JP7131192B2 - Device for monitoring the state of a running vehicle - Google Patents

Description

TECHNICAL FIELD The present invention relates to a monitoring device, program, and method for monitoring the state of a running vehicle.

In order to run a vehicle comfortably, it is important that the tire pressure is adjusted. This is because if the air pressure falls below the proper value, the durability, wear resistance, steering stability, etc. of the tire may deteriorate, hydroplaning may easily occur, and fuel consumption may deteriorate. Conventionally, a system for automatically detecting tire pressure (Tire Pressure Monitoring System; TPMS) has been studied. Information that the tire is decompressed can be used, for example, to warn the driver.

Methods for detecting tire pressure include methods that directly measure tire pressure, such as by attaching a pressure sensor to the tire, and methods that indirectly evaluate tire pressure using other index values. There is As one of such indirect evaluation methods, a resonance frequency method (RFM) is known. RFM utilizes the change in tire resonance frequency characteristics included in the wheel speed signal due to pressure reduction.

However, due to limited computer resources onboard, it is sometimes difficult to analyze changes in tire resonance frequency characteristics using fast Fourier transform (FFT). For this reason, in the conventional method, as shown in Patent Document 1, for example, the resonance frequency is estimated without conversion to the frequency domain. Specifically, in Patent Literature 1, a wheel speed signal acquired as time-series data is subjected to time-series analysis based on a K-order autoregressive (AR) model. Specifically, parameters _θ =[a ₁ , a ₂ , .

where y(t) is the wheel speed at time t, ε is white noise, and K is the order of the model. The resonance frequency is estimated as the frequency corresponding to the poles of the transfer function representing the AR model.

JP 2009-274639 A

However, if there are noise peaks within the bandwidth to which the AR model is applied that are different from the tire resonance peak, the resonance frequency derived based on the AR model will be dragged by such noise peaks, resulting in the true tire resonance frequency. It deviates from the resonance frequency. Therefore, in order to avoid noise peaks, it is conceivable to set a narrow bandwidth to which the AR model is applied. However, this means that the search range for the resonance frequency is narrowed, and as a result, the resonance frequency may not be calculated correctly. (magnitude of change in resonance frequency) is reduced. Therefore, it is desirable to cancel the influence of noise while maintaining the search range of the resonance frequency so that the resonance frequency can be accurately calculated. It should be noted that this is not limited to cases where tire pressure reduction is detected, but it is required to monitor the state of the vehicle while it is running based on the resonance frequency of the sensing data regarding the vibration state of the vehicle that is acquired while the vehicle is running. It can apply to all situations. Examples of such situations other than tire pressure detection are situations such as monitoring gains/frequencies of sprung and unsprung resonances related to ride comfort, and vehicle vibration gains/frequencies such as shimmy and shake. , the situation of judging the state of the road surface on which the vehicle travels, the state of tire wear, and the like are assumed.

The present invention provides a monitoring device, method, and program capable of accurately calculating the resonance frequency by canceling the influence of noise from sensing data regarding the vibration state of a running vehicle while maintaining the search range of the resonance frequency. intended to provide

A monitoring device according to a first aspect of the present invention is a monitoring device for monitoring the state of a running vehicle, wherein time-series sensing data relating to the vibration state of the running vehicle is passed through a fixed band-pass filter. a first filtering unit for extracting a first frequency band component of the sensing data; and passing the first frequency band component through a variable bandpass filter or a variable bandstop filter movable within the first frequency band. By doing so, a second filtering unit that extracts or removes components of a second frequency band having a narrower bandwidth than the first frequency band, a value that specifies the second frequency band, and at least the components of the second frequency band and a peak determination unit that determines a resonance frequency of the sensing data based on one.

A monitoring device according to a second aspect of the present invention is the monitoring device according to the first aspect, wherein the sensing data is rotation information of wheels mounted on the vehicle.

A monitoring device according to a third aspect of the present invention is the monitoring device according to the second aspect, wherein the second frequency band has a bandwidth of 4 Hz to 12 Hz.

A monitoring device according to a fourth aspect of the present invention is the monitoring device according to the second aspect or the third aspect, wherein the bandwidth of the first frequency band is 4 Hz to 10 Hz than the bandwidth of the second frequency band wide.

