基于增量反馈RBM的电能替代终端能源评价模型

Electric power substitute end-use energy evaluation model based on incremental feedback RBM

针对电能替代终端能源评价方法中存在的效率低、精确度不高等问题,提出一种基于深度学习思想的电能替代终端能源的评价模型。重新设计最优RBM训练算法,提出增量反馈的模型优化迭代策略,优化模型训练参数的初始化赋值,提升电能替代终端能源评价效率和精确度。实验结果表明,对比其它优化算法模型,增量反馈的RBM在抽取和表现特征方面,能够很好将样本的本质特征体现出来,取得了更加高效和准确的评价结果。

Aiming at the situation of poor efficiency and low accuracy in the domain of the electric power substitution of end-use energy,an evaluation model based on the thought of deep learning was proposed.An optimal RBM training algorithm was redesigned and the incremental feedback iteration strategy was put forward,and the assignment of the value initialization was optimized and the efficiency and accuracy of evaluation was improved in electric power substitute end-use energy.The simulation results show that using the method has good capacity of extracting and expressing features.It can also reflect the essential features and achieve more effective and accurate evaluation comparing with other different modern optimization algorithms.