The invention discloses an optimization design method of a radial-flow-type hydraulic turbine. In the design method, a unitary thermal optimization design, a three-dimensional modeling method of a through-flow part and a complete machine optimization platform are utilized, wherein the optimization platform comprises four modules, namely nozzle blade and impeller blade parameterization, a coevolution genetic algorithm, a self-adaption approximation model, and autocall of CFD (computational fluid dynamics). Through the parameterization, characteristic variables describing impeller blades, nozzle blade patterns and installation angle variation are extracted. The optimization target is to enhance the overall efficiency and expansion ratio of the hydraulic turbine simultaneously under a complete machine environment. The optimization platform can be used for reducing the calculated amount and accelerating the convergence by virtue of the following measures: the approximation model is built and updated by a dynamic sampling strategy, and enough prediction accuracy is obtained by virtue of less CFD calculation; and a complicated multivariable optimization problem is decomposed into a plurality of relatively independent and interactive subproblems by virtue of the coevolution genetic algorithm, so that not only can the characteristics of the original problem be maintained, but also the calculated amount is reduced effectively.