潮汐发电中调和分析与人工神经网络预测潮位

潮汐电站发电量与每日潮差的大小有关，通过对潮水位的预测，可以预估潮汐电站的发电量以及编制适合的发电计划。对现有的经典调和分析法和新兴起的人工神经网络潮位预测方法进行了比较分析，总结了两种方法的发展现状和存在的问题。分析表明，人工神经网络方法预测精度更高，和实测数据对比其具有更小的误差；调和分析误差稍大，但是满足国家海洋局规定的允许误差。调和分析法和人工神经网络法的预测结果均可应用于潮汐发电当中。

Tide Prediction in Tidal Power Generation by Harmonic Analysis and Artificial Neural Network

The output of tidal power generation depends on the tidal range. Therefore we can estimate the power generation of a tidal power station and plan the generation schedule by forecasting tidal levels. The article reviews the existing harmonic analysis methods and emerging artificial neural network tide prediction methods, and then summarizes the development and existing problems. It can draw a conclusion that artificial neural network methods can make the results more accurate, and the error is smaller compared with the actual data. While the harmonic analysis error is larger, but it still meets the permissible error stipulated by the state oceanic administration, so both methods can be used in certain areas.