Abstract: Here internal structures were added into the curved side wall and vertical seam, in order to improve the fish passage efficiency of fishways. Auxiliary fish passage devices were arranged on the inner wall and bottom plate of the fishway. Six layouts were also combined to explore the effect on the fish upstream behavior using simulation and model testing. The results showed that the success rates of upstream migration were 36.7%, 43.3%, and 73.3%, respectively, for the combinations of barriers in schemes 1, 2, and 3. Because scheme 1 was set as more barriers, the flow velocity in the mainstream decreased rapidly to cause the vortex-like flows, where the fish was unable to identify the upstream direction. There was the a low flow velocity in the downstream half of the mainstream, leading to difficultly ascending. A group of barriers was removed perpendicular to the direction of flow in scheme 2. The mainstream flow was connected from the top to the bottom, indicating the fish upstream migration. The sScheme 3 was equipped with the a central connection and only one group of barriers on the side wall. The highest success rate was achieved for fish upstream migration. The wide distribution of mainstream flow velocity and large flow velocity differentiation were provided more selective areas for fish with the different physical abilities, thus improving the fish upstream migration. There were the hindering and disturbing effects of circular columns on the flow in the tank chamber, indicating the variation in the mainstream flow path and flow velocity distribution. There was the an increase in the mainstream branches with the upstream flow velocity decreased. The test fish was easily fieasily found the correct path. The maximum mainstream flow velocity was greatly contributed to the better better-ascending fish. An optimal number of recirculation zones were was also maintained to ensure the rest for upstream fish. The combinations (schemes 4, 5, and 6) of straight circular columns were introduced in the scheme 3 to further improve the effect of fish upstream migration. The success rates were 56.7%, 73.3%, and 90.0%, respectively, for the upstream migration. The internal disturbance was significant in the area that composed of internal barriers in scheme 4, resulting in a narrower middle mainstream width and lower flow velocity than on both sides. There was the a low flow velocity in the tail of the tank chamber. There was no benefit to inducing fish upstream migration. In scheme 5, the width of the middle mainstream increased, while the flow velocity in on both sides decreased. The lower flow velocity and longer distribution distance in the higher flow velocity area were promoted the fish upstream migration, while reducing the swimming difficulty. There was the a cylindrical diameter with one-half of the inlet width of the tank chamber in scheme 6. The highest success rate was then achieved for the test fish upstream migration. The middle mainstream flow velocity increased in the tank chamber. The mainstream was much clearer to fill with internal barriers. There was the a diffusion from the middle mainstream to both sides. The width of mainstream distribution increased for the fish upstream migration. In the recommended solution, the fish swimming was roughly followed the mainstream upstream, providing the multiple mainstream flow paths for fish. The pass efficiency of fish was improved with the large discrimination of flow velocity, the reasonable attenuation of mainstream flow velocity, and the flow pattern with the fewer backflow zones.