【摘要】 首先描述基于数据挖掘的舰船电力负荷预测系统架构,然后按照此架构进行系统实现,并结合舰船电力负载预测的特点,利用遗传算法获取较好的搜索空间,这样可以避免BP神经网络算法陷入局部最优的情况。通过对比实验结果可知,本文所采用的遗传算法和BP神经网络相结合的优化算法预测能力强,拟合度高。更多还原

【Abstract】 Firstly,this paper described system schema of ship power load forecasting based on data mining. Then follow this schema to realize system. Combine with the features of ship power load forecasting,use genetic algorithm to obtain better search space. It could avoid BP neural network algorithm to fall into local optima. Finally,by comparing the experimental results,optimization algorithm was used in this paper based on genetic algorithm and BP neural networks,which had predictive ability and high degree of fitting.更多还原