JP5444929B2 - Chassis dynamometer system - Google Patents

Description

  The present invention relates to a chassis dynamometer system using a roller angular velocity estimator designed by H∞ control or μ design method.

  An electric inertia control system is employed to electrically compensate a mechanical inertia component of a power measurement target on the load side or drive side of the dynamometer measurement system. As this electric inertia control system, Patent Document 1 is known.

  Chassis dynamometer systems and drive train bench system dynamometer systems also have shaft torque detection characteristics, dynamometer angular speed detection characteristics, inverter response characteristics, etc. in addition to mechanical resonance characteristics. Without considering the detection delay factor, stable control with higher response cannot be achieved.

  For example, Patent Document 2 discloses a design that allows a highly responsive electric inertia control circuit to be designed using a generalized plant model in consideration of the above-described characteristics.

JP 2004-361255 A JP 2008-286614 A

  A conventional problem will be described with reference to FIG.

  FIG. 13 shows a chassis dynamometer system including a dynamometer 101, a dynamometer speed detector 102, a shaft 103, a shaft torque meter 104, a roller 105, a control unit 106, and an inverter 107. The roller 105 is rotated through a shaft 103 connected to the dynamometer 101 and the roller 105.

  The speed detector 102 is attached to the dynamometer 101 and detects the rotational speed of the dynamometer 101. The angular velocity is obtained from the number of rotations per unit time.

  The shaft torque meter 104 is a shaft torque meter that detects the shaft torque acting on the shaft 103 from, for example, the amount of distortion in the torsional direction.

  The control unit 106 outputs a control signal to the inverter 107 based on the calculation result based on the detection signals of the speed detector 102 and the shaft torque meter 104.

  The inverter 107 converts DC power from a DC power source (not shown) into AC power using a switching element such as an IGBT (not shown), and converts predetermined AC power into dynamometer 101 based on a signal from the control unit 106. To supply.

  The chassis dynamometer system shown in FIG. 13 shows a state in which the speed detector 102 can be attached only to the dynamometer 101 for some reason. In such a case, the dynamometer angular velocity signal has an anti-resonance point to be described later in detail. Therefore, it is difficult to increase the response such as ASR (speed control) using dynamometer angular velocity or ALR (running resistance control) by electric inertia control.

  The present invention has been made based on the above-described problems, and provides a chassis dynamometer system for speed control and running resistance control with high response.

In order to solve the above problems, the present invention provides a roller, a dynamometer that rotates the roller by rotating a common shaft with the roller, a shaft torque detector that detects the torque of the shaft, In a chassis dynamometer system comprising a dynamometer for detecting an angular velocity, and a controller for controlling an inverter for supplying electric power to the dynamometer, the control means includes a shaft torque detected by the shaft torque detector. a roller angular velocity estimation circuit that generates a roller angular velocity estimation signal is the estimated angular velocity of the roller on the basis of the detection signal and the dynamometer angular velocity detection signal the dynamometer angular velocity detector detects the roller angular velocity estimation signal and said roller reaches power is the torque to be output by the power meter performs PID control on the difference of roller angular velocity command is should do angular velocity And a speed control circuit for generating a torque command, chassis and mechanical models of the dynamometer, a first weighting function to weight the signal corresponding to the surface driving force of said roller, torque the power meter to be output A second weighting function for weighting a signal corresponding to a dynamometer torque command, a third weighting function for weighting a signal corresponding to shaft torque observation noise that is an observation noise of the shaft torque detector, and the power A fourth weighting function that weights a signal corresponding to dynamometer observation noise that is observation noise of the meter speed detector, and the roller generated based on each signal weighted by the first and second weighting functions. The difference signal between the signal corresponding to the angular velocity and the estimated angular velocity of the roller generated based on the signals weighted by the first to fourth weighting functions is weighted. A fifth weighting function; a sixth weighting function that weights a signal corresponding to the torsional torque of the shaft generated based on each signal weighted by the first and second weighting functions; A seventh weight function for weighting a signal corresponding to the angular velocity of the dynamometer generated based on each signal weighted by the second weight function, and an eighth weight for weighting a signal corresponding to the angular velocity of the roller A mechanical system model of the chassis dynamometer includes a first transfer function having a characteristic of the moment of inertia of the roller, a second transfer function having a characteristic of a spring stiffness of the shaft, and the dynamometer A third transfer function having a characteristic of moment of inertia, and the first to third transfer functions receive signals based on the respective signals weighted by the first and second weight functions. The signal corresponding to the angular velocity of the roller is generated using the first transfer function and the second transfer function, and the signal corresponding to the torsional torque of the shaft is the first transfer function, the second transfer function. 2 and the third transfer function, and a signal corresponding to the angular velocity of the dynamometer is generated using the second transfer function and the third transfer function. A signal corresponding to a surface driving force, a signal corresponding to a dynamometer torque command which is a torque to be output by the dynamometer, a signal corresponding to shaft torque observation noise which is an observation noise of the shaft torque detector, and the dynamometer angular velocity In the generalized plant in which the signal corresponding to the dynamometer observation noise which is the observation noise of the detector is a disturbance input and each weighted output of the fifth to eighth weight functions is a control amount, the first to eighth A roller angular velocity is estimated based on a shaft torque detection signal and a dynamometer angular velocity detection signal by obtaining a coefficient of a state equation of the controller using a weight function by a control system design method called H∞ control or μ design method. A roller angular velocity estimator is designed, the designed roller angular velocity estimator is applied to the roller angular velocity estimation circuit, and the controller adds a signal corresponding to the torsional torque of the shaft to the third weight function output. The control input is a signal corresponding to the obtained shaft torque detection signal and a signal corresponding to the dynamometer angular velocity detection signal obtained by adding a signal corresponding to the angular velocity of the dynamometer to the fourth weight function output. The estimated angular velocity of the roller is output as an observation amount, and the control amount as the fifth weight function output is a signal corresponding to the angular velocity of the roller and the control amount. The roller angular velocity estimation circuit outputs the estimated roller angular velocity output from the controller as a roller angular velocity estimation signal .

