| The aircraft needs a fuel control system to provide safe and reliable fuel transportation,and the ground test bench is usually used to simulate the aircraft fuel control system.The core of the test bench is the electro-hydraulic servo system.The motion accuracy of the system determines the accuracy of the test data and the safety of the test bench operators.Therefore,improving the control technology of the gantry servo system is of great significance to the aerospace industry and national defense research.This dissertation takes the two-dimensional gantry manufactured by a research institute as the main research object,analyzes it,and studies the classical predictive control algorithm.However,because the fuel test bench is a strong nonlinear object,there are still some difficulties in establishing an accurate predictive control model,which leads to the complex and large amount of calculation and calculation of the rolling optimization process of predictive control.For these difficulties,this paper uses the adaptive fuzzy neural network as the prediction model of the fuel test bench,and uses the improved LM algorithm as the rolling optimization algorithm of the predictive control.In this paper,BP,RBF and fuzzy neural network are used to construct the multi-step prediction model of fuel test bench respectively.The fuzzy neural network adopts the advanced RAF algorithm,thus realizing the autonomous generation of fuzzy neural network rules.Through simulation experiments,the results show that the adaptive fuzzy neural network has excellent performance.At the same time,the LM algorithm is used as the rolling optimization algorithm,and a fuzzy neural network prediction controller based on the LM algorithm is constructed.The controller is used to simulate the fuel gantry system.The simulation results show that the adaptive fuzzy neural network prediction controller designed in this paper has excellent accuracy,stability and anti-interference ability for the angle adjustment of the gantry system.Finally,this paper also implements the software system of the fuel test bench controller,which includes demand analysis,overall planning,detailed scheme design and the final control system implementation and simulation.This paper mainly takes the methodology of software engineering as the theoretical guidance,introduces the object-oriented thinking,and uses the QT framework to write the control software of the fuel test bench. |