be calculated by using Lyapunov Exponent， The fonctionat relation between the forecasting point and its following time series can also be got instead by getting the relationship between the near points of forecasting point and its following points The local forecasting is widely used because of its high preciseness and tenderness．Linear fimction expressions are often used as the forecasting models，but forecasting models are more accurate by using nonlinear models．How to get nonlinear models is the purpose of this paper A new method based on chaos theory and genetic programming algorithm for short—term load forecasting，local genetic programming algorithms，is proposed in the paper．This algorithm has the ability automatically to find out the correla