Predictive model of tractor driving posture considering front and rear view. (September 2019)

• Record Type: Journal Article
• Authors: Li, Haoyang; Zhao, Dongwei; Ma, Xuechao; Jin, Xiaoping
• Is Part Of: Biosystems engineering. Volume 185(2019)
• Page Start: 64
• Page End: 75
• Publication Date: 2019-09
• Subjects: Tractor driving posture prediction -- Forward driving posture -- Backward working posture -- Multiple stepwise regression; Bioengineering -- Periodicals; Agricultural engineering -- Periodicals; Biological systems -- Periodicals; Génie rural -- Périodiques; Systèmes biologiques -- Périodiques; 631
• DOI: 10.1016/j.biosystemseng.2019.03.010
• Languages: English
• ISSNs: 1537-5110

Abstract

Abstract : Recently, tractor research and development has focused on ergonomic performance design and proposed improved requirements for driving comfort and safety. The digital human model is important for virtual ergonomics design because whether or not the driving posture of the model matches reality significantly influences the design results. Because of the poor accuracy and integrity of predictions for tractor driving posture, simulations for the front and rear driving postures of high-power tractors were analysed. Fifteen dependent variables were determined by simplifying driving posture according to the representative and measurability of the joint angle of the human body and the effective segment size. The data from 54 subjects in the forward driving and backward working postures were measured using a portable three-coordinate measuring machine. The key factors influencing each dependent variable were determined through comprehensive correlation, scatter plot, and linear regression analyses. Correlation analysis of the dependent variables revealed that regression of the dependent variable was arranged in descending order. The regression prediction model of each dependent variable was established by analysing the lower-level regression variables while introducing strong correlation and high-level variables into the model. To verify the accuracy and rationality of the model, statistical tests and experimental verification were carried out, respectively. The model was accurate (p < 0.001). To further verify the fitting effect of the model, data of five other subjects were compared. The relative error of the variables was less than 5%, which proved that the prediction model has high accuracy and was aligned as expected. Highlights: Predictive model established for tractor driving posture. Model considers both front and rear view of tractor driving posture. Model can increase posture accuracy of digital human model in virtual design. Validation implies that the average error of the model is small (<5%).