CN105629211A - Signal processing method for vehicle lane change assistance system based on combined waveform for multi-target detection and vehicle lane change assistance system - Google Patents

Abstract

A signal processing method of a combined waveform automobile lane change auxiliary system for multi-target detection and the automobile lane change auxiliary system belong to the field of signal processing and are used for solving the problem that the existing automobile lane change auxiliary system is difficult to solve the multi-target resolving, and the technical key points are as follows: the waveform comprises a first section of constant frequency wave CW1, a second section of sawtooth wave FMCW1, a third section of constant frequency wave CW2 and a fourth section of sawtooth wave FMCW 2. The effect is as follows: due to the adoption of the four-segment waveform, a multi-target detection function can be realized, a real target can be detected, and a false target can be removed.

Description

Signal processing method for vehicle lane change assistance system based on combined waveform for multi-target detection and vehicle lane change assistance system

technical field

The invention belongs to the field of signal processing, and in particular relates to a signal processing method of a vehicle lane-changing auxiliary system with combined waveforms.

Background technique

With the progress of modern society, the use of cars and the rate of use have shown rapid growth, but with the problems caused by cars, there are more and more problems, and even more serious, such as the consumption of energy by cars, environmental hazards, and traffic safety. Among them, the problem of traffic safety has attracted more and more people's attention. The corresponding car collision avoidance function, car lane change assist function, automatic parking function, automatic cruise function, etc. have been applied to the car.

The vehicle lane change assist system provides an effective basis for the driver to change the road, and can effectively reduce the traffic accidents caused by the vehicle lane change. With the continuous development of sensor technology, the millimeter wave radar sensor is small in size and light in weight. The advantage of being able to be used in relatively severe rainy and snowy weather is being increasingly used in automobiles. At the same time, linear frequency modulated continuous wave (LFMCW) radar has high velocity resolution and distance resolution, so linear frequency modulated continuous wave (LFMCW) is more widely used in millimeter wave radar systems. However, when the linear frequency modulation continuous wave radar detects moving targets, if it only adopts a single waveform design, such as a single constant frequency wave, triangular wave, sawtooth wave and other waveforms, it is difficult to solve the multi-target problem.

Contents of the invention

In order to better solve the problem of multi-target resolution in the existing vehicle lane change assistance system, the present invention provides a signal processing method of a multi-target detection combined waveform vehicle lane change assistance system to realize multi-target solution Calculate.

In order to achieve the above object, the technical solution of the present invention is: a signal processing method of a combined waveform vehicle lane change assist system for multi-target detection. Three sections of constant frequency wave CW2, the fourth section of sawtooth wave FMCW2, the method includes the following steps:

S1. Carry out FFT calculation for each segment of waveform and IQ data collected by A/D;

S2. Threshold detection is performed on the complex modulus values after FFT transformation of each segment of the waveform, and the position of the crossing threshold point is output.

The present invention also relates to a vehicle lane change assisting system, which executes the above method.

Beneficial effects: due to the adoption of four-segment waveforms, the multi-target detection function can be realized, real targets can be detected, and false targets can be removed.

Description of drawings

Fig. 1 Frequency change diagram of constant frequency wave CW and sawtooth wave FMCW within a frequency sweep period;

Figure 2 (R, V) space diagram of a single target;

Figure 3 (R, V) space diagram of multiple targets;

Figure 4 is a flow chart of the signal processing of the vehicle lane change assist system based on groups and waveforms.

detailed description

Embodiment 1: a signal processing method of a combined waveform vehicle lane change assist system for multi-target detection, the waveform includes the first section of constant frequency wave CW1, the second section of sawtooth wave FMCW1, the third section of constant frequency wave CW2, the first section of constant frequency wave CW2, Four sawtooth waves FMCW2, wherein, the first constant frequency wave CW1 and the third constant frequency wave CW2 are constant frequency bands, the second sawtooth wave FMCW1 and the fourth sawtooth wave FMCW2 are sawtooth wave bands, and the method includes the following steps:

S1. to each section waveform, the IQ data that A/D collects, carry out FFT calculation; As the optimization of technical scheme, it can adopt following method: to the first section constant frequency wave CW1 in channel 1, the second section sawtooth wave FMCW1, the third section of CW2 and the fourth section of sawtooth wave FMCW2, the IQ data collected by A/D, select 256 points of data with high linearity in each section, and perform 256 points of FFT respectively, for the first section of constant frequency in channel 2 Wave CW1 and the second segment of sawtooth wave FMCW1, IQ data collected by A/D, select 256 points of data with high linearity in each segment, and perform 256 points of FFT respectively. Among them, high linearity should be better expressed by good linearity. Linearity means that there is a certain linear relationship in the collected data, and poor linearity means that the data in this segment does not show a certain linear relationship. What is characteristic of the nonlinear relationship. In order to analyze the data more accurately, only the part showing a linear relationship in the collected data is selected for FFT analysis.

