JP5862436B2 - Control device for electric vehicle - Google Patents

Description

  The present invention relates to a control device for an electric vehicle.

  Conventionally, a drive torque target value is calculated by performing a filtering process to remove or reduce the natural vibration frequency component of the vehicle torque transmission system on the drive torque request value of the drive motor calculated from the accelerator opening or the vehicle speed. A motor control device for an electric vehicle that controls the current of the drive motor so that the torque of the drive motor matches the drive torque target value is known (see Patent Document 1).

  Further, in the vehicle vibration damping control device described in Patent Document 2, the motor rotation speed is estimated from the drive torque target value in consideration of the motor characteristic model with respect to the drive torque target value determined by the method described in Patent Document 1. Torque command value calculated by calculating the value and passing the deviation from the actual motor rotational speed through a filter composed of a bandpass filter whose center frequency is the natural vibration frequency of the driving force transmission system and an inverse system of the motor characteristic model The final drive torque target value is calculated by adding This eliminates the effects of road gradients, torque transmission system disturbances, motor characteristic model errors, etc., and also eliminates or reduces the natural vibration frequency component of the vehicle torque transmission system, thereby reducing the damping effect and steep torque. Can be achieved at the same time.

JP 2001-45613 A JP 2003-9566 A

  However, in a scene that accelerates from coasting or deceleration, a dead zone where the drive motor torque is not transmitted to the drive shaft torque of the vehicle occurs due to the effect of gear backlash. Even if the control described in Patent Literature 1 or Patent Literature 2 is used, vibration of the output shaft torque is generated.

  An object of the present invention is to suppress the vibration of the drive shaft torque even in the case where a dead zone where the motor torque is not transmitted to the drive shaft torque of the vehicle occurs in the drive force transmission system of the vehicle.

  The control device for an electric vehicle according to the present invention applies a filtering process to the motor torque command value to reduce the natural vibration frequency of the driving force transmission system of the vehicle having a dead zone where the motor torque is not transmitted to the driving shaft torque of the vehicle, and performs the filtering process. The motor torque is controlled in accordance with the final torque command value obtained by the application.

  According to the present invention, the natural vibration frequency of the driving force transmission system of a vehicle having a dead zone can be reduced even during acceleration from coasting or deceleration, and the vibration of the driving shaft torque can be suppressed.

FIG. 1 is a block diagram illustrating a main configuration of an electric vehicle including a control device for an electric vehicle according to the first embodiment. FIG. 2 is a flowchart showing a flow of processing of motor current control performed by the electric motor controller. FIG. 3 is a diagram illustrating an example of an accelerator opening-torque table. FIG. 4 is an example of a control block diagram for performing processing for setting the second torque command value Tm2 ^* . FIG. 5 is a diagram modeling a vehicle driving force transmission system. FIG. 6 is a diagram illustrating the characteristics of the transfer function H (s). FIG. 7 is a block diagram of the vibration suppression control FF calculation unit. FIG. 8 is a diagram illustrating an example of a control result by the control device for the electric vehicle according to the first embodiment. FIG. 9 shows an example of a control result by the control device described in Japanese Patent Laid-Open No. 2003-9566 that does not consider the influence of backlash when performing the process of reducing the natural vibration frequency component of the driving force transmission system of the vehicle. FIG. FIG. 10 is an example of a control block diagram for performing a process of setting a second torque command value Tm2 ^* in the control apparatus for an electric vehicle according to the second embodiment. FIG. 11 is a block diagram of an error compensation calculation unit in the second embodiment. FIG. 12 is an example of a control block diagram for performing a process of setting a second torque command value Tm2 ^* in the control apparatus for an electric vehicle in the third embodiment. FIG. 13 is a diagram illustrating a configuration of a vibration suppression control FF calculation unit according to the third embodiment. FIG. 14 is a diagram illustrating another configuration of the vibration suppression control FF calculation unit according to the third embodiment. FIG. 15 is a detailed block diagram of the vibration suppression control FB calculation unit in the third embodiment. FIG. 16 is a block diagram showing a configuration in which the limit value St (θ) of the drive shaft twist angle calculated by the vibration suppression control FF calculation unit is used as an input of the transfer function Gps (s). FIG. 17 is a diagram illustrating an example of a control result by the control device for the electric vehicle according to the third embodiment.

-First embodiment-
FIG. 1 is a block diagram illustrating a main configuration of an electric vehicle including a control device for an electric vehicle according to the first embodiment. The control apparatus for an electric vehicle according to the present invention includes an electric motor as a part or all of the drive source of the vehicle, and can be applied to an electric vehicle that can be driven by the driving force of the electric motor. And can be applied to fuel cell vehicles.

  The electric motor controller 2 uses, as digital signals, signals indicating the vehicle state such as the vehicle speed V, the accelerator opening θ, the rotor phase α of the electric motor (three-phase AC motor) 4, and the currents iu, iv, iw of the electric motor 4. Based on the input signal, a PWM signal for controlling the electric motor 4 is generated. Further, a drive signal for the inverter 3 is generated according to the generated PWM signal.

