Sizing And Power Split Strategy Of A Modular Connected Machines System For Electrified Vehicles

Modular Connected Machines(MCM)system is one multi-machine system where several Electrical Machines(EM)are integrated into a module.It has wider high-efficiency area,better fault tolerance ability and abundant operation modes,and is promising in the field of electrified vehicles.The combination of multiple EMs can increase the degrees of freedom for control and optimization.However,it also leads to many complex problems,such as sizing of the system,coordinated control and energy management of multiple EMs.The objective of this Ph D subject is to develop different MCMs for electrified vehicles.Some works will be done in this thesis around the two basic key technologies,i.e.,MCM sizing and power split strategies.To apply a MCM system in an electrified vehicle,analyze the interactions between each EM,and develop the control of the system,Energetic Macroscopic Representation(EMR)is used organize the model of the MCM-based vehicle in this thesis.EMR can easily obtain the tuning and control path of the system,which more clearly reflects the energy flow inside the system.The use of EMR emphasizes the important couplings which exist between the different parts of the system.It focuses the modelling on the coupling devices,which distribute energy.The inversion of the model leads to a general architecture of the control scheme.It defines the closed-loop control of every state variables.In order to size a MCM,and determine the design parameters of the EMs,it is necessary to acquire their efficiency maps.If each EM is designed by FEM,it would spend a lot of time.This thesis studies the efficiency map evaluation method for Induction Machine and Permanent Magnet Synchronous Machine(PMSM),which is essential to form the objective function for MCM sizing.The method calculates the EM loss based on rated loss.Only 4 basic parameters are needed and no EM design is required.It can complete the map calculation quickly,which satisfies the requirement of MCM sizing.Moreover,in comparison with the experimental efficiency maps,the errors of energy consumptions of the vehicle with the evaluated efficiency maps are less than 3.2%,which satisfies the requirements of MCM sizing for calculation accuracy.To optimize the MCM and determine the design specifications of each EM,this thesis proposed a MCM sizing method,aiming at expanding MCM high-efficiency area,increasing system torque density and reducing the usage amount of Permanent Magnets(PM).The sizing of series and parallel MCM is completed,and 2 MCM topologies are proposed based on planetary gear-train.The MCM with one IM and one PMSM reduces the PM volume by 44% then a single PMSM.Meanwhile,the system has a larger highefficiency area.The MCM based on planetary gear-train changes the coupling mode of the EMs,and improves the energy saving potential.It has been proved that MCM system has considerable energy-reduction ability for pure EVs and series HEVs.Series MCM is more efficient and compact than the parallel MCM.Generally,a 2-EM MCM has a higher average energy redaction rate.In order to improve the overall system efficiency,two MCM power split strategies based on ANFIS(Adaptive Network based Fuzzy Interface System)and model predictive control are proposed.The velocity prediction model based on wavelet transform and neural network improves the prediction accuracy.The operation characters of the EMs are improved by nonlinear regression and interpolation,using the historical and predicted torques.The experiment proves that in comparisons with the optimal efficiency strategy,the ANFIS-based strategy improves the EM working characters to some extent,with energy consumption increased by 2.45%.The predictive-based strategy improves the EMs working condition,the noise and vibrations are reduced significantly,and the energy consumption is increased only by 1.4%.Hence,the proposed predictive-based strategy is more efficient and reliable for MCM power split.