Modeling and attitude control of aircraft with center of gravity variations

Active center of gravity (C.G.) control is a key technology of vehicle management system, which has been extensively used in various types of aircraft. With active C.G. control system, aircraft may deviate from its known dynamics due to the C.G. offset, thus bringing new problems to the existing flight control system. A new design approach of the attitude control system of aircraft with C.G. variations is proposed in this paper. First of all, a 6-DoF flight dynamics model of aircraft with C.G. variations is built. Taking the effects of the C.G. shift into consideration, equations of motion expressed in the body-fixed reference frame are derived using the Newton´s second law. The resulting model can accurately reflect the effects of C.G. offset. Based on the above model, an attitude controller is designed to provide consistent performance under varying C.G. locations. This controller consists of four modules, expected attitude characteristic module, inverse dynamics module, actuator allocation module and in-flight C.G. identification module. In the expected attitude characteristic module, a second- order linear characteristic equation of dynamics of those Euler angles is expected because it is the moment that dominates the aircraft´s rotation and directly determines the second-order derivatives of the Euler angles. In the inverse dynamics module, both the dynamical equations of rotation and in-flight estimation of C.G. position are used to derive the moment command from Euler angular acceleration command. In the actuator allocation module, the deflection commands of the actuators are determined according to the identified control derivatives to achieve the required moments. In the last module, the estimation of C.G. location can be obtained by adaptive weighted data fusion based on optimal weight distribution principle. Simulations are carried out under various possible conditions of C.G. offset to validate the proposed design method. The results demonstrate good c- ontrol performance and effectiveness of this approach.