KR100402089B1 - Method and device for receiving frequency-modulated signal - Google Patents

Abstract

A train detection method with high immunity to noise in order to detect trains with or without demodulation because the level of the received signal is reduced due to the axle short circuit to detect trains using frequency-modulated train detection signals. Before demodulation, a combination of a gain process for amplifying the amplitude of the signal, a limiter process for limiting the amplitude of the output signal of the gain process for a specific level, and a limiter process are performed in combination to set the lower limit of the level of the demodulated signal arbitrarily. It is a method of filter processing to remove harmonics generated because the threshold value of train detection is arbitrarily set.

Description

Frequency modulated signal receiving method and apparatus {METHOD AND DEVICE FOR RECEIVING FREQUENCY-MODULATED SIGNAL}

In signal control of railroads, as a method of detecting trains, there is a method of electrically dividing tracks of tracks into sections called track circuits and detecting the presence or absence of trains in each track circuit. In such a track circuit, when a signal is transmitted from one end and the signal is received at the other end and the train is in the target track circuit, the wheel of the train short-circuits the left and right tracks so that the received signal decreases, thereby changing the level of the received signal. Is observed and the presence or absence of the train is determined by comparing the reception level with any threshold value that determines that the train is present.

The reception level is different in the length of each track circuit and is affected by the surrounding environment, the level of the state where no train is present is also different, and the state of the effect that the effect of short circuit caused by the wheel of the train is given to the reception level is also different. Therefore, it is necessary to enable arbitrary limit values to be set individually for each track circuit.

Since the reception level value is influenced by the entire signal component of the frequency band determined by the filter characteristics, it is difficult to exclude the influence as a signal noise used by these apparatuses when there is other equipment or the like that uses the frequency band in the vicinity. Do. When the influence of such a noise is large, for example, whether or not the train is in a range within the target track circuit, when the received signal level becomes greater than the train detection level due to the influence of noise, the presence or absence of the train is determined. There is a possibility of misjudgment.

As a method of using a frequency modulated signal as a signal for detecting such a train, there is the invention of Japanese Patent Laid-Open No. 9-125261. According to this application, if a frequency modulated signal is used, it is included in the received signal in addition to the received signal level. The information government can be used to determine the presence of trains, making it possible to accurately determine whether noise is present.

However, with respect to the received signal, the condition that changes with or without the train is only a change in the level of the received signal. Therefore, in the case of using the frequency modulated signal, the receiver decides whether to demodulate by changing the amplitude in order to perform the demodulation process. It is necessary to use a demodulation method having an amplitude dependency as shown in FIG. As a demodulation processing method that satisfies such a condition, for example, a method using PLL processing can be considered. In the PLL process, the amplitude of the input signal is set in advance at the time of design. When the input amplitude is small, the synchronous frequency range becomes small, and when this frequency range cannot satisfy the modulation range of the frequency modulated signal, the synchronization condition cannot be satisfied. Qualitatively represented. From this, the demodulation lower limit level, which is the received signal level when the synchronization condition is not satisfied and the demodulation cannot be obtained, can be obtained by calculation, and the PLL processing is calculated so that the train detection level matches the demodulation lower limit level.

However, since the train detection level is different for each track circuit, it is necessary to also set the demodulation lower limit level separately for the track circuit. However, it is practically difficult to produce PLL processing in which the demodulation limit level is individually set for all track circuits.

(Initiation of invention)

An object of the present invention is to improve the productivity of a train detection apparatus that is more resistant to noise.

The above object is a method of arbitrarily demodulating a demodulation process and arbitrarily setting a train detection level which is a threshold value of a signal level of train determination. Demodulation by performing a gain process for amplifying the amplitude, a limiter for limiting the amplitude of the output signal of the gain process to a constant value equal to the design input amplitude of the demodulation process, and a filter process for removing harmonics generated by the limiter process. This is achieved by setting the lower limit of the possible signal level to the same value as the train detection level by changing the gain value.

Further, the filter processing is set such that the sampling frequency of the filter processing is one multiple of the radix of the sampling frequency with respect to the frequency of the processing target, and sets the passband of the filter to a value wider than the signal band required for demodulation. By making the passband one of the radix of the sampling frequency, the harmonic effect of the radix aberration can be most effectively removed among the harmonics generated by the limiter process.

