Installation and establishment of appropriate flow measurement structures, their proper operation, and collection of water consumption data play a crucial role in empowering network managers' decision-making, ensuring fair water distribution and transmission, and ultimately achieving water conservation. Flap or hanging gates are simple and inexpensive devices for both automatic control and flow measurement. In this research, the laboratory results obtained from a rectangular hinged gate structure, which is placed in a channel according to hydraulic and geometric specifications and leads to flow rate measurement, were used to develop machine learning models. To estimate the flow rate in this type of channel, models including Group Method of Data Handling (GMDH), Support Vector Machines (SVM), and Random Forest (RF) were employed. To this end, parameters such as water depth, channel width and gate width, length, thickness, and weight were considered as input variables, and flow rate as the output (response) variable for modeling. The results showed that the Root Mean Square Error (RMSE) values for GMDH-, SVM-, and RF-based models were 0.024, 0.011, and 0.041, respectively, and the Coefficient of Determination (R2) values were 0.981, 0.996, and 0.955, respectively. A comparison between past research and the present results indicated the superiority of the SVM-based model over the other developed models. Water depth to channel width was identified as the most significant input data for the developed models