S2. Threshold detection is performed on the complex modulus values after FFT transformation of each segment of the waveform, and the position of the crossing threshold point is output. As a preferred technical solution, it can adopt the following method: in channel 1, the first section of constant frequency wave CW1 has n1 points to cross the threshold, then the position matrix A _CW1 of the corresponding first section of constant frequency wave crossing the threshold point =[ a ₁ , a ₂ ,…a _n1 ], the second section of sawtooth wave FMCW1, n2 points pass the threshold, then the position matrix of the corresponding second section of sawtooth wave FMCW1 crossing the threshold point is B _FMCW1 =[b ₁ ,b ₂ , ...b _n2 ], the third section of constant frequency wave CW2, there are n3 points passing the threshold, then the corresponding position matrix of the third section of constant frequency wave CW2 passing the threshold point is C _CW2 =[c ₁ ,c ₂ ,...c _n3 ] , the fourth sawtooth wave FMCW2, there are n4 points crossing the threshold, then the corresponding position matrix of the fourth sawtooth wave FMCW2 passing the threshold point is D _FMCW2 =[d ₁ ,d ₂ ,…d _n4 ], in channel 2, For the first section of constant frequency wave CW1, n5 points pass the threshold, then the position matrix of the corresponding first section of constant frequency wave CW1 passing the threshold point is E _CW1 =[e ₁ ,e ₂ ,…e _n5 ], the second section of sawtooth Wave FMCW1 has n6 points that pass the threshold, and the corresponding position matrix of the second sawtooth wave FMCW1 crossing the threshold is G _FMCW1 =[g ₁ ,g ₂ ,...g _n6 ]. If the position point passing the threshold is equal to 1, it is considered as a DC component, and it is not used as a target judgment, and the position point is directly eliminated.

This embodiment describes a waveform with a center frequency of 24 GHz or 77 GHz, based on a combination of a CW signal modulated by a constant frequency wave and an FMCW signal modulated by a sawtooth wave (as a preferred form of the above waveform), and the modulation waveform is used to realize A signal processing method for an automobile lane change assist system. The vehicle lane change assist system designed by this method can realize the calculation of the relative distance, relative speed and direction angle of the target within the coverage of the millimeter-wave radar beam behind the vehicle. The combined waveform designed by FMCW can realize the multi-target detection problem, which makes the system have better accuracy and rapidity for multi-target detection, and ensures the safety of the driver when changing roads while driving.

Waveform design and waveform analysis:

This embodiment provides a specific waveform diagram at a working frequency of 24.128 GHz at a center frequency f. The first segment of the waveform is a constant frequency wave CW1, and the operating frequency is 24.128 GHz. The second segment of the waveform is a rising sawtooth wave FMCW1, and the operating frequency changes. The range changes from 24.128GHz to 24.278GHz, the bandwidth is 150MHz, the third section is constant frequency wave CW2, the working frequency is 24278GHz, the fourth section is sawtooth wave FMCW2, the working frequency range is from 24.278GHz to 24.128GHz. The period T of each segment is 5ms. The frequency changes of the constant frequency wave CW and the sawtooth wave FMCW within a frequency sweep period are shown in Figure 1.

The reasons for choosing this design waveform in this embodiment are:

(1) Improve the calculation accuracy of relative speed and relative distance.

In the first section of waveform - constant frequency wave CW1, according to the characteristics of constant frequency wave, the Doppler frequency value fd1 caused by speed can be obtained, and according to the second section of waveform - sawtooth wave FMCW1, the difference frequency of the target can be calculated Frequency value fo1. The relative velocity value v1 of the target can be calculated from the Doppler frequency value fd1 obtained by the first waveform, and the Doppler frequency value obtained by the first waveform and the difference frequency value fo1 obtained by the second waveform can be calculated. Calculate the relative distance R1 of the target.

Similarly, the third waveform - constant frequency wave CW2, can also obtain the target's Doppler frequency value fd2, and according to the fourth segment waveform - sawtooth wave FMCW2, the target's difference frequency value fo2 can be calculated. The relative velocity value v2 of the target can be calculated from the Doppler frequency value fd2 obtained by the third waveform, and the Doppler frequency value obtained by the fourth waveform and the difference frequency value fo2 obtained by the second waveform can be calculated. Calculate the relative distance R2 of the target.

The relative distance and relative speed of the real target can be finally obtained by using the four-segment combined waveform and processing according to the multi-target matching algorithm in the later stage.

Through four combined waveforms, relative speed v1 and relative speed v2, relative distance R1 and relative distance R2 are obtained. Theoretically, v1=v2, R1=R2, but due to factors such as the complexity of the equipment use environment, it may cause interference and other factors, resulting in a deviation between the measured v1 value and v2 value, and there is also a deviation between R1 and R2. If a single Using v1 or v2 as the relative speed of the target, and R1 or R2 as the relative speed of the target will cause inaccuracy in obtaining the relative speed and distance of the target. As shown in Figure 2, it is the (R, V) space diagram of a single target. It can be seen from the figure that four straight lines determine an intersection point, and the accuracy of the obtained relative distance value and relative speed value is higher than that determined by two straight lines. At the same time, even if the value calculated by one band is relatively inaccurate, the relative distance value and relative speed value determined by the other three straight lines will be stronger than the relative distance value and relative speed determined by two straight lines value, so that the system has a certain anti-interference ability and robust characteristics. Therefore, through the four bands, the accuracy of target speed and distance calculation can be effectively improved, and the system has anti-interference ability and robust characteristics;

(2) Due to the use of four-segment waveforms, the multi-target detection function can be realized, and real targets can be detected and false targets can be removed.

Because the relative speed and relative distance of the real target have nothing to do with the frequency modulation period T, but the distance and speed calculation values of the false target are related to the frequency modulation period T. Therefore, for the real target, its real value always exists in different periods, while the false value changes with the period T in different periods. As shown in Figure 3, it is the (R, V) space diagram of multiple targets. As can be seen from Figure 2, the four bands will more accurately obtain the relative distance and relative velocity of a target. Detect real targets and remove false targets.

Through the above two proofs, the waveform designed in this embodiment can effectively improve the accuracy of the calculation of the relative distance value and relative speed value of the target, and at the same time make the system have anti-interference characteristics and robust characteristics, and can effectively It can effectively detect multiple targets, and can effectively obtain real targets and remove false targets.

