摘要：【中文】一种电气系统包括可再充电能量存储系统(RESS)和控制器。RESS包括连接到电压总线的第一和第二电池组，每个电池组具有相应的多个电池单元和相应的电池单元平衡电路。所述RESS还包括开关，所述开关选择性地将所述电池组彼此连接或断开以实现串联和并联模式。所述控制器通过检测所请求的串并联模式转变来执行一种方法。响应于所述电池组的充电状态或电池组电压相对于彼此存在阈值不平衡，所述控制器使用所述电池单元平衡电路的断开/闭合状态控制并且可能使用具有PWM受控开关和电路元件的开关块来平衡所述充电状态/电压。所述控制器可以在平衡时执行所述请求的模式转变。 【EN】An electrical system includes a Rechargeable Energy Storage System (RESS) and a controller. The RESS includes first and second battery packs connected to the voltage bus, each battery pack having a respective plurality of cells and a respective cell balancing circuit. The RESS further includes a switch that selectively connects or disconnects the battery packs from each other to achieve series and parallel modes. The controller performs a method by detecting a requested series-parallel mode transition. In response to a threshold imbalance of the state of charge or battery pack voltages with respect to each other, the controller uses open/closed state control of the cell balancing circuit and possibly a switch block with PWM controlled switches and circuit elements to balance the state of charge/voltages. The controller may perform the requested mode transition at equilibrium.

摘要：【中文】用于机动车辆预测性电荷规划和动力系控制的智能车辆系统和逻辑、制造/操作此系统的方法及具有智能电荷规划和动力系控制能力的电驱动车辆。智能电动车辆的基于AI的预测性电荷规划的系统和方法使用机器学习(ML)驾驶员模型，其利用可用交通、位置和道路地图信息估计车辆速度和推进扭矩需求以导出给定行程的总能耗。智能混合动力车辆的基于AI的预测性动力系控制的系统和方法将ML驾驶员模型与深度学习技术一起使用导出具有可用交通、地理位置、地理空间和地图数据的由预览路线限定的驾驶循环简档。利用收集的数据发展ML生成的驾驶员模型来复制驾驶员行为并预测预览路线的驾驶循环简档，包括预测的车辆速度、推进扭矩和加速器/制动踏板位置。 【EN】An intelligent vehicle system and logic for predictive charge planning and powertrain control of a motor vehicle, a method of manufacturing/operating such a system, and an electrically driven vehicle having intelligent charge planning and powertrain control capabilities. Systems and methods for AI-based predictive charge planning for intelligent electric vehicles use a Machine Learning (ML) driver model that estimates vehicle speed and propulsion torque requirements using available traffic, location, and road map information to derive total energy consumption for a given trip. Systems and methods for AI-based predictive powertrain control of intelligent hybrid vehicles use an ML driver model with deep learning techniques to derive a driving cycle profile defined by a preview route with available traffic, geographic location, geospatial, and map data. The ML generated driver model is developed using the collected data to replicate the driver behavior and predict a driving cycle profile for the preview route, including predicted vehicle speed, propulsion torque, and accelerator/brake pedal position.