  The inverter 3 includes, for example, two switching elements (for example, power semiconductor elements such as IGBTs and MOS-FETs) for each phase, and turns on / off the switching elements in accordance with a drive signal, thereby removing from the battery 1. The supplied direct current is converted into alternating current, and a desired current is passed through the electric motor 4.

  The electric motor 4 generates a driving force by the alternating current supplied from the inverter 3, and transmits the driving force to the left and right driving wheels 9 a and 9 b via the speed reducer 5 and the driving shaft 8. Further, when the vehicle is driven and rotated by the drive wheels 9a and 9b, the kinetic energy of the vehicle is recovered as electric energy by generating a regenerative driving force. In this case, the inverter 3 converts an alternating current generated during the regenerative operation of the electric motor 4 into a direct current and supplies the direct current to the battery 1.

  The current sensor 7 detects three-phase alternating currents iu, iv, iw flowing through the electric motor 4. However, since the sum of the three-phase alternating currents iu, iv, and iw is 0, any two-phase current may be detected, and the remaining one-phase current may be obtained by calculation.

  The rotation sensor 6 is, for example, a resolver or an encoder, and detects the rotor phase α of the electric motor 4.

  FIG. 2 is a flowchart showing a flow of processing of motor current control performed by the electric motor controller 2.

  In step S201, a signal indicating the vehicle state is input. Here, the vehicle speed V (km / h), the accelerator opening θ (%), the rotor phase α (rad) of the electric motor 4, the rotational speed Nm (rpm) of the electric motor 4, and the three-phase AC flowing through the electric motor 4 Currents iu, iv, iw, and a DC voltage value Vdc (V) between the battery 1 and the inverter 3 are input.

  The vehicle speed V (km / h) is acquired by communication from a vehicle speed sensor (not shown) or another controller such as a brake controller (not shown). Alternatively, the vehicle rotational speed ωm is multiplied by the tire dynamic radius R, and the vehicle speed v (m / s) is obtained by dividing by the gear ratio of the final gear, and unit conversion is performed by multiplying by 3600/1000 to obtain the vehicle speed V (Km / h) is obtained.

  The accelerator opening θ (%) is acquired from an accelerator opening sensor (not shown), or is acquired by communication from another controller such as a vehicle controller (not shown).

  The rotor phase α (rad) of the electric motor 4 is acquired from the rotation sensor 6. The rotational speed Nm (rpm) of the electric motor 4 is obtained by dividing the angular speed ω (electrical angle) of the rotor by the number of pole pairs of the electric motor 4 to obtain a motor rotational speed ωm (rad / s) is obtained by multiplying the obtained motor rotational speed ωm by 60 / (2π). The angular velocity ω of the rotor is obtained by differentiating the rotor phase α.

  Currents iu, iv, iw (A) flowing through the electric motor 4 are acquired from the current sensor 7.

  The DC voltage value Vdc (V) is obtained from a power supply voltage value transmitted from a voltage sensor (not shown) provided on a DC power supply line between the battery 1 and the inverter 3 or a battery controller (not shown).

In step S202, a first torque command value Tm1 ^* is set. Specifically, the first torque command value Tm1 ^* is set by referring to the accelerator opening-torque table shown in FIG. 3 based on the accelerator opening θ and the vehicle speed V input in step S201.

In step S203, the first torque command value Tm1 ^* and the motor rotation speed ωm set in step S202 are input, and the driving force transmission system vibration (the driving shaft 8 of the driving shaft 8 is not sacrificed) without sacrificing the response of the driving shaft torque. A second torque command value Tm2 ^* that suppresses torsional vibration or the like is set. Details of the method of setting the second torque command value Tm2 ^* will be described later.

In step S204, the d-axis current target value id ^* and the q-axis current target value iq ^* are obtained based on the second torque command value Tm2 ^* , the motor rotation speed ωm, and the DC voltage value Vdc set in step S203.

In step S205, current control is performed to match the d-axis current id and the q-axis current iq with the d-axis current target value id ^* and the q-axis current target value iq ^* obtained in step S204, respectively. For this reason, first, the d-axis current id and the q-axis current iq are obtained based on the three-phase AC current values iu, iv, iw input in step S201 and the rotor phase α of the electric motor 4. Subsequently, d-axis and q-axis voltage command values vd and vq are calculated from a deviation between the d-axis and q-axis current command values id ^* and iq ^* and the d-axis and q-axis current id and iq.

  Next, three-phase AC voltage command values vu, vv, vw are obtained from the d-axis and q-axis voltage command values vd, vq and the rotor phase α of the electric motor 4. Then, PWM signals tu (%), tv (%), and tw (%) are obtained from the obtained three-phase AC voltage command values vu, vv, and vw and the DC voltage value Vdc. The electric motor 4 can be driven with a desired torque indicated by the torque command value by opening and closing the switching element of the inverter 3 by the PWM signals tu, tv, and tw obtained in this way.

