基于改进PSO-SVR模型的港口潮汐预报算法

Port tide forecasting algorithm based on improved PSO-SVR model

基于调和分析的传统潮汐预测方法仅考虑了引潮力引起的线性因素,忽略了风力、气压等非线性因素的影响,因此存在预测精度不高的问题。为了解决上述问题,本文提出一种粒子群算法(Particle Swarm Optimization, PSO)优化支持向量回归(Support Vector Regression, SVR)模型的港口潮汐预报算法。首先,采用SVR对港口潮汐水位变化进行建模,针对SVR核参数和惩罚因子设置难题,利用PSO算法进行全局寻优,自动确定SVR最优参数组合;其次,为了解决PSO算法易早熟,迭代后期计算效率下降的问题,在PSO算法中加入变异因子,提升算法效率;最后,利用伊莎贝尔(Isabel)港口潮汐数据进行仿真验证。结果表明,相较于单一调和分析模型和SVR模型,该方法的预测性能和鲁棒性均更高,具有良好的应用前景。

Traditional tidal prediction methods based on harmonic analysis include only the linear factors caused by tidal force, and ignore the influence of wind, pressure and other nonlinear factors. In order to solve this problem, a port tide forecasting algorithm based on support vector regression(SVR) model optimized by particle swarm optimization(PSO) is proposed in this paper. Firstly, SVR is used to model the change of port tidal water level. As it is difficult to set the SVR kernel parameters and penalty factors, the PSO algorithm is used for global optimization to determine the optimal parameter combination of SVR automatically. Secondly, in order to solve the problem of premature convergence of PSO algorithm and to improve computational efficiency in the later iteration, mutation factor is added to PSO algorithm to improve the efficiency of the algorithm. Finally, the tidal data of Isabel port is used for simulation verification. The results show that the proposed method can get higher prediction performance and robustness than the single harmonic analysis model and SVR model, and has good application prospects.