A piano key weir (PK weir) is a non-linear, labyrinth-type weir that benefits of a high discharge capacity, and is well suited for low head dams. Determination of the discharge coefficient (Cd) is considered as one of the most important issues, which plays a substantial role in reducing structural and financial damages caused by floods. The main aim of the present study is to experimentally investigate the variations of PK weirs discharge coefficient (Cd) through altering the geometric parameters. The obtained results revealed that in modified PK weirs (by an 11.5% increase in weir height, changing the crest shape, and fillet installation), the Cd values were about 5–15% more than those of the standard PK weirs. The Cd values of the non-contracted weirs were increased by increasing the inlet/outlet width ratio by 1.4, while this relation was adverse for contracted weirs. In the modified PK weirs, the submergence would occur faster than the standard weirs, while the complete submergence would occur later. Moreover, robust kernel-based approaches (kernel extreme learning machine and support vector machine) were successfully employed to the extensive experimental dataset by taking into consideration the Cd as a function of dimensionless geometric variables of PK weirs. The obtained results showed that the ratio of the upstream hydraulic head (H0) to total weir height (P) plays a significant role in the modeling process
