Power Load Forecasting by Using Quantum Neural Network

• Main Authors: Chang-Teh Lee, 李昌德
• Other Authors: Rey-Chue Huang
• Format: Others
• Language: zh-TW
• Published: 2004
• Online Access: http://ndltd.ncl.edu.tw/handle/69606706693514989599

碩士 === 義守大學 === 電機工程學系 === 92 === In this thesis, the non-stationary signal prediction by using quantum neural network (QNN) is proposed. The signals with fuzziness are expected to be classified clearly for enhancing the learning efficiency of neural network due to the hidden units with various graded levels in QNN structure. As we know, power load forecasting is a type of non-stationary signal processing. Such an environment of signal information is generally complex and ill defined; also the behavior of power load is time-varying and dynamic. It sometimes might make NN have an ill learning and then cause it to have an unsatisfactory performance. Therefore, in order to improve the accuracy of NN’s prediction when it is used to deal with such signals with high non-stationary, uncertainty and fuzziness, several pre-analysis works might needed. However, it is usually a very difficult work to design analysis system, if too many unknown or uncertain factors are involved in the signals desired to process. In this research, QNN power load forecasting is developed. This model is expected to precisely capture the complex relationships among load and its possible influencing factors, such as weather information, time of day, season of year and so on. By passing the complex signal’s pre-analysis work, the QNN model can effectively be trained and then have a better performance than traditional NN has.