Flow Induced Oscillations (FIO) of tandem cylinders can be enhanced to harness hydrokinetic energy by varying the system parameters. In general, the Converter consists of two mass–spring–damper oscillators subjected to transverse FIOs and specifically Vortex Induced Vibrations and galloping. These FIOs are strongly influenced by variations of the inflow velocity, damping, stiffness, mass and in-flow center-to-center spacing L between two tandem cylinders. In turn, those influence the harnessed power and efficiency of the Converter. In previous experiments, the interactions between the cylinders were proven to be beneficial for the synergy of the cylinders. In this paper, modeling of tandem-cylinder converters is studied considering the Converter parameters, aiming at enhancing the cylinder synergy resulting in increased harnessed power by using a backpropagation (BP) neural network. The main conclusions are: (1) The surrogate model is constructed by a BP network using the experimental data to reduce excessive experimentation or computational inaccuracy. The harnessed power at different flow velocities is computed by the present model and is found to be consistent with experimental results not included in the modeling. (2) Increasing the damping ratio (0.20–0.30) of two tandem cylinders is conducive to improve the power efficiency, but has little effect on power harvesting. (3) In galloping, the harnessed power and its corresponding efficiency for the case of L/D = 1.57 perform at a higher level than that of bigger spacing ratios.