Monthly-scale runoff forecast model based on PSO-SVR

Abstract The current methods used in the Lubbog reservoir runoff forecast generally have shortcomings such as low accuracy and stability. Aiming at these problems, this paper constructs a PSO-SVR mid-and-long term model, it uses particle swarm optimization algorithm (PSO) to find penalty coefficient C, insensitivity ε gamma parameter of Gaussian radial basis kernel function support vector regression machine (SVR). results demonstrates that average relative errors model is relatively small, which are all within reasonable range; qualification rates for most monthly forecasts above 80%. Experimental indicate compared with multiple analysis, has higher accuracy, stronger stability, credibility. It certain practical value provides reference related research.