A monitoring device according to a fifth aspect of the present invention is the monitoring device according to any one of the first aspect to the fourth aspect, wherein the peak determination unit determines the resonance frequency based on a value specifying the second frequency band. A frequency is determined as the center frequency of the second frequency band.

A monitoring device according to a sixth aspect of the present invention is the monitoring device according to any one of the first aspect to the fourth aspect, wherein the peak determination unit includes: The resonance frequency is determined by time-series analysis of the components based on an autoregressive model.

A monitoring device according to a seventh aspect of the present invention is the monitoring device according to any one of the first aspect to the fourth aspect, wherein the second filtering unit filters the components of the first frequency band to the variable bandstop Filtering removes the components of the second frequency band. The peak determination unit defines a bandpass filter that passes a signal in a band including at least the center frequency of the second frequency band and having a narrower bandwidth than the first frequency band, and the components of the first frequency band are defined as: The resonance frequency is determined by passing through the defined band-pass filter and performing a time-series analysis of the components passed through the defined band-pass filter based on an autoregressive model.

A monitoring device according to an eighth aspect of the present invention is the monitoring device according to any one of the first aspect to the seventh aspect, wherein the pressure of a tire mounted on the vehicle is reduced based on the determined resonance frequency. It further includes a reduced pressure detection unit that detects the

A monitoring program according to a ninth aspect of the present invention is a monitoring program for monitoring the state of a running vehicle, and causes a computer to execute the following.
(1) extracting a component of a first frequency band of the sensing data by passing time-series sensing data relating to the vibration state of the vehicle during running through a fixed band-pass filter; Extracting or removing components in a second frequency band narrower than the first frequency band by passing the components through a variable bandpass filter or a variable bandstop filter that is movable within the first frequency band. (3) determining the resonance frequency of the sensing data based on at least one of a value specifying the second frequency band and a component of the second frequency band;

A monitoring method according to a tenth aspect of the present invention is a monitoring method for monitoring the state of a running vehicle, and includes the following.
(1) extracting a component of a first frequency band of the sensing data by passing time-series sensing data relating to the vibration state of the vehicle during running through a fixed band-pass filter; Extracting or removing components in a second frequency band narrower than the first frequency band by passing the components through a variable bandpass filter or a variable bandstop filter that is movable within the first frequency band. (3) determining the resonance frequency of the sensing data based on at least one of a value specifying the second frequency band and a component of the second frequency band;

According to the present invention, first, sensing data relating to the vibration state of a running vehicle is narrowed down to components in the first frequency band, which is the resonance frequency search range, using a fixed band-pass filter. Next, apply a variable bandpass filter or a variable bandstop filter to the same component, which is movable within the first frequency band, to further narrow the first frequency band to a second frequency band that is narrower than this and contains the resonance frequency. . Therefore, the range (second frequency band) in which the true peak of the resonance frequency exists can be narrowed down while maintaining the search range (first frequency band) of the resonance frequency. As a result, the resonance frequency can be accurately calculated by canceling the influence of noise from the sensing data.

The schematic diagram which shows a mode that the monitoring apparatus which concerns on one Embodiment of this invention was mounted in the vehicle. FIG. 2 is a block diagram showing the electrical configuration of the monitoring device; 4 is a flowchart showing the flow of pressure reduction detection processing; 7 is a graph of resonance frequency values according to a comparative example; 4 is a graph of resonance frequency values according to Example 1. FIG. 7 is a graph of resonance frequency values according to Example 2. FIG.

Hereinafter, a monitoring device, a program, and a method for monitoring the state of a running vehicle according to one embodiment of the present invention will be described with reference to the drawings.

<1. Configuration of monitoring device>
FIG. 1 is a schematic diagram showing how a monitoring device 2 is mounted on a vehicle 1. As shown in FIG. The monitoring device 2 is implemented as a control unit mounted on the vehicle 1 and monitors the state of the vehicle 1 during travel. The vehicle 1 is a four-wheeled vehicle and includes a front left tire FL, a front right tire FR, a rear left tire RL and a rear right tire RR. The monitoring device 2 has a function of detecting decompression of these tires FL, FR, RL, RR. An alarm to that effect is issued via the display device 3 . The details of the flow of such pressure reduction detection processing will be described later.

