Multi-parameter and multi-objective collaborative optimization of a suspended monorail vehicle addressing its...

Area of Science:

• Vehicle Dynamics and Control
• Computational Engineering
• Optimization Algorithms

Background:

• Traditional vehicle dynamic analysis often relies on linearization, which can oversimplify complex, nonlinear systems.
• Accurate modeling of physical vehicle characteristics is crucial for predicting performance improvements.
• Efficient and reliable optimization methods are needed for complex coupled systems.

Purpose of the Study:

• To develop a high-precision, rapid computation method for vehicle dynamic performance optimization.
• To preserve the strongly coupled nonlinear dynamic properties of vehicle systems.
• To validate the effectiveness of the proposed optimization approach on an experimental vehicle.

Main Methods:

• Twin modeling technology to accurately reflect physical vehicle dynamics.
• Big data search for multi-parameter, multi-objective collaborative optimization.
• Sobol sensitivity analysis to identify key dynamic parameters.
• An improved multi-start parallel simulated annealing algorithm for enhanced efficiency and reliability.

Main Results:

• Accurate reflection of actual vehicle dynamic characteristics using twin modeling.
• Identification of key parameters influencing vehicle dynamics via sensitivity analysis.
• Significant improvements in the dynamic performance of the experimental vehicle.
• Enhanced computational efficiency and reliability of optimization results.

Conclusions:

• The proposed method achieves high-precision, rapid computation of vehicle dynamic performance.
• It effectively preserves nonlinear dynamic properties, overcoming limitations of traditional linearization.
• Demonstrates significant engineering application value in rail transportation dynamics optimization.