JP5814024B2 - Vehicle vibration reduction system and vehicle vibration reduction method - Google Patents

Description

  The present invention relates to a vehicle vibration reduction system and a vehicle vibration reduction method by adaptive control.

  Conventionally, an engine damping system has been studied in order to reduce transmission of vibrations in a predetermined direction to various parts of a vehicle among vibrations generated in an engine mounted on the vehicle. The engine damping system includes a sensor for detecting vibration attached to the engine, and an active vibration isolator (active control mount, hereinafter referred to as ACM) for supporting and damping the engine, and a signal from the sensor. The vibration control force of the ACM is controlled based on the above.

  On the other hand, adaptive control technology using an adaptive digital filter is widely used for the purpose of reducing sound and vibration, and many methods such as Filtered-X LMS (Least Mean Square) have been proposed. Furthermore, various structures such as an FIR (Finite Impulse Response) filter and a SAN (Single-frequency Adaptive Notch) filter have been proposed as the structure of the adaptive filter.

  In these adaptive controls, the adaptive filter is updated according to a predetermined law so as to minimize the signal for detecting the sound and vibration to be controlled, thereby reducing the sound and vibration. And while adaptive control is functioning well, sound and vibration become minimum values when the adaptive filter converges to the optimal solution. The filter update coefficient weighted by the filter update amount at the time of update is an important parameter related to the speed at which the filter converges to the optimal solution, that is, control followability and control stability. However, the optimum filter update coefficient varies depending on the amplitude characteristic with respect to the frequency of sound and vibration and the structure of the filter. For this reason, the filter update coefficient is determined in a format such as a reference table, for example, based on data obtained by actually performing measurement on amplitude characteristics of various values.

  Conventionally, in adaptive control, a certain amount of convergence time has been required for the adaptive filter to converge to an optimal solution. For example, when an adaptive filter is used to reduce vibrations generated in a passenger car, this convergence time has caused a sense of discomfort for the passenger to ride. Therefore, in order to shorten the convergence time and ensure the stability of the control, for example, the sampling frequency and the filter learning coefficient are changed according to the engine speed that is the source of vibration and sound. Various update algorithms have been considered.

As such an update algorithm, for example, in adaptive control using LMS (Least Mean Square), when the transfer characteristic C is set from the actuator to the control point, and the adaptive filter is configured by an FIR type digital filter with i taps, The update formula of the adaptive filter is described as Equation (1).
w (n + 1) = w (n) + 2μ · e (n) · r (n−i) (1)

  Here, e is a signal to be controlled at the control point, and r is a signal obtained by convolution of a reference signal created by detecting a signal related to the generation state of the target signal and the transfer characteristic C. For example, in engine vibration control, r is a signal obtained by convolving a sine wave or cosine wave having a control frequency calculated from the rotation pulse of the crankshaft of the engine with the transfer characteristic C. Further, μ is an update coefficient of the filter, and is a parameter that contributes to the convergence speed to the optimum value of the filter and the stability of control. If the update coefficient μ is increased, the convergence performance is improved, but unstable control is likely to occur. If the update coefficient μ is decreased, more stable control is realized, but the convergence time is increased.

  Since the gain characteristics of sound and vibration usually have frequency characteristics, it is considered desirable to have an update coefficient corresponding to the frequency in order to achieve both control stability and high-speed tracking. . For example, when adaptive control is applied to active control of engine vibration, the engine shows various vibration levels at various rotational speeds depending on the specifications, so there is an optimum value for the update coefficient corresponding to the engine rotational speed. Is easily imagined. Further, since the convergence speed of the filter is important when the engine is accelerated, it is desirable to make the value of the update coefficient μ variable depending on the degree of acceleration. In view of this, a configuration has been proposed in which the relationship between the engine speed, the amount of change thereof, and the convergence coefficient α is stored as a table, and the convergence coefficient α is thereby changed (see, for example, Patent Document 1). Here, the convergence coefficient α corresponds to the update coefficient μ. That is, conventionally, adaptive control has been performed on the basis of the update coefficient μ that is predetermined and stored in accordance with the amount of change in the rotational speed of the engine, as in the table described in Patent Document 1. That is, according to the configuration according to Patent Document 1, for example, when the amount of change in the rotational speed of the engine increases, the update coefficient μ is set to a large value so that the followability of convergence of the filter coefficient is improved, Control can be performed.

