The smart skin of an aircraft is realized by integrating distributed sensors, actuators, and controllers into the composite skin, thereby enabling it to monitor its own state and detect damages. The physical field inversion algorithm plays a key role in the signal processing of the smart skin. However, due to factors such as the low sensor density, traditional inversion algorithms exhibit limited accuracy. In order to enhance the monitoring precision of the smart skin, a back propagation(BP) neural network-improved grey wolf optimizer(IGWO) inversion algorithm, which combined a BP neural network with an IGWO-optimized Kriging model, was proposed. A prototype of the smart skin was subsequently fabricated, and wind tunnel tests were conducted to validate the proposed algorithm. The results demonstrate that the BP-IGWO inversion algorithm achieves higher accuracy and superior detail representation compared to traditional inversion algorithms, and can better monitor the state of smart skin.