Intelligent Car Cockpit Comfort Evaluation Model Based on SVM

Abstract

With the popularization of intelligent cars, users’ understanding of the value of cars gradually changes from a travel tool to a “third living space”, and cabin comfort is becoming a criterion for evaluating the goodness of cars. In this paper, we start from the physical environment and human-computer interaction environment that affect the comfort of the intelligent cockpit of a car and establish a comprehensive comfort evaluation model of the intelligent cockpit of a car based on the support vector machine (SVM) algorithm in machine learning by conducting experiments on the comfort evaluation of the intelligent cockpit of a car and compare it with several classical machine learning algorithms. The mean square error ( $MSE$ ) of the model based on the SVM algorithm is 0.00096, and the coefficient of determination ( $R^{2}$ ) reaches 0.966, which is better than several other algorithms. The results show that the established evaluation model has good generalization ability and can evaluate the comprehensive comfort of the intelligent cockpit of the car, thus helping the cockpit to make timely and accurate comfort adjustments to ensure the occupant’s riding experience. This project provides a reference direction for the comprehensive evaluation of cockpit comfort, which is of great significance for the future development of intelligent cockpit comfort. In addition, the comfort model can be applied to a variety of comfort evaluation scenarios, which has great practical value.

Keywords

• SVM
• machine learning
• intelligent car cockpit
• comfort
• comprehensive evaluation