JP-A-5-61485

  In a vehicle such as a passenger car, the frequency characteristic of vibration changes from a predetermined value in accordance with deterioration of the vehicle, temperature, and the like. Accordingly, there has been a demand for appropriate control corresponding to vibration generated in an actual vehicle, instead of control using an update coefficient determined according to a frequency characteristic assumed in advance.

  However, as described in Patent Document 1, when the update coefficient μ is changed based on a table having a predetermined value, the update coefficient μ is set according to a predetermined table. For this reason, for example, when the frequency characteristic of vibration generated at a predetermined position in the vehicle is changed due to deterioration of the engine support device or the like, appropriate control using an update coefficient optimized for the changed frequency characteristic is performed. Could not do.

  Therefore, the present invention dynamically optimizes the update coefficient of the adaptive filter corresponding to the frequency characteristics of vibration in adaptive control, and can stably compensate for vibrations occurring at a predetermined position in the vehicle. An object is to provide a vibration reduction system and a vehicle vibration reduction method.

In order to achieve the above object, the first invention provides:
At least one excitation device for generating an excitation force;
A device for detecting vibration at a predetermined position in the vehicle;
A device for measuring the rotational speed of the engine;
A vehicle vibration reduction system comprising: a control device that performs adaptive control by changing an update coefficient of the adaptive filter;
The control device generates a control signal based on a rotation speed signal from a device that measures the rotation speed of the engine, a signal obtained from a device that detects the vibration, and the update coefficient. A control signal generation unit that updates the update coefficient based on fluctuations of the control signal or a signal determined by the control signal, and the at least one excitation is generated by the control signal. Controlling the apparatus to generate a vibration force to reduce vibration generated at a predetermined position in the vehicle ,
The fluctuation of the control signal or a signal determined by the control signal defines a current value input to the vibration exciter, and the controller inputs the update coefficient to the exciter within a predetermined time. The current value is updated based on a difference between the maximum amplitude value and the minimum amplitude value of the amplitude value of the current value .

A second invention is the vehicle vibration reduction system according to the first invention.
The control device further updates the update coefficient based on an average value or a mean square value of amplitude values of the current values input to the vibration generator within the predetermined time. is there.
A third invention is the vehicle vibration reduction system according to the first invention,
The control device uses the update coefficient as an average of each peak value of a first evaluation target value based on a current value input to an active control mount that supports the engine and a positive current amplitude value output from the control device. It is characterized by determining together with the 2nd evaluation object value which is a value.

The fourth invention is:
At least one excitation device for generating an excitation force;
A device for detecting vibration at a predetermined position in the vehicle;
A device for measuring the rotational speed of the engine;
A vehicle vibration reduction method in a system including a control device that performs adaptive control by changing an update coefficient of an adaptive filter,
Generating a reference signal having a signal frequency that is a real number multiple of the frequency of the rotational speed signal based on the rotational speed signal from the apparatus for measuring the rotational speed of the engine; A step of generating a control signal based on the reference signal and the update coefficient, and generating a vibration force in the at least one vibration device by the control signal, thereby reducing a vibration generated at a predetermined position in the vehicle. And a step of updating the update coefficient based on a fluctuation of the control signal or a signal determined by the control signal.

  According to the present invention, in the adaptive control, the update coefficient of the adaptive filter can be dynamically optimized in accordance with the frequency characteristic of vibration, and vibration generated at a predetermined position in the vehicle can be compensated stably.

1 is a schematic diagram illustrating an example of a vehicle vibration reduction system according to the present invention. It is a block diagram which shows an example of the basic control system of the vibration reduction system of the vehicle shown in FIG. It is a flowchart which shows an example of the setting method of the update coefficient of the filter in the vibration reduction system of the vehicle which concerns on this invention. It is a figure for demonstrating the flowchart shown in FIG. It is a figure for demonstrating the evaluation object value in the flowchart shown in FIG. It is a figure which shows the change of the electric current value input into an oscillating device when the value of an update coefficient is made into various values.

