Offshore gangways play an important role in the operation, maintenance, and personnel transfer of offshore wind power platforms. With the gradual deepening of research on offshore gangways, the requirements for control accuracy are becoming increasingly high. A control algorithm has been proposed to compensate for joint torque disturbances caused by waves during the operation of offshore gangways, in order to achieve joint torque control under the influence of disturbances. RBF neural network has good approximation ability to nonlinear functions, can fit the uncertainties caused by wave disturbance and model uncertainty, and uses different radial basis functions according to its characteristics, and uses improved sparrow search algorithm to optimize the value of RBF neural network basis function parameters, so that the neural network has better fitting accuracy. The stability of the control system was verified by constructing Lyapunov functions. Adopting neural network adaptive control law to compensate for wave disturbances and fitting errors by setting robust terms, thereby improving the control accuracy of offshore gangways on complex sea surfaces. The designed controller and neural network were simulated and validated, and the actual environment of the offshore gangway was simulated in adams. The designed controller has good control performance, and the designed neural network has good fitting effect on torque disturbances caused by waves.