Decompression of tires FL, FR, RL, and RR is detected based on wheel rotation information. A wheel speed sensor 6 is attached to each of the tires FL, FR, RL, RR (more precisely, to the wheels to which the tires FL, FR, RL, RR are attached). Each of the wheel speed sensors 6 receives wheel speed signals, i.e., signals representing the rotational speeds of the tires FL, FR, RL, and RR (hereinafter referred to as sensing data) as rotation information of the wheels mounted on the vehicle 1 while the vehicle 1 is running. ) is detected at a predetermined sampling period. In this embodiment, the sensing data sampling period is 4 milliseconds (250 Hz in terms of sampling frequency). The wheel speed sensor 6 is connected to the monitoring device 2 via a communication line 5a, and the sensing data sensed by the wheel speed sensor 6 from time to time is transmitted to the monitoring device 2 in real time via the communication line 5a. . The time-series sensing data acquired in this manner represents the vibration state of the vehicle 1 during travel.

As the wheel speed sensor 6, any sensor can be used as long as it can detect the rotational speed of the running tires FL, FR, RL, and RR. For example, a sensor that measures the rotation speed from the output signal of an electromagnetic pickup can be used, or a sensor that generates electricity using rotation like a dynamo and measures the rotation speed from the voltage at this time can be used. can also be used. The mounting position of the wheel speed sensor 6 is also not particularly limited, and can be appropriately selected according to the type of sensor as long as the rotational speed can be detected.

FIG. 2 is a block diagram showing the electrical configuration of the monitoring device 2. As shown in FIG. The monitoring device 2 is a control unit mounted on the vehicle 1 as hardware, and as shown in FIG. 15. The I/O interface 11 is a communication device that realizes communication with external devices such as the wheel speed sensor 6 and the alarm indicator 3 . A program 7 for controlling the operation of each part of the vehicle 1 is stored in the ROM 13 . The CPU 12 virtually operates as a first filtering section 21 , a second filtering section 22 , a peak determining section 23 and a decompression detecting section 24 by reading and executing the program 7 from the ROM 13 . Details of the operations of the units 21 to 24 will be described later. The storage device 15 is composed of a hard disk, a flash memory, or the like. Note that the storage location of the program 7 may be the storage device 15 instead of the ROM 13 . The RAM 14 and the storage device 15 are appropriately used for calculation of the CPU 12 .

The alarm indicator 3 can be implemented in any form, such as a liquid crystal display element or a liquid crystal monitor, as long as it can inform the user that decompression is occurring. For example, the alarm indicator 3 can be configured by arranging four lamps respectively corresponding to the four-wheel tires FL, FR, RL, and RR in accordance with the actual arrangement of the tires. Alternatively, only one lamp may be provided, and it may be configured to light when decompression is detected in any tire. The installation position of the alarm indicator 3 can also be selected as appropriate, but it is preferable to install it in a position that is easy for the driver to understand, such as on the instrument panel. When the control unit (monitoring device 2) is connected to a car navigation system, it is possible to use a monitor for car navigation as the alarm indicator 3. When a monitor is used as the alarm indicator 3, the alarm can be an icon or character information displayed on the monitor.

<2. Decompression detection process>
Hereinafter, the decompression detection process for detecting decompression of the tires FL, FR, RL and RR will be described with reference to FIG. The pressure reduction detection process shown in FIG. 3 is started, for example, when the vehicle 1 starts running, and ends when the running stops. This process is a pressure reduction detection process based on the resonance frequency method (RFM), and RFM utilizes the fact that the resonance frequency f _R of the tire changes when the air pressure of the tire changes. Therefore, in this process, as will be described later, the resonance frequency f _R of the tire is calculated, and tire decompression detection is performed based on this. In the following description, for the sake of simplification, a description will be given of how the pressure reduction of the left front tire FL is detected, but the same processing is performed for the remaining tires FR, RL, and RR.

First, in step S1, the first filtering unit 21 acquires sensing data representing the rotational speed of the tire FL, which is transmitted from the wheel speed sensor 6 every moment. The sensing data acquired here is received via the I/O interface 11 .

