Aiming at the problem of communication delay in multiple unmanned aerial vehicles （UAVs） under a typical mountainous wind field， a time-varying delay distributed adaptive control algorithm based on temporal correlation is proposed. This algorithm holds an improved time-varying delay control capability and an enhanced anti-disturbance ability under wind fields. First， a typical mountainous wind field model is constructed based on multi-compound winds. Then， the formation model of multiple UAVs is established. Second， an adaptive delay control algorithm based on temporal correlation is proposed， utilizing the partial autocorrelation function （PACF） to achieve temporal regulation of both the delays and the state itself. In the control protocol， a formation control compensation vector is introduced， and the target state is input into the control protocol through a feedback gain matrix， allowing for more precise tracking of the target state. This algorithm also uses Lur’e nonlinearity to handle the dual coupling effects caused by the nonlinear characteristics of the wind field and the UAV itself， which removes the dependence on nonlinear bounded treatments. The adaptive law is applied to the algorithm to estimate the noise disturbance under the wind field. A non-smooth function is used to simulate unknown environmental disturbance， and an adaptive gain is designed to compensate for the dynamic characteristics of the system. Finally， the effectiveness of the proposed control protocol is verified through numerical simulations， demonstrating that the algorithm can improve the stability and adaptability of UAV formation under mountainous wind fields.