Investigation On Variable Geometry Optimal Impulse Adaptive Control Of Air Turbine Rocket Engine Based On Genetic Algorithm

Air Turbine Rocket(ATR)is a turbo-based composite cycle engine with broad flight packages,excellent mobility,high thrust and greater impact,which make it a promising development prospect in the application of adjacent space aircraft.However,because the flight package line of the ATR engine is larger and the control variable is more,it is important to make full use of the high mobility in the entire flight line.and the adaptive adjustment of the controller parameters and flight conditions is critical.In this paper,the adaptive control method of ATR engine is studied by genetic algorithm,and the main research is as follows:Firstly,the model of the computational model and combustion calculation model of the ATR engine system is build.Based on the internal flow characteristics,the key component models such as variable geometry inlet,compressor,gas generator,turbine,combustion chamber and variable geometric nozzle are constructed,and the ATR engine is established by the measurement method,and the calculation of steady state and dynamic energy is carried out,and the influence of variable geometry parameters and controller parameters on the dynamic energy of the ATR engine is analyzed.Based on the analysis of the characteristics of ATR flow,combustion gas flow and gas flow,the calculation of thermodynamic parameters of the industrial matter in each flow road is simplified,and the simplified model and the original model calculation result are compared and verified.The simulation results show that the closed-loop control method,which adjusts the flow rate of the gas generator to control the speed and adjusts the nozzle throat area to control the turbine expansion ratio,can obtain the dynamic performance of the ATR engine efficiently and accurately under the condition of multivariable coupling,and avoid the tedious iterative process;The model can reduce the calculation time by 80 percent by simplifying the chemical balance calculation of the industrial and the compressor and the turbine characteristic diagram interpolation.Secondly,based on the ATR engine steady-state model,the physical speed and compressor working point is used as the optimization variable,the calculation is obtained by the thrust and the ratio of the shock,and the two are the target functions of the genetic algorithm,and the ATR engine performance optimization model is established.The optimization results of different flight conditions are analyzed,and the two conflicts are found that after the optimal solution of Pareto,the designer needs to balance the actual application scene and preferences to determine the appropriate design party case.The population size and iteration number of genetic algorithms have a significant effect on the optimization results,and the appropriate increase of the population size helps to broaden the scope of the optimal solution of the Pareto.Finally,according to the controller performance design objective,a multiobjective optimization method for ATR engine controller parameters is established by using fast non-dominated sorting genetic algorithm(NSGA-II),taking the weighted formation of several performance evaluation indexes as the objective function.This method can obtain a set of optimal controller parameters—Pareto frontier,and use TOPSIS based on entropy weight method to process it.Finally,the optimal controller parameters of the current transition process are obtained.After the acceleration process is completed,the controller parameters are automatically replaced with the optimized parameters.Compared with continuing to use the controller parameters in the acceleration process,a more rapid,stable and accurate control process is achieved in the take-off section of the ATR engine.