In-vehicle human–machine interface (HMI) mainly refers to the T-shaped panel system with instruments, centre console, gear lever and other components installed. For intelligent vehicles, the high level of intelligent interconnection may to some extent make drivers lack situational safety awareness and reduce the usability of the system. Thus, this study attempted to establish a relationship between design features and system usability of the in-vehicle panels. From the perspective of visual ergonomics, the panels were deconstructed into design features to determine 36 samples to be studied. After dividing each sample into four areas of interest (AOI), eye movement and subjective preference data were collected to quantify the user experience. Artificial neural network (ANN) and support vector machine (SVM) were used in the study. Nevertheless, conventional learning algorithms often underwent deficiencies in accuracy and robustness in the detection of multifarious kinds of panels. Therefore, the parameters of the two models were tuned to deal with the noise common in user experience data. The determinant coefficients, mean-square errors and mean relative errors of the two models showed that the SVM model had a higher accuracy, smaller error and was more stable in the learning of user experience of HMI design features, which could provide a method for the layout design and evaluation of T-shaped instrument panel.

车载人机界面（HMI）主要是指安装有仪表、中控台、变速杆等部件的T形面板系统。对于智能汽车而言，高水平的智能互联可能会在一定程度上使驾驶员缺乏情境安全意识，降低系统的可用性。因此，本研究试图建立车载面板的设计特征和系统可用性之间的关系。从视觉人体工程学的角度，将面板解构为设计特征，以确定36个待研究的样品。在将每个样本分为四个感兴趣区域 (AOI) 后，收集眼球运动和主观偏好数据以量化用户体验。研究中使用了人工神经网络（ANN）和支持向量机（SVM）。尽管如此，传统的学习算法在检测各种面板时往往存在准确性和鲁棒性方面的不足。因此，对两个模型的参数进行了调整，以处理用户体验数据中常见的噪声。两种模型的行列式系数、均方误差和平均相对误差表明，SVM模型在学习HMI设计特征的用户体验方面具有更高的准确度、更小的误差和更稳定的特性，为人机界面设计特征的学习提供了一种方法。 T形仪表盘的布局设计与评价.