【摘要】 为提高负荷预测精度,提出了一种新的4层模糊神经网络短期负荷预测模型.该模型将模糊逻辑和神经网络的长处融合在一起,使模糊推理和解模糊均通过神经网络来实现.选取的隶属函数使神经网络权值有一定的知识表示意义,并通过模糊化层将输入特征量转化为模糊量.在模糊推理层提出了两种不同的算法来完成模糊推理,然后从中确定出模糊取小算法预测效果更好.最后在输出层通过适当的解模糊得到确切的预测输出值.仿真结果表明了该方法的有效性.更多还原

【Abstract】 In order to enhance the load forecasting precision,a short-term load forecasting model based on four layers fuzzy neural networks is presented.By fusing the strong points of fuzzy logic and neural networks,the fuzzy inference and defuzzification of this model were both realized by neural networks.The selected membership function made neural network weight values have definite knowledge meaning,and the input characteristic variables were translated into fuzzy variables by fuzzy layer.On the fuzzy inference layer,two different fuzzy inference algorithms were put forward to accomplish the fuzzy inference,and it was confirmed that the fuzzy which had got smaller inference algorithm could achieve better forecasting effect.Finally,the reliable forecasting output values were gained by proper defuzzification on the output layer.The simulation results showed the validity of this method.更多还原