基于优化神经网络的电力调度自动化系统入侵检测算法

Approach of intrusion detection for power dispatching automation system based on improved neural network

电力调度自动化系统是由用来监视、测量和控制参与发电设备所组成的系统,它的安全运行是电力系统正常运转的基础。提出一种基于优化的神经网络的电力调度自动化系统入侵检测算法,通过采用粒子群优化算法对传统的BP神经网络参数进行优化,并使用多种类型的攻击数据对参数优化的BP神经网络进行训练,使基于该优化算法的检测模型可有效检测入侵电力调度自动化系统的典型攻击。另外,将人工误操作视为一种特殊类型的攻击,通过对此类攻击发生时系统重要运行特征的训练学习,算法也可实现对此类型攻击的识别检测。实验结果对比验证了算法的可行性和有效性。基于本算法的模型可为电力调度自动化系统的安全提供有力保障。

Electric power dispatching automation system is used to monitor, measure and control the devices that participated in generating electricity, which is the key factor to ensure the running of power systems. This paper proposes an approach of intrusion detection based on improved neural network. By using particle swarm optimization to optimize the parameters of BP neural network and training lots of data of several typical attacks, the improved BP neural network will be able to detect the most typical attacks effectively. In addition, we take human errors as a special attack. Some important running features of the system are extracted through the process of this attack to train the BP neural network. Thus, the improved BP neural network can detect human errors either. The experiment results show the feasibility and effectiveness of the proposed method. The model based on the proposed way can provide powerful guarantee for the safety of power dispatching automation system.