Step S1 is repeated until a certain amount or more of sensing data representing the rotational speed is accumulated, and then the process proceeds to step S3. In step S3, the first filtering unit 21 removes the DC component (0 Hz component) by passing the accumulated time-series sensing data on the rotation speed through a high-pass filter. Step S3 serves as pre-processing for active noise control in subsequent step S5.

As indicated by the dotted line in FIG. 3, in step S2 after step S1, the sensing data representing the rotational speed acquired in step S1 may be converted into sensing data representing rotational acceleration. Rotational acceleration is obtained by differentiating the rotational speed. In this case, in subsequent processing, sensing data representing rotational acceleration is used instead of sensing data representing rotational speed. Also, in this case, step S3 can be omitted. This is because the rotational acceleration is obtained by differentiating the rotational speed with respect to time, and the DC component has already been removed.

In subsequent step S4, the first filtering unit 21 down-samples (decimates) the time-series sensing data that has passed through step S3. In this embodiment, as described above, the sampling frequency of the wheel speed sensor 6 is 250 Hz, and the Nyquist frequency is half that, 125 Hz. Therefore, the effective band before downsampling is 0 to 125 Hz. In this embodiment, the number of samples is thinned out by 0.5 times. Therefore, the time-series sensing data that has passed through step S3 is passed through a low-pass filter that removes frequencies above 62.5 Hz to derive sensing data with an effective band of 0 to 62.5 Hz.

In subsequent step S5, the first filtering unit 21 applies active noise control (ANC) to the time-series sensing data that has passed through step S4. As a result, periodic noise can be removed from the sensing data as much as possible, and the accuracy of the resonance frequency f _R of the tire FL determined in step S8 can be improved.

As a reference signal for ANC, past sensing data representing the rotational speed of the tire for which the resonance frequency f _R is to be calculated (hereinafter referred to as self-tire) can be used. Alternatively, as a reference signal for ANC, it is possible to use sensing data of a tire that is coaxial with the own tire and on the opposite side (for example, if the front left tire FL is the front right tire FR). In this case, a combination of the current sensing data and the most recent predetermined amount of past sensing data can be used. Alternatively, as a reference signal for ANC, it is possible to use sensing data of a tire on the same side as the own tire but on the opposite side (for example, in the case of the left front tire FL, the left rear tire RL). It should be noted that when noise in the sensing data of the rear tires is eliminated by referring to the sensing data of the front tires, the past sensing data of the front tires is referred to. On the other hand, when noise in the sensing data of the front tires is eliminated by referring to the sensing data of the rear tires, the future sensing data of the rear tires is referred to. This is easy to understand when considering the scene of eliminating the noise that occurs when going over a seam on the road surface, because the front tires first go over a specific seam, and then the rear tires go over it. Alternatively, noise can be removed more effectively by applying multiple ANC that appropriately combines ANCs based on these multiple reference signals. ANC can mainly remove periodic noise such as engine noise.

In subsequent step S6, the first filtering unit 21 extracts the components of the first frequency band by passing the sensing data that has passed through step S5 through a fixed bandpass filter. The first frequency band is predetermined, and is preferably set so as to include a change range of the resonance frequency f _R that changes due to pressure reduction and pressure increase of the tire FL. From this point of view, the first frequency band in this embodiment is a band around the torsional resonance frequency of the tire FL (typically around 30 Hz to 50 Hz). From the same point of view, the bandwidth of the first frequency band is preferably 8 Hz to 22 Hz, more preferably 10 Hz to 20 Hz.

In subsequent step S7, the second filtering unit 22 extracts the second frequency band component by passing the sensing data that has passed through step S6, ie, the first frequency band component, through a variable bandpass filter. In this embodiment, the bandwidth of the second frequency band is narrower than that of the first frequency band and is predetermined. The bandwidth of the second frequency band is preferably between 4 Hz and 12 Hz. Also, the bandwidth of the first frequency band is preferably 4 Hz to 10 Hz wider than the bandwidth of the second frequency band.

The variable bandpass filter used in step S7 is a type of bandpass filter, and adaptively changes the center frequency of the filter so that the signal gain (spectrum area) passing through the filter is maximized. is a filter. As an adaptation algorithm for this filter, for example, the LMS algorithm can be used. The second filtering unit 22 moves the variable band-pass filter within the first frequency band to determine the second frequency band that maximizes the signal gain of the sensing data passing through the same filter. A second frequency band component is extracted.