FIG. 4 is an example of a control block diagram for performing processing for setting the second torque command value Tm2 ^* . The vibration suppression control calculation unit 400 that sets the second torque command value Tm2 ^* includes a vibration suppression control feedback calculation unit 401 (hereinafter referred to as a vibration suppression control FB calculation unit 401) and a vibration suppression control feedforward calculation unit 402 ( Hereinafter, the vibration suppression control FF calculation unit 402) and an adder 49 are provided.

  The vibration suppression control FB calculation unit 401 includes a control block 41 having a transfer characteristic Gm (s) / Gp (s), a control block 42 having a transfer characteristic Gp (s), and H (s) / Gp (s ), A subtracter 44, and an adder 45.

  Gp (s) is a linear plant model indicating a transfer characteristic between the motor torque input to the vehicle and the motor rotation speed, and does not consider a dead zone due to gear backlash. Gm (s) is an ideal model indicating the transfer characteristic between the motor torque input to the vehicle and the response target of the motor rotation speed. The transfer characteristic H (s) is set so that the difference between the denominator order and the numerator order of the transfer characteristic H (s) is greater than or equal to the difference between the denominator order and the numerator order of the transfer characteristic Gp (s).

The control block 41 inputs the first torque command value Tm1 ^* and outputs FF _out2 .

The control block 42 receives a third torque command value Tm3 ^*, which will be described later, and outputs an estimated value ωm ^ of the motor rotation speed.

  The subtractor 44 calculates the difference between the estimated value ωm ^ of the motor rotation speed and the actual motor rotation speed ωm.

  The control block 43 inputs the difference between the estimated value ωm ^ of the motor rotation speed and the actual motor rotation speed ωm, and outputs an estimated disturbance d ^.

The adder 45 obtains a third torque command value Tm3 ^* by adding the output FF _out2 from the control block 41 and the estimated disturbance d ^.

Next, a linear plant model Gp (s) indicating a transfer characteristic between the motor torque input to the vehicle and the motor rotation speed will be described. FIG. 5 is a diagram in which a driving force transmission system of a vehicle is modeled, and parameters in the figure are as follows.
Jm: Motor inertia Jw: Drive wheel inertia M: Vehicle mass Kd: Torsional rigidity of drive system Kt: Coefficient of friction between tire and road surface N: Overall gear ratio r: Tire load radius ωm: Motor rotation speed Tm ^* : Motor torque command value Td: Torque of driving wheel F: Force applied to vehicle V: Speed of vehicle ωw: Angular velocity of driving wheel From FIG. 5, the following equation of motion can be derived.

Based on the equations of motion (1) to (5), when a linear plant model Gp (s) indicating the transfer characteristic of the motor rotational speed ωm is obtained from the motor torque command value Tm ^* , the following equations (6) to (14) are obtained. Become.

  Equation (6) is modified and expressed as the following equation (15). In a general vehicle, when the poles and zeros of the transfer function of Equation (15) are examined, one pole and one zero show extremely close values. This corresponds to the fact that α and β in Equation (15) show extremely close values. Here, ζp is a damping coefficient of the drive torsional vibration system, and ωp is a natural vibration frequency.

  Therefore, by performing pole-zero cancellation (approximate α = β) in equation (15), as shown in equation (16) below, transfer characteristic Gp (s) in which the denominator is the third order and the numerator is the second order. Configure.

  Next, the transfer function H (s) will be described. H (s) is a feedback element that reduces only vibration when a band-pass filter is used. At this time, the greatest effect can be obtained by setting the filter characteristics as shown in FIG. That is, the transfer function H (s) has substantially the same attenuation characteristics on the low-pass side and the high-pass side, and the torsional resonance frequency of the drive system is close to the center of the passband on the logarithmic axis (log scale). Is set to For example, when H (s) is composed of a first-order high-pass filter and a first-order low-pass filter, the frequency fp is the torsional resonance frequency of the drive system, k is an arbitrary value, and the equation (17) Configure.

However, τ _L = 1 / (2πf _HC ), f _HC = k · f _p , τ _H = 1 / (2πf _LC ), and f _LC = f _p / k.

  Next, the ideal model Gm (s) and the transfer function Gm (s) / Gp (s) indicating the response target of the motor rotation speed to the torque input to the vehicle will be described. When the ideal model Gm (s) is expressed by Expression (18), Gm (s) / Gp (s) is expressed by Expression (19).

Where ζm and ωm are the damping coefficient and natural vibration frequency of the ideal model, respectively.

  Then, the control content performed in the vibration suppression control FF calculating part 402 is demonstrated.

  If the wheel model is omitted from the vehicle model of FIG. 5 and the dead zone due to backlash is expressed by the difference between the linear function and the saturation function, the equation of motion of the vehicle can be approximated by the following equations (20) to (23).