Embodiment 2: have the technical scheme identical with embodiment 1, more specifically: described method also comprises step: S3. Calculate and obtain the Doppler frequency value of constant frequency band, thus obtain Doppler frequency matrix.

Embodiment 3: It has the same technical solution as Embodiment 1 or 2, more specifically: the method also includes the step: S4. Calculate the beat frequency value of the sawtooth band, thereby obtaining the beat frequency matrix.

Embodiment 4: have the technical scheme identical with any one of embodiment 1-3, more specifically: also comprise step: S5. calculate relative velocity matrix according to the Doppler frequency value obtained. As an optimization of the technical solution, according to the Doppler frequency value matrix calculated in each section, the specific method for calculating the velocity matrix of the target is: the formula for calculating the velocity is Wherein, c is the speed of light, c=3×10 ⁸ , and f is the center frequency f=24.128 GHz. According to the Doppler frequency matrix FA _CW1 =[fd _a1 ,fd _a2 ,…fd _an1 ] obtained from the first constant frequency wave CW1, the velocity matrix is VA _CW1 =[v _a1 ,v _a2 ,…v _an1 ], According to the Doppler frequency matrix FC _CW1 =[fd _c1 ,fd _c2 ,...fd _cn3 ] obtained from the second constant frequency wave CW2, the velocity matrix is VC _CW1 =[v _c1 ,v _c2 ,...v _cn3 ].

Embodiment 5: have the same technical scheme as any one of Embodiment 1-4, more specifically: also include step: S6. According to the calculated Doppler frequency matrix and difference frequency matrix, calculate the relative distance matrix.

Embodiment 6: have the technical scheme identical with any one of embodiment 1-5, more specifically: this method also comprises step: S9. carry out the calculation of the azimuth angle of multi-target.

Embodiment 7: have the same technical scheme as any one of Embodiment 1-6, more specifically: also include step: S7. Find the real distance of multiple targets.

Embodiment 8: have the technical scheme identical with any one of embodiment 1-7, more specifically: also comprise step: S8. carry out the matching and/or S10 of the real distance value of multi-target and its corresponding speed value. According to the real distance value of multiple targets, match the azimuth angles of multiple targets.

Embodiment 9: have the technical scheme identical with any one of embodiment 1-8, more specifically: the concrete step that calculates the Doppler frequency value of constant frequency band is: in channel 1, first section constant frequency wave The position matrix A _CW1 of the CW crossing threshold point =[a ₁ ,a ₂ ,…a _n1 ], calculate the Doppler frequency value at the corresponding point according to the following rules, and obtain the Doppler frequency matrix as FA _CW1 =[fd _a1 , fd _a2 ,...fd _an1 ], similarly, for channel 1, the position matrix C _CW2 =[c ₁ ,c ₂ ,...c _n3 ] of the third segment of constant frequency wave CW2 crossing the threshold point, the corresponding point is calculated according to the following rules The Doppler frequency value above, the obtained Doppler frequency matrix is FC _CW1 =[fd _c1 , fd _c2 ,...fd _cn3 ].

The rule is, if the number of points is 1≤x _i ≤128 (1≤i≤n), it is judged that the target is approaching, and the Doppler frequency at the corresponding point If the number of points is 128< _xi ≤256 (1≤i≤n), it is judged that the target is far away, and the Doppler frequency on the corresponding point ${\mathrm{fd}}_i=-\frac{(256-(x_i-1\left)\right)*f_{\mathrm{the\; s}}}{256},(1<i\&le;\mathrm{no});$

Embodiment 10: have the same technical scheme as any one of Embodiments 1-9, more specifically: the specific steps for calculating the difference frequency value of the sawtooth wave band are: In channel 1, the second section of sawtooth wave FMCW1 passes through The position matrix B _FMCW1 of the threshold point =[b ₁ ,b ₂ ,...b _n2 ], calculate the beat frequency matrix FB _FMCW1 =[fo _b1 ,fo _b2 ,...fo _bn2 ] at the corresponding point according to the following rules, similarly, For channel 1, the position matrix D _FMCW2 = [d ₁ , d ₂ ,...d _n4 ] of the fourth segment of the sawtooth wave FMCW2 crossing the threshold point, calculate the difference frequency matrix FD at the corresponding point according to the following rules _FMCW1 = [fo _d1 ,fo _d2 ,...fo _dn4 ].

The rule is, that is, if the number of points is 1≤y _j ≤128 (1≤j≤n), the difference frequency value on the corresponding point If the number of points is 128<y _j ≤256 (1≤j≤n), the beat frequency value on the corresponding point ${\mathrm{fo}}_j=-\frac{(256-({\mathrm{the\; y}}_j-1\left)\right)*f_{\mathrm{the\; s}}}{256},(1<j\&le;\mathrm{no}).$

Embodiment 11: It has the same technical solution as any one of Embodiments 1-10, more specifically: according to the calculated Doppler frequency matrix and difference frequency matrix, the specific steps for calculating the relative distance matrix are: calculating The distance formula is Among them, T is the action time of each waveform, T=5ms, B is the frequency modulation bandwidth, B=150MHz, fd is the Doppler frequency value, fo is the difference frequency value.

