KR101967812B1 - A radar signal processing device using a weighted Kalman filter and a target detection system using the same - Google Patents

Abstract

The present invention relates to a radar signal processing apparatus and a target detection system using the same, and more particularly, to a radar signal processing apparatus and a target detection system using the same, Digital Beam Former (DBF); A second window for applying a window function for reducing a leakage error occurring during fast Fourier transform on the synthesized signals; A Fast Fourier Transformer (FFT) for performing a fast Fourier transform on the output signal of the second window; A weight calculator for calculating a weight according to power of bit frequencies of an output signal of the fast Fourier transformer; A constant false alarm rate (CFAR) for selecting specific bit frequencies by applying a variable detection threshold to the output signal of the fast Fourier transformer; An estimator for estimating information including the distance, velocity and azimuth of the target from the selected bit frequencies; And a Kalman filter for tracking the target in consideration of the weighted power calculated by the weight calculator, and a target detection system using the radar signal processing apparatus.

Description

Technical Field [0001] The present invention relates to a radar signal processing apparatus using a weighted Kalman filter and a target detection system using the Kalman filter and a target detection system using the Kalman filter.

The present invention relates to a radar signal processing apparatus and a target detection system using the same, and more particularly, to a radar signal processing apparatus using a weighted Kalman filter for applying a weight to a Kalman filter to enable more accurate position tracking, and a target detection system using the same .

Generally, a radar apparatus is a device that detects a target by detecting a parameter such as a distance, a speed, and an azimuth of a target from a frequency component of a received signal by transmitting an electromagnetic wave and receiving the reflected wave to detect a target such as a ship, an automobile, , Collision prevention, and the like.

There are various kinds of radar apparatuses such as pulse radar and continuous wave radar depending on the type of radio wave.

Among these, a frequency modulation continuous wave (FMCW) radar, which is a kind of continuous wave radar, has various advantages over a pulse radar system in which a pulse width is narrow and a high output waveform is transmitted and received.

Kalman (Kalman) filter has been used as a technique to track a target by using FMCW radar, which is a type of continuous wave radar.

The Kalman filter plays a role of tracking the position of an actual object using positional information in which a false alarm is present.

Such a Kalman filter uses the calculated positional information and the positional information at the previous time. In this case, the false alarmed object should be removed and only the position of the actual object should be tracked.

However, it is possible to track the position more precisely by using the weight information by using the power information of the detected object in the radar system.

Domestic Publication No. 2016-0141946 Korean Publication No. 2015-0094240 Domestic Publication No. 2017-0000835

SUMMARY OF THE INVENTION The present invention provides a radar signal processing apparatus using a weighted Kalman filter and a target detection system using the Kalman filter, in which a weight is applied to a Kalman filter to enable more accurate position tracking.

According to an aspect of the present invention, there is provided a radar signal processing apparatus for receiving a signal reflected from a target received by a plurality of antennas, each of which receives a signal reflected from a target, as digital complex signals of a baseband, An ADC symbol formatter for dividing serial bits into ADC symbols, rearranging the ADC symbols in the order of arrangement of the antennas, and outputting the ADC symbols; A DC offset remover for removing a DC component from the rearranged ADC symbol; A first window for applying a window function to lower a side-lobe level of the interference signal for the ADC symbol from which the DC component is removed; A digital beamformer (DBF) for synthesizing the signals received for each antenna based on the output signal of the first window and outputting the synthesized signals for each beam; A second window for applying a window function for reducing a leakage error occurring during fast Fourier transform on the synthesized signals; A Fast Fourier Transformer (FFT) for performing a fast Fourier transform on the output signal of the second window; A weight calculator for calculating a weight according to power of bit frequencies of an output signal of the fast Fourier transformer; A constant false alarm rate (CFAR) for selecting specific bit frequencies by applying a variable detection threshold to the output signal of the fast Fourier transformer;

An estimator for estimating information including the distance, velocity and azimuth of the target from the selected bit frequencies; And a Kalman filter for tracking the target in consideration of the weighted power calculated from the weight calculator.

The Kalman filter of the radar signal processing apparatus of the present invention is characterized in that a position estimated by the estimator is determined when the weighted power inputted from the weight calculator is equal to or larger than a predetermined value.