By the way, if the largest peak in the first frequency band is the tire resonance frequency f _R , the finally determined second frequency band follows the resonance frequency f _R and is near the center frequency of the second frequency band. A resonant frequency f _R appears. Therefore, by applying a variable bandpass filter to the components of the first frequency band, the range in which the resonance frequency f _R exists can be expanded while maintaining the width of the search range (first frequency band) for the resonance frequency f _R . can be narrowed down to a narrower range (second frequency band). Further, since the second frequency band is narrower than the first frequency band, the probability that an extra noise peak is included in the band of interest is reduced, and the tire resonance frequency f _R becomes less susceptible to noise.

Also, the faster the tire rotation speed, the higher the vibration frequency from the road surface, so the gain of the sensing data based on the wheel speed tends to increase as the frequency increases. This produces noise that depends on the running speed of the vehicle 1, and this noise should be removed when calculating the resonance frequency f _R as an index value for determining the state of the vehicle. In this regard, as described above, in the present embodiment, the second frequency band is narrower than the first frequency band, so that the amount of this type of noise included in the band of interest is reduced, and pressure reduction is detected in step S8. The tire resonance frequency f _R determined as the index value of is less susceptible to noise.

In subsequent step S8, the peak determining unit 23 determines the resonance frequency f _R of the tire FL based on the components of the second frequency band of the sensing data that have passed through the variable bandpass filter. In this embodiment, the peak determining unit 23 performs time series analysis on the components of the second frequency band using a Kalman filter based on a second-order autoregressive (AR) model, and sets the peak frequency as the resonance frequency f _R . Alternatively, the components of the second frequency band are time-series analyzed by a Kalman filter based on a higher-order (third-order or higher) autoregressive (AR) model. Subsequently, from the parameters of the AR model estimated in this way and the sensing data used for the estimation, an input signal that can be assumed to have been given to the model is restored. Then, after applying appropriate signal processing such as a band-pass filter to this input signal and output signal (sensing data), the system is processed by a Kalman filter based on a second-order Autoregressive Moving Average (ARMA) model. decide. Then, the peak frequency at this time may be determined as the resonance frequency f _R . Since the method of calculating the resonance frequency by a high-order AR model and a high-order ARMA model is publicly known, detailed description is omitted here. See US Pat. No. 102,077.

In subsequent step S9, the reduced pressure detection unit 24 compares the resonance frequency f _R as an index value for detecting reduced pressure calculated in step S7 with the threshold value stored in the storage device 15, and the former is the latter. If so, the process proceeds to step S10. On the other hand, if the latter is larger, the process returns to step S1.

In step S10, the decompression detector 24 issues a decompression warning. Specifically, the decompression detection unit 24 lights the alarm indicator 3 mounted on the vehicle 1 to notify the driver that the tire FL is decompressed. After step S9 ends, the reduced pressure detection process ends.

<3. Variation>
Although one embodiment of the present invention has been described above, the present invention is not limited to the above-described embodiment, and various modifications are possible without departing from the scope of the invention. For example, the following changes are possible. Also, the gist of the following modified examples can be combined as appropriate.

<3-1>
In the above embodiment, the resonance frequency is determined using the wheel speed signal (or the sensing data representing the rotational acceleration obtained by differentiating the wheel speed signal) in order to detect the tire pressure reduction. However, the above algorithm for determining the resonance frequency from sensing data regarding the vibration state of the running vehicle can also be applied in other situations. For example, in order to determine ride comfort, vehicle vibration (shimmy, shake, etc.), road surface conditions on which the vehicle is running, tire failure (including wear), etc., peak There are situations where it is required to derive the frequency and peak level, and the method of steps S3 to S8 can also be applied in such situations. In addition, acceleration sensors ( There are situations where it is required to derive the resonance frequency from the output signal of the vibration sensor), and the method of steps S3 to S8 can also be applied in such situations.