According to channel 1, the Doppler frequency matrix FA _CW1 =[fd _a1 , fd _a2 ,... fd _an1 ] obtained by the first constant frequency wave CW1 and the difference frequency matrix FB FMCW1 obtained by the second sawtooth wave _FMCW1 =[ fo _b1 ,fo _b2 ,…fo _bn2 ], all the elements in the Doppler matrix and all the elements in the beat frequency matrix are paired one by one to calculate the relative distance matrix, and the calculated relative distance matrix is Where r _aibj (1≤i≤n1,1≤j≤n2) represents the difference frequency between the i-th element in the Doppler frequency matrix obtained by the first constant frequency wave CW1 and the second sawtooth wave FMCW1 The distance value obtained by calculating the jth element in the frequency matrix. In the same way, the Doppler frequency matrix FC _CW1 =[fd _c1 , fd _c2 ,...fd _cn3 ] obtained for the second section constant frequency wave CW2 and the difference frequency matrix FD FMCW2 obtained by the fourth section sawtooth wave _FMCW2 =[fo _d1 ,fo _d2 ,…fo _dn4 ], the above-mentioned processing is also carried out, and finally the relative distance matrix is obtained as Among them, r _cicj (1≤i≤n3,1≤j≤n4) represents the difference frequency between the i-th element in the Doppler frequency matrix obtained by the third segment of constant frequency wave CW2 and the fourth segment of sawtooth wave FMCW2 The distance value obtained by calculating the jth element in the frequency matrix.

Embodiment 12: It has the same technical solution as any one of Embodiments 1-11, more specifically: the specific steps for calculating the azimuth angle of multiple targets are:

(1) First calculate the phase value corresponding to each threshold point.

In channel 1, the position matrix A _CW1 of the first section of constant frequency wave CW1 passing the threshold point =[a ₁ ,a ₂ ,…a _n1 ], calculate the phase value at the corresponding point according to the following calculation method, and obtain the phase matrix as ψA _CW1 =[ψ _a1 ,ψ _a2 ,…ψ _an1 ], in channel 1, the position matrix B of the threshold point of the second sawtooth wave FMCW1 _FMCW1 =[b ₁ ,b ₂ ,…b _n2 ], according to the following calculation method to calculate the corresponding The beat frequency value at the point ψB _FMCW1 =[ψ _b1 ,ψ _b2 ,...ψ _bn2 ]. In channel 2, the position matrix E _CW1 of the first section of constant frequency wave CW2 passing the threshold point =[e ₁ ,e ₂ ,…e _n5 ], calculate the phase value at the corresponding point according to the following calculation method, and obtain the phase matrix as ψE _CW1 =[ψ _e1 ,ψ _e2 ,…ψ _en5 ], in channel 2, the position matrix G _FMCW1 of the second sawtooth wave FMCW1 crossing the threshold point =[g ₁ ,g ₂ ,…g _n6 ], according to the following calculation method to calculate the corresponding The beat frequency value at the point ψG _FMCW1 =[ψ _g1 ,ψ _g2 ,...ψ _gn6 ].

Wherein, the method of calculating the phase is to obtain a matrix of respective complex values after calculating the FFT of each segment of the waveform. The method of calculating the phase value according to the characteristics of complex numbers is assuming that the complex number is c=a+j*b=cosθ+j*sinθ, then $ta\mathrm{no}\mathrm{\&psi;}=\frac{\mathrm{the\; s}i\mathrm{no}\mathrm{\&theta;}}{co\mathrm{the\; s}\mathrm{\&theta;}}=\frac{b}{a},$ get the phase value of the complex number $\mathrm{\&psi;}=\arctan \left(\frac{\mathrm{the\; s}i\mathrm{no}\mathrm{\&theta;}}{co\mathrm{the\; s}\mathrm{\&theta;}}\right)=ta\mathrm{no}\left(\frac{b}{a}\right).$

(2) Calculate the phase difference.

Calculate the phase difference between the first constant frequency wave CW1 of channel 1 and the first constant frequency wave CW2 of channel 2, and the phase difference matrix is obtained as Calculate the phase difference between the second sawtooth wave FMCW1 of channel 1 and the second sawtooth wave FMCW1 of channel 2, and the phase difference matrix is obtained as

The formula for calculating the specific phase difference is: $ta\mathrm{no}\left({\mathrm{\&Delta;\&psi;}}_{aiej}\right)=ta\mathrm{no}({\mathrm{\&psi;}}_{ej}-{\mathrm{\&psi;}}_{ai})=\frac{ta\mathrm{no}\left({\mathrm{\&psi;}}_{ej}\right)-ta\mathrm{no}\left({\mathrm{\&psi;}}_{ai}\right)}{1+ta\mathrm{no}\left({\mathrm{\&psi;}}_{ej}\right)ta\mathrm{no}\left({\mathrm{\&psi;}}_{ai}\right)},1\&le;i\&le;\mathrm{no}1,1\&le;j\&le;5$ ${\mathrm{\&Delta;\&psi;}}_{aiej}=\arctan ({\mathrm{\&psi;}}_{ej}-{\mathrm{\&psi;}}_{ai})=\arctan (\frac{ta\mathrm{no}\left({\mathrm{\&psi;}}_{ej}\right)-ta\mathrm{no}\left({\mathrm{\&psi;}}_{ai}\right)}{1+ta\mathrm{no}\left({\mathrm{\&psi;}}_{ej}\right)ta\mathrm{no}\left({\mathrm{\&psi;}}_{ai}\right)},(1\&le;i\&le;\mathrm{no}1,1\&le;j\&le;5).$

(3) Calculate the azimuth.

After obtaining the phase difference matrix, according to the formula azimuth angle formula, Among them, d=7.5mm is the antenna spacing, and λ=12.4mm.

