（Lower Changjiang River Bureau of Hydrological and Water Resources
Survey, Nanjing 210011, China）
Abstract: In the lower tidal reach of the Yangtze River, the mean velocity
at a cross-section can be calculated by using the ADCP real-time
index velocity in the on-line flow measurement system, which can be
solved by establishing a neural network model. Based on the neural
network toolbox of MATLAB software, this paper introduces the
detailed steps and solving methods of establishing the
corresponding model in the toolbox, taking nine tidal measurements
of the Nanjing Hydrological Experimental Station from May 2014 to
December 2016 as an example. The calculation results show that the
model is good fitting effect and can meet the accuracy requirement.
Therefore, it has a good application prospect in the study of
calculating the mean velocity at a cross-section of tidal reach
based on the ADCP real-time index velocity.