Further, the Kalman filter of the radar signal processing apparatus of the present invention is characterized in that the position is corrected by tracking the position estimated by the estimator when the weighted power input from the weight calculator is less than a predetermined value.

Further, the weight calculator of the radar signal processing apparatus of the present invention provides the calculated weighted power to a certain false alarm probability detector, and the certain false alarm probability detector converts the detection threshold value by reflecting the inputted weighted power.

The Kalman filter of the radar signal processing apparatus according to the present invention is characterized in that when the value obtained by dividing the weighted power inputted from the weight calculator by the detection threshold value of the certain false alarm probability detector is equal to or greater than a first predetermined value, And the first constant value is an arbitrary value of 2 or more.

Further, in the radar signal processing apparatus according to the present invention, when the value obtained by dividing the weighted power input from the weight calculator by the detection threshold value of the predetermined false alarm probability detector is less than the first predetermined value and equal to or greater than the second predetermined value Wherein the second predetermined value is smaller than the first predetermined value, the first predetermined value is an arbitrary value of at least 2, the second predetermined value is at least 1, 1 < / RTI > constant value.

The Kalman filter of the radar signal processing apparatus according to the present invention is characterized in that when the value obtained by dividing the weighted power inputted from the weight calculator by the detection threshold value of the predetermined false alarm probability detector is less than a second predetermined value smaller than the first predetermined value The first constant value is an arbitrary value of 2 or more, and the second constant value is an arbitrary value of 1 or more and smaller than the first constant value.

Meanwhile, a target detection system according to the present invention includes a multi-array antenna including a plurality of antennas each receiving a signal reflected from a target;

An RF receiver for converting a signal received by each antenna of the multiple array antenna into complex signals of a baseband; Analog-to-digital converters (ADCs) for converting the baseband complex signals into digital signals; And a radar signal processing apparatus for processing the digital signals to estimate information on the target and performing target tracking using a Kalman filter to which a weight based on the weighted power of the bit frequency is applied.

According to the present invention as described above, a weight is applied to the Kalman filter to enable more accurate position tracking.

1 is an overall block diagram of a target detection system in accordance with an embodiment of the present invention.
FIG. 2 is a diagram illustrating a frequency change occurring during one cycle of a frequency modulated continuous waveform (FMCW) used in the target detection system according to one embodiment.
3 is a block diagram illustrating the structure of a radar signal processing apparatus using a weighted Kalman filter according to an embodiment of the present invention.
4 is a diagram for explaining an error of a target tracking when a target changes direction in an intersection.

BRIEF DESCRIPTION OF THE DRAWINGS The present invention is capable of various modifications and various embodiments, and specific embodiments will be described in detail below with reference to the accompanying drawings.

The following examples are provided to aid in a comprehensive understanding of the methods, apparatus, and / or systems described herein. However, this is merely an example and the present invention is not limited thereto.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings. In the following description, well-known functions or constructions are not described in detail since they would obscure the invention in unnecessary detail. The following terms are defined in consideration of the functions of the present invention, and may be changed according to the intention or custom of the user, the operator, and the like. Therefore, the definition should be based on the contents throughout this specification. The terms used in the detailed description are intended only to describe embodiments of the invention and should in no way be limiting. Unless specifically stated otherwise, the singular form of a term includes plural forms of meaning. In this description, the expressions "comprising" or "comprising" are intended to indicate certain features, numbers, steps, operations, elements, parts or combinations thereof, Should not be construed to preclude the presence or possibility of other features, numbers, steps, operations, elements, portions or combinations thereof.

It is also to be understood that the terms first, second, etc. may be used to describe various components, but the components are not limited by the terms, and the terms may be used to distinguish one component from another .

1 is an overall block diagram of a target detection system in accordance with an embodiment of the present invention.

Referring to FIG. 1, a target detection system 100 according to an embodiment of the present invention may include a transmitter 110, a processor 130, and a receiver 150.

The transmitter 110 generates a baseband frequency modulated continuous waveform (FMCW) and converts it into a signal of a certain frequency band (for example, 77 GHz band), and transmits a signal of a certain frequency band to a target .