As a result, it is possible to reduce the increase in the total throughput of the reception processing due to the increase in the filter processing.

In addition, the productivity of the train detection device with high resistance to noise can be increased.

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a signal receiving apparatus using demodulation processing of a frequency modulated signal having an amplitude dependency, and more particularly to a train detecting apparatus using a track circuit in a railway.

1 is a view showing the configuration of a train detecting apparatus in an embodiment of the present invention.

2 is a diagram showing a relationship between a processing configuration of a receiving apparatus and a reception signal level in an embodiment of the present invention.

Fig. 3 is a diagram showing the relationship between the input level of the limiter process and the design value of the input amplitude of the PLL process of the output level in the embodiment of the present invention.

4 is a diagram illustrating a relationship between a train detection level and a demodulation unit in an embodiment of the present invention.

5 is a diagram showing frequency characteristics of filter processing in an embodiment of the present invention.

Fig. 6 is a diagram showing the relationship between the frequency characteristics of the filter process and the harmonic harmonics in the embodiment of the present invention.

7 is an embodiment when the present invention is applied to a plurality of track circuits.

8 is a device configuration when applied to the water quality detection device in the tank.

Hereinafter, an embodiment of the present invention will be described. Here, as an apparatus to which the present invention is applied, a case where a receiving apparatus of a train detecting apparatus using a track circuit in a railroad is considered, and a PLL process is used as a frequency-modulated signal demodulation processing having an amplitude dependency will be described as an example.

1 shows a system configuration according to the present method. The track on which the train travels consists of one or more track circuits. Each track circuit is connected to a transmitter 2 for transmitting a train detection signal as a frequency modulation signal at one end and a receiver 3 for receiving a train detection signal at the other end. Thus, the train detection device 1 is connected to this transmission device and the reception device through a transmission path such as a network.

Before explaining the outline of the processing between the devices in the system, the processing principle performed by the system will be described.

When the train is within the range of the track circuit, the track is shorted by the axle of the train, so the received signal level of the train detection signal received by the receiver is compared with the train not being within the range of the track circuit and thus low. It becomes a level. At this time, the reception signal level, which is a threshold value when the reception signal level is lowered and it is determined that a train exists, is called a train detection level.

Since the train detection level needs to set an arbitrary value for each track circuit, a method for arbitrarily setting the demodulation lower limit level, which is a lower limit that can be demodulated in accordance with the train detection level, for each track circuit individually, The gain processing section, the limiter processing section, and the harmonic rejection filter processing section are processed in this order.

First, the gain processing unit amplifies a signal that matches the train detection level with the demodulation lower limit level, and receives, as gain information, an amplification factor necessary for signal amplification from the gain information storage unit.

The limiter processor limits the amplitude so that the PLL processor does not receive excessive input by the amplification factor.

And the harmonic removal filter processing part removes the harmonic component generate | occur | produced by a limiter process.

Detection of train assumes the following procedures. First, the transmission information creating unit of the train detection apparatus creates the transmission information and transmits it to the transmission apparatus via the network. This transmission information is also transmitted to the reception information combination unit in the train detection apparatus.

Next, the transmitting apparatus converts the transmission information received through the network into a train detection signal in the frequency modulation processing unit and transmits it to the track circuit.

The reception device 3 first removes noise included in the reception signal by the noise reduction filter processing unit 31 with respect to the train detection signal received through the track circuit. Next, the gain processing section 32 amplifies the received signal using the gain information from the gain information storage section 33. Next, the limiter processor 34 limits the amplitude of the received signal. Next, the harmonic component removal filter processing part 35 removes a harmonic component contained in the train detection signal. Next, the modulation component of the train detection signal received by the PLL processing part 36 is detected. Finally, the reception information creating unit 37 creates the reception information from the modulation component and transmits the reception information to the train detection apparatus through a transmission path such as a network.

Finally, the train detecting apparatus combines the transmission information and the reception information in the reception information combination unit, detects the presence or absence of trains within the track circuit range, provides the detection result to the display unit, and displays the detection result. do.

