Serial combinational optimization method for double wishbone suspension’s pseudo damage improvement
Suo Zhang,Yunkai Gao,De Gao,Ting Pan,James Yang
DOI: https://doi.org/10.1007/s00158-023-03579-9
IF: 4.279
2023-01-01
Structural and Multidisciplinary Optimization
Abstract:The durability of commercial vehicles is heavily dependent on the design of their suspension systems’ hard points. The design optimization of hard points involves a large number of design parameters. However, traditional surrogate model-assisted multi-objective optimization (MOO) methods face challenges for high-dimensional problems due to the expensive computational cost required to generate accurate metamodels. To mitigate the aforementioned limitations, this study proposes a serial combinational MOO design framework based on a validated multi-body dynamic model, which focuses on the hard point optimization of a double wishbone independent suspension. The proposed method involves two optimization stages. At Optimization Stage I (i.e., warm-up stage), Taguchi-based grey relational analysis is performed to determine a preliminary solution based on the data from the design of experiment in a large design space. At Optimization Stage II, a hybrid surrogate model with the highest average R -squared value is chosen based on the reduced design space obtained from Optimization Stage I. Furthermore, the non-dominated sorting genetic algorithm II is applied to search for optimal points in a relatively small design space. Subsequently, a unique optimal solution is obtained by employing the improved multi-objective decision-making model based on entropy weight and Technique for Order of Preference by Similarity to Ideal Solution methods. When the improvement of the optimal solution fails to meet the desired objective, Stage I and iteration follow. Several representative functions are utilized as demonstration examples to verify the developed optimization method. Then, the proposed method is applied to solve a typical engineering optimization problem, namely, suspension optimization with a large design space. Results show that the two-stage method considerably reduces the pseudo damage of the suspension when compared to the conventional one-stage optimization method.