The transmitter 110 may include a frequency modulated continuous waveform generator (FMCW Generator) 113 and an RF transmitter (RF transmitter)

The processor 130 controls the operations of the transmitter 110 and the receiver 150. The processor 130 provides a start signal to the frequency modulated continuous waveform generator 113 of the transmitter 110 and provides the receiver 150 with a synchronization signal indicating the start point of each chirp interval of the frequency modulated continuous waveform FMCW lt; RTI ID = 0.0 > (sync) < / RTI >

A frequency modulated continuous waveform generator (FMCW Generator) 113 can generate a base frequency frequency modulated continuous waveform (FMCW). The frequency modulated continuous waveform generator 113 can generate a frequency modulated continuous waveform FMCW whose frequency is modulated, for example, as shown in FIG. The frequency modulated continuous waveform FMCW generated by the frequency modulated continuous waveform generator 113 will be described with reference to FIG.

The RF transmitter 115 can convert the baseband frequency modulated continuous waveform FMCW generated by the frequency modulated continuous waveform generator 113 into a signal of a certain frequency band (for example, 77 GHz band) . A signal in a certain frequency band can be radiated to the target via a transmitting radar.

The frequency modulated continuous waveform FMCW emitted through the transmitting radar may be received and reflected by the target and through the multiple array antenna 151 of the receiver 150. [

The receiver 150 receives the signals reflected from the target for each antenna of the multi-array antenna 151, and based on the signals received for each antenna, information on the target (e.g., distance to the target, And the azimuth angle of the target). In addition, the receiver 150 may calculate bit frequencies for each segment based on signals received for each antenna, and estimate information about the target using the calculated bit frequencies.

The receiver 150 includes a multi-array antenna 151, RF receivers 153, analogue-to-digital converters (ADCs) 155, and a radar signal processor ) ≪ / RTI >

The multi-array antenna 151 may include a plurality of antennas each receiving a signal reflected from the target.

The RF receivers 153 can convert signals received for each antenna of the multiple array antenna 151 into complex signals of baseband. At this time, the signal received through each antenna of the multiple array antenna 151 is multiplied by a frequency-modulated continuous wave (FMCW) transmission signal of 77 GHz band provided by the RF transmitter 115 to generate a baseband signal having I channel and Q channel components It can be converted into a complex signal.

The analog-to-digital converter (ADC) 155 can convert the baseband complex signals (i.e., analog signals) converted by the RF receiver 153 into digital signals. At this time, the number of ADCs 155 may be proportional to the number of antennas included in the multiple array antenna 151.

A radar signal processor 157 may process the digital signals converted by the ADCs 155 to estimate information about the target. The radar signal processing unit 157 can calculate the distance, the velocity, and the azimuth of the target from the reflected signal that has undergone the ADC conversion in the ADC 255. The radar signal processor 157 is connected to the plurality of ADCs 155 through a serial bus interface. Accordingly, the converted signals (digital signals) from the ADCs 155 can be input to the radar signal processor 157 one bit at a time.

FIG. 2 is a diagram illustrating a frequency change occurring during one cycle of a frequency modulated continuous waveform (FMCW) used in the target detection system according to one embodiment.

Referring to FIG. 2, in a short range mode (SRR) of a frequency modulated continuous waveform (FMCW) during a cycle, a waveform for short range target detection is generated. In a long range mode (LRR) It can be seen that a waveform for detecting the target is generated. At this time, the waveform for the near target and the waveform for the far target can be generated continuously.

In FIG. 2, the waveform for proximity target detection may be composed of five chirps having different frequency change slopes with respect to time. In addition, the waveform for the remote target detection can be composed of seven pairs of different frequency gradient slopes with respect to time.

The null portion in FIG. 2 may have a value of 0.2 ms, for example, as a time interval for the processor to control the transmitter and the receiver before transmitting the frequency modulated continuous waveform (FMCW).

3 is a block diagram illustrating the structure of a radar signal processing apparatus using a weighted Kalman filter according to an embodiment of the present invention.

