MC & OS-based CMAC compound control of inertially stabilized platform for aerial remote sensing

To improve the pointing accuracy and stability of inertially stabilized platform (ISP) control system for aerial remote sensing, a modified cerebellar model articulation controller (CMAC) compound control method is proposed based on the modified credit (MC) method and optimized smoothness (OS) method (MC & OS). Based on the number of the memory unit activation during the CMAC study process, different credit degrees are set to the memory units, which enhances the study efficiency and control precision of the CMAC, and avoids the precision degradation even collapse of the control system due to the diffusion of over-study in the system dynamic tracking process. Then, aiming at the problem of large system output fluctuation of conventional CMAC algorithm, the optimized weight algorithm is put forward to improve the smoothness of system output, enhance the stability of CMAC, and avoid the damage of the motor and transmission system caused by system output fluctuation. Finally, simulation analysis and experiment verification on a certain 3 axis inertial experiment platform in the laboratory were conducted to verify the proposed method. The experiment results show that after adopting the MC & OS-based CMAC compound control method, the control precision, response speed and output smoothness stability of the ISP control system are all improved effectively; the angular horizontal tracking error (RMS) under the moving base vehicle condition is 0.0216°, which is decreased by 55.09% and 30.55%, respectively compared with those for the PID control method and conventional CMAC control method.