Hereinafter, an embodiment of the present invention will be described with reference to the drawings.
FIG. 1 is a schematic diagram showing an example of a vehicle vibration reduction system according to the present invention. In FIG. 1, an engine 12 is supported by two active control mounts 3a and 3b (hereinafter referred to as ACM) incorporating electromagnetic actuators and a plurality of engine mounts (not shown). 3b has a function as a vibration device as well as a function of supporting the engine 12, and functions to actively generate a vibration force to reduce vibration at a predetermined position in the vehicle. The engine 12 is provided with a sensor 6 for measuring the engine speed. A sensor 9a that detects vibration is attached to the steering 4 of the vehicle 10, and a sensor 9b that detects vibration is attached to the floor of the driver's seat. In addition, a control device 11 that controls the damping force of the ACMs 3a and 3b by a control signal is disposed at the front position of the driver's seat (for example, in the instrument panel). The control device 11 performs adaptive control by changing the update coefficient of the adaptive filter.

  The sensor 9 that detects vibration functions to detect vibration (for example, acceleration) at a predetermined position in the vehicle in real time, and is located at a place where a passenger feels or a vibration source (a vehicle body, the vicinity of a vibration device). It is preferable to be disposed. In the above-described example, the steering wheel 4 and the driver's seat are disposed on the floor portion. As the sensor 9, for example, an acceleration sensor, a load sensor, or the like can be used. Further, the control device 11 may be arranged at any location in the vehicle.

  FIG. 2 is a block diagram showing an example of a basic control system of the vehicle vibration reduction system shown in FIG. The control device 11 includes a control signal generation unit configured by the adaptive filter 17 and an update coefficient update unit configured by the LMS calculation unit 18. In FIG. 2, C represents an actual transfer characteristic from the ACMs 3a and 3b to the sensor 9a, and C 'represents an estimated transfer characteristic (measured impulse response).

  The adaptive filter 17 constituting the control signal generation unit generates a control signal based on the rotational speed signal from the device that measures the rotational speed of the engine, the signal obtained from the device that detects vibration, and the update coefficient. . Specifically, the adaptive filter 17 adjusts the phase of the reference signal x (n) from the synthesizer 16 and the signal indicating the gain and outputs the control signal y (n) of the ACMs 3a and 3b. The synthesizer 16 generates a reference signal (basic frequency signal and its harmonic signal) x (n) from an engine speed signal (for example, a rotation pulse signal) output from the sensor 6 that measures the engine speed. The adaptive filter 17 reads the acceleration signal (error signal) e (n) detected by the sensor 9 as a gain. Furthermore, the adaptive filter 17 also uses the update coefficient of the adaptive filter 17 updated (set) by the method described with reference to FIG. 3 when outputting the control signal y (n).

  The LMS calculation unit 18 constituting the update coefficient update unit updates the above-described update coefficient based on a control signal or a signal variation determined by the control signal. Specifically, the update coefficient of the adaptive filter is determined by a method described later with reference to FIG. Here, the control signal corresponds to the voltage value applied by the control device 11, and the signal determined by the control signal corresponds to the current value that actually flows by this voltage value.

  An outline of the operation of the control device 11 including the functional unit as described above will be described below. First, when the control device 11 receives a rotation pulse signal, which is an engine speed signal, from the sensor 6 that measures the engine speed, the synthesizer 16 generates a signal frequency based on the rotation pulse signal. A reference signal x (n) that is a real number multiple of (fundamental frequency) is generated. And the control apparatus 11 controls the LMS calculating part 18, reads the reference signal x (n), reads the error signal e (n) from the sensor 9, and also determined by the method demonstrated with reference to FIG. Using the update coefficient of the adaptive filter 17, the filter coefficient of the adaptive filter 17 is updated using the LMS algorithm. The adaptive filter 17 outputs a control signal y (n) to the ACMs 3a and 3b based on the updated filter coefficient.

