Optimization of system parameters for gas-generator engines

For this paper, software to optimize the system parameters for gas-generator cycle engines was developed. The gas-generator cycle engine was decomposed into several principal components, such as the thruster chamber, turbopumps and gas generators. Numerical simulation codes were developed for each component and the system. According to the different relationships within the components, these modularized codes could be integrated arbitrarily. A generic optimization procedure of system parameters for the gas-generator cycle engine was developed. According to the design mission, optimal engine system parameters could be decided for different components assembling forms by using this procedure. One hybrid algorithm, which combined the genetic algorithm with gradient-based algorithm, was utilized during optimization procedure to increase the globality of the optimization exploration. Several traditional multi-objective methods, such as changing weight method, e-constraint method, and neighborhood cultivation genetic algorithm were used to obtain the Pareto frontier solutions. Physical programming method and fuzzy related degree method were used to obtain optimal design based on preference of the designer. Comparative results presented show the impact of different methods on the design variables and objectives.