Aiming at the problem of trajectory tracking accuracy degradation in large-scale tank derusting robots during vertical wall operation due to wind disturbances and sudden curvature changes, a strategy based on fuzzy model predictive control was proposed. Firstly, the kinematic model of a four-wheeled robot was constructed and discretized to derive a linear discrete equation within one sampling period. A prediction model and a multi-objective optimization function weere established, and the optimal solution was obtained via quadratic optimization. The position error and heading angle error undergo fuzzification and normalization processing, and were subsequently used as input variables to the fuzzy controller. An improved weighted center average method was employed for defuzzification, which outputed adaptively adjusted parameters for the prediction horizon, control horizon, and weight matrices, thereby achieving real-time optimization of the MPC controller parameters. Simulation verification conducted using MATLAB/Simulink demonstrates that compared to traditional MPC(model predictive control), the proposed fuzzy MPC method reduces the displacement error by 49% and its peak value by 14.65%, while reducing the heading angle error by 46.7% and its peak value by 17.4%, the effectiveness of the proposed Fuzzy-MPC algorithm is verified, which provides an effective and feasible optimization strategy for high-precision trajectory tracking control of wall-climbing robots.