Urban water consumption forecast based on PQPSO-LSSVM
Xingtong Zhu, Jianping Chen · 2013
It is well known that accurate forecast of urban water consumption has significance for water supply system. In order to improve the accuracy of prediction, we proposed a novel forecast method based on quantum particle swarm optimization algorithm (QPSO) and least squares support vector machine (LSSVM). Firstly, an improved quantum particle swarm optimization algorithm is proposed. The proposed algorithm is encoded by qubit phase adjust the inertia weight factor and global factors according to the particle's fitness value, which is defined as PQPSO. Secondly, the parameters of LSSVM are selected by PQPSO. Finally, urban water consumption is predicted by the proposed method. The experimental results show that prediction accuracy and computational speed are better than the method based on SVM, LSSVM. Therefore, the proposed method is an effective tool for urban water consumption forecasting.