Calculate the azimuth angle between the first constant frequency wave CW1 of channel 1 and the first constant frequency wave CW2 of channel 2, and the azimuth matrix is obtained as Calculate the azimuth angle between the second sawtooth wave FMCW1 of channel 1 and the second sawtooth wave FMCW1 of channel 2, and obtain the phase difference matrix as

Embodiment 13: have the same technical solution as any one of Embodiments 1-12, more specifically: the specific steps for finding the true distance of multiple targets are:

(1) Since the Doppler value obtained by the first section of constant frequency wave CW1 is in one-to-one correspondence with the difference frequency obtained by the second section of sawtooth wave FMCW1, that is, in In the first row of the relative distance matrix, only one value is the true distance value of the target, and the other distance values are the distance values of false targets generated by matching the Doppler value with the wrong beat frequency value. Similarly, there is also a one-to-one correspondence between the Doppler value obtained from the third constant frequency wave CW2 and the beat frequency value obtained from the fourth sawtooth wave FMCW2.

(2) Since the value of the real target has nothing to do with the cycle, but the value of the false target has a relationship with the cycle, that is, in the matrix of the relative distance matrix R _AB and R _CD , the distance values of the real target are equal, and the false target The target's distance value is completely different. As long as you find a value with equal distance in the two matrices, it is the distance value of the real target.

(3) Since the dimensions of the relative distance matrix R _AB and the relative distance matrix R _CD may be different, this is due to the fact that when the threshold detection is performed, the number of target points detected after each band passes the threshold is caused by a certain difference . The principle of processing is to search for the true value of the distance based on the matrix whose sum of the rows and columns of the relative distance matrix R _AB and the relative distance matrix R _CD is the smallest.

(4) When a distance true value is found in the relative distance matrix R _AB and the relative distance matrix R _CD , record the row value and column value of the matrix where the true value is located, and at the same time, obtain the row and column values of the matrix where the true value is located All the values of the columns are removed, and all the true values are searched in turn until the dimension of the matrix as the reference is 0.

Embodiment 14: have the technical scheme identical with any one of embodiment 1-13, more specifically:

The specific steps of speed value matching are: use the row value where the found real distance value is located, as a relative speed matrix VA _CW1 =[v _a1 ,v _a2 ,...v _an1 ] or VC _CW1 =[v _c1 ,v _c2 ,... v _cn3 ] column value of the matrix, find the speed value corresponding to the real target distance value, so that the matching of the distance and speed of the real target is completed;

The specific steps of azimuth matching are: according to the seventh step, when the true value of the distance is found in the relative distance matrix R _AB , record the row value and column value of the relative distance matrix where the true value of the target is located, and in the relative velocity matrix VA When the true velocity value is found in _CW1 , record the column values of the relative velocity matrix. Use the row value where the found true distance value is located, for all the azimuth angle values in the same row of the azimuth matrix θ _AE and the column value of the true speed value, and for all the azimuth angle values in the row of the same value for the azimuth matrix θ _BG , to find the value of the same azimuth angle, after finding the azimuth angle corresponding to the target, record the row value and column value of the matrix where the azimuth angle is located, and at the same time, in the matrix where the azimuth angle obtained by θ _AE and θ _BG is located, the corresponding row Remove all the values of the sum and column, according to the number of rows where the real target is located in the relative distance matrix R _AB , search the corresponding direction angles of the targets in turn until the direction angles corresponding to all the real targets are found.

This embodiment provides a combined waveform design scheme that can realize multi-target detection, and at the same time provides a theoretical analysis that can realize multi-target detection, and provides a waveform design when designing other waveforms to achieve multi-target detection train of thought;

The lane change assistance system designed in this embodiment can realize the detection of the relative distance and relative speed of multiple targets, and at the same time can realize the detection function of the target direction angle, and realize the distinction of the spatial orientation of multiple targets.

This embodiment provides a detailed signal processing process, including the calculation of the relative velocity of multiple targets, the calculation of the relative distance, the matching method of the relative distance and relative velocity of the real target, the calculation of the phase difference, the calculation of the direction angle, and The matching method of the real target direction angle and other processing procedures and related formulas. This part provides a signal processing method for those who design lane change assistance systems.

Embodiment 15: A signal processing method of a combined waveform vehicle lane change assist system for multi-target detection, the waveforms include the first section of constant frequency wave CW1, the second section of sawtooth wave FMCW1, the third section of constant frequency wave CW2, the first section of constant frequency wave Four sections of sawtooth wave FMCW2, the method comprises the following steps:

S1. Carry out FFT calculation for each segment of waveform and IQ data collected by A/D;

S2. Threshold detection is performed on the complex modulus value after FFT transformation of each segment of waveform, and the position of the threshold crossing point is output;

S3. Calculate and obtain the Doppler frequency value of the constant frequency band.

S4. Calculate and obtain the beat frequency value of the sawtooth band.

S5. Calculate the relative velocity matrix according to the obtained Doppler frequency value.

S6. Calculate and obtain a relative distance matrix according to the calculated Doppler frequency matrix and beat frequency matrix.

S7. Find the real distance of multiple targets.

S8. Matching the real distance values of multiple targets with their corresponding speed values and/or S10. Matching the azimuth angles of multiple targets according to the real distance values of multiple targets.

S9. Calculating the azimuth angles of multiple targets.

