Content area
Full text
1. Introduction
In this study, the four-wheel steering of a car is taken as an example. When the current wheel is turned, the intersection of the left and right wheels of the front axle falls on the extension line of the rear wheel. In this case, the steering will be the smoothest. In this example, we selected angle (F1), arm length (F2), toe in and out (F3), width (F4) and length (F5) as experimental factors, hoping to satisfy the Ackerman principle.
In research methods, we can generally use the design of experiment method or mathematical programming to solve such problems (Babu et al., 2014; Koladia, 2014; Jeang, 2015). However, the disadvantage of mathematical programming is that processing cannot take place when the reaction equation is unknown. Moreover, it is not possible to satisfy the product design with the concept of probability and it is impossible to reflect the order of each experimental factor in terms of its relevance. This study applied the response surface methodology (RSM) assigned to the experimental design. Its advantage is that combined with statistical regression and mathematical techniques, the design can not only be completed with a small number of experiments, it is also suitable for the problem of multi-factor experiments. More importantly, an approximate reaction equation can be provided which reflects the effect of each factor on the response (Kleijnen, 2008). At the same time, it is possible to draw both a contour plot and response surface plot and to arrange the importance and order of each experimental factor. Finally, optimization methods can be used to predict reaction values.
This study is comprised of five sections. Section 1 offers the introduction. Section 2 describes the background information needed for this research and presents the proposed approach and application for illustrations. Section 3 presents the results of the proposed method with illustrations. Section 4 provides a discussion. Finally, a conclusion is given in Section 5.
2. Background
The proposed approach starts with the formulation for the example discussion. In addition to the formulation, RSM and four-wheel steering structure are also introduced in the following subsections.
2.1 Response surface methodology
RSM combines mathematical applications, statistical analysis and experimental design (Box and Wilson, 1951; Box and Behnken, 1960; Montgomery, 2017). It is...