2 shows a level diagram regarding the relationship between the received signal level and the demodulation lower limit level.

First, the procedure for matching a train detection level and a demodulation lower limit level is shown. This corresponds to the characteristic of FIG. Since the characteristics of the PLL processing unit depend on the amplitude of the received signal which is the input signal for the PLL processing, the PLL processing unit is initially designed by determining the amplitude of the input signal and inducing a demodulation lower limit level. For example, when the amplitude of the input signal is designed as 1.0, the demodulation lower limit level is 0.316. Here, if the design value 1.0 of the amplitude of the input signal is 0 [dB], the PLL processing can demodulate at a signal level of -10 [dB] or more.

The gain processing section first sets an amplification factor for each track circuit to match the demodulation lower limit level of the PLL process and the train detection level, and stores the gain information in the gain information storage section as gain information. For example, when the train does not exist within the range of the track circuit, the reception level is 3.16 (equivalent to +10 [dB]) and the amplitude at the train detection level is 0.0316 (equivalent to -30 [dB]). If the lower demodulation level is 0.316 (-10 [dB]), it is better to set 10.0 (equivalent to +20 [dB]) as gain information.

In the limiter processing section, the amplitude is limited with respect to the received signal to which the gain processing section is added, so that the maximum amplitude of the input signal of the PLL processing section is a value that matches the amplitude at the time of designing the PLL processing section. For example, when the amplitude of the train detection level is 0.0316 (-30 [dB]), the lower demodulation level is 0.316 (-10 [dB]), and the amplification factor is 10.0 (+20 [dB]), the range of the track circuit In the state where the reception level is 3.16 (+10 [dB]) when no train is present, the amplitude of the output signal of the gain processing section is 3.16 (+30 [dB]). On the other hand, since the set value of the input amplitude is 1.0 (0 [dB]), the PLL processing section becomes an excessive level input, making it difficult for the PLL processing to operate in accordance with the design value. On the other hand, the limiter processing unit suppresses the value exceeding 1.0 (0 [dB]), which is the set value of the input amplitude of the PLL processing, to 1.0 (0 [dB]) to avoid excessive level input to the PLL processing unit. For this reason, the relationship between the input amplitude and the output amplitude of the limiter processing unit is equal to the input amplitude and output amplitude in the range below the amplitude when the PLL processing unit is designed. It is the same value as the amplitude at design time. This relationship is shown in FIG.

The harmonic elimination filter processing unit then removes harmonics generated by the limiter processing unit. Since the output signal of the limiter processing unit is close to the square wave when the reception level exceeds the amplitude at the time of designing the PLL processing unit, the harmonic component is included.

When the output signal of the limiter processing unit is input to the PLL processing unit in the state including the harmonics, the harmonic component becomes a noise component to the PLL processing unit, and the characteristics designed by the PLL processing unit cannot be satisfied. Therefore, by removing the harmonic components from the harmonic elimination filter processing unit, the PLL processing unit performs the processing according to the design.

As a result, the PLL processing unit can receive an input signal in which the demodulation lower limit level and the train detection level coincide.

Next, consider the case where the train exists and does not exist within the range of the track circuit. For example, the amplitude at the train detection level is 0.0316 (-30 [dB]), the amplification factor is 10.0 (+20 [dB]), and the setting value of the input amplitude when designing the PLL processing unit is 1.0 (0 [dB]). In the following, the demodulation lower limit level is 0.316 (-10 [dB]).

First, consider the case where the train is in the track circuit. For example, it is assumed that the reception level is 0.0177 (-35 [dB]). The characteristic at this time corresponds to (2) of FIG. The reception level is smaller than 0.0316 of the train detection level. At this time, the output of the gain processing unit is 0.177 (-15 [dB]), and is smaller than 0.316 (-10 [dB]).

