The article titled “Estimation of Acceleration Amplitude of Vehicle by Back Propagation Neural Networks” [1] was found to contain material from Yildirim and Uzmay [2], which was cited. Details of the similarity are as follows:(i)In the Abstract, the text “This paper investigates the variation … waved stone block paved, and country roads” and the sentence “This method is conceptually straightforward, and it is also applicable to other type vehicles for practical purposes.”(ii)In the Introduction, the text “Recently, improving comfort and safety conditions … by random theory based on statistics” and most of the text in “A solving method of low-frequency vehicle… paper is concluded with Section 6.”(iii)In “Random Vibration Theory,” most of the text.
The authors clarified that their article differs from Yildirim and Uzmay. In Table 2, the authors used nine methods of training for the BP neural network, while Yildirim and Uzmay used only a radial basis function in this area. The input of Figure 3 includes four parameters (velocity, damping ratio, the natural frequency of vehicle shock absorber, and road condition) but in Yildirim and Uzmay there is only one parameter (velocity).
In the BP model, the authors used three functions (newff, newcf, and newelm) for the first time in this area. Additionally, Yildirim and Uzmay only used a Gaussian function in the hidden layer of RBFNN, while the authors used three activation functions in Table 3.