Embodiment 16: A signal processing method of a combined waveform vehicle lane change assistance system for multi-target detection, the waveforms include the first section of constant frequency wave CW1, the second section of sawtooth wave FMCW1, the third section of constant frequency wave CW2, the first section of constant frequency wave Four sections of sawtooth wave FMCW2, the method comprises the following steps:

S1. Carry out FFT calculation for each section of waveform and IQ data collected by A/D: the first section of constant frequency wave CW1, the second section of sawtooth wave FMCW1, the third section of CW2 and the fourth section of sawtooth wave in channel 1 FMCW2, IQ data collected by A/D, select 256 points of data with high linearity in each segment, and perform 256 points of FFT respectively, for the first segment of constant frequency wave CW1 and the second segment of sawtooth wave FMCW1 in channel 2, A/ For the IQ data collected by D, select 256 points of data with high linearity in each segment, and perform 256 points of FFT respectively;

S2. Perform threshold detection on the complex modulus value after FFT transformation of each segment of waveform, and output the position of the threshold crossing point: suppose that in channel 1, the first segment of constant frequency wave CW1, if there are n1 points passing the threshold, then the corresponding first segment of constant frequency The position matrix of the wave passing the threshold point is A _CW1 =[a ₁ ,a ₂ ,…a _n1 ], the second segment of the sawtooth wave FMCW1, if n2 points pass the threshold, then the corresponding position of the second segment of the sawtooth wave FMCW1 passing the threshold point The matrix is B _FMCW1 = [b ₁ ,b ₂ ,...b _n2 ], the third section of constant frequency wave CW2, if n3 points pass the threshold, the corresponding position matrix of the third section of constant frequency wave CW2 passing the threshold point is C _CW2 =[c ₁ ,c ₂ ,...c _n3 ], the fourth section of sawtooth wave FMCW2 has n4 points that pass the threshold, then the corresponding position matrix of the fourth section of sawtooth wave FMCW2 that crosses the threshold is D _FMCW2 =[d ₁ ,d ₂ ,...d _n4 ], assuming that in channel 2, the first section of constant frequency wave CW1 has n5 points passing the threshold, then the corresponding position matrix of the first section of constant frequency wave CW1 passing the threshold point is E _CW1 =[e ₁ , e ₂ ,…e _n5 ], the second sawtooth wave FMCW1, there are n6 points passing the threshold, then the position matrix of the corresponding second sawtooth wave FMCW1 passing the threshold point is G _FMCW1 =[g ₁ ,g ₂ ,…g _n6 ]. If the position point passing the threshold is equal to 1, it is considered to be a DC component, and it is not used as a target judgment, and the position point is directly eliminated;

S3. Calculate the Doppler frequency value of the constant frequency band: in channel 1, the position matrix A _CW1 of the first constant frequency wave CW crossing the threshold point =[a ₁ ,a ₂ ,…a _n1 ], calculate the corresponding according to the following rules The Doppler frequency value on the point, the obtained Doppler frequency matrix is FA _CW1 =[fd _a1 , fd _a2 ,...fd _an1 ], similarly, for channel 1, the third constant frequency wave CW2 passes the threshold point Position matrix C _CW2 =[c ₁ ,c ₂ ,...c _n3 ], calculate the Doppler frequency value at the corresponding point according to the following rules, and obtain the Doppler frequency matrix as FC _CW1 =[fd _c1 ,fd _c2 ,...fd _cn3 ].

The rule is, if the number of points is 1≤x _i ≤128 (1≤i≤n), it is judged that the target is approaching, and the Doppler frequency at the corresponding point If the number of points is 128< _xi ≤256 (1≤i≤n), it is judged that the target is far away, and the Doppler frequency on the corresponding point ${\mathrm{fd}}_i=-\frac{(256-(x_i-1\left)\right)*f_{\mathrm{the\; s}}}{256},(1<i\&le;\mathrm{no});$

S4. Calculate the difference frequency value of the sawtooth wave band: in channel 1, the position matrix B _{FMCW1 of the threshold point of the second sawtooth wave FMCW1} =[b ₁ ,b ₂ ,...b _n2 ], calculate the corresponding point according to the following rules The beat frequency matrix FB _FMCW1 =[fo _b1 ,fo _b2 ,...fo _bn2 ], similarly, for channel 1, the position matrix D _FMCW2 of the fourth sawtooth wave FMCW2 passing the threshold point =[d ₁ ,d ₂ , ...d _n4 ], calculate the beat frequency matrix FD _FMCW1 =[fo _d1 , fo _d2 ,...fo _dn4 ] at the corresponding point according to the following rules.

The rule is, that is, if the number of points is 1≤y _j ≤128 (1≤j≤n), the difference frequency value on the corresponding point If the number of points is 128<y _j ≤256 (1≤j≤n), the beat frequency value on the corresponding point ${\mathrm{fo}}_j=-\frac{(256-({\mathrm{the\; y}}_j-1\left)\right)*f_{\mathrm{the\; s}}}{256},(1<j\&le;\mathrm{no});$

S5. Calculate the relative velocity matrix according to the obtained Doppler frequency value: the calculation velocity formula is Wherein, c is the speed of light, c=3×10 ⁸ , and f is the center frequency f=24.128 GHz. According to the Doppler frequency matrix FA _CW1 =[fd _a1 ,fd _a2 ,…fd _an1 ] obtained from the first constant frequency wave CW1, the relative velocity matrix is VA _CW1 =[v _a1 ,v _a2 ,…v _an1 ] , according to the Doppler frequency matrix FC _CW1 =[fd _c1 ,fd _c2 ,…fd _cn3 ] obtained from the second constant frequency wave CW2, its relative velocity matrix is VC _CW1 =[v _c1 ,v _c2 ,…v _cn3 ].