FIG. 3 is a flowchart showing a filter update coefficient setting method according to the present invention. Here, the first evaluation target value E ₁ described in detail with reference to FIG. 5 is used as the evaluation target value for determining whether or not the update coefficient μ is appropriate. The LMS calculation unit 18 sets the update coefficient μ ₀ (step S01). Further, the LMS calculation unit 18 calculates a first evaluation target value E ₁ to be described later with reference to FIG. 5 based on the current value input to the ACM 3a (step S02). And the LMS calculating part 18 compares the preset threshold value with the evaluation object value calculated by step S02 (step S03). In step S03, it is determined whether it is stable or unstable within a range allowed by the control using the currently set update coefficient. When the threshold value is equal to or greater than the first evaluation target value E _1, a value obtained by adding a predetermined amount of value Δμ preset for adjusting μ to the number of updates μ is set as a new update coefficient μ ₁ (step S04). ). The adaptive filter 17 is updated again with this new update coefficient μ ₁ .

Then, the LMS computing unit 18 calculates the first evaluation target value E ₁ again based on the current value input to the ACM 3a (step S05). Then, the LMS calculation unit 18 compares the preset threshold value again with the first evaluation target value E ₁ calculated in step S05 (step S06). As a result, when the threshold value is less than the first evaluation target value E ₁ , the LMS calculation unit 18 subtracts a predetermined amount of value Δμ from the update coefficient μ _{1 and} sets the value as a new update coefficient μ ₂ (step S07). ).

On the other hand, LMS arithmetic unit 18, if the threshold is determined to be the first smaller than the evaluation object value E ₁ in step S03, the update coefficient mu ₀ by subtracting a predetermined amount of values [Delta] [mu, the value new update coefficient mu ₁ (step S08). Then, LMS arithmetic unit 18, as in step S05, calculates a first evaluation object value E ₁ based on the current value inputted to ACM3a (step S09). Furthermore, LMS arithmetic unit 18, similarly to step S06, again, it compares the preset threshold, the first and the evaluation object value E ₁ calculated in step S09 (step S10).

FIG. 4 is a diagram for explaining the flowchart shown in FIG. At step S03 in FIG. 3, (1) if the threshold is determined to be equal to the first evaluation object value E ₁ or more, each time the processing from step S04~ step S06 is performed once, a black circle one step (one The value of the first evaluation target value E ₁ increases and approaches the threshold T. And it is determined to have reached the optimum value O ₁ at the stage before exceeding the threshold T. On the other hand, at step S03 of FIG. 3, (2) if the threshold is determined to be the first smaller than the evaluation object value E _1, the solid circle one step (one every time processing from step S08~S10 is performed once The value of the first evaluation target value E ₁ decreases and approaches the threshold value T.

FIG. 5 is a diagram for explaining the first evaluation target value E ₁ in the flowchart shown in FIG. 3 and the second evaluation target value E ₂ , which is an evaluation target value different from the first evaluation target value E ₁ . is there. The vertical axis in the figure is the amplitude value of the current value output from the control device 11 (hereinafter referred to as the current value amplitude value). The current value amplitude value is a value defined by the control signal y (n) output from the adaptive filter 17 and varies according to the variation of the control signal. The first evaluation target value E ₁ is obtained by subtracting the minimum value f _min of the current value amplitude value from the maximum value f _max of the current value amplitude value within a predetermined time (for example, 100 milliseconds, 200 milliseconds, 300 milliseconds). Which is given by the formula (f _max −f _min ).

In order to stably control the vibration generated in the vehicle, it is ideal that the current value amplitude value is constant. Therefore, the first evaluation object value E ₁ is preferably as small as possible. However, in reality, the value of the first evaluation object value E ₁ is a case similar small, both of the maximum value f _max and the minimum value f _min of the current amplitude value is extremely large difference When it is small (when it is controlled more than necessary or when control is oscillating) and when both the maximum value f _max and the minimum value f _min of the current amplitude value are extremely small (the difference is small) Since the value of the update coefficient is too small, the effect of reducing the vibration generated in the vehicle differs from the case where the effect of reducing the control is small although it is stable. In view of this, when it is determined to be stable, that is, when the threshold value is equal to or greater than the first evaluation target value E ₁ , by increasing the value of the update coefficient μ within the allowable stable region, Realize stable and high-speed control. On the other hand, when the threshold value is less than the first evaluation target value E ₁ , it exceeds the allowable stable region, so the value of the update coefficient μ is decreased to stabilize the control.