According to the above configuration, the roller angular velocity estimation signal, which is the roller angular velocity estimated by the roller angular velocity estimation circuit based on the shaft torque and the dynamometer angular velocity, is input to the speed control circuit. The speed control of the chassis dynamometer is less affected by the anti-resonance point by controlling the inverter based on the dynamometer torque command generated by the speed control circuit based on the roller angular speed estimation signal. For this reason, speed control of a highly responsive chassis dynamometer system is possible.
In the generalized plant model, the coefficient of the state equation of the roller angular velocity estimator is obtained by H∞ control or μ design method using the first to eighth weight functions. Thus, a roller angular velocity estimator that estimates the roller angular velocity based on the two signals of shaft torque detection and dynamometer angular velocity detection can be designed, and a roller angular velocity estimation circuit to which this roller angular velocity estimator is applied can be configured.

Further, the control means includes a shaft torque control circuit having a resonance suppression characteristic in which a signal generated by performing PID control on a difference between the signal generated by the speed control circuit and the shaft torque detection signal is used as a dynamometer torque command. It is characterized by having.

  According to the above configuration, the resonance point is suppressed by the shaft torque control circuit having the resonance suppression characteristic generating the dynamometer torque command based on the output signal of the speed control circuit. That is, since the resonance point is suppressed in addition to being less susceptible to the influence of the anti-resonance point, it is possible to control the speed of the chassis dynamometer system with higher response.

A roller surface driving force estimation circuit configured to subtract the shaft torque detection signal from the generated roller angular velocity estimation signal to generate a roller surface driving force estimation signal that is an estimated surface driving force of the roller; And an inertia setting circuit that generates a roller angular velocity command based on a difference signal between the roller surface driving force estimation signal and a running resistance command that is a command value of the running resistance of the vehicle.

  According to the above configuration, the roller angular velocity estimation signal generated by the roller angular velocity estimation circuit is input to the roller surface driving force estimation circuit, and the roller surface driving force estimation circuit and the EIC (set inertia value calculation equivalent to vehicle inertia) circuit are provided. A roller angular velocity command is generated via The speed control circuit controls the inverter based on the dynamometer torque command generated based on the roller angular velocity command and the roller angular velocity estimation signal, so that the electric inertia control of the chassis dynamometer is less affected by the anti-resonance point. Become. For this reason, running resistance control by electric inertia control of a highly responsive chassis dynamometer system is possible.

According to the first aspect of the present invention, the roller angular velocity estimation signal, which is the angular velocity of the roller estimated by the roller angular velocity estimation circuit based on the shaft torque and the dynamometer angular velocity, is input to the speed control circuit. The speed control of the chassis dynamometer is less affected by the anti-resonance point by controlling the inverter based on the dynamometer torque command generated by the speed control circuit based on the roller angular speed estimation signal. For this reason, speed control of a highly responsive chassis dynamometer system is possible.
In the generalized plant model, the coefficient of the state equation of the roller angular velocity estimator is obtained by H∞ control or μ design method using the first to eighth weight functions. Thus, a roller angular velocity estimator that estimates the roller angular velocity based on the two signals of shaft torque detection and dynamometer angular velocity detection can be designed, and a roller angular velocity estimation circuit to which this roller angular velocity estimator is applied can be configured.