On the other hand, since the maximum value of the output amplitude of the limiter processing unit is 1.0 (0 [dB]) at the same value as the set value of the input amplitude of the PLL processing unit, the output of the gain processing unit is not affected. For this reason, since the harmonic component does not generate | occur | produce the output of a limiter process part, a harmonic removal filter process part is also not affected. As a result, a signal of 0.177 (-15 [dB]), which is a signal having a level smaller than amplitude 0.316 (-10 [dB]), is input to the PLL processing unit. On the other hand, since the lower demodulation level of the PLL processing section is 0.316 (-10 [dB]), demodulation cannot be performed. In short, a signal having an amplitude smaller than the train detection level is demodulated. Therefore, since the reception information creating unit does not create the reception information, the transmission information and the reception information do not match in the reception information combination unit of the train detection apparatus. The presence of a train in the track circuit is detected.

Next, consider the case where it does not exist. For example, it is assumed that the reception level is 3.16 (+10 [dB]). The characteristic at this time corresponds to (3) of FIG. The reception level is greater than 0.0316 (-30 [dB]) of the train detection level. At this time, the output of the gain processing unit is 3.16 (+30 [dB]), which is a value larger than 0.316 (-10 [dB]). On the other hand, since the maximum value of the output amplitude of the limiter process is 1.0 (0 [dB]), which is the same value as the set value of the input amplitude of the PLL processing unit, the level exceeding 1.0 (0 [dB]) among the outputs of the gain processing unit. The amplitude is limited by the influence, and the output signal of the limiter processing part becomes a signal similar to a square wave, and its amplitude is 1.0 (0 [dB]). For this reason, the harmonic component is contained in the output of a limiter process part. However, since the harmonic component is removed by the harmonic rejection filter processing section, the output signal of the harmonic rejection filter processing section becomes a signal from which harmonics have been removed, and the amplitude becomes the same value as that of the output signal of the limiter processing section, and the maximum is 1.0 (0 [dB]). )to be. As a result, a signal having a value greater than 0.316 (-10 [dB]) and smaller than 1.0 (0 [dB]) is input to the PLL processing unit. On the other hand, the lower demodulation level of the PLL processing section is 0.316 (-10 [dB]), and the PLL processing section can demodulate because the set value of the input amplitude is 1.0 (0 [dB]). In short, demodulation is performed on signals with amplitudes greater than the train detection level. For this reason, since the reception information creating unit creates the reception information, the transmission information and the reception information coincide in the reception information combination unit of the train detecting apparatus. Accordingly, the case where no train exists in the track circuit is detected.

In short, demodulation can be performed qualitatively without demodulation at a signal level below the train detection level but at a signal level above the train detection level. As described above, by using a method in which the gain processing unit, the limiter processing unit, and the harmonic rejection filter processing unit are matched to the front end of the PLL processing unit, it becomes possible to set whether to demodulate in accordance with the train detection level. 4 shows the relation between reception level and demodulation.

In this configuration, since the value to be set for each track circuit is as much as the amplification factor of the gain processing, the PLL processing unit can have a common design, thereby increasing productivity. An example in the case of applying to a plurality of track circuits is shown in FIG. 7.

Since the harmonic filter processing unit removes harmonic components of the input signal generated in the limiter processing unit, it is required to reduce the influence of the aberration harmonics, which are mainly square wave components.

In general, in the case of performing the filter processing by digital processing, it is known that if the frequency of one fourth of the sampling frequency is set to the center frequency of the pass band, the required characteristic can be obtained at the lowest throughput.

However, since an aliasing phenomenon occurs at half the sampling frequency, if a passband is set as the input frequency at one quarter of the sampling frequency, the passband at one quarter of the sampling frequency The same characteristics are also present in the band centered on a frequency of three quarters of the sampling frequency. If a square wave is input for this filter characteristic, the necessary attenuation is obtained because the frequencies of the aberration harmonics overlap in the passband at the frequency of one quarter of the sampling frequency and the passband at three quarters of the sampling frequency. Can't.

For this reason, it is necessary to set the input frequency to a band which does not coincide with one quarter of the sampling frequency. However, a filter having a band with a large difference in one quarter of the sampling frequency as a pass band generally requires more throughput. Since this is associated with an increase in the cost by increasing the throughput of the digital circuit, it is desired to make the difference as small as possible.

The square wave component generated by the limiter process is mainly set to the aberration harmonic, and the value of the input frequency is set to a multiple of one of the radix of the sampling frequency, and the passband is a sampling frequency centered on the input frequency. By setting the band to be smaller than one-half of the number and setting the stopband to all bands other than one-numbered of the sampling frequency centered on the input frequency, filter processing for efficiently removing harmonics is realized.

