基于分形理论的内部过电压识别技术

Internal overvoltage recognition technology based on fractal theory

电力系统中的内部过电压类型多种多样,其产生原因错综复杂.目前,对过电压智能化分析的整体水平还不高,数据的分析只能依靠人工的方法,由生产技术人员凭经验并依据过电压波形的时间、幅值、频率等参数进行分析.但仅仅依靠这几个参数,有时不能对内部过电压进行完全正确的定性.对常见的几种内部过电压进行ATP-EMTP(电磁暂态计算软件)仿真,并提出基于分形理论的过电压信号识别,计算出过电压波形的分形维数.仿真结果表明,随着过电压发生条件的不同及参数的变化,同种过电压的分形维数在一定的范围之内;不同种类的过电压的分形维数在不同的范围.

There are many kinds of overvoltages and the reasons for these are different. At present, the only way to analyze the overvoltage data are relying on production and technical personnel, who analyze it rely on experience and the time, amplitude, frequency of overvoltage waveform. But sometime the overvoltages can＇t be determined only according to these parameters. Several common internal overvoltages are simulated by ATP-EMTP （alternative transients program-electormagnetic transient program） and proposed to calculate the fractal dimensions of overvoltage signal waveform according to fraetal theory for the first time. The fractal dimensions are calculated from the simulation results. The simulation and fractal dimension calculation results show that the fractal dimensions of the same kind overvoltage existed in a certain range; the ranges of fractal dimensions are different if the overvoltages are different.