  According to the invention of claim 2, the resonance point is suppressed by the shaft torque control circuit having the resonance suppression characteristic generating the dynamometer torque command based on the output signal of the speed control circuit. That is, since the resonance point is suppressed in addition to being less susceptible to the influence of the anti-resonance point, it is possible to control the speed of the chassis dynamometer system with higher response.

  According to the invention of claim 3, the roller angular velocity estimation signal generated by the roller angular velocity estimation circuit is input to the roller surface driving force estimation circuit, and the roller surface driving force estimation circuit and EIC (set inertia value calculation corresponding to vehicle inertia) are calculated. ) A roller angular velocity command is generated via a circuit. The speed control circuit controls the inverter based on the dynamometer torque command generated based on the roller angular velocity command and the roller angular velocity estimation signal, so that the electric inertia control of the chassis dynamometer is less affected by the anti-resonance point. Become. For this reason, running resistance control by electric inertia control of a highly responsive chassis dynamometer system is possible.

The Bode diagram of the transfer function G1 (s). The Bode diagram of the transfer function G2 (s). The block diagram of the generalized plant model used for the design of the roller angular velocity estimation circuit in this invention. The block diagram of the mechanical system model in a generalized plant model. The block diagram of the designed roller angular velocity estimation circuit in this invention. The block diagram of the conventional speed control circuit. 1 is a block diagram of a speed control circuit in Embodiment 1. FIG. FIG. 4 is a block diagram of a speed control circuit in Embodiment 2. The block diagram of the conventional electric inertia control circuit. The internal block diagram of a roller surface driving force estimation circuit. FIG. 6 is a block diagram of an electric inertia control circuit in Embodiment 3. FIG. 6 is a block diagram of an electric inertia control circuit in Embodiment 4. The block diagram of a chassis dynamometer system.

Hereinafter, a chassis dynamometer system according to an embodiment of the present invention will be described in detail with reference to the drawings.
<About anti-resonance point>
As described above, it is difficult to increase the response of ASR and ALR due to the influence of the antiresonance point. With respect to this, a two-inertia model obtained from a chassis dynamometer such as that shown in FIG. It explains using.

  The equation of motion of the two-inertia model is as follows: Jroller is the moment of inertia of the roller 105, Jdy is the moment of inertia of the dynamometer 101, Ksh is the torsional rigidity of the shaft 103, Tdy is the torque of the dynamometer 101, and ωroller is the angular velocity of the roller 105. , Ωdy is defined as the angular velocity of the dynamometer 101, and Tsh is defined as the torsional torque of the shaft 103.

It becomes.

  When the expressions (1) to (3) are Laplace transformed, the expressions (1) to (3) are respectively

Equations (4) to (6) are obtained as follows (transfer function s is a Laplace operator including the following), and the angular velocity ωroller of roller 105 and the angular velocity ωdy of dynamometer 101 are respectively solved:
A transfer function G1 (s) (hereinafter referred to as G1 (s)) that outputs an angular velocity ωroller of the roller 105 in response to the input of Tdy of the dynamometer torque 101 is as follows.

The transfer function G2 (s) (hereinafter referred to as G2 (s)) that outputs the angular velocity ωdy of the dynamometer 101 in response to the input of the torque Tdy of the dynamometer 101 is

It becomes.

  FIG. 1 shows a Bode diagram for G1 (s), and FIG. 2 shows a Bode diagram for G2 (s). 1 and 2, for the frequency of 1 to 100 Hz, the upper part shows the change of Magnitude (dB) and the lower part shows the change of Phase (deg).

  In the case of FIG. 1, Magnitude decreases from around −63 dB as the frequency increases, and Phase is approximately −90 deg. However, from near 50 Hz, Magnitude suddenly rises to near -55 dB as it approaches 75 Hz, and suddenly falls after exceeding 75 Hz, and a resonance point appears near 75 Hz. Also, Phase changes by approximately -180 deg from approximately -90 deg to approximately -270 deg around 75 Hz.

  On the other hand, in the case of FIG. 2, Magnitude is changed to 10 Hz and the resonance point appears in the vicinity of 75 Hz as in FIG. However, Magnitude is rapidly increased to about -135 dB at around 11.2 Hz and then rapidly rising, and an anti-resonance point appears at around 11.2 Hz. Phase changes from approximately −90 deg to approximately 90 deg around 11.2 Hz where the anti-resonance point appears, and Phase changes from approximately 90 deg to approximately −90 deg around 75 Hz as approximately −180 deg. Yes.

