In this paper, the fuzzy control problem of unmanned vehicles is studied. In the unmanned vehicle systems, the lateral offset, yaw rate and heading offset are sampled and sent to the control center, and may unfortunately be dropped due to transmission conflicts. To solve the packet loss issues, a fuzzy model is first established to approximate the nonlinear lateral dynamics of the unmanned vehicles. The measured signals are assumed to be sampled in a random way, induced by which, the phenomenon of probabilistic packet loss is considered. Then, a method of input delay considering double stochasticity delay is used to model a probability distribution of the packet loss problem under stochastic sampling. Furthermore, relaxed sufficient conditions are given to ensure the robust mean square exponential stability and \(H_{\infty } \) performance of the closed-loop system. Finally, the effectiveness of the lateral controller design is verified via a simulation example.