This study investigates the speed tracking control problem for the new hydraulic transformer controlled system based on the active regulating common pressure rail with both input delay and strong interference. Firstly, the Pade approximation approach is adopted to approximate and compensate for the input delay of the valve plate rotating controlled by the hydraulic swing motor. Then, the Levant filter is introduced to solve the ‘explosion of complexity’ caused by repeated differentiation of the virtual control signals, and the compensation mechanism is constructed to compensate for the filtering errors. Besides, due to the strong non‐linear characteristics of the proposed system, the pi‐sigma fuzzy neural network is employed as the function approximator to address the non‐linear terms. Based on finite‐time command control technology and the barrier Lyapunov functionals, a robust controller is synthesised to ensure a better error convergence performance while without violating the state constraint. Finally, the effectiveness of the controller is verified by comparison with the traditional control method and simulation result.