3, a radar signal processing apparatus 200 using a weighted Kalman filter according to an embodiment of the present invention includes an ADC symbol formatter 210,

A DC offset remover 220, a first window 230, a digital beamformer (DBF) 240, a second window 250, a Fast Fourier Transform (FFT) A weight calculator 260, a constant false alarm rate (CFAR) 270, an estimator 280, and a Kalman filter 290. [

The ADC symbol formatter 210 collects the serial bits input from the ADCs 155 and divides the serial bits into ADC symbols, rearranges the ADC symbols according to the arrangement order of the antennas, and outputs the ADC symbols .

The DC offset remover 220 can remove the DC component from the ADC symbols rearranged in the order of the antennas by the ADC symbol formatter 210.

The first window 230 applies a first window function to lower the side-lobe level of the interference signal for the ADC symbol from which the DC component is removed by the DC offset remover 220 . The first window 230 may reduce the side lobe level of the interfering signal using a first window function, such as, for example, a Hamming window or a Chebyshev window. The first window 230 applies a first window function prior to forming a digital beam in the digital beamformer 240 so that the first window 230 is a ' (Digital Beam Former (DBF) Window) '.

Generally, in a radar system using a frequency-modulated continuous waveform, it is necessary to estimate a mutation frequency and an additional Doppler spectrum according to each distance in order to remotely detect a moving target, and a baseband or intermediate frequency band A fast Fourier transform (FFT) technique can be used.

As is well known, the characteristics of the system may cause the radar antenna to be given a significantly shorter dwell time at which to acquire the reflected signal of the target. In this case, serious side effects such as side-lobes of the interfering signal may be obscured by the information of adjacent adjacent signals, which may lead to serious performance deterioration.

That is, when the reception time of the reflected wave reflected from the target is relatively short, a side leaf of a strong interference signal such as a clutter leaks to the adjacent Doppler filter, and a signal to be detected can be hidden. Here, clutter refers to a reflection obstacle such as echo caused by reflected waves generated from ground, sea surface, raindrops, etc. in a radar.

Accordingly, in one embodiment, it is possible to facilitate the detection of the bit frequency by lowering the side lobe level of the interference signal using various window functions.

The digital beamformer (DBF) 240 synthesizes signals received for each antenna based on the output signal of the first window 230, and outputs the synthesized signals on a beam-by-beam basis. The digital beamformer (DBF) 240 multiplies a gain using a predetermined weight value for each antenna output signal of the first window 230, and rotates a phase to synthesize a signal received in a specific direction . More specifically, the digital beamformer (DBF) 240 performs a digital operation (for example, an operation of multiplying each signal received by each antenna by a phase) on a signal received for each antenna using a preset weight value So that the received signals can be combined for each antenna. In this case, the predetermined weight value may be a weight for a signal received for each antenna.

The second window 250 may apply a second window function to reduce a leakage error occurring during fast Fourier transform on the signals synthesized by the digital beam former (DBF) 240.

Fast Fourier transform is a type of Discrete Fourier Transform (DFT), which can not actually receive data over a sufficient time, i.e., infinite time, due to memory capacity limitations. Therefore, the frequency domain data has a discrete property. Due to the discontinuity of the frequency, an error occurs between the time signal and the actual frequency in the spectrum. At this time, an error occurring between the time signal and the actual frequency is referred to as a leakage error.

The second window 250 may reduce the leakage error using a second window function, such as a Hamming window, a Hanning window, or a Chebyshev window, for example. Since the second window 250 applies the second window function prior to the fast Fourier transform in the Fast Fourier Transform (FFT) 260, the second window 250 may be referred to as an FFT window (Fast Fourier Transform (FFT)).

A Fast Fourier Transform (FFT) 260 can perform fast Fourier transform on the output signal of the second window 250, which represents a process of converting a bit frequency signal to a frequency axis.

The weight calculator 265 calculates the weighted power of the peak frequency signal output from the fast Fourier transformer 260 and provides it to the Kalman filter 290.

The result of fast Fourier transform performed by the fast Fourier transformer 260 is represented by a complex number, and the weight calculator 265 calculates a weighted power value for an arbitrary bit frequency x by Equation (1) below.