Furthermore, in order to improve the suppression effect of the vibration generated in the vehicle, a method of determining the update coefficient μ using the second evaluation target value E ₂ in addition to the first evaluation target value E ₁ can be considered. Second evaluation object value E ₂ is the average value of peak values of the positive current amplitude shown in FIG. 5 (convex value). Here, instead of the average value of each peak value (convex value) of the positive current amplitude value, the second evaluation object is obtained by calculating the average value (square mean value) of each peak value (convex value) after squaring the current waveform. it may be used as the value _{E 2.} By evaluating the second evaluation object value E _2, the case described above, both values is large difference between the maximum value f _max and the minimum value f _min of the current amplitude value is small, the maximum current value amplitude value It is possible to determine when both the value f _max and the minimum value f _min are small and the difference is small. Evaluation method, for example, as in the first evaluation object value E _1, may be compared with a predetermined threshold value. In this case, in the flowchart of FIG. 3, the process from step S04 to step S06 or the process from step S08 to step S10 is executed once, and the second evaluation target value E ₂ is evaluated before entering the iterative process. Based on the result, the value of Δμ in step S04 or step S08 during the iterative process may be changed.

In FIG. 6, when the value of the update coefficient μ is set to various values (μ _a = 0.01, μ _a = 0.1, μ _c = 1, μ _d = 10), it is input to the vibration device. It is a figure which shows the change of the value (current value index | exponent) which indexed the current value. Here, the values of μ _{a to} μ _d are not the values actually used, but the values actually used are values obtained by multiplying the values of μ _{a to} μ _d by 0.001. As shown in FIGS. 6A to 6D, the current value index changes greatly according to the value of the update coefficient μ. As described above, the case illustrated in FIG. _6A corresponds to the case where both the maximum value f _max and the minimum value f _min of the current value amplitude value are small and the difference is small as described with reference to FIG. To do. The case shown in FIG. 6B corresponds to the case where both the maximum value f _max and the minimum value f _min of the current value amplitude value are large and the difference is small. 6 (a) and 6 (b), the control by the control device 11 is stable, FIG. 6 (c) is somewhat stable, and in the case of FIG. 6 (d), the control is unstable. Furthermore, towards the control of applying the update coefficient mu _b showing the change in the current value index as in FIG. 6 (b), applying the update coefficient mu _a showing the change in the current value index as shown in FIG. 6 (a) The vibration reduction effect is higher than that of the control.

Furthermore, control is in the applied control update coefficient mu _c showing the change in the current value index as, according to the update coefficient mu _a showing the change in the current value index as shown in FIG. 6 (a) FIG. 6 (c) In some cases, the vibration reduction effect is higher, but the stability of the control is inferior when compared with the control to which the update coefficient μ _b indicating the change in the current value index as shown in FIG. _6B is applied.
Further, in the control using the update coefficient μ _c indicating the change in the current value index as shown in FIG. 6D, the control is unstable and the vibration reduction effect is also shown in FIGS. 6A to 6C. It is inferior to the other controls shown in. As described above, the stability of the control and the vibration reduction effect are greatly different depending on the value of the update coefficient μ.
Therefore, the stability of vibration control and the vibration reduction effect can be optimized by updating the value of the update coefficient μ to the optimum value in a timely manner by the method described with reference to FIGS. Can do.

  Thus, according to the vibration reduction system according to the present embodiment, the control device 11 that performs adaptive control by changing the update coefficient of the adaptive filter allows the current input to the ACMs 3a and 3b that constitute the vibration generator. Since the update coefficient is determined based on the amplitude value, the update coefficient of the adaptive filter can be automatically optimized to control the vibration generated at a predetermined position of the vehicle. For this reason, even if the vibration generated at a predetermined position of the vehicle changes from the mode assumed at the time of design according to the temperature of the vehicle or aging deterioration, the update coefficient of the adaptive filter is dynamically optimized and the vibration is appropriately performed. Can be controlled.