  Here, the resonance point and the anti-resonance point will be described. In the Bode diagram when vibrating with a sine wave, the response amplitude (Magnitude) has such a characteristic that it repeats a peak (resonance point) and a valley (anti-resonance point). At the resonance point, when a signal is applied to the input of some object, the output signal becomes a signal larger than the input signal. At the anti-resonance point, when a signal is applied to the input, the output signal becomes a signal smaller than the input signal. Therefore, when a force is applied at the resonance point, vibration occurs, but at the anti-resonance point, the force is hardly transmitted.

  When the dynamometer angular velocity is used in ASR or ALR (running resistance control), control with high response becomes difficult due to the influence of such an anti-resonance point.

In order to reduce the influence of this anti-resonance point, the present invention uses a roller angular velocity estimation circuit that estimates a roller angular velocity based on shaft torque and dynamometer angular velocity signals. The method will be described below.
<Roller angular velocity estimator design method>
A design method of a roller angular velocity estimator designed by the H∞ control or μ design method for estimating the angular velocity of the roller 105 based on signals from the velocity detector 102 and the shaft torque meter 104 will be described with reference to FIGS. To do.

  FIG. 3 is an overall view of a generalized plant model used for designing a roller angular velocity estimator.

  The generalized plant model includes weight functions 11 to 14 represented as Gw1 (s) to Gw4 (s), weight functions 21 to 24 represented as Gz1 (s) to Gz4 (s), a controller 10, Gmec ( s), a two-inertia mechanical system model 15, a transfer function 16 expressed as Gtm (s), a transfer function 17 expressed as Genc (s), adders 18 and 19, and a subtracter 20. The Further, w1 to w4 are disturbances, z1 to z4 are control amounts, u1 and u2 are control inputs, and y1 is an observation amount.

  The disturbance w1 is a roller surface driving force, the disturbance w2 is a dynamometer torque command, the disturbance w3 is an axial torque observation noise, and the disturbance w4 is a dynamometer angular velocity noise.

  Here, the roller surface driving force is a signal corresponding to the vehicle driving force applied to the surface of the roller 105. The dynamometer torque command is a signal corresponding to the torque that the dynamometer 101 should output. The shaft torque observation noise is a signal corresponding to noise when the shaft torque meter 104 observes the shaft torque of the shaft 103. The dynamometer angular velocity observation noise is a signal corresponding to noise when the velocity detector 102 observes the angular velocity of the dynamometer 101.

  The weighting function 11 is a transfer function with a predetermined constant, weights the disturbance w1, and the torque 105 of the roller 105 due to the vehicle driving force applied to the surface of the roller 105. As T, it outputs to the 2-inertia mechanical system model 15 mentioned later.

  The weighting function 12 is a transfer function with a predetermined constant, weights the disturbance w2, and the torque J2. T is output to the two-inertia mechanical system model 15 as T.

  The weighting function 13 is a predetermined constant or a transfer function having a characteristic that the gain increases at high frequencies, weights the disturbance w3, and outputs it to the adder 18.

  The weighting function 14 is a predetermined constant or a transfer function having a characteristic that the gain increases at high frequencies, weights the disturbance w4, and outputs it to the adder 19.

  The 2-inertia mechanical system model 15 expresses the mechanical characteristics of the chassis dynamometer as a transfer function. T and J2. T is input, and the angular velocity of the roller 105 is J1. w, the torsional torque of the shaft 103 connecting the dynamometer 101 and the roller 105, K12. T and the angular velocity of the dynamometer 101 J2. Output w.

  The two-inertia mechanical system model 15 will be further described with reference to FIG. The configuration includes an adder 15a, subtracters 15b and 15c, and transfer functions 15d to 15f.

  The adder 15 a is a signal corresponding to the rotational moment torque of the roller 105 output from the weight function 11. T and a signal from a transfer function 15e described later are added and output to the transfer function 15d.

  The transfer function 15d is expressed as 1 / (J1 · s), where J1 is a roller inertia moment. The signal output from the adder 15 a is processed by the transfer function, and is a signal corresponding to the angular velocity of the roller 105. Output as w and output to the subtractor 15c.

  The subtractor 15b outputs the J2. A signal of a transfer function 15e described later is subtracted from the signal of T and output to the transfer function 15f.

  The transfer function 15f is expressed as 1 / (J2 · s), where J2 is a dynamometer moment of inertia. The signal output from the subtractor 15b is processed for its transfer function, and is a signal corresponding to the angular velocity of the dynamometer 101. Output as w and output to the subtractor 15c.

  The subtractor 15c subtracts the signal of the transfer function 15d from the signal of the transfer function 15f, and outputs it to the transfer function 15e.

  The transfer function 15e is expressed as K12 / s, where K12 is the spring stiffness of the shaft 103. The signal output from the subtractor 15 c is processed for its transfer function, and is a signal corresponding to the torsional torque of the shaft 103. Output as T and output to the adder 15a and subtractor 15b.

