Due to the absence of vacuum sources for electric vehicle, the vacuum booster is replaced by electronic brake boosting (EBB) system. In order to improve the driver’s brake feeling based on EBB, this work fully exploits the flexible and variable power-assisted characteristics, and designs different transmission ratios according to the driver’s target braking strength. To identify the driver’s braking strength, an improved radial basis function (IRBF) neural network, combining self-organization method and supervised learning method, is proposed to establish the relationship between the driver’s braking strength and the characteristic parameters. Based on this, the variable transmission ratio strategy is designed, and its main optimized parameters are optimized by means of multi-objective optimization algorithm to provide the driver with a satisfactory brake feeling. The strategies under fixed and variable transmission ratios are simulated and analyzed in low-speed with gentle-brake and high-speed with emergency-brake. The simulation results show that, compared with the fixed transmission ratio, the proposed variable transmission ratio shows excellent performances in both brake feeling and brake safety.

中文翻译：

---

基于驾驶员目标制动强度的改善制动感的变传动比策略

由于电动汽车没有真空源，真空助力器被电子制动助力（EBB）系统取代。为了改善驾驶员基于EBB的制动感觉，本工作充分利用灵活可变的助力特性，根据驾驶员的目标制动强度设计不同的传动比。为了识别驾驶员的制动强度，提出了一种改进的径向基函数（IRBF）神经网络，结合自组织方法和监督学习方法，建立驾驶员制动强度与特征参数之间的关系。在此基础上，设计了可变传动比策略，其主要优化参数通过多目标优化算法进行优化，为驾驶员提供满意的制动感觉。分别在低速平缓制动和高速紧急制动情况下对固定和可变传动比下的策略进行了仿真和分析。仿真结果表明，与固定传动比相比，所提出的可变传动比在制动感觉和制动安全性方面均表现出优异的性能。