Multi-objective Genetic Algorithm-based Axial Flow Compressor Aerodynamic Optimization Design Study

Aerodynamic optimization design technology is an important research direction in turbomachinery design. Aerodynamic optimization technology for compressor design plays an important role in the development of high performance axial flow compressor, multiobjective genetic algorithm are currently being widely used in turbomachine optimization design.It study one dimension and quasi-three dimensional design method of multistage axial flow compressor, build loss model for design and loss and deviation model for off-design respectively. The result shows that these models relative reflect flow phenomena on axial flow compressor. Point to the character of compressor optimization design, multistage axial flow compressor optimization method adopt as a fast elitist non-dominated sorting genetic algorithm (NSGA-II).For one-dimension optimization problem, stage pressure ratio and flow angles at the stator trailing edges, solidities at the stator and rotor trailing edges are considered as design parameters, To demonstrate feasibility of the present method, a multiobjective optimization of a six-stage compressor design was performed for maximization of the overall isentropic efficiency and the total pressure ratio, It obtained dozens of reasonable and uniformly distributed Pareto-optimal solutions that outperformed the baseline design in both objectives. Detailed observation of the Pareto-optimal designs revealed that aerodynamic parameter follow axial distribution of optimization scheme is reasonable. For quasi-three-dimensional optimization design problem, aiming at a two-stage axial fan and a three-stage axial flow compressor with IGV, maximization of the overall isentropic efficiency and the total pressure ratio are objectives of optimization, It obtained hundreds of reasonable and uniformly distributed Pareto-optimal solutions that outperformed the baseline design in both objectives. Detailed observation of the Pareto-optimal designs revealed that aerodynamic parameter follow radial distribution of optimization scheme is reasonable.The results confirm that the optimization design method of axial flow compressor based on multiobjective genetic algorithm was practical.