S6. Calculate the relative distance matrix according to the calculated Doppler frequency matrix and difference frequency matrix: the formula for calculating the distance is Among them, T is the action time of each waveform, T=5ms, B is the frequency modulation bandwidth, B=150MHz, fd is the Doppler frequency value, fo is the difference frequency value. According to channel 1, the Doppler frequency matrix FA _CW1 =[fd _a1 , fd _a2 ,... fd _an1 ] obtained by the first constant frequency wave CW1 and the difference frequency matrix FB FMCW1 obtained by the second sawtooth wave _FMCW1 =[ fo _b1 ,fo _b2 ,…fo _bn2 ], all the elements in the Doppler matrix and all the elements in the beat frequency matrix are paired one by one to calculate the relative distance matrix, and the calculated relative distance matrix is Where r _aibj (1≤i≤n1,1≤j≤n2) represents the difference frequency between the i-th element in the Doppler frequency matrix obtained by the first constant frequency wave CW1 and the second sawtooth wave FMCW1 The distance value obtained by calculating the jth element in the frequency matrix. In the same way, the Doppler frequency matrix FC _CW1 =[fd _c1 , fd _c2 ,...fd _cn3 ] obtained for the second section constant frequency wave CW2 and the difference frequency matrix FD FMCW2 obtained by the fourth section sawtooth wave _FMCW2 =[fo _d1 ,fo _d2 ,…fo _dn4 ], the above-mentioned processing is also carried out, and finally the relative distance matrix is obtained as Among them, r _cicj (1≤i≤n3,1≤j≤n4) represents the difference frequency between the i-th element in the Doppler frequency matrix obtained by the third segment of constant frequency wave CW2 and the fourth segment of sawtooth wave FMCW2 The distance value obtained by calculating the jth element in the frequency matrix;

S7. Find the real distance of multiple targets:

(1) Since the Doppler value obtained by the first section of constant frequency wave CW1 is in one-to-one correspondence with the difference frequency obtained by the second section of sawtooth wave FMCW1, that is, in In the first row of the relative distance matrix, only one value is the true distance value of the target, and the other distance values are the distance values of false targets generated by matching the Doppler value with the wrong beat frequency value. Similarly, there is also a one-to-one correspondence between the Doppler value obtained from the third constant frequency wave CW2 and the beat frequency value obtained from the fourth sawtooth wave FMCW2.

(2) Since the value of the real target has nothing to do with the cycle, but the value of the false target has a relationship with the cycle, that is, in the matrix of the relative distance matrix R _AB and R _CD , the distance values of the real target are equal, and the false target The target's distance value is completely different. As long as you find a value with equal distance in the two matrices, it is the distance value of the real target.

(3) Since the dimensions of the relative distance matrix R _AB and the relative distance matrix R _CD may be different, this is due to the fact that when the threshold detection is performed, the number of target points detected after each band passes the threshold is caused by a certain difference . The principle of processing is to search for the true value of the distance based on the matrix whose sum of the rows and columns of the relative distance matrix R _AB and the relative distance matrix R _CD is the smallest.

(4) When a distance true value is found in the relative distance matrix R _AB and the relative distance matrix R _CD , record the row value and column value of the matrix where the true value is located, and at the same time, obtain the row and column values of the matrix where the true value is located All the values of the columns are removed, and all the true values are searched in turn until the dimension of the matrix as the reference is 0.

S8. Carry out the matching of the real distance value of multiple targets and its corresponding speed value and/or S10. According to the real distance value of multiple targets, perform the matching of multiple target azimuth angles:

Velocity value matching: Use the value of the row where the found real distance value is located, as a relative velocity matrix VA _CW1 =[v _a1 ,v _a2 ,…v _an1 ] or VC _CW1 =[v _c1 ,v _c2 ,…v _cn3 ] matrix to find the speed value corresponding to the distance value of the real target, so as to complete the matching of the distance and speed of the real target;

Azimuth matching: when the true value of the distance is found in the relative distance matrix R _AB , record the row value and column value of the relative distance matrix where the true value of the target is located, and when the true value of the speed is found in the relative speed matrix VA _CW1 , record the relative Column values of the velocity matrix. Use the row value where the found true distance value is located, for all the azimuth angle values in the same row of the azimuth matrix θ _AE and the column value of the true speed value, for all the azimuth angle values in the same row of the azimuth matrix θ _AE , to find the value of the same azimuth angle, after finding the azimuth angle corresponding to the target, record the row value and column value of the matrix where the azimuth angle is located, and at the same time, in the matrix where the azimuth angle obtained by θ _AE and θ _BG is located, the corresponding row Remove all the values of the sum and column, according to the number of rows where the real target is located in the relative distance matrix R _AB , search the corresponding direction angles of the targets in turn until the direction angles corresponding to all the real targets are found.

S9. Carry out the calculation of the azimuth angle of multiple targets:

(1) First calculate the phase value corresponding to each threshold point.

In channel 1, the position matrix A _CW1 of the first section of constant frequency wave CW1 passing the threshold point =[a ₁ ,a ₂ ,…a _n1 ], calculate the phase value at the corresponding point according to the following calculation method, and obtain the phase matrix as ψA _CW1 =[ψ _a1 ,ψ _a2 ,…ψ _an1 ], in channel 1, the position matrix B of the threshold point of the second sawtooth wave FMCW1 _FMCW1 =[b ₁ ,b ₂ ,…b _n2 ], according to the following calculation method to calculate the corresponding The beat frequency value at the point ψB _FMCW1 =[ψ _b1 ,ψ _b2 ,...ψ _bn2 ]. In channel 2, the position matrix E _CW1 of the first section of constant frequency wave CW2 passing the threshold point =[e ₁ ,e ₂ ,…e _n5 ], calculate the phase value at the corresponding point according to the following calculation method, and obtain the phase matrix as ψE _CW1 =[ψ _e1 ,ψ _e2 ,…ψ _en5 ], in channel 2, the position matrix G _FMCW1 of the second sawtooth wave FMCW1 crossing the threshold point =[g ₁ ,g ₂ ,…g _n6 ], according to the following calculation method to calculate the corresponding The beat frequency value at the point ψG _FMCW1 =[ψ _g1 ,ψ _g2 ,...ψ _gn6 ].

