COMPARATIVE STUDY BETWEEN PROPULSION CONTROL SYSTEM FAILURES OF AN ELECTRICAL VEHICLE PILOTED BY FOC AND BY DTC USING DUAL-INDUCTION-MOTORS STRUCTURE

This paper deals with a comparative study using numerical simulations between the failures effect caused by the speed sensor faults for a propulsion control system (PCS) of an electrical vehicle (EV) using dualinduction motors structure. The PCS strategies are achieved on two types of controls where the first one is done from a flux-oriented control (FOC) and the second one is conducted from a direct torque control (DTC). For an electric vehicle, we will often guarantee service continuity, in spite of the occurred faults such as an offset fault in speed sensor and a zero-feedback sensor speed fault which both are essentially needed in the structure of the PCS-EV. The occurred fault cited above might influence one of the dual induction motors which could be conducted an unbalance in the dual used motors and from which the control of the vehicle might be also lost. The results of the realized numerical simulations on the EV conducted by the PCS demonstrate clearly the impact of the so-called-fault. Thereafter, we can also appreciate the robustness using each used control propulsion system in despite of the occurred speed sensor fault.


INTRODUCTION
The demand for alternative energy sources to reduce CO2 emission is one of the greatest challenges nowadays [1]. Electric vehicle offers the most promising solution to reduce vehicular emissions [2]. The principle of an electric vehicle (EV) operation is converting electricity to mechanical energy by the electrical motors, such induction motors. The conventional mechanical structures using reduction gears and mechanical differential can raise the weight of the EV and could cause in greater energy consumption. The alternative solution was to replace it with a propulsion control system (PCS) using an electric differential system assured by dual motors operating at different speeds. For control of the propulsion system we use two types of robust control, flux-oriented control (FOC) and direct torque control (DTC). Among these two used controls, each of them presents its advantages and its drawbacks from the point of view of implementation and performance. In this case, by a comparative study when the speed sensor faults are occurred, we would be able to derive the advantage of one control over the other in terms of performance and robustness of which this electrical differential ultimately depends on it.
Since the faults will be occurred, the PCS failure, so the performances of EV, decrease and it can reach a loss of the control. The control of propulsion system is predisposed to many faults, which may affect the system performance. Concerning speed sensor faults, we can consider both faults as follows:  offset fault (measurement offset).  zero feedback fault (no feedback information).
The paper is organized as follows. Sections 2 and 3 presents both controls of the induction motor used for the PCS whereas concern successively the field-oriented control and the direct torque control. Dual-motor structure of the proposed PCS is shown in section 4. Speed sensor faults are presented in section 5. Section 6 exposes the effected simulations results. The conclusion is done in section 7.

FILED-ORIENTED CONTROL (FOC)
The well-known field orientation control (FOC) strategy provides a linear and decoupled control between the flux and torque of an induction machines.
Then the rotor flux orientation process is given by the imposed zero constraints of quadrate rotor flux component [3]. The induction motor model after flux orientation is given as follows in its conventional nomenclature indicated above.
= ∫( + Ω) (7) = (8) Figure 1 shows a general block diagram implementation of FOC on an induction motor. Torque and rotor flux are controlled indirectly by stator current components , (direct and quadrate) respectively.

DIRECT TORQUE CONTROL (DTC)
The direct torque control of induction motor, assumed as one of the high-performance control systems not needed a transformation rotation, consists of estimators, hysteresis comparators for electromagnetic torque, stator flux and an optimal switching table [4] [5].
The stator flux components equations in stationary reference frame are given by The stator current is described by the complex form: = + (10) The location of the estimated flux is defined as: The torque of induction motor is given by the following relation: = ( − ) (12) Figure 2 shows a general block diagram implementation of DTC-IM intended to EV application.

