Abstract: For the problem that is caused by surplus torque of aircraft rudder electro-hydraulic loading system is not easy to inhibit, compound controller structure function and control strategy were proposed that combined rubber-metal buffer spring, the displacement of steering gear and speed feed forward of aircraft rudder and the output force feedback use single neuron PID based on dynamic RBF neural network on-line identification. Ant clustering algorithm was used to optimize the dynamic RBF neural network to identify the object online to obtain Jacobian information, then use the single neuron PID controller complete control parameters online self-tuning control. The simulation results show that this method can achieve rapid and accurate load for simulated aircraft rudder power load under laboratory conditions, and can guarantee the stability of the system and has strong robustness.

Key words: aircraft rudder electro-hydraulic loading system, surplus torque, single neuron PID controller, dynamic RBF neural network, ant clustering algorithm