Abstract Dams, regarded as a form of hydraulic engineering, are purposed to intercept river or canal flow to either elevate water levels or regulate discharge rates. The discharge of a dam is subject to influence by various factors, including precipitation, temperature, and regulatory policies. Nevertheless, in real-world scenarios, the precise value of the expected dam flow rate is challenging to ascertain through simple arithmetic operations. In this study, a novel dam flow model has been formulated. Through the utilization of the DDPG algorithm and deep learning techniques applied to data from the current month, the parameters of the dam control differential equations have been determined, and the transfer function after Laplace transformation has been established. Consequently, only the current month’s fluctuations in the water level of Lake Huron need to be determined to derive the corresponding dam flow rate for that month.