PROPULSION CONTROL SYSTEM STRUCTURE
For conventional vehicle, when a turn is reached, the speed difference of the rear wheels is regulated by a mechanical differential in order to avoid vehicle slipping. This differential allows one wheel to rotate more quickly than the other. Besides the mechanical means, the differential action of a PCS when cornering can be electrically provided by electric motors operating at different speeds [6] [7]. Figure 3 presents the layout of the proposed propulsion control structure in which a dual induction motors are used for driving separately the rear wheels of the vehicle via fixed gearing.
For a direction order of the steering wheels complementary signals are transmitted to the motors through power converters verifying the following equations [6]: Therefore, from the system of equations (13) above, the speed difference Ω and the vehicle speed Ω 0 may be done simply by: (Ω 1 + Ω 2 ) (14) Figure 4 illustrates the functional block-diagram of the proposed electric differential. During a vehicle turn, the inner wheel will rotate slower than the outer one. Then the PCS, placed at the rear wheels of the vehicle, propels electrically the vehicle and assures the required differential speed between the wheels according to Eq. (13). The control speed is separately realized by the PIcontrollers for each induction motor and without a speed differential control loop. The vehicle speed Ω o is not controlled but deduced from the motor'scontrolled speed Ω 1 and Ω 2 . Also, as a result, using s-Laplace operator, we can write the following matrix relation: Evidently, the decoupled control of differential speed is accomplished if and only if we have 1 (s) = 2 (s). This indicates that the two closed-loop controls of both motors must be identical else we cannot control, separately and in the same time, the differential speed Ω and the vehicle speed Ω 0 . Therefore, the successfully reached control of the vehicle is obtained when the H-matrix is written as: [ [ Ω 0 * Ω * ] (18)

SPEED SENSOR FAULT
There are three main kinds of faults in induction motor drive systems: electrical faults, mechanical faults, and sensor faults [8]. On the controller side, sensor faults are one of the most common problems in industrial applications [9]. In this paper we study two types of speed sensor faults presented below in the next subsections.

Offset Sensor Speed Fault
An offset fault means a perturbation caused by the difference between the real speed and its value measured by the sensor. This fault is simulated as shown in figure 5, by injecting a constant value to the real speed, so we can write that follows [10]:  The obtained EV performances with the occurred sensor speed faults are illustrated in figures 7-8 and where the drawn dashed line showing the step reference speed from which the vehicle starts at t=0 s and continues to work during tests in despite to turning manoeuvres and occurred sensor speed faults. However, at t = 3s, a load torque is applied on each motor. This last case might be occurred for example when EV wheels are stopped by a strong obstacle. After there, at t = 4s the vehicle will turn right which results in the increase of the left wheel speed compared to the right one, i.e Ω left > Ω Right . This result proves the validity of the electric differential principle resulting from the proposed PCS when the vehicle accomplishes the turning maneuvers.
Then, to test severely the PCS-EV control performances, at t =7 s a speed sensor fault has been occurred. As we can see from figures 7a-b and 8a-b, the control system reacts very well to the load perturbation and the reference trajectory is more tracked even in the presence of speed sensor faults. Contrary to this, for the zero-feedback fault (Figures 7c-8c) the vehicle leaves its reference trajectory, so the control of the differential action is gravely lost accompanied by the loss of the aim control vehicle direction. Figures 9 and 10, both present with the same above tests, the variation of electromagnetic torque versus time according to the chosen speed direction of the vehicle. For the positive occurred offset sensor speed fault, we can see a big negative pic which affects the developed torque response on the FOC case, but with small impact on the DTC case. For the negative offset fault, the impact shows a similar precedent case, but with a positive torquepic. Figures 11 and 12 illustrate and compare respectively the stator current responses versus time of one motor of the PCS-EV using FOC and DTC schemes. It can be seen that the current increases considerably for a FOC than in the DTC where it is noted a minimal impact.

CONCLUSION
This paper presents a comparative study between the failures effect caused by the speed sensor faults on the proposed propulsion control system of the electrical vehicle (PCS-EV).
To drive this vehicle, two conventional controls are used. First one is a field-oriented control (FOC) and the second one is a direct torque control (DTC). Two types of speed sensor faults have been occurred on the PCS-EV structure: offset faults according to positive and negative values, and also a zero-feedback fault. The analysis time responses of the simulation tests carried out the PCS-EV show that DTC scheme presents improved control performances than FOC scheme with the presence of the offset faults. In other hand, we can note that the great impact of zero-feedback on the FOC scheme relatively to DTC scheme where the stator current strongly increased in FOC scheme. However, note that in both cases of the control schemes the speed of motor is lost. That's means unfortunately the loss of the systematic control on the electrical vehicle driven by the PCS structure.