Institute: District of Columbia
Year Established: 2015 Start Date: 2015-03-01 End Date: 2016-02-28
Total Federal Funds: $10,000 Total Non-Federal Funds: $33,217
Principal Investigators: Nian Zhang, Nian Zhang
Abstract: Prediction of river stage has been an interesting research topic in hydrologic engineering, from water supply, flood forecasting, and water quantity and quality management viewpoint. For example, long-term evaluation on the potential impact of climate change on the water supply resources of the Washington, D.C., metropolitan area is very important. In order to predict the future river stage based on the current information is very much necessary for water supply managers. We are proposing a research area: a) to develop a Multi-Objective Hybrid PSO-EA Feature Selection Algorithm with LS-SVM classifier. The USGS real-time water data will be used as time series input. Specifically, a simultaneous parameter optimization algorithm for the least squares support vector machine (LS-SVM) predictor using the hybrid particle swarm optimization (PSO) and evolutionary algorithm (EA) will be developed. In addition, an effective LS-SVM classifier that can predict future discharge values based on more than one input (i.e. previous discharge values) will be developed; b) to compare the performance of a variety of promising computational intelligence methods that can efficiently address the water quantity and quality problems in the watershed. They include the recurrent neural networks, and the stand-alone LS-SVM method without the proposed hybrid particle swarm optimization and evolutional algorithm; c) to generate a flow-duration curve (FDC) for specific locations of the Potomac River, and analyze their accumulated impact to the Potomac River.