I-WOA(improved whale optimization algorithm) was proposed to solve the multi-objective ATO speed curve optimization problem of high-speed trains in complex wind-blown sand environment. Firstly, a dynamic model of wind-sand coupled train was constructed, sand particle size parameter was introduced, aerodynamic resistance was considered comprehensively, and multi-objective optimization function including energy consumption index, punctuality index and comfort index was established. Through multi-strategy algorithm improvement: improve tent chaos mapping, design nonlinear convergence factor a, ε-elite dimensional reverse learning strategy fusion Levy flight, improve I-WAO convergence speed, global search ability, out of the region optimal ability. Finally, the simulation experiment based on CRH3C train shows that under the condition of 1.0 g/m³ sand concentration and 12.5 km/h cross wind speed, the I-WOA algorithm can shorten the running time by 6.12%, reduce the energy consumption by 6.63%, and improve the comfort level by 0.74% compared with the traditional WOA. When the 200 km/h temporary speed zone is added, I-WOA still has a clear advantage. The simulation results show that I-WOA can effectively solve the multi-objective optimization problem under strong disturbance environment through multi-strategy cooperative optimization, and provides a practical solution for ATO control of high-speed trains under complex sand environment.