基于优化卷积神经网络的非侵入式负荷识别方法

Research on non-intrusive load identification method based on optimized convolutional neural network

针对居民家用电器使用环境安全问题,采用非侵入式负荷监测技术进行负荷监测。负荷识别方法是负荷监测的主要难题,采用优化卷积神经网络对负荷识别方法进行研究以克服负荷识别这一难题。基于U-I轨迹负荷特征,以卷积神经网络为基础,分析了卷积神经网络对U-I轨迹图自学习式提取特征识别过程;并采用粒子群优化算法,对卷积神经网络超参数配置进行优化,解决传统人工操作导致分类性能不确定性的问题,从而使卷积神经网络分类性能趋近于最优。研究表明:优化卷积神经网络具有很好的分类性能,可精确识别电器类型,完善负荷监测技术,为家用电器提供安全的使用环境,保障居民生命财产安全。

To solve the environmental safety problems of residentshousehold appliances,nonintrusive load monitoring technology is used for load monitoring.The load identification method is the main challenge in load monitoring.The optimized convolutional neural network is applied to study the load identification method to overcome the load identification problem.Based on the U-I trajectory load characteristics and the convolutional neural network,the self-learning feature extraction recognition process of the U-I trajectory diagram is analyzed by the convolutional neural network.The particle swarm optimization algorithm is employed to optimize the hyperparameter configuration of the convolutional neural network to solve the problem of uncertainty in classification performance caused by traditional manual operations,thereby making the classification performance of the convolutional neural network converge to the optimum.Research shows that the optimized convolutional neural networks have good classification performance,can accurately identify electrical appliance types,and improve load monitoring technology.It can provide a safe environment for the use of household appliances and protect the lives and properties of residents.