Font Size: a A A

Research On Torque Control System Of Induction Motor For Hybrid Electric Vehicles (HEVs)

Posted on:2008-03-05Degree:DoctorType:Dissertation
Country:ChinaCandidate:Y LiuFull Text:PDF
GTID:1102360218453577Subject:Control theory and control engineering
Abstract/Summary:PDF Full Text Request
A Hybrid Electric Vehicle (HEV) is an automobile in which the propulsion comprises bothan Internal Combustion Engine (ICE) and an Electric Motor (EM). HEVs have the characterto pollute less and consume less fuel and will become a new research field for the automobileindustry in future years. As a core part of Hybrid Electric Vehicle (HEV), the motor drivesystem requires fast and accurate control of electromagnetic torque which plays crucial rolesin drive system of a HEV. Therefore, it is especially necessary for basic and applicationresearch to develop a torque control system of induction motor for HEVs.Good performance of the electromagnetic torque is dependent upon reliable operation ofthe feedback sensors such as position sensors, current sensors and voltage sensors. In event ofsensor failures, HEVs may accept permanently reduced drive performance and even bringabout fatal traffic accidents. In practical applications, the most common failure of controlsystems is sensor faults. Therefore, it is also necessary to develop a torque control system ofinduction motor for HEVs with sensor failures. The induction motor (IM) is widely used inHEV applications due to its higher efficiency, better reliability, no maintenance, and next tolower cost. In this thesis, the induction motor drive system in HEV applications is taken as theresearch object. Systemic theory analysis and simulation experiments on its torque control arecarried out with complex and variable operations or sensor-fault operations. The researchwork of this thesis consists of the following major parts:1. The HEVs are operated at various operating conditions and transient, it could then beexpected that the motor parameters vary considerably and frequently, and that the keycharacter of the IM is variable parameters of model. Moreover, large errors can be noticedundertaking the use of the traditional IM model. Considering the above problems, a novelmodeling methodology for induction motor is presented wherein motor parameters of variousoperating conditions are estimated from transient data information using an off-line method,and a correlation analysis is employed to map the parameters to operating conditions. As aresult, it is possible to simultaneously estimate multi-parameters of the IM online. Theeffectiveness of the proposed motor model is verified by the comparison with experimentaldata. The effective motor model is the basis of design of the motor controller applied inHEVs. 2. Considering that the classical Field Oriented Control (FOC) is sensitive to parametervariations and needs tuning many control parameters, a new robust PI controller is proposedbased on the standard Field Oriented Control (FOC). The PI regulator is used to generate thecontrol inputωe (synchronous frequency) instead of iqs (stator current in the dq rotatingreference frame), which not only needs tuning less control parameters, but also shows abetter performance in the rotor flux (λr) magnitude and its orientation tracking, and is robustto model uncertainty. Simulation results show that the proposed torque control system has agood tracking performance of torque. In order to minimize the consumed energy and runlonger distance after a battery charge, it has become a research focus how to optimize theefficiency while higher accurate, higher reliable and higher robust torque control is applied tothe IM in HEV applications. In this dissertation, the optimized rotor flux reference is derivedfrom the expression of the minimum power losses. The motor drive system consumes lessenergy with the new optimal flux reference. The new scheme of the torque control system andthe maximum efficiency control is designed with rotor flux inner loop and electromagnetictorque outer loop. The simulation results, which are obtained from the representativeoperations of a HEV, show the effectiveness of the proposed method.3. Because many speed sensorless control schemes show poor robustness with respect tothe IM parameters uncertainty, an adaptive speed-flux observer based on sliding mode controlis developed, which not only improves the robustness of the observer, but also allows for thesimultaneous observations of rotor fluxes and speed, decreases the quantities of observers andgurantees the accuracy of the observer at medium or high speed. As for serious speedestimation problems in accuracy and convergence speed at low speed or low frequency, theabove speed-flux observer is replaced with a designed stator frequency observer to improvethe performance of motor drive system at low speed. Simulation results show that theobservers are satisfying with good performance of torque control during the motor drivesystem is operated over a wide speed range.4. In order to enhance the torque control performance of the motor drive system in eventof phase current sensor loss, the anaylsis of the relationship between Idc, phase currents andswitching states is carried out, and a new algorithm using a dc link current sensor andknowledge of switching states to reconstruct all three inverter phase currents is proposed Thegenerator of switching state information is designed based on Digital Current Hystersis-BandRegulation scheme. In order to reduce error resulting from measurement noise and switchingcontrol, the IIR digital filter is designed to improve the accuracy of the reconstructed currents.The algorithm proposed is independent of the motor model parameters and the load applied.Therefore, it can be applied to any 3-phase motor drive systems operating under balanced phase current regulation. The effectiveness of the proposed new method is verified bysimulation results.5. Since there are different sensor failures in induction motor drive systems for HybridElectric Vehicles (HEVs), the optimal control strategies to various sensor failures arediscussed. An active fault-tolerant reconstructed controller is integrated into the torque controlsystem. The proposed torque control system sustains the best control performance as possiblegiven the complement of remaining sensors, which enables Field Oriented Control (FOC)carried out in most cases of sensor failures. In order to obtain a smooth transition in terms oftorque responses in the event of sensor failure, a smooth-fuzzy controller is developed tocompensate the existing stator voltage drift between different control schemes. Simulationsshow that the proposed torque control system can ensure the performance of a HEV when ithas troubles with sensors.
Keywords/Search Tags:Hybrid electric vehicle, Induction motor torque control, Fuzzy control, Sliding mode control, Sensor failure, Reconstruction of phase current
PDF Full Text Request
Related items