灌区明渠糙率及其计算方法

Open Channel Roughness of Irrigation District and its Calculation Methods

明渠糙率是影响灌区输配水和排水过程的重要参数,获取准确的糙率值是提高明渠设计合理性、实现灌区水动力多过程高精度模拟、提升灌区用水管理质量的重要前提。为提高灌区明渠的糙率计算质量,从定量的视角回溯了糙率表征的发展历史,阐述了各类糙率计算公式的来源及其之间的逻辑关系,分析了工程和物理影响要素,其中,工程要素主要包括渠道断面尺寸、壁面粗糙度、水工建筑物形式、渠道平顺情况以及渠底比降等,物理要素主要包括傅汝德数和湍流状态。在此基础上,首先回顾了糙率的原型观测方法,因为其是其他糙率获取方法的基础,进而对系统辨识方法、基于水动力方程反算的糙率优化方法、基于数据同化的糙率同步求解方法等研究工作进行了阐述,讨论了GPU并行计算技术、大数据分析技术和web技术快速发展的背景下糙率计算方法的发展趋势,在GPU海量并发线程背景下,着重分析了人工智能技术强悍的图形图像识别能力与渠道水动力过程基础波形及其非线性叠加水波波形以及水工建筑物几何信息之间的结合方法,指出该方法具有能实时获取不同空间平均糙率的显著优点,具有巨大的理论与应用潜力。

Open channel roughness is an important parameter affecting the water transmission and distribution and drainage process in the irrigation district.Obtaining accurate roughness value is an important premise to improve the rationality of open channel design,realize the high-precision simulation of multiprocess hydrodynamic,and improve the quality of water management in the irrigation district.In order to improve the roughness calculation quality of open channel in irrigation district,this paper traces back the development history of roughness characterization from a quantitative perspective,expounds the sources of various roughness calculation formulas and their logical relationship,and analyzes the engineering and physical influence factors,including channel section size,wall roughness,hydraulic structure form,channel smoothness and channel bottom gradient,etc.The physical elements mainly include Froude number and turbulence state.On this basis,firstly,the prototype observation method of roughness is reviewed,because it is the basis of other roughness acquisition methods.Then,the research work such as system identification method,roughness optimization method based on inverse calculation of hydrodynamic equation and roughness synchronous solution method based on data assimilation are described,and the GPU parallel computing technology is discussed.The development trend of roughness calculation method under the background of the rapid development of big data analysis technology and web technology is also discussed.Under the background of massive concurrent threads of GPU,this paper focuses on the combination method between the powerful graphic image recognition ability of artificial intelligence technology and the basic waveform of channel hydrodynamic process and its nonlinear superimposed water wave waveform,as well as the geometric information of hydraulic structures,It is pointed out that this method has the remarkable advantage of obtaining different spatial average roughness in real time,and has great theoretical and application potential.