  Where Tc is a virtual viscous friction torque around the drive shaft, Cd is a virtual viscous friction coefficient around the drive shaft, Jv is an equivalent moment of inertia of the vehicle and wheels, and St () is a saturation function (limit). When b and the lower limit are −b, St (a) is a function indicating the characteristics of the following equation.

z = (ωm / N−ωw) / s, and based on the equations of motion (20) to (23), a non-linear plant model Gp ′ () indicating a transfer characteristic from the motor torque command value Tm ^* to the motor rotational speed ωm. When s) is obtained, it can be expressed by the following equations (25) to (29).

  Further, the ideal model Gm (s) indicating the response target of the motor rotation speed with respect to the torque input to the vehicle is expressed by Expression (30).

However, ζm is the attenuation coefficient of the target response, and ωm is the natural vibration frequency, which agrees with the value of equation (19).

  From Expression (25) and Expression (30), a filter that reduces the natural vibration frequency component of torque transmission of a vehicle having a dead zone is represented by Expression (31).

  Further, in order to perform equivalent conversion with respect to Expression (31), Expressions (32) and (33) are defined with y as an output, u as an input, and x as an internal variable.

  From the equations (32) and (33), the equation (31) can be equivalently transformed as the following equations (34) and (35).

  The variable z can be expressed by the following equations (36) and (37) from the equations (20) to (23).

  Assuming that x = z / γ, ζp = ζm, and Tm = u in the equation (36), the equation (34) can be expressed by the following equation (38).

Therefore, the block diagram of the vibration suppression control FF calculation unit 402 is expressed by FIG. 7 from the equations (35) and (38). That is, the vibration suppression control FF calculation unit 402 calculates a control block 71 having a transfer characteristic of 1 / (s ² + 2ζm · ωm · s + ωm ² ), a limiter 72 having a characteristic represented by the equation (24), and ωm ² . Control block 73 for multiplication, adder 74, control block 75 for multiplication by ωm ² , adder 76, and (s ² + 2ζp · ωp · s + ωp ² ) / (s ² + 2ζm · ωm · s + ωm ² ) A control block 77 having characteristics, a control block 78 for multiplying ωp ² , and a subtractor 79 are provided.

At this time, by setting the upper and lower limit values of the saturation function St (x) (limiter 72) of the vibration suppression control FF calculation unit 402 to ± θ _dead / 2γ, complicated calculation (for example, initialization, condition determination, switching) The dead zone compensation amount can be calculated without performing the above. However, θ _dead is a value obtained by converting the backlash amount of the gear into the drive shaft.

  FIG. 8 is a diagram illustrating an example of a control result by the control device for the electric vehicle according to the first embodiment. FIGS. 8A to 8D show temporal changes in torque command value, drive shaft torque, motor rotation speed, and acceleration, respectively.

  When the accelerator is stepped on at the time t1 from the coast state, and the target torque command value suddenly rises, the torque command value is output so as to suppress the vibration frequency component of the driving force transmission system having a dead zone due to backlash. Thus, in the control device described in Japanese Patent Laid-Open No. 2003-9566 described with reference to FIG. 9, it is possible to suppress the overshoot of the drive shaft torque and the vibration of the acceleration.

  FIG. 9 shows an example of a control result by the control device described in Japanese Patent Laid-Open No. 2003-9566 that does not consider the influence of backlash when performing the process of reducing the natural vibration frequency component of the driving force transmission system of the vehicle. FIG. FIGS. 9A to 9D show temporal changes in the torque command value, drive shaft torque, motor rotation speed, and acceleration, respectively.

  When the accelerator is stepped on at the time t1 from the coast state and the target torque command value rises rapidly, the torque is not transmitted between the time t2 and the time t3 due to the dead band due to the gear backlash, so the drive shaft torque is zero. Become. The drive shaft torque is transmitted from time t3, but the drive shaft torque is overshooted by generating a torque command value on the acceleration side in order to correct the dead zone error. As a result, acceleration vibration (shock) occurs between time t3 and time t4.

  As mentioned above, according to the control apparatus of the electric vehicle in 1st Embodiment, it is a control apparatus which sets the motor torque command value based on vehicle information, and controls the torque of the motor connected to a driving wheel, Comprising: A filtering process for reducing the natural vibration frequency of the driving force transmission system of the vehicle having a dead zone that is not transmitted to the driving shaft torque is applied to the motor torque command value, and the motor torque is controlled according to the final torque command value after the filtering process. As a result, the natural vibration frequency component of the driving force transmission system of the vehicle having a dead zone can be reduced even during acceleration from coasting or deceleration, and vibration of the driving shaft torque can be suppressed. Accordingly, it is possible to realize a steep acceleration performance that is smooth and does not impair the response without causing an occupant shock or unpleasant vibration.

