Modular Cascade Machines (MCM) system is a multi-machine system for the traction systems of electrified vehicles. Currently, MCM is composed of several machines of the same type, generally, several Permanent Magnet Synchronous Machines (PMSM). This paper proposes a hybrid MCM system, which is composed of a PMSM and an Induction Machine (IM). This combination aims to fully benefit the properties of different machines. This hybrid MCM is compared with other traction systems, namely, PMSM-based MCM and single PMSM drive. For a fair comparison, the design specifications of all the traction systems are optimally sized by a specific sizing method, which can maximize the system high-efficiency area. Then, different traction systems are designed by Finite Element Method (FEM), in order to have accurate material consumptions and efficiency maps. The comparison study shows that the proposed hybrid MCM is competitive for energy saving. Therefore this new drive saves about 5% of energy in comparison with a single PMSM derive. The hybrid MCM energy saving is very close to the one of a PMSM-based MCM. Moreover, the hybrid MCM system reduces the PM volume (reduced by 53%) while increases the usage amount of winding copper, steel and aluminum, due to the existence of IM. Fortunately, the cost estimation proves that the hybrid MCM (not include inverters) has almost the same cost as the single PMSM. Hence the proposed hybrid MCM is a competitive alternative of the single PMSM than the PMSM-based MCM.

A Modular Cascade Machine (MCM) system is composed of several electric machines instead of a unique one. The MCM system is expected for application to electric vehicles (EVs). This paper aims to compare a MCM-based EV with a classical single drive EV in normal operations. The Energetic Macroscopic Representation formalism is used to organize the system model and control of both studied EVs. An optimal energy management strategy of the MCM-based EV is developed for minimizing its energy consumption. Simulation results demonstrate the interest of the MCM-based EV in terms of energy saving. Different driving cycles are analyzed and the comparison of both vehicles highlights the interest of this new multi-machine system despite the increase of its mass and cost.

Modular cascade machines (MCM), as a new multi-machine system, are expected to be applied for electric vehicles (EVs). This paper focuses on the sizing of MCM systems. Expanding MCM high-efficiency area and reducing MCM mass are selected as optimization targets for sizing. The proposed machine efficiency map and MCM mass evaluation methods are the means to achieve this goal without machine pre-design. The efficiency evaluation method enables to evaluate machines efficiency maps only with four basic design specifications (rated speed, rated torque, overload factor, and speed expansion ratio). Neither finite element (FE) nor experimental data (inductance and resistance) are used. Its calculation is fast and can be finished within several milliseconds. The method is validated by comparisons with the experimental efficiency maps of two prototypes. The proposed methods are proven having an accuracy that is acceptable for the MCM definition and pre-design. The exemption of machine pre-design saves a lot of time during sizing stage. The sizing optimization is realized by genetic algorithm. And the comparisons of different sizing are conducted with the same vehicle model. Moreover, general sizing rules for the similar EVs are proposed. For the studied EV, the number of machines is proposed as a function of traction power. The sized specifications could be used for more accurate machine design by the FE method in future. The sizing method could be extended for the other kind of multi-machine systems.

Battery thermal management (BTM) is crucial for maximizing the battery performance and maintaining the battery health. Model Predictive Control (MPC) stands as a promising technique for achieving this objective. However, conventional MPC approaches employing fixed parameters often restrict the control performance. This paper aims to adaptively tune the MPC parameters for an optimal BTM system using reinforcement learning. To achieve this, the nonlinear MPC controller is linearized at nominal operating points and transformed into an adaptive MPC controller. Leveraging a soft actor-critic reinforcement learning agent, the MPC decisive parameters, including weights of the cost function, lengths of the prediction horizon and control horizon, are regulated online. Simulation results demonstrate that the proposed scheme achieves a 5.8% reduction in energy consumption for BTM, an 18.8% decrease in constraints violation, and a substantial 59% reduction in MPC execution time compared to an MPC controller with fixed parameters. Additionally, the feasibility of the proposed reinforcement learning-driven adaptive MPC has been verified on a dSPACE machine. These outcomes underscore the significant potential of MPC optimization for enhancing battery management systems.

Optimal energy management for electrified vehicles can be achieved with a prior understanding of the velocity profile, whose prediction accuracy is influenced by the stochastic uncertainty of real driving cycles. This paper proposes a new state of charge planning method for plug-in hybrid electric trucks. The strategy eliminates the need for velocity prediction and relies solely on some readily available signals, such as estimated remaining distance and travel time. This exemption from velocity prediction avoids expensive computational costs, making it possible to use a more affordable processor. To test the strategy, four random driving cycles representing two different vocational uses are selected. The results show that the proposed strategy only increases fuel consumption by 1.3% for unknown urban and long-haul delivery cycles compared to the optimal strategy. Additionally, a sensitivity study reveals the robustness of the method on inaccurate navigation signals. These findings demonstrate that the proposed method is efficient and adaptive, making it suitable for existing truck applications without the need for additional hardware.

Multi-machine systems are increasingly utilized in electrified vehicles, offering enhanced opportunities for system optimization. This paper investigates the energy-saving potentials of two dual-machine systems designed for electric vehicles: one employing speed-addition and the other torque-addition. In the speed-addition system, two electrical machines are linked to the sun and the ring of a planetary geartrain. In the torque-addition system, both machines share the same shaft. To ensure a fair assessment, a unified co-design that involves sizing, design, modelling, control, and energy management is adopted. Each machine is designed using the finite element method, following optimal sizing specifications with uniform materials, topologies, and adherence to identical physical constraints. Comparative trip analyses employing optimal energy management strategies reveal distinct outcomes. The speed-addition system achieves an energy saving up to 3% compared to a conventional single-drive vehicle. In contrast, the torque-addition system demonstrates a more substantial energy saving up to 12%. The findings confirm that the dual-machine system using torque addition is notably more efficient and advantageous for automotive applications under the same constraints, outperforming the speed addition configuration in terms of efficiency, cost, volume, mass, and complexity.