<3-2>
In the above embodiment, the resonant frequency f _R was determined by applying the AR model to the components in the second frequency band. However, the peak determining unit 23 may determine the resonance frequency f _R based on the value specifying the second frequency band determined in step S7. For example, the resonance frequency f _R may be determined as the center frequency of the second frequency band (including frequencies in the vicinity thereof).

<3-3>
Alternatively, the resonance frequency f _R may be determined by the following method. First, the second filtering section 22 removes the components of the second frequency band by passing the components of the first frequency band through a variable band-stop filter instead of the variable band-pass filter. A variable band-stop filter is a type of band-stop filter, and the center frequency of the filter is adaptively adjusted using the LMS algorithm etc. so that the signal gain (spectrum area) removed by the filter is maximized. It is a filter that changes. The second filtering unit 22 determines a second frequency band that maximizes the signal gain of the sensing data removed by the filter while moving the variable band-pass filter within the first frequency band. remove the components of the second frequency band. Subsequently, the peak determination unit 23 may determine the resonance frequency f _R based on the thus determined value specifying the second frequency band. For example, the resonance frequency f _R may be determined as the center frequency of the second frequency band (including frequencies in the vicinity thereof).

<3-4>
Alternatively, the resonance frequency f _R may be determined by the following method. First, the second filtering section 22 removes the components of the second frequency band by passing the components of the first frequency band through a variable band-stop filter instead of the variable band-pass filter. Subsequently, the peak determination unit 23 includes at least the center frequency of the second frequency band (typically, the center frequency of the second frequency band is the center frequency) and has a narrower bandwidth than the first frequency band Define a bandpass filter that passes signals of (which may be the same as the second frequency band). Then, the component of the first frequency band is passed through the bandpass filter, and the component passed through the bandpass filter is subjected to time-series analysis based on the AR model and the ARMA model in the same manner as in step S8, thereby obtaining the resonance frequency f _R . may decide.

<3-5>
Even if a variable band-stop filter that removes the components of the second frequency band is used instead of a variable band-pass filter that allows the components of the second frequency band to pass as in the modification 3-4, the second frequency band It is possible to extract components. In such a case, the resonance frequency f _R can be determined based on the AR model and the ARMA model, as in the above embodiment. As another example, the resonant frequency f _R may be determined by the following method. First, the second filtering section 22 removes the components of the second frequency band by passing the components of the first frequency band through a variable band-stop filter. Subsequently, the peak determination unit 23 subtracts the sensing data that has passed through the variable band-stop filter and has the components of the second frequency band removed from the sensing data immediately before passing through the variable band-stop filter, thereby obtaining the obtain the components of the second frequency band. Then, the resonance frequency f _R may be determined by time-series analysis of the sensing data thus acquired based on the AR model and the ARMA model, as in step S8.

Examples 1 and 2 of the present invention and a comparative example are described below. First, wheel speed signals (sensor data) were collected from wheel speed sensors attached to the right front wheel and the left front wheel, which are the driving wheels of the vehicle, and then steps S3 to S6 according to the above embodiment were executed. At this time, in the ANC of step S5, the past sensing data of the own tire was used as the reference signal. In step S6, a fixed bandpass filter was used that passes the band from 33 Hz to 47 Hz.

An AR model and an ARMA model similar to those in step S8 were applied to the sensing data after execution of steps S3 to S6 as described above to calculate the resonance frequency (comparative example).

On the other hand, steps S7 and S8 were further executed on the sensing data after execution of steps S3 to S6 to calculate the resonance frequency (Example 1). In step S7 here, a variable bandpass filter with a bandwidth of 10 Hz was used, moving in the band from 33 Hz to 47 Hz.

Further, step S7 is further executed on the sensing data after execution of steps S3 to S6, and then, as in modification 3-2, the center frequency of the second frequency band is determined as the resonance frequency (embodiment 2). . Here also step S7 used a variable bandpass filter with a bandwidth of 10 Hz moving in the band from 33 Hz to 47 Hz.

Resonance frequencies were calculated by the methods of Comparative Example and Examples 1 and 2 with respect to wheel speed signals when the tire pressure was normal, 15% decompression and 30% decompression. The results of Comparative Example are shown in FIG. 4, the results of Example 1 are shown in FIG. 5, and the results of Example 2 are shown in FIG. The horizontal axis of each graph represents the number of pieces of sensing data (that is, the time axis), and the dashed-dotted line in each graph represents the average value at normal pressure.