For example, consider the case where the input frequency is set to 2 [kHz], the band required for demodulation is ± 300 [Hz] centered on the input frequency, and the pass band of the filter is set to a band of 7/7 of the sampling frequency. . 5 shows filter characteristics. In this example, the sampling frequency is 7 [kHz] and the center frequency of the pass band is 2 [kHz].

In this case, the width of the pass band should be less than one seventh of the sampling frequency, but the band required for demodulation has a bandwidth of ± 300 [Hz] centered on an input frequency of 2 [kHz]. There is no damage to the band required for demodulation. Here, the pass band is a band of +/- 300 [Hz] centering on 2 [kHz]. On the other hand, since the stop band is all bands other than one seventh of the sampling frequency centered on the input frequency 2 [kHz], it is necessary to set all bands except the band of 1 [kHz] centered on 2 [kHz]. There is. Here, a frequency band of 1.5 [kHz] or less and a frequency band of 2.5 [kHz] or more are used as stop bands.

On the other hand, since there exists a pass band by an alias, at frequencies exceeding 3.5 [kHz] which is half of a sampling frequency, frequency characteristics of 3.5 [kHz] or less are reproduced in a repetitive state. For this reason, the same characteristics as the pass band exist in the band of 3.5 [kHz] or more. The passband characteristics are ± 300 [Hz] centered on 5.0 [kHz], which is a band of less than ± 300 [Hz] centered on 2.0 [kHz] which is the passband, and 3.5 [kHz] which is half of the sampling frequency. ] Is the following band. On the other hand, since the stopband characteristics are similarly repeated, the range from 3.5 [kHz] to 4.5 [kHz] corresponding to 1/2 of the sampling frequency and from 5.5 [kHz] to 7 [kHz] which is the sampling frequency Yes.

Harmonic components, on the other hand, exist primarily at odd values of the input frequency. Here, since the input frequency is 2/7 of the sampling frequency, the harmonics are present in the band of the multiple of the band of 2/7 of the sampling frequency. The following bands exist up to the 15th harmonic.

The first harmonic is the same frequency as the input frequency and is a frequency component to be processed by the PLL process. For this reason, it corresponds to the pass band in this filter process. This frequency component is in the range of ± 300 [Hz] around 2 [kHz], which is the same as the input frequency, and corresponds to 2/7 of the sampling frequency.

The third harmonic is in the range of ± 300 [Hz] around 6 [kHz], which is three times the input frequency of 2 [kHz], which is 1 [kHz] for the sampling frequency of 7 [kHz]. Is equivalent to being in the range of ± 300 [Hz] around For this reason, it corresponds to a seventh band of the sampling frequency.

The fifth harmonic is in the range of ± 300 [Hz] around 10 [kHz], which is 5 times the input frequency of 2 [kHz], which is 3 [kHz] at the sampling frequency of 7 [kHz]. Is equivalent to being in the range of ± 300 [Hz] around For this reason, it corresponds to a band of three thirds of the sampling frequency.

The seventh harmonic is in the range of ± 300 [Hz] around 14 [kHz], which is 7 times the input frequency of 2 [kHz], which is 0 [kHz] at the sampling frequency of 7 [kHz]. Is equivalent to being in the range of ± 300 [Hz] around For this reason, it corresponds to a band of sevenths of the sampling frequency.

The ninth harmonic is in the range of ± 300 [Hz] around 18 [kHz], which is 9 times the input frequency of 2 [kHz], which is 3 [kHz] at the sampling frequency of 7 [kHz]. Is equivalent to being in the range of ± 300 [Hz] around For this reason, it corresponds to a band of three thirds of the sampling frequency.

The 11th harmonic is in the range of ± 300 [Hz] centered on 22 [kHz], which is 11 times the input frequency of 2 [kHz], which is 1 [kHz] at the sampling frequency of 7 [kHz]. Is equivalent to being in the range of ± 300 [Hz] around For this reason, it corresponds to a seventh band of the sampling frequency.

