基于元胞自动机的城市地表径流流向优化算法

在考虑城市地表复杂水力梯度特征的前提下,提出了一种基于元胞自动机的城市径流流向优化(UCA)算法,通过坡面流模拟试验选出不同流态的最优水量分配权重,并将UCA算法与D8、4+4N以及FMFD算法进行不同数学曲面上的单位汇流面积对比试验,选择相对误差、均方根误差和平均绝对误差对算法性能进行评价。结果表明:UCA算法中适用于城市陡坡汇聚流态和缓坡漫散流态模拟的最优水量分配权重取值分别为2.0和1.1,在凸坡、凹坡、山脊和直坡面上,UCA算法的单位汇流面积模拟结果与理论值的相对误差和均方根误差均最小;UCA算法既适合模拟城市平坦地形上的缓坡漫散流,又适合模拟城市水力梯度较大地形上的陡坡汇聚流,可为城市地表径流模拟提供更优的求解思路。

Optimal algorithm for urban surface flow direction based on Cellular Automata

This paper proposed an optimal algorithm for the calculation of urban runoff direction based on Cellular Automata(UCA). The optimal water distribution weights that is suitable for different flow patterns are selected by the flow simulation experiments on slopes. The UCA, D8, 4+4N and FMFD algorithms are used to calculate the Specific Catchment Area(SCA)on four mathematical surfaces, and the results of four algorithms are compared. The relative error(RE), root mean square error(RMSE)and mean absolute error(MAE)are chosen to evaluate the performance of four algorithms. The results showed that the optimal water distribution weights of UCA algorithm for the converge flow and the diverge flow are 2. 0 and 1. 1 respectively, and the RE and RMSE between the UCA simulated and theoretical values of SCA are the smallest on convex slopes, concave slopes, ridges and plane slopes. The UCA algorithm is suitable for the diverge flow on urban flat terrain and the converge flow on terrains with steep hydraulic gradient, which can provide a better solution for the simulation of urban runoff.