基于SMDP强化学习的太阳能供电信息容侵控制方法

An Intrusion Tolerance Control Method of Solar Power Supply Information Based on SMDP Reinforcement Learning

针对目前方法未能使用小波去噪法处理太阳能供电信息而导致对其进行容侵控制时,存在控制正确率低、控制时间长以及控制的误报率高的问题,提出基于SMDP强化学习的太阳能供电信息容侵控制方法。寻找数据中的缺失值并对其进行插补处理,通过插补结果,使用小波去噪算法完成数据的去噪;使用SMDP强化学习算法强化数据,建立太阳能供电信息的容侵控制模型;将去噪后的供电信息数据输入模型中,依据输出结果实现太阳能供电信息数据的容侵控制。实验结果表明,运用该方法进行数据容侵控制的正确率高、控制时长短、误报率低。

When the current method performs intrusion tolerance control on solar power supply information,the wavelet denoising method fails to denoise the solar power supply information,thus the method has low control accuracy,long control time and high false alarm rate of control. To this end,a new of solar power supply information intrusion tolerance control based on SMDP reinforcement learning is proposed in this paper.This method finds missing values in the data and performs interpolation processing on them. Through the interpolation results,the wavelet denoising algorithm is used to complete the denoising of the data;the SMDP reinforcement learning algorithm is used to strengthen the data,and the intrusion tolerance control model of solar power supply information is established;Finally,the denoised power supply information data is put into the model,and the intrusion tolerance control of the solar power supply information data is realized according to the output results. The experimental results show that this method helps to control data intrusion tolerance with high accuracy,shorter control time and lower false alarm rate.