基于量子加权GRU神经网络的电力系统短期负荷预测

Short-term Load Forecasting of Power System Based on Quantum Weighted GRU Neural Network

为提升电力系统短期负荷预测精度,提出粒子群算法优化量子加权门控循环单元神经网络模型。首先,将量子加权神经元融入门控循环单元神经网络中,构建量子加权门控循环单元神经网络预测模型,利用量子信息处理机制,提高该神经网络的非线性逼近能力和泛化能力。然后,使用全局优化能力较强的改进粒子群优化算法对所提出模型的参数进行寻优,构建权重矩阵进行负荷预测。最后,通过实际电网算例进行仿真,仿真结果表明,本文提出的粒子群优化量子加权门控循环单元神经网络预测模型的预测精度较高。

To improve the short-term load forecasting accuracy of power system,a particle swarm optimized quantum weighted gated recurrent unit(QWGRU-PSO)neural network model is proposed in this paper.First,the quantum weighted neuron is integrated into the GRU neural network,and the quantum weighted GRU neural network prediction model is constructed.By means of the quantum information processing mechanism,the neural network can obtain better nonlinear approximation and generalization capabilities.Then,the improved particle swarm optimization(PSO)algo⁃rithm with a strong global optimization capability is used to optimize the parameters of the proposed model and construct a weight matrix for load forecasting.Through the simulation of an actual power grid as an example,it is verified that the proposed QWGRU-PSO neural network prediction model has a higher prediction accuracy.