  In particular, in the control apparatus for an electric vehicle according to the first embodiment, a vibration suppression control FF calculation unit 402 that inputs a motor torque command value and performs a feedforward calculation, and a vehicle state quantity (motor angular velocity) indicating the state of the vehicle. Vibration suppression control FB calculation unit 401 (feedback calculation means) that performs feedback calculation based on the detected value and the estimated value of the vehicle state quantity, and addition that calculates the final torque command value based on the feedforward calculation result and the feedback calculation result 49 (final torque command value calculation means), and the filtering process constitutes a part of the feedforward calculation. As a result, it is possible to exert a vibration suppressing effect on a non-linear vehicle driving force transmission system having a dead zone.

  Further, the control block 47 that performs the filtering process includes a control block 77 that is a linear filter and a limiter 72 that limits the upper and lower limit values. Therefore, it is necessary to change the filter configuration depending on whether or not there is a dead band (presence or absence of backlash). There is no need for complicated calculations (initialization, condition determination, switching, etc.).

  By setting the upper and lower limit values of the limiter 72 to values based on the backlash amount of the gear, the backlash correction amount can be calculated.

  Since the vibration suppression control FB calculation unit 401 designed based on the linear vehicle model is configured not to be affected by the control block 47 designed based on the nonlinear vehicle model, the vibration suppression control FB calculation unit 401 can remove disturbances and motor characteristic model errors. Particularly, the vibration suppression control FB calculation unit 401 is designed based on a linear vehicle model, and includes a control block 41 (feedforward compensator) that inputs a motor torque command value and performs feedforward calculation. Perform feedback computation using output. Thereby, feedback control can be performed appropriately.

-Second Embodiment-
FIG. 10 is an example of a control block diagram for performing a process of setting a second torque command value Tm2 ^* in the control apparatus for an electric vehicle according to the second embodiment. The vibration suppression control calculation unit 400A that sets the second torque command value Tm2 ^* includes a vibration suppression control FB calculation unit 401A, a vibration suppression control FF calculation unit 402, and an error compensation calculation unit 403. Note that control blocks having the same configuration as that shown in FIG. 4 are denoted by the same reference numerals, and detailed description thereof is omitted.

The vibration suppression control FF calculation unit 402 illustrated in FIG. 10 has the same configuration as the vibration suppression control FF calculation unit 402 illustrated in FIG. 4, and by performing the processing described using the equations (20) to (38), 1 torque command value Tm1 ^* is input and FF _out1 is output.

The adder 49 adds the FF _out1 output from the vibration suppression control FF calculation unit 402 and the estimated disturbance d ^ output from the vibration suppression control FB calculation unit 401A.

The adder 100 adds an output from the adder 49 and an output FF _out3 from an error compensation calculation unit 403, which will be described later, and outputs a second torque command value Tm2 ^* as an addition result.

  The vibration suppression control FB calculation unit 401A includes a control block 42 having a transfer characteristic of Gp (s), a control block 43 having a transfer characteristic of H (s) / Gp (s), and a subtractor 44.

The control block 42 inputs a second torque command value Tm2 ^*, which will be described later, and outputs an estimated value ωm ^ of the motor rotation speed.

  The subtractor 44 calculates the difference between the estimated value ωm ^ of the motor rotation speed and the actual motor rotation speed ωm.

  The control block 43 inputs the difference between the estimated value ωm ^ of the motor rotation speed and the actual motor rotation speed ωm, and outputs an estimated disturbance d ^.

Error compensation operation unit 403A includes a non-linear plant model used in the feed-forward operation, to compensate for the error in the linear plant model used for the feedback operation, the first type the torque command value Tm1 ^*, the FF _out3 Output.

For simplification, with respect to the equations (19), (35), and (38), the natural vibration frequency of the target response is ωm = ωp, the input is Tm1 ^* , the output of the linear FF compensator is y1, If the output of the FF compensator is y2 and the internal variable is x, they can be expressed by the following equations (39) to (41), respectively.

Therefore, the output FF _out3 of the error compensation calculation unit 403 can be expressed by the following equations (42) and (43).

From the above equation, a block diagram of the error compensation calculation unit 403 is represented in FIG. That is, the error compensation calculation unit 403 multiplies the control block 111 having a transfer characteristic of 1 / (s ² + 2ωp · s + ωp ² ), the limiter 112 having the characteristic represented by the equation (24), and ωp ^2. 113, a control block 114 having a transfer characteristic of 2 (1-ζp) ωp · s / (s ² + 2ωp · s + ωp ² ), and an adder 115.

  As mentioned above, according to the control apparatus of the electric vehicle in 2nd Embodiment, it is the structure which uses the output of the damping control FF calculating part 402 which is also a filtering means as an input of the damping control FB calculating part 401A, Comprising: Designed based on the model, the output of the feedforward calculation means that does not include filtering processing as a part of the component, and the output of the damping control FF calculation unit 402 that is the feedforward calculation means that includes filtering processing as a part of the component And an error compensation calculation unit 403 for adding the difference to the motor torque command value. Thereby, feedback control can be performed appropriately.