  Returning to FIG. 3, the transfer function 16 is a characteristic of the shaft torque meter 104 when detecting the shaft torque, and is a signal from the two-inertia mechanical system model 15. The transfer function is processed at T and output to the adder 18.

  The transfer function 17 is a characteristic of the speed detector 102 that detects the dynamometer angular velocity, and is a signal from the two-inertia mechanical system model 15. The transfer function is processed for w and output to the adder 19.

  The adder 18 adds the signal from the weight function 13 and the signal from the transfer function 16 and outputs shaft torque detection, which is a signal corresponding to the detection signal of the shaft torque meter 104, to the controller 10 as a control input u1.

  The adder 19 adds the signal from the weight function 14 and the signal from the transfer function 17 and outputs the dynamometer angular velocity detection corresponding to the detection signal of the velocity detector 102 to the controller 10 as the control input u2. .

  The controller 10 generates an observation amount y1 that is an estimated value of the angular velocity of the roller 105 based on the shaft torque detection that is the control input u1 and the dynamometer angular velocity detection that is the control input u2, and outputs the observation amount y1 to the subtracter 20. The coefficients of the internal state equation are determined by weight functions 11 to 14 and weight functions 21 to 24 described later.

  The subtracter 20 is a signal of the 2-inertia mechanical system model 15 J1. The roller angular velocity estimation signal, which is a signal from the controller 10, is subtracted from w and output to the weighting function 21.

  The weighting function 21 is a predetermined constant or a transfer function having a characteristic that the gain becomes low at a high frequency. The transfer function is processed on the signal from the subtracter 20, and the weighted roller speed Observer which is the control amount z1 is used. Generate a deviation.

  The weight function 22 is a predetermined constant or a transfer function having a characteristic that the gain increases at high frequencies, and is a signal from the two-inertia mechanical system model K12. This transfer function processing is performed on T to generate a weighted shaft torque that is a controlled variable z2.

  The weight function 23 is a transfer function having a predetermined constant or a characteristic that gain increases at high frequencies, and is a signal from the two-inertia mechanical system model J2. This transfer function processing is performed on w to generate a weighted dynamometer angular velocity that is a control amount z3.

  The weight function 24 is a predetermined constant or a transfer function having a characteristic that the gain increases at a high frequency, and is a signal from the two-inertia mechanical system model 15. This transfer function processing is performed on w to generate a weighted roller angular velocity that is a control amount z4.

  In such a generalized plant model, the coefficient of the state equation of the controller 10 is obtained by the H∞ control or μ design method using the weight functions 11 to 14 and 21 to 24, thereby detecting the shaft torque and the dynamometer angular velocity. A roller angular velocity estimator that estimates the roller angular velocity based on the two signals can be designed. Further, by applying the roller angular velocity estimator designed in this way, a roller angular velocity estimation circuit can be configured.

Further, when determining the coefficient of the state equation of the controller 10, not only the weight functions 11 to 14 and 21 to 24, but also other transfer functions of the two-inertia mechanical system model 15 and / or the transfer function 16 of the detector, 17 may be used. By doing so, it is possible to design a roller angular velocity estimator that takes into consideration the resonance characteristics of the mechanical system of the chassis dynamometer, the detection characteristics of the speed detector 102, the detection characteristics of the shaft torque meter 104, and the like.

  FIG. 5 is a block diagram of the roller angular velocity estimation circuit 50. The roller angular velocity estimation circuit 50 receives shaft torque detection and dynamometer angular velocity detection, and outputs a roller estimated angular velocity. Each embodiment using the roller angular velocity estimation circuit 50 will be described below.

  Here, the two-inertia system configuration shown in FIGS. 5 to 12 is a block diagram of a chassis dynamometer including the components indicated by reference numerals 101 to 105 and 107 in FIG. And the like are block diagrams of a control circuit provided in the control unit 106.

  This embodiment relates to speed control of a chassis dynamometer system. This embodiment will be described in comparison with the conventional example shown in FIG. FIG. 6 is a block diagram of a circuit configuration in speed control of the chassis dynamometer system, which includes an ASR 51 and a two-inertia mechanical configuration 52. The ASR 51 corresponds to a speed control circuit.

  The ASR 51 performs speed control by PID control shown in FIG. The ASR 51 receives a dynamometer angular velocity command that is an angular velocity that the dynamometer 101 should reach and a dynamometer angular velocity detection signal that is a signal from the two-inertia mechanical configuration 52. Then, the ASR 51 performs PID control on these difference signals and outputs them as a dynamometer torque command to the inverter 107, that is, the two-inertia system mechanical configuration 52.

