| 摘要: |
| 【目的】解决传统优化算法在新安江模型参数估计中存在的早熟、收敛速度慢和易陷入局部最优等问题。【方法】在标准PSO算法的基础上，引入小生境和交叉选择算子，对寻优过程中粒子的个体历史最好位置进行多样化处理，提出基于小生境和交叉选择算子的粒子群算法(NCSPSO)，建立基于NCSPSO算法的新安江模型参数估计数学模型，并给出具体求解步骤。最后将该方法在具体流域的洪水预报中进行应用。【结果】NCSPSO算法计算时间短，参数估计精度大大提高，且预报结果均达到了规范要求。【结论】NCSPSO算法为新安江模型参数估计提供了一条新途径。 |
| 关键词: 参数估计 新安江模型 小生境 交叉选择 粒子群算法 |
| DOI： |
| 分类号: |
| 基金项目:国家“863”计划项目(2006AA01A126)；国家自然科学基金项目(50979088)；陕西省国际合作重点项目(2008KW-32) |
|
| Application of optimization algorithm for parameter calibration with the Xin’anjiang model |
|
|
| Abstract: |
| 【Objective】The study was conducted to overcome the disadvantages of classical optimization algorithm used to solve Xinanjiang model parameter estimation problems,such as premature convergence,poor convergence and liable to fall in local optima.【Method】Particle swarm optimization algorithm based on the niche,crossover and selection operators (NCSPSO),which carries out diversification treatment with history best position of particle in the process of optimization,was presented in this paper.The mathematical model and procedures for Xin’anjiang model parameter estimation by using NCSPSO were proposed in detail.The mathematical model was applied to specific basin.【Result】Study results show that this method has higher efficiency and precision in parameter estimation,and flood forecast result can meet standard of demand.【Conclusion】NCSPSO will be a new method for Xin’anjiang model parameter estimation. |
| Key words: parameter calibration Xin’anjiang model niche crossover and selection particle swarm optimization algorithm |
