Evaporation and seepage are the most important forms of water loss in an open channel. Evaporation loss depends on the area of free surface while, seepage loss is function of channel geometry. The design that achieves minimum water loss is determined for trapezoidal and rectangular cross sections using Lagrangian method with Manning’s equation as a constraint. The water loss function which comprises evaporation and seepage loss was used. Matlab code is created to solve the equations of optimization by trial and error. An approach to determine the bed width and water depth from known discharge, side slope, longitudinal slope, manning’s coefficient, evaporation rate, and soil type is proposed. Design charts which determine bed width and water depth that ensure minimum water loss are presented. Moreover, artificial neural networks are used to get the best design for any value of evaporation rate. Using the created design charts and neural networks optimum design can be determined easily, accurately and quickly.

蒸发和渗漏是明渠中最重要的水分流失形式。蒸发损失取决于自由表面的面积，而渗漏损失是通道几何形状的函数。使用曼宁方程作为约束条件的拉格朗日方法确定梯形和矩形横截面的水损失最小的设计。使用了包括蒸发和渗漏损失的失水函数。创建Matlab代码以通过反复试验来求解优化方程。提出了一种根据已知流量，侧坡度，纵向坡度，曼宁系数，蒸发速率和土壤类型确定河床宽度和水深的方法。给出了确定床宽和水深以确保最小水损失的设计图。而且，对于任何蒸发速率值，都使用人工神经网络来获得最佳设计。使用创建的设计图和神经网络，可以轻松，准确，快速地确定最佳设计。