The 13th harmonic is in the range of ± 300 [Hz] centered on 26 [kHz], which is 13 times the input frequency of 2 [kHz], which is 2 [kHz] at the sampling frequency of 7 [kHz]. Is equivalent to being in the range of ± 300 [Hz] around For this reason, it corresponds to a band of 2/7 of the sampling frequency.

The 15th harmonic is in the range of ± 300 [Hz] around 30 [kHz], which is 15 times the input frequency of 2 [kHz], which is 2 [kHz] at the sampling frequency of 7 [kHz]. Is equivalent to being in the range of ± 300 [Hz] around For this reason, it corresponds to a band of 2/7 of the sampling frequency.

6 shows the relationship between harmonics and filter characteristics. Radix harmonics that match the passband become harmonics after the thirteenth harmonic.

This effect is obtained from Fourier transform by making the input signal a perfect square wave.

From the Fourier transform, the thirteenth harmonic level with respect to the first harmonic can be found to be about 24 [dB] when the first harmonic is 0 [dB]. For example, if the attenuation of the passband of the filter process is 0 [dB] and the output amplitude of the limiter is 1.0 at maximum, the amplitude of the first harmonic component is 1.0, whereas the component of the 13th harmonic is The amplitude is converted to 0.07 or less. As a result, the output amplitude of the filter process is at most about 1.07 due to the influence of the thirteenth harmonic, so that harmonics can be effectively removed.

On the other hand, since the pass band exists at two-sevenths of the sampling frequency, the difference between the one-fourth of the pass band and the sampling frequency is small and the increase in filter throughput can be suppressed low. This shows that harmonics can be effectively removed.

By using the above-described configuration, it was possible to improve the productivity of the train detector having high resistance to noise.

As another application example of this system, the apparatus structure in the case of applying to the water quality detection apparatus in a tank is shown in FIG. The light emitting device and the light receiving device are provided in pairs in the water tank, the light emitting device is connected to the transmitting device, and the light receiving device is connected to the receiving device. The transmitter and receiver connect to the water quality detector through the network.

The water quality detection device displays a warning on the display unit when the light intensity generated by the light emitting device is attenuated by the transmittance of the water quality with respect to the water quality in the water tank, and the light intensity received by the light receiving device is below a certain value. Instruct the water quality control unit to improve the water quality.

Since the threshold value of the water quality detection varies depending on the installation distance of the light emitting device and the light receiving device, etc., it is necessary to determine each light emitting device and the light receiving device individually.

Here, it is apparent that by applying the present method to the receiver, the threshold value of the water quality detection can be set to an arbitrary value by setting the threshold value of the water quality detection as the gain information of each receiver.

As described above, the apparatus according to the present invention can be applied to signal control of railroads as a train detecting device with high resistance to noise and good productivity.

Claims (3)

A frequency modulated signal receiving method for demodulating a frequency modulated signal having an input amplitude dependency, A gain process for setting the amplification factor of the input signal to set the lower limit level to be demodulated to match the demodulation lower limit level of the demodulation process; A limiter process for suppressing a level above a reference value of the input signal of the demodulation process within a reference value, Filter processing to remove harmonics generated by the limiter processing, A frequency modulated signal receiving method which performs demodulation processing after the filter processing. The method of claim 1, The sampling frequency when the filter processing is a digital processing is set to be a multiple of one of the base of the sampling frequency with respect to the frequency of the processing target, The passband of the filter including the bandwidth required for demodulation is set to be a band less than one-th of the sampling frequency centered on a multiple of one-th of the base of the sampling frequency, A frequency modulated signal receiving method, wherein the stop band is any band other than one band of the odd number of the sampling frequency centered on one multiple of the odd number of the sampling frequency. A frequency modulation receiver for demodulating a frequency modulated signal having an amplitude dependence, A gain processing unit which sets the lower limit level of the input signal to be amplified to be equal to the lower limit level of the demodulation process; A limiter processing unit for suppressing a level equal to or greater than a reference value of the input signal of the demodulation process within a reference value; A filter processing unit for removing harmonics resulting from the processing by the limiter processing unit, And a demodulation processing section for demodulating the signal processed by the filter processing section.