From these graphs, it is easy to understand when looking at the case of 30% reduced pressure, but Examples 1 and 2 have higher reduced pressure sensitivity than Comparative Example (the difference between normal pressure and 30% reduced pressure is large). Therefore, it was confirmed that, based on the resonance frequencies calculated by the methods of Examples 1 and 2, it is possible to detect a reduced pressure with sufficient determination accuracy.

1 vehicle 2 monitoring device (computer)
7 program 21 first filtering unit 22 second filtering unit 23 peak determination unit 24 decompression detection unit

Claims (12)

A monitoring device for monitoring the state of a running vehicle,
a first filtering unit that extracts a component of a first frequency band of the sensing data by passing time-series sensing data relating to the vibration state of the running vehicle through a fixed band-pass filter;
By passing the components of the first frequency band extracted by the first filtering unit through a variable band-pass filter or a variable band-stop filter that can be moved within the first frequency band, the band is wider than the first frequency band. a second filtering unit that extracts or removes components of a narrow second frequency band;
A peak determination unit that determines the resonance frequency of the sensing data based on at least one of a value specifying the second frequency band and a component of the second frequency band,
monitoring equipment. The sensing data is rotation information of wheels mounted on the vehicle,
A monitoring device according to claim 1 . The second frequency band has a bandwidth of 4 Hz to 12 Hz.
3. A monitoring device according to claim 2. The bandwidth of the first frequency band is 4 Hz to 10 Hz wider than the bandwidth of the second frequency band,
4. A monitoring device according to claim 2 or 3. The peak determination unit determines the resonance frequency as the center frequency of the second frequency band based on the value specifying the second frequency band,
5. A monitoring device according to any one of claims 1-4. The peak determination unit determines the resonance frequency by time-series analysis of the components of the second frequency band that have passed through the variable bandpass filter based on an autoregression model.
5. A monitoring device according to any one of claims 1-4. The second filtering unit removes the components of the second frequency band by passing the components of the first frequency band through the variable band-stop filter,
The peak determination unit
defining a bandpass filter that passes signals in a band that includes at least the center frequency of the second frequency band and has a narrower bandwidth than the first frequency band;
passing the components of the first frequency band through the defined bandpass filter;
Determining the resonance frequency by time-series analysis based on an autoregression model of the component that has passed through the defined bandpass filter,
5. A monitoring device according to any one of claims 1-4. The monitoring device according to any one of claims 1 to 7, further comprising a decompression detector that detects decompression of tires mounted on the vehicle based on the determined resonance frequency. 9. The variable bandpass filter according to any one of claims 1 to 6 and 8, wherein the center frequency of the variable bandpass filter is adaptively changed so that the spectral area of the signal passing through the variable bandpass filter is maximized. monitoring equipment. 9. Any one of claims 1 to 5 and 7 and 8, wherein the variable band-stop filter adaptively changes its center frequency so as to maximize the spectral area of the signal removed by the variable band-stop filter. A monitoring device according to any one of the preceding claims. A monitoring program for monitoring the condition of a running vehicle,
extracting a component of a first frequency band of the sensing data by passing time-series sensing data relating to the vibration state of the running vehicle through a fixed band-pass filter;
By passing the extracted components of the first frequency band through a variable band-pass filter or a variable band-stop filter movable within the first frequency band, a second frequency band having a narrower bandwidth than the first frequency band extracting or removing frequency band components;
A monitoring program causing a computer to determine a resonance frequency of the sensing data based on at least one of a value specifying the second frequency band and a component of the second frequency band. A monitoring method for monitoring the condition of a running vehicle, comprising:
extracting a component of a first frequency band of the sensing data by passing time-series sensing data relating to the vibration state of the running vehicle through a fixed band-pass filter;
By passing the extracted components of the first frequency band through a variable band-pass filter or a variable band-stop filter movable within the first frequency band, a second frequency band having a narrower bandwidth than the first frequency band extracting or removing frequency band components;
and determining a resonant frequency of the sensing data based on at least one of a value identifying the second frequency band and a component of the second frequency band.

Images