-Third embodiment-
FIG. 12 is an example of a control block diagram for performing a process of setting a second torque command value Tm2 ^* in the control apparatus for an electric vehicle in the third embodiment. The damping control calculation unit 400B that sets the second torque command value Tm2 ^* includes a damping control FB calculation unit 401B and a damping control FF calculation unit 402B.

  A feedforward calculation process performed by the vibration suppression control FF calculation unit 402B will be described. If the dead zone due to backlash is expressed by the difference between the linear function and the saturation function, the equation of motion of the vehicle is expressed by the following equations (44) to (49).

Here, each parameter is as follows.
Jm: Motor inertia Jw: Drive wheel inertia (for one axis)
Kd: torsional rigidity Kt of the drive shaft: tires and a coefficient N _al Friction road: overall gear ratio r: tire load radius .omega.m: motor angular velocity Omegadaburyu: driving wheel velocity Tm: motor torque Td: drive wheel torque F: driving force ( (For two axes)
V: body speed θ: torsion angle of drive shaft However, St (θ) in equation (47) is a saturation function and is defined by the following equation (50).

Θ _BL in the equation (50) is the gear backlash amount in the overall from the motor to the drive shaft.

  From the equations (44) to (49), the transmission characteristics from the torque command value to the drive shaft torsion angle can be expressed by the following equations (51) to (53).

  However, p1, p0, a3, a2, a1, and a0 in formulas (52) and (53) are represented by formulas (54) to (59), respectively. Ζp is a damping coefficient of the drive torque transmission system, and ωp is a natural vibration frequency of the drive torque transmission system.

  Accordingly, the drive shaft torque is expressed by the following equation (60) from equations (47) and (51).

  Here, the normative response of the drive shaft torque is defined by the following equations (61) and (62).

Here, ζm and ωm are the damping coefficient and natural vibration frequency of the ideal model, respectively.

  When obtaining a torque command value such that the drive shaft torque Td and the motor torque Tm coincide with each other, the following equations (63) and (64) are obtained.

Therefore, a linear filter G _INV (s) that reduces the natural vibration frequency component of the torque transmission of the vehicle, a filter Gt (s) that calculates the drive shaft torsion angle, a saturation function (limiter), and a wheel inertia of the drive shaft torsion angle The configuration of the vibration suppression control FF calculation unit 402B is represented in FIG. 13 by the filter Fs (s) that compensates for the phase shift due to the tire friction force. That is, the vibration suppression control FF calculation unit 402B has a control block 131 having a transfer characteristic G _INV (s), a control block 132 having a transfer characteristic Gt (s), and a characteristic represented by the equation (50). A limiter 133, a control block 134 having a transfer characteristic of Fs (s), an adder 135, an adder 136, and a subtractor 137 are provided.

  By substituting equation (63) into equation (51), equivalent conversion can be made with the following equation (65).

Therefore, the vibration suppression control FF calculation unit 402B includes a linear filter G _INV (s) for reducing the natural vibration frequency component of the torque transmission of the vehicle, a filter Gtm (s) for calculating the ideal response of the drive shaft torsion angle, and saturation. A function (limiter), a wheel inertia of a drive shaft torsion angle, and a filter Fs (s) that compensates for a phase shift due to tire frictional force can also be configured as shown in FIG. According to the configuration illustrated in FIG. 14, the vibration suppression control FF calculation unit 402B includes a control block 131 having a transfer characteristic of G _INV (s), a limiter 133 having a characteristic represented by Expression (50), and Fs (s ), A control block 134 having a transfer characteristic, an adder 136, a subtractor 137, a control block 140 having a transfer characteristic Gtm (s), and an adder 141.

  Next, feedback calculation processing performed by the vibration suppression control FB calculation unit 401B will be described. FIG. 15 is a detailed block diagram of the vibration suppression control FB calculation unit 401B. The vibration suppression control FB calculation unit 401B has a control block 151 having a transfer characteristic of Gp (s), a control block 152 having a transfer characteristic of Gps (s), and a transfer characteristic of H (s) / Gp (s). A control block 153, an adder 154, a subtractor 155, a control block 156 having a transfer characteristic Gt (s), a limiter 157 having a characteristic represented by the equation (50), and a transfer Fs (s). A control block 158 having characteristics and an adder 159 are provided.

  Gp (s) is a linear plant model indicating the transfer characteristic of the motor rotation speed with respect to the motor torque input to the vehicle, and Gps (s) is a transfer function for calculating the backlash compensation amount of the motor rotation speed. Gt (s) is a filter that calculates the drive shaft twist angle. Fs (s) is a filter that compensates for the phase shift caused by the wheel inertia of the drive shaft twist angle and the tire friction force, and is specifically expressed by equations (52) and (53).