  In the two-inertia mechanical configuration 52, the dynamometer 101 is operated by the AC power from the inverter 107 controlled based on the dynamometer torque command output from the ASR 51, and the dynamometer detected by the speed detector 102 at that time. A dynamometer angular velocity detection signal having an angular velocity of 101 is output to the ASR 51.

  Next, the present embodiment will be described with reference to FIG.

  A significant difference from the configuration of FIG. 6 is that the roller angular velocity estimation circuit 50 described above is provided. Thereby, the ASR 51 and the two-inertia machine configuration 52 are as follows.

  The ASR 51 receives a signal from the roller angular velocity estimation circuit 50 instead of the dynamometer angular velocity detection signal from the two-inertia mechanical configuration 52. Further, the angular velocity command in this case is a roller angular velocity command that is an angular velocity that the roller 105 should reach.

  In addition to detecting the dynamometer angular velocity that is the angular velocity of the dynamometer 101, the two-inertia mechanical configuration 52 also outputs shaft torque detection that is the torque of the shaft 103 detected by the shaft torque meter 104. These signals are output to the roller angular velocity estimation circuit 50.

  The roller angular velocity estimation circuit 50 generates a roller angular velocity estimation signal that is an estimated angular velocity of the roller 105 based on shaft torque detection and dynamometer angular velocity detection that are signals from the two-inertia mechanical configuration 52, and outputs the roller angular velocity estimation signal to the ASR 51.

  Thus, the roller angular velocity estimation circuit 50 estimates the angular velocity of the roller 105 based on the shaft torque detection and the dynamometer angular velocity detection from the two-inertia mechanical configuration 52, and the roller angular velocity estimation signal is input to the ASR 51. Then, the dynamometer 101 is less affected by the anti-resonance point by controlling the inverter 107 that supplies power to the dynamometer 101 by the dynamometer torque command generated by the ASR 51 based on the roller angular velocity estimation signal. For this reason, speed control of a highly responsive chassis dynamometer system is possible.

  This embodiment also relates to speed control of the chassis dynamometer system. This embodiment will be described with reference to FIG. The same components as those in the first embodiment are denoted by the same reference numerals and description thereof is omitted. ATR 53 corresponds to the shaft torque control circuit.

  A significant difference from the configuration of the first embodiment is that an ATR 53 is provided.

  The ATR 53 has resonance suppression characteristics and performs shaft torque control by PID control. The ATR 53 receives the dynamometer torque command in the first embodiment output from the ASR 51 and the shaft torque detection signal output from the two-inertia mechanical configuration 52. Then, PID control is performed on the difference signal, and a dynamometer torque command is output to the 2-inertia machine configuration 52.

  Here, the resonance suppression characteristic means that in the case of the frequency characteristic from the dynamometer torque command to the shaft torque detection, the resonance suppression characteristic is controlled to be a characteristic such that the gain in the resonance frequency band is 0 dB.

  In this way, by using the roller angular velocity estimation circuit 50 and using the angular velocity signal input to the ASR 51 as the roller angular velocity estimation signal, the dynamometer 101 becomes less susceptible to the anti-resonance point. Furthermore, the resonance point is suppressed by the ATR 53 having resonance suppression characteristics by generating a dynamometer torque command based on the output signal of the ASR 51 and the shaft torque detection. For this reason, since the resonance point is suppressed in addition to being less susceptible to the influence of the anti-resonance point, it is possible to control the speed of the chassis dynamometer system with higher response than that of the first embodiment.

This embodiment relates to running resistance control of a chassis dynamometer system. This embodiment will be described in comparison with the conventional example shown in FIG. FIG. 9 is a block diagram of a circuit configuration in the running resistance control of the chassis dynamometer system, which includes an ASR 51, an inertial system configuration 52, a roller surface driving force estimation circuit 54, a subtractor 55, and an EIC 56. Since the ASR 51 and the two-inertia mechanical configuration 52 are the same as those in the first embodiment, the description thereof is omitted. The EIC 56 corresponds to an inertia setting circuit.

  The roller surface driving force estimation circuit 54 generates a roller surface driving force estimation signal that is a signal of an estimated surface driving force of the roller 105 from dynamometer angular velocity detection and shaft torque detection that are signals from the two-inertia mechanical configuration 52. It is. The roller surface driving force estimation circuit 54 will be further described with reference to FIG. FIG. 10 illustrates the roller surface driving force estimation circuit 54 using a transfer function. The roller surface driving force estimation circuit 54 includes transfer functions 54a and 54b and a subtractor 54c.

  The transfer function 54a is expressed as Jroller · s / G (s). Jroller is the moment of inertia of the roller 105, and G (s) is an arbitrary transfer function having a relative order of the first order or higher. Further, the transfer function 54a processes the transfer function on the dynamometer angular velocity detection signal from the two-inertia mechanical configuration 52, and outputs it to the subtractor 54c.

