Research On Object-Oriented Modeling And Simulation For Aeroengine And Control System

Modeling and simulation of aeroengine and control system is one of the essentials to develop advanced aerospace propulsion system with lower cost, less risk and higher efficiency. This dissertation studied the simulation models of aeroengines and control systems that are widely used in the semi-physical simulation for the design of aeroengine digital controllers. Research work includes aeroengine simplified mathematical model, object-oriented engine component-level model, engine control system model, startup model on ground and at high altitude, and control algorithm. Firstly, based on the engine ground test data, a simplified engine mathematical model was developed without detailed component characteristics. Interpolated algorithm and transient coefficient method were used in order to predict the engine steady-state and transient performance. The simplified model has a flexible framework and could simulate the whole process of engine from near zero state to maximum state on ground. Simulation results show that the model has a satisfied precision, and the calculated steady-state error is less than 2% while the transient error is less than 5%.A general simulation platform for aeroengine and control system performance prediction was built up with object-oriented analysis and design (OOAD) method. Through deeply investigation into the realization mechanism of OOAD in engine performance prediction, a total framework was carried out for building a general engine performance simulation platform. The framework includes a reasonable base classes hierarchy, some important assistant tool classes, an interactive algorithm and software interface, which greatly increases the program code-reused rate and reduces the utilization of global variables. Owing to the OOAD technology, the designed engine simulation platform has a user-friendly interface that supports drag-and-drop operation to build engine model. So it is very easy to use and capable of simulating engine under different working conditions.Based on the total simulation framework, codes of all simulation classes, interactive algorithms and software interfaces were programmed using Visual C++ language. A full-state full-envelop real-time component-level model (FFRCM) for a twin-spool, mixing flow, non-afterburner, turbofan engine was developed, which shows the validity of the designed simulation platform. Besides, a number of control system classes were coded through enlarging and inheriting the engine simulation classes, based on which a general engine control system simulation model was developed Simulation Results show the validity of the control system simulation model.The development of FFRCM, especially the startup model is another emphasis of this thesis. Due to the unavailable engine component characteristics at low rotor speed, The absented one third above-idle component characteristics and the whole startup characteristics were obtained through extrapolated method. Then a series of research work were done in order to improve precision, reduce calculation time and ensure model convergence in the full envelope, which include corrections of component characteristics, air bleed modeling, steady-state guess vector calculation, dynamic volume approach, etc. Specially, in order to develop a component-level startup model, corrections of combustor efficiency and component total pressure recovery coefficients, heat transfer modeling of high temperature components, calculation of startup model guess vector were all studied. A component-level startup model for a twin-spool turbofan was then developed, which can calculate all of the engine state parameters during startup process both on ground and at high altitude. Simulation results of the FFRCM show that the steady-state, above-idle transient and startup transient calculated errors are less than 3%, 5%, 10% separately.PID control algorithm is still widely used in present aeroengine control system for its simple structure and high reliability. As a new intelligent system, artificial immune system provides a new way to solve complex control problems. In the end, a novel immune feedback algorithm was proposed by introducing immune feedback mechanism to PID control system. Simulation results of aeroengine rotor speed show that the proposed immune feedback algorithm has less overshoot, faster response, stronger ability of anti-jamming and adaptiveness.The simplified model and the FFRCM for a twin-spool turbofan engine have been used in the semi-physical tests of a certain engine digital controller, and achieve good appraisement.