遗传禁忌神经网络模型及其在电力负荷预测中的应用

Neural network models optimized by hybrid genetic-Tabu search and their application to power load forecast

针对神经网络权值学习算法的不足,结合遗传算法和禁忌算法各自的优势,建立了遗传禁忌神经网络预测模型。实际应用表明,与遗传算法相比,该模型具有收敛速度快、预测精度高等优点,具有满意的预测效果。

To solve the defects of the weights learning algorithm of neural network, the genetic algorithm and the Tabu search are combined to construct the neural network model. Application shows that compared with the genetic algo- rithm, the model has the advantages of quick convergence and high forecast accuracy.