Optimum Design Of Electro Hydraulic Loading System For Aircraft Servo

In the field of aircraft control,the steering gear is an important flight control servo element.In the laboratory conditions,the loading system is usually used to simulate the changes in the load of the aircraft and its servo system.It is mainly used to check and test the static and dynamic technical indicators.The electro-hydraulic loading system of the aircraft is a typical passive force servo control system in which the presence of excess force seriously affects the control performance and loading accuracy of the loading system,and many important performance indicators of the loading system are related to the excess force,so how to overcome the excess force is a key issue that must be addressed during the design of a load system.In this paper,based on the research status of electro-hydraulic loading system of aircraft steering gear at home and abroad,the electro-hydraulic servo valve,hydraulic cylinder,buffer spring,position sensor and force sensor are established according to the structure and working principle of the loading system.Mathematical model,the mechanism of excess force is analyzed,and the main factors influencing the loading performance are determined.Then,the key components of the electro-hydraulic loading system of the aircraft steering gear are optimized and optimized.The parallel pilot electro-hydraulic servo valve structure and the composite buffer cylinder structure are designed and the respective mathematical models are established.In order to further reduce the redundant force,the double-valve hydraulic cylinder synchronous control structure is designed,and the whole mathematical model of the electro-hydraulic loading system of the aircraft steering gear is obtained.Finally,according to the established mathematical model,BP neural network is used as the control algorithm to solve the control problem of the system.In addition,the BP neural network will reduce the control precision of the system,and the PID controller combined with RBF and BP neural network is studied.The correctness and validity of the conclusion are verified by simulation experiments.