The vibration suppression control FB calculation unit 401B receives the post-vibration control torque command value Tm2 ^* and the output FF _out of the vibration suppression control FF calculation unit 402B, and receives a linear plant model Gp (s), a filter Gt (s), and a filter Fs. (S), a saturation function (limiter), and a transfer function Gps (s), a motor rotational speed estimated value ωm ^ is calculated. Further, the difference between the calculated motor rotation speed estimated value ωm ^ and the actual motor rotation speed ωm is input, and the output FB _out of the vibration suppression control FB calculation unit 401B is calculated from the transfer function H (s) / Gp (s). . The transfer characteristic H (s) is set such that the difference between the denominator order and the numerator order of the transfer characteristic H (s) is equal to or greater than the difference between the denominator order and the numerator order of the transfer characteristic Gp (s). As described above, the filter characteristics as shown in FIG. 6 are set.

  Further, as shown in FIG. 16, the drive shaft twist angle limit value St (θ) calculated by the vibration suppression control FF calculation unit 402B may be used as the input of the transfer function Gps (s).

  Hereinafter, a linear plant model Gp (s) indicating the transfer characteristic of the motor rotation speed with respect to the motor torque input to the vehicle and a transfer function Gps (s) for calculating the backlash compensation amount of the motor rotation speed will be described.

  Expressions (44) to (49) are subjected to Laplace conversion to obtain transfer characteristics from the torque command value to the motor angular velocity, and can be expressed by the following expressions (66) to (68).

However, each parameter in the formulas (67) and (68) is expressed by the following formulas (69) to (74).

  Formula (67) is arranged and expressed as the following formula (75). In a general vehicle, when the poles and zeros of the transfer function of Expression (75) are examined, one pole and one zero show extremely close values. This corresponds to the fact that α and β in Equation (75) show extremely close values. Here, ζp and ωp are the damping coefficient and natural vibration frequency of the drive torsional vibration system, respectively.

  Therefore, by performing pole-zero cancellation (approximate with α = β) in equation (75), the (second order) / (third order) transfer characteristic Gp (s) is configured as shown in the following equation (76). To do.

  Note that the transfer function H (s) is expressed by Expression (17).

  FIG. 17 is a diagram illustrating an example of a control result by the control device for the electric vehicle according to the third embodiment. FIGS. 17A to 17D show changes over time in the torque command value, motor rotation speed, drive shaft torque, and acceleration, respectively.

  When the accelerator is stepped on from the coast state at time t1 and the target torque command value suddenly rises, the electric vehicle control device according to the first embodiment uses the wheel inertia of the drive shaft torsion angle in the feedforward calculation. Since the phase delay due to the tire friction force is not taken into consideration, the drive shaft torque slightly overshoots from time t3 to time t4 (see FIG. 8). In addition, the feedback calculation does not take into account the influence of backlash, and thus cannot completely suppress overshoot.

  On the other hand, in the control apparatus for an electric vehicle according to the third embodiment, the motor torque command value is subjected to filtering processing for reducing the natural vibration frequency of the driving force transmission system of the vehicle having the dead zone and suppressing the phase shift due to the wheel characteristics. Therefore, the occurrence of overshoot of the drive shaft torque is suppressed between time t3 and time t4. Thereby, acceleration vibration (shock) is also suppressed between time t3 and time t4.

  As described above, according to the control apparatus for an electric vehicle in the third embodiment, the filtering process for reducing the natural vibration frequency of the driving force transmission system of the vehicle having the dead zone and suppressing the phase shift due to the wheel characteristics is used as the motor torque command value. Therefore, even when accelerating from coasting or deceleration, it is possible to reduce the natural vibration frequency component of the driving force transmission system of the vehicle having a dead zone, to suppress the vibration of the driving shaft torque, and the phase due to the wheel characteristics Deviation can also be suppressed.

  The vibration suppression control FF calculation unit 402B includes a linear filter (control block 131) that reduces the natural vibration frequency component of the driving force transmission system of the vehicle, a filter (control block 132) that calculates the drive shaft torsion angle, A limiter (limiter 133) that limits the lower limit value, a wheel inertia of the drive shaft torsion angle, and a filter (control block 134) that compensates for phase shift due to tire frictional force. Thus, when there is no disturbance or model error, the natural vibration frequency component of the vehicle driving force transmission system having a phase shift due to wheel characteristics and a dead zone due to gears can be reduced only by feedforward compensation. In addition, since it is not necessary to change the filter configuration depending on the presence or absence of the dead zone, it is not necessary to perform complicated calculations (initialization, condition determination, switching, etc.).

  In addition, since the controlled object model of the vibration suppression control FB calculation unit 401B includes at least a linear filter and a limiter that limits the upper and lower limit values, disturbances other than backlash and modeling errors can be compensated.

  Furthermore, the backlash correction amount can be calculated by setting the upper and lower limit values of the limiters 133 and 157 to values based on the gear backlash amount.