  The transfer function 54b is expressed as 1 / G (s), and G (s) is the same as that of the transfer function 54a. The transfer function is processed on the shaft torque detection signal from the two-inertia machine configuration 52 and output to the subtractor 54c.

  The subtractor 54c subtracts the signal of the transfer function 54b from the signal of the transfer function 54a, generates a roller surface driving force estimation signal, and outputs it to the subtractor 55.

  The subtractor 55 subtracts a running resistance command, which is a command value for running resistance of the vehicle, from the roller surface driving force estimation signal and outputs the difference to the EIC 56.

  The EIC 56 is expressed as 1 / (EIC_J · s), and EIC_J is a set inertia value corresponding to the inertia of the vehicle. The signal input to the EIC 56 is multiplied by 1 / EIC_J and further integrated. This transfer function is processed on the signal from the subtractor 55 and output to the ASR 51 as a dynamometer angular velocity command.

  Next, the present embodiment will be described with reference to FIG.

  A significant difference from the configuration of FIG. 9 is that the roller angular velocity estimation circuit 50 described above is provided. As a result, the roller surface driving force estimation circuit 54, the EIC 56, and the ASR 51 are as follows.

  The roller surface driving force estimation circuit 54 receives a roller angular velocity estimation signal output from the roller angular velocity estimation circuit 50 instead of receiving the dynamometer angular velocity detection output from the two-inertia mechanical configuration 52.

  The EIC 56 generates a roller angular velocity command.

  Similarly, the ASR 51 receives the roller angular velocity estimation signal output from the roller angular velocity estimation circuit 50 instead of the dynamometer angular velocity detection signal output from the two-inertia mechanical configuration 52.

  Thus, the roller angular velocity estimation circuit 50 estimates the roller angular velocity based on the shaft torque detection and the dynamometer angular velocity detection from the two-inertia mechanical configuration 52, and the roller angular velocity estimation signal is the ASR 51 and the roller surface driving force estimation circuit 54. Has been entered. A roller angular velocity command is generated via the roller surface driving force estimation circuit 54, the EIC 56, and the like. Furthermore, the dynamometer 101 is influenced by the anti-resonance point by controlling the inverter 107 that supplies electric power to the dynamometer 101 by the dynamometer torque command that the ASR 51 generates based on the roller angular velocity command and the roller angular velocity estimation signal. It becomes difficult to receive. For this reason, running resistance control by electric inertia control of a highly responsive chassis dynamometer system is possible.

  This embodiment also relates to running resistance control of the chassis dynamometer system. This embodiment will be described with reference to FIG. The same components as those in the third embodiment are denoted by the same reference numerals, and description thereof is omitted.

  A significant difference from the configuration of the third embodiment is that an ATR 53 is provided. ATR 53 is the same as that described in the second embodiment. Further, the same components as those described in the third embodiment are denoted by the same reference numerals, and the description thereof is omitted.

  The ATR 53 receives the dynamometer torque command in the third embodiment output from the ASR 51 and the shaft torque detection signal output from the two-inertia mechanical configuration 52. Then, PID control is performed on the difference signal, and a dynamometer torque command is output to the 2-inertia machine configuration 52. In addition, the ATR 53 has a resonance suppression characteristic as described above, thereby suppressing the resonance point.

  In this way, by using the roller angular velocity estimation circuit 50 and using the angular velocity signal input to the ASR 51 and the roller surface driving force estimation circuit 54 as the roller angular velocity estimation signal, the roller resonance velocity estimation signal is less affected by the anti-resonance point. Further, the resonance point is suppressed by the ATR 53 having resonance suppression characteristics generating a dynamometer torque command based on the output signal of the ASR 51 and the shaft torque detection signal. For this reason, since the resonance point is suppressed in addition to being less susceptible to the influence of the anti-resonance point, it is possible to perform the running resistance control by the electric inertia control of the chassis dynamometer system having a higher response than that of the third embodiment. Become.

Although the present invention has been described in detail only for the specific examples described above, it is obvious to those skilled in the art that various changes and modifications are possible within the scope of the technical idea of the present invention. Such variations and modifications are naturally within the scope of the claims.