The present invention is not limited to the above-described embodiment, and many variations or modifications are possible. For example, in FIG. 3, the LMS calculation unit 18 calculates the first evaluation target value E ₁ based on the current value input to the ACM 3 a that is the front-side ACM, and updates the value of the update coefficient μ. However, also for the current value input to the ACM 3b which is the rear ACM, it is possible to calculate the first evaluation target value E ₁ and update the update coefficient μ for the ACM 3b to perform vibration control. Thereby, it is possible to perform vibration control using the optimum update coefficient μ for each of the front-side ACM 3a and the rear-side ACM 3b connected to the engine 12.

  For example, in the above-described embodiment, one or two ACMs are provided as a vibration device for applying a vibration force, but three or more ACMs may be provided. The vibration device is not limited to the ACM, but may be an active mass damper or a torque rod type. The place where the vibration device is provided is not limited to the lower part of the engine, but may be provided between the suspension device and the vehicle body. By providing the vibration device between the suspension device and the vehicle body, the vibration of the vehicle body can be reduced.

  In the above embodiment, the filter coefficient update processing of the adaptive filter is performed using the LMS algorithm. For the filter coefficient update processing, a complex LMS algorithm (Complex Least Mean Square Algorithm), a Normalized LMS algorithm (Normalized Least algorithm) is used. Mean Square Algorithm), Projection Algorithm, SHARF Algorithm (Simple Hyperstable Adaptive Recursive Filter Algorithm), RLS Algorithm (Recursive Least Square Algorithm), FLMS Algorithm (Fast Least Mean Square Algorithm), and Adaptive Filter Using DCT (Adaptive) It can be said that various algorithms such as Filter using Discrete Cosine Transform, SAN filter (Single Frequency Adaptive Notch Filter), neural network, genetic algorithm, etc. can be used. Needless to say.

3a, 3b Active control mount (ACM)
4 Steering 5 Engine mount 9, 9a, 9b Sensor 10 Vehicle 11 Control device 12 Engine 13 Seat 14 Car body 16 Synthesizer 17 Adaptive filter 18 LMS calculation unit

Claims (4)

At least one excitation device for generating an excitation force;
A device for detecting vibration at a predetermined position in the vehicle;
A device for measuring the rotational speed of the engine;
A vehicle vibration reduction system comprising: a control device that performs adaptive control by changing an update coefficient of the adaptive filter;
The control device generates a control signal based on a rotational speed signal from a device that measures the rotational speed of the engine, a signal obtained from the device that detects the vibration, and the update coefficient. And an update coefficient update unit that updates the update coefficient based on fluctuations in the control signal or a signal determined by the control signal, and the excitation force is applied to the at least one excitation device by the control signal. Control to reduce vibration generated at a predetermined position in the vehicle ,
The fluctuation of the control signal or a signal determined by the control signal defines a current value input to the vibration exciter, and the controller inputs the update coefficient to the exciter within a predetermined time. The vehicle vibration reduction system is updated based on a difference between a maximum amplitude value and a minimum amplitude value of the amplitude values of the current values . The control device further updates the update coefficient based on an average value or a mean square value of amplitude values of the current values input to the excitation device within the predetermined time. Item 4. The vehicle vibration reduction system according to Item 1 . The control device uses the update coefficient as an average of each peak value of a first evaluation target value based on a current value input to an active control mount that supports the engine and a positive current amplitude value output from the control device. The vehicle vibration reduction system according to claim 1, wherein the vehicle vibration reduction system is determined in combination with a second evaluation target value that is a value. At least one excitation device for generating an excitation force;
A device for detecting vibration at a predetermined position in the vehicle;
A device for measuring the rotational speed of the engine;
A vehicle vibration reduction method in a system including a control device that performs adaptive control by changing an update coefficient of an adaptive filter,
Generating a reference signal having a signal frequency that is a real number multiple of the frequency of the rotational speed signal based on the rotational speed signal from the apparatus for measuring the rotational speed of the engine; and a signal obtained from the apparatus for detecting the vibration, A step of generating a control signal based on the reference signal and the update coefficient, and generating a vibration force in the at least one vibration device by the control signal, thereby reducing a vibration generated at a predetermined position in the vehicle. And a step of updating the update coefficient based on a fluctuation of the control signal or a signal determined by the control signal.

Images