To address the significant load swing caused by unknown external forces during the hoisting process of bridge cranes, a neural network Sliding Mode Controller (SMC) with parameter optimization based on an improved honey badger algorithm is proposed. First, a Radial Basis Function (RBF) neural network is employed to effectively approximate the unknown external forces in the dynamic model of the crane system. The output of this approximation is then fed into the SMC to ensure the asymptotic stability of the crane system during the hoisting process. Next, the Chebyshev chaos population mechanism and Gaussian mutation strategy are introduced to improve the standard honey badger algorithm, enhancing its optimization ability for the rate parameters of the RBFSMC and thereby reducing system oscillations. Finally, through comparative simulations with SMC, RBFSMC, and particle swarm optimization-RBFSMC controllers, the proposed controller is shown to achieve better positioning, output driving force, and anti-swing control performance, demonstrating its superior effectiveness in precise positioning and sway suppression for the crane system.