(1)

P _x = 20 log ₁₀ (r _x ² + i _x ² )

Here, P _x is the weighted power of the bit frequency x, r _x is the real part of the complex value of the bit frequency x, and i _x is the imaginary part of the complex value of the bit frequency x. That is, the complex value C _x = r _x + j i _x (where j ² = -1) corresponding to the x-bit frequency.

A constant false alarm rate (CFAR) 270 may select a specific bit frequency by applying a variable detection threshold to the output signal of the fast Fourier transformer 260. The CFAR 270 removes the clutter while maintaining a constant false alarm rate by applying a variable detection threshold in a variable noise and clutter environment and plays a role of selecting a meaningful bit frequency do.

(CFAR) 270 may be a cell-averaging CFAR, an order statistics CFAR, or a GO-CFAR, depending on the method of selecting the detection threshold T _x . -CFAR), and SO-CFAR (Smallest Of-CFAR).

Here, the constant false alarm rate (CFAR) 270 calculates a detection threshold value using the weighted power provided by the weight calculator 265 when calculating the detection threshold value T _x .

Estimator 280 may estimate information including the distance, velocity, and azimuth of the target from selected bit frequencies at a given false alarm probability detector (CFAR)

In the following embodiments, it is assumed that the number of receiving antennas, that is, the number of antennas included in the multi-array antenna is 8, the ADC supports 4 channels, and 16-bit sampling is performed .

On the other hand, the Kalman filter 290 optionally performs continuous tracking of the detected target.

The Kalman filter is an optimal estimation technique that finds target state variables using probabilistic models and measurements of the target. That is, the state value is estimated through the time propagation by the state equation and the improvement by the measurement equation, and the error covariance of the actual state value and the estimated state value is minimized. A Kalman filter driven by white noise with a linear target and a normal distribution is known as an unbiased unbiased estimator with a minimum covariance. The Kalman filter calculates the final estimated value by appropriately correcting the predicted value with the prediction error of the measured value by selecting the most probable value as the estimated value according to the probability distribution of the estimated value.

That is, based on the target model, the state of the next time point and the value of the error covariance are predicted, and a new estimated value is calculated by compensating the difference between the measured value and the predicted value. This estimate is then the final product of the Kalman filter. The process then continues to be repeated to update the new estimate continually. This feature enables the continuous tracking of the target using a Kalman filter.

However, as shown in FIG. 4, when the target is switched at an intersection or the like, an error may occur if the position is estimated by the Kalman filter.

In order to prevent this, the Kalman filter 290 of the present invention refers to the weighted power provided from the weight calculator 265, and when the weighted power becomes equal to or greater than a predetermined value, that is, The position of the corresponding target is adopted as the actual position without modification.

The Kalman filter 290 refers to the weighted power provided from the weight calculator 265. When the weighted power becomes less than a predetermined value, that is, when the weighted power of the beat frequency becomes less than a predetermined value, Modify and adopt the modified position.

Of course, the Kalman filter 290 may perform a more precise process in consideration of the detection threshold value T _x by the certain false alarm probability detector (CFAR) 270.

That is, the Kalman filter 290 compares the weighted power P _x with the detection threshold value T _x to determine the position information detected by the radar without using the estimated position when P _x / T _x is greater than or equal to N, (N is an arbitrary value of 2 or more).

The Kalman filter 290 uses the estimated position (where M is any value greater than or equal to 1 and less than N) when P _x / T _x is less than M.

In other cases, the Kalman filter 290 calculates a weighted average weighted average using the values of P _x and T _x to calculate the weighted estimated position.

(2)

H ' _x = ((NM-α) H _x + α D _x ) / (NM)

Here, H ' _x is the estimated position to which the weight is applied, H _x is the estimated position of the Kalman filter,? Is P _x / T _x -M, and D _x is the radar measurement position.

According to the present invention as described above, a weight is applied to the Kalman filter to enable more accurate position tracking.

The foregoing description is merely illustrative of the technical idea of the present invention, and various changes and modifications may be made by those skilled in the art without departing from the essential characteristics of the present invention. Therefore, the embodiments of the present invention are not intended to limit the scope of the present invention but to limit the scope of the present invention. The scope of protection of the present invention should be construed according to the following claims, and all technical ideas within the scope of equivalents thereof should be construed as being included in the scope of the present invention.