DESCRIPTION OF SYMBOLS 10 ... Controller 11-14, 21-24 ... Weight function 15 ... 2 Inertial mechanical system model 15d, 15e, 15f, 16, 17, 54a, 54b ... Transfer function 15a, 18, 19 ... Adder 15b, 15c, 20, 54c, 55 ... subtractor 50 ... roller angular velocity estimation circuit 51 ... ASR
52 ... Two-inertia machine configuration 53 ... ATR
54 ... Roller surface driving force estimation circuit 56 ... EIC
101 ... Dynamometer 102 ... Speed detector 103 ... Shaft 104 ... Shaft torque meter 105 ... Roller 106 ... Controller 107 ... Inverter

Claims (3)

A roller, a dynamometer that rotates the roller by rotating a common shaft with the roller, a shaft torque detector that detects torque of the shaft, a dynamometer angular velocity detector that detects an angular velocity of the dynamometer, and the power In a chassis dynamometer system equipped with a control means for controlling an inverter that supplies power to the meter,
The control means includes
A roller angular velocity estimation circuit that generates a roller angular velocity estimation signal that is an estimated angular velocity of the roller based on a shaft torque detection signal detected by the shaft torque detector and a dynamometer angular velocity detection signal detected by the dynamometer angular velocity detector ;
And a speed control circuit for generating a dynamometer torque command is the roller angular velocity estimation signal and said roller torque to be output by the power meter performs difference to the PID control of roller angular speed command is the angular velocity to be reached ,
Weighting a mechanical system model of the chassis dynamometer, a first weight function that weights a signal corresponding to the surface driving force of the roller, and a signal corresponding to a dynamometer torque command that is a torque to be output by the dynamometer A second weighting function, a third weighting function that weights a signal corresponding to the shaft torque observation noise that is the observation noise of the shaft torque detector, and a dynamometer observation noise that is the observation noise of the dynamometer angular velocity detector And a signal corresponding to the angular velocity of the roller generated based on the signals weighted by the first and second weight functions, and the first to fourth signals. A fifth weighting function for weighting a difference signal from the estimated angular velocity of the roller generated based on each weighted signal of the weighting function, and the first and second weighting functions. Is generated based on each signal weighted by a sixth weighting function weighted to a signal corresponding to the torsional torque of the shaft generated based on each weighted signal, and the first and second weighting functions Using a seventh weighting function that weights a signal corresponding to the angular velocity of the dynamometer and an eighth weighting function that weights a signal corresponding to the angular velocity of the roller ,
The chassis dynamometer mechanical system model includes a first transfer function having a characteristic of the moment of inertia of the roller, a second transfer function having a characteristic of the spring stiffness of the shaft, and a characteristic of the moment of inertia of the dynamometer. A signal corresponding to the angular velocity of the roller, the first to third transfer functions being input with signals based on the signals weighted by the first and second weight functions. Is generated using the first transfer function and the second transfer function, and signals corresponding to the torsional torque of the shaft are the first transfer function, the second transfer function, and the third transfer function. A signal generated using a transfer function and corresponding to an angular velocity of the dynamometer is generated using the second transfer function and the third transfer function;
A signal corresponding to a surface driving force of the roller, a signal corresponding to a dynamometer torque command which is a torque to be output by the dynamometer, a signal corresponding to a shaft torque observation noise which is an observation noise of the shaft torque detector, and In the generalized plant in which the signal corresponding to the dynamometer observation noise which is the observation noise of the dynamometer angular velocity detector is a disturbance input, and each weighted output of the fifth to eighth weight functions is a controlled variable,
Using the first to eighth weight functions, the coefficient of the state equation of the controller is obtained by a control system design method called H∞ control or μ design method, so that a shaft torque detection signal and a dynamometer angular velocity detection signal are obtained. A roller angular velocity estimator for estimating the roller angular velocity based on the roller angular velocity estimator, and applying the designed roller angular velocity estimator to the roller angular velocity estimation circuit;
The controller includes a signal corresponding to the shaft torque detection signal obtained by adding a signal corresponding to the torsion torque of the shaft to the third weight function output, and an angular velocity of the dynamometer as the fourth weight function output. And a signal corresponding to the dynamometer angular velocity detection signal obtained by adding a signal corresponding to
The control amount that is the fifth weight function output is a deviation between a signal corresponding to the angular velocity of the roller and an estimated angular velocity signal of the roller output from the controller,
The chassis dynamometer system, wherein the roller angular velocity estimation circuit outputs an estimated angular velocity of the roller output from the controller as a roller angular velocity estimation signal . The control means includes a shaft torque control circuit having a resonance suppression characteristic in which a signal generated by performing PID control on a difference between the signal generated by the speed control circuit and the shaft torque detection signal is used as a dynamometer torque command. The chassis dynamometer system according to claim 1. The control means subtracts the shaft torque detection signal from the generated roller angular velocity estimation signal to generate a roller surface driving force estimation circuit that is an estimated surface driving force of the roller;
2. An inertia setting circuit that generates a roller angular velocity command based on a difference signal between the roller surface driving force estimation signal and a running resistance command that is a command value of a running resistance of a vehicle. 2. The chassis dynamometer system according to 2.

Images