  The present invention is not limited to the embodiment described above.

2. Electric motor controller (motor torque control means)
6 ... Rotation sensor (vehicle state quantity detection means)
41 ... Control block 49 ... Adder (final torque command value calculation means)
401, 401A ... Damping control FB calculation unit (feedback calculation means, vehicle state quantity estimation means)
402 ... Damping control FF calculation unit (feed forward calculation means)
403 ... Error compensation calculation unit (model error compensation means)

Claims (11)

In a control device for an electric vehicle that sets a motor torque command value based on vehicle information and controls the torque of a motor connected to a drive wheel,
Filtering means for applying a filtering process to the motor torque command value to reduce the natural vibration frequency of the driving force transmission system of the vehicle having a dead zone where the motor torque is not transmitted to the driving shaft torque of the vehicle;
Motor torque control means for controlling motor torque in accordance with a final torque command value obtained by filtering the motor torque command value;
Feedforward calculation means for inputting the motor torque command value and performing feedforward calculation for suppressing vibration of the driving force transmission system of the vehicle;
Vehicle state quantity detecting means for detecting a vehicle state quantity indicating the state of the vehicle;
Vehicle state quantity estimating means for estimating the vehicle state quantity;
Feedback calculation means for performing feedback calculation for suppressing vibration of the driving force transmission system of the vehicle based on the detected value of the vehicle state quantity and the estimated value of the vehicle state quantity;
A final torque command value calculating means for calculating the final torque command value based on the feedforward calculation result and the feedback calculation result;
The filtering means is a part of a component of the feedforward calculation means.
A control apparatus for an electric vehicle. In the control apparatus of the electric vehicle according to claim 1 ,
The filtering means includes a linear filter and a limiter that limits the upper and lower limit values.
A control apparatus for an electric vehicle. In the control apparatus of the electric vehicle according to claim 2 ,
The upper and lower limit values of the limiter are values based on the backlash amount of the gear.
A control apparatus for an electric vehicle. In the control apparatus of the electric vehicle according to claim 1 ,
The feedback calculation means designed based on the linear vehicle model is configured not to be affected by the filtering means designed based on the nonlinear vehicle model.
A control apparatus for an electric vehicle. In the control apparatus of the electric vehicle according to claim 4 ,
The feedback calculation means includes a feedforward compensator that is designed based on a linear vehicle model and performs a feedforward calculation by inputting the motor torque command value, and the feedback calculation using the output of the feedforward compensator I do,
A control apparatus for an electric vehicle. In the control apparatus of the electric vehicle according to claim 4 ,
Model error compensation calculation means for compensating for an error between the linear vehicle model and the nonlinear vehicle model is further provided.
A control apparatus for an electric vehicle. In the control apparatus of the electric vehicle according to claim 1,
The filtering means performs a filtering process on the motor torque command value for reducing a natural vibration frequency of a driving force transmission system of a vehicle having the dead zone and suppressing a phase shift due to wheel characteristics.
A control apparatus for an electric vehicle. The control apparatus for an electric vehicle according to claim 7 ,
Before SL feedforward calculation means includes a linear filter for reducing natural vibration frequency component of the driving force transmission system of a vehicle, a filter for calculating a twist angle drive shaft, and a limiter for limiting the upper and lower limit values, the torsional angle of the drive shaft A wheel inertia and a filter that compensates for a phase shift due to tire friction force,
A control apparatus for an electric vehicle. The control apparatus for an electric vehicle according to claim 8 ,
The controlled object model of the feedback calculation means is composed of at least a linear filter and a limiter that limits the upper and lower limit values.
A control apparatus for an electric vehicle. The control apparatus for an electric vehicle according to claim 8 or 9 ,
The upper and lower limit values of the limiter are values based on the backlash amount of the gear.
A control apparatus for an electric vehicle.   In a control method for an electric vehicle that sets a motor torque command value based on vehicle information and controls the torque of a motor connected to a drive wheel,
  Applying a filtering process to the motor torque command value to reduce a natural vibration frequency of a driving force transmission system of a vehicle having a dead zone where the motor torque is not transmitted to the driving shaft torque of the vehicle;
  Controlling the motor torque in accordance with a final torque command value obtained by filtering the motor torque command value;
  Inputting the motor torque command value and performing a feedforward calculation for suppressing vibration of a driving force transmission system of the vehicle;
  Detecting a vehicle state quantity indicating a state of the vehicle;
  Estimating the vehicle state quantity;
  Performing a feedback calculation for suppressing vibration of a driving force transmission system of the vehicle based on the detected value of the vehicle state quantity and the estimated value of the vehicle state quantity;
  Calculating the final torque command value based on the feedforward calculation result and the feedback calculation result;
Prepared,
  The step of performing the filtering process is a part of the step of performing the feedforward calculation.
  An electric vehicle control method characterized by the above.