110: Transmitter
111: Timing controller
113: Frequency Modulation Continuous Waveform Generator (FMCW Generator)
115: RF transmitter
150: receiver
151: Multi-array antenna
153: RF receiver (RF receiver) 253
155: Analog-to-Digital Converter (ADC)
157: Radar Signal Processor (Radar Signal Processor)
210: ADC Symbol Formatter
220: DC Offset Remover
230: 1st window
240: Digital Beam Former (DBF)
250: second window
260: Fast Fourier Transform (FFT)
265: Weight calculator
270: Constant False Alarm Rate (CFAR)
280: Estimator
290: Kalman filter

Claims (8)

A plurality of antennas for receiving signals reflected from the target, a plurality of antennas for receiving the signals reflected from the target, and a plurality of baseband digital complex signals for dividing the input serial bits into ADC symbols, An ADC symbol formatter for rearranging and outputting in the order of the symbols;
A DC offset remover for removing a DC component from the rearranged ADC symbol;
A first window for applying a window function to lower a side-lobe level of the interference signal for the ADC symbol from which the DC component is removed;
A digital beamformer (DBF) for synthesizing the signals received for each antenna based on the output signal of the first window and outputting the synthesized signals for each beam;
A second window for applying a window function for reducing a leakage error occurring during fast Fourier transform on the synthesized signals;
A Fast Fourier Transformer (FFT) for performing a fast Fourier transform on the output signal of the second window;
A weight calculator for calculating a weighted power of bit frequencies of an output signal of the fast Fourier transformer;
A constant false alarm rate (CFAR) for selecting specific bit frequencies by applying a variable detection threshold to an output signal of the FFT;
An estimator for estimating information including the distance, velocity and azimuth of the target from the selected bit frequencies; And
And a Kalman filter for tracking the target in consideration of the weighted power calculated from the weight calculator. The method according to claim 1,
Wherein the Kalman filter determines a position estimated by the estimator when the weighted power input from the weight calculator is equal to or greater than a predetermined value. The method according to claim 1,
Wherein the Kalman filter tracks the position estimated by the estimator and corrects the position to compensate for the difference from the measured value when the weighted power input from the weight calculator is less than a predetermined value. A radar signal processing device using the same. The method according to claim 1,
Wherein the weight calculator provides the calculated weighted power to a certain false alarm probability detector and the predetermined false alarm probability detector uses a weighted Kalman filter to transform the detection threshold value by reflecting the input weighted power. The method according to claim 1 or 4,
Wherein the Kalman filter determines a position estimated by the estimator when the value obtained by dividing the weighted power inputted from the weight calculator by the detection threshold value of the predetermined false alarm probability detector is equal to or greater than a first predetermined value, A radar signal processing apparatus using a weighted Kalman filter having a constant value of 2 or more. The method according to claim 1 or 4,
Wherein the Kalman filter calculates a corrected position by applying a weighted average method when the value obtained by dividing the weighted power inputted from the weight calculator by the detection threshold value of the certain false alarm probability detector is less than the first predetermined value and is equal to or greater than the second predetermined value Wherein the second constant value is smaller than the first constant value, the first constant value is an arbitrary value greater than or equal to 2, the second constant value is an arbitrary value smaller than or equal to 1 and smaller than the first constant value, A radar signal processing device using a filter. The method according to claim 1 or 4,
Wherein the Kalman filter corrects a position estimated by the estimator if the value obtained by dividing the weighted power inputted from the weight calculator by the detection threshold value of the predetermined false alarm probability detector is less than a second predetermined value smaller than the first predetermined value, Wherein the first constant value is an arbitrary value equal to or greater than 2, and the second constant value is equal to or greater than 1 and smaller than the first constant value. A multi-array antenna including a plurality of antennas each receiving a signal reflected from the target;
An RF receiver for converting a signal received by each antenna of the multiple array antenna into complex signals of a baseband;
Analog-to-digital converters (ADCs) for converting the baseband complex signals into digital signals; And
And a radar signal processing device for processing the digital signals to estimate information about the target and perform target tracking using a Kalman filter to which a weight based on the power of a bit frequency is applied.

Images