A two-variable control and optimization method for imbalance of high pressure compressor based on improved genetic...

Area of Science:

• Mechanical Engineering
• Computational Engineering
• Aerospace Engineering

Background:

• High-pressure compressors are critical in many industries, but low quality rates in one-time assembly pose significant challenges.
• Rotor imbalance after assembly directly impacts compressor performance and operational lifespan.
• Optimizing assembly parameters is crucial for achieving high-quality, reliable compressor systems.

Purpose of the Study:

• To develop and validate an improved genetic algorithm (GA) for optimizing the post-assembly imbalance of high-pressure compressor rotors.
• To enhance the quality rate of one-time compressor assembly through intelligent optimization techniques.
• To reduce post-assembly imbalance by adjusting rotor blade arrangement and inter-rotor assembly phase.

Main Methods:

• An improved genetic algorithm (GA) was employed, incorporating a four-sector staggered distribution for initial population generation and an elite retention strategy.
• Adaptive adjustment of crossover and mutation probabilities based on fitness function values was implemented.
• A threshold termination condition was added to ensure rapid convergence and efficient search for optimal solutions.

Main Results:

• Simulations demonstrated a 99.46% reduction in imbalance using the improved GA, outperforming traditional GA methods.
• Experimental validation showed significant reductions in imbalance on two correction surfaces, reaching 640 and 760 g·mm.
• The improved GA achieved 86.7% and 87.1% better results compared to zero-degree assembly methods.

Conclusions:

• The improved genetic algorithm effectively minimizes post-assembly imbalance in high-pressure compressors.
• This optimization approach leads to substantial improvements in assembly quality and efficiency.
• The method offers a robust solution for enhancing the reliability and performance of high-pressure compressor systems.