Wherein, the method of calculating the phase is to obtain a matrix of respective complex values after calculating the FFT of each segment of the waveform. The method of calculating the phase value according to the characteristics of complex numbers is, assuming that the complex number is c=a+j*b=cosθ+j*sinθ, then $ta\mathrm{no}\mathrm{\&psi;}=\frac{\mathrm{the\; s}i\mathrm{no}\mathrm{\&theta;}}{co\mathrm{the\; s}\mathrm{\&theta;}}=\frac{b}{a},$ get the phase value of the complex number $\mathrm{\&psi;}=\arctan \left(\frac{\mathrm{the\; s}i\mathrm{no}\mathrm{\&theta;}}{co\mathrm{the\; s}\mathrm{\&theta;}}\right)=ta\mathrm{no}\left(\frac{b}{a}\right).$

(2) Calculate the phase difference.

Calculate the phase difference between the first constant frequency wave CW1 of channel 1 and the first constant frequency wave CW2 of channel 2, and the phase difference matrix is obtained as Calculate the phase difference between the second sawtooth wave FMCW1 of channel 1 and the second sawtooth wave FMCW1 of channel 2, and the phase difference matrix is obtained as

The formula for calculating the specific phase difference is: $ta\mathrm{no}\left({\mathrm{\&Delta;\&psi;}}_{aiej}\right)=ta\mathrm{no}({\mathrm{\&psi;}}_{ej}-{\mathrm{\&psi;}}_{ai})=\frac{ta\mathrm{no}\left({\mathrm{\&psi;}}_{ej}\right)-ta\mathrm{no}\left({\mathrm{\&psi;}}_{ai}\right)}{1+ta\mathrm{no}\left({\mathrm{\&psi;}}_{ej}\right)ta\mathrm{no}\left({\mathrm{\&psi;}}_{ai}\right)},1\&le;i\&le;\mathrm{no}1,1\&le;j\&le;5$ ${\mathrm{\&Delta;\&psi;}}_{aiej}=\arctan ({\mathrm{\&psi;}}_{ej}-{\mathrm{\&psi;}}_{ai})=\arctan (\frac{ta\mathrm{no}\left({\mathrm{\&psi;}}_{ej}\right)-ta\mathrm{no}\left({\mathrm{\&psi;}}_{ai}\right)}{1+ta\mathrm{no}\left({\mathrm{\&psi;}}_{ej}\right)ta\mathrm{no}\left({\mathrm{\&psi;}}_{ai}\right)},(1\&le;i\&le;\mathrm{no}1,1\&le;j\&le;5).$

(3) Calculate the azimuth.

After obtaining the phase difference matrix, according to the formula azimuth angle formula, Among them, d=7.5mm is the antenna spacing, and λ=12.4mm.

Calculate the azimuth angle between the first constant frequency wave CW1 of channel 1 and the first constant frequency wave CW2 of channel 2, and the azimuth matrix is obtained as Calculate the azimuth angle between the second sawtooth wave FMCW1 of channel 1 and the second sawtooth wave FMCW1 of channel 2, and obtain the phase difference matrix as

Embodiment 17: An automobile lane-changing assisting system, performing the signal processing method of the multi-target detection combined waveform automobile lane-changing assisting system described in any one of the technical solutions of embodiments 1-16.

The group and waveform of constant frequency wave and sawtooth wave designed in this embodiment, as well as the selection of parameters involved in the waveform are not limited to the parameters disclosed in this embodiment. Those skilled in the art can select different design parameters according to specific application scenarios, or It is to improve the waveform, etc.; within the technical scope disclosed in the present invention, any equivalent replacement or change according to the technical solution and the inventive concept of the present invention shall be covered within the scope of protection of the present invention.

Claims (15)

1. A signal processing method of a combined waveform vehicle lane change assist system for multi-target detection, characterized in that: the waveform includes the first section of constant frequency wave CW1, the second section of sawtooth wave FMCW1, and the third section of constant frequency wave CW2 , the fourth section sawtooth wave FMCW2, the method comprises the following steps: S1. Carry out FFT calculation for each segment of waveform and IQ data collected by A/D; S2. Threshold detection is performed on the complex modulus values after FFT transformation of each segment of the waveform, and the position of the crossing threshold point is output. 2. The method according to claim 1, further comprising the step: S3. Calculate the Doppler frequency value of the constant frequency band. 3. The method according to claim 2, further comprising the step: S4. Calculate the beat frequency value of the sawtooth band. 4. The method according to claim 2 or 3, further comprising the step: S5. Calculate the relative velocity matrix according to the obtained Doppler frequency value. 5. The method according to claim 3, further comprising the step of: S6. Calculate the relative distance matrix according to the calculated Doppler frequency matrix and beat frequency matrix. 6. The method according to claim 5, further comprising step: S9. Carrying out calculation of azimuth angles of multiple targets. 7. The method according to claim 5 or 6, further comprising the step: . S7. Finding the real distance of multiple targets. 8. The method according to claim 7, further comprising the step of: S8. matching the real distance values of multiple targets with their corresponding speed values and/or S10. Matching of target azimuth. 9. method as claimed in claim 2 is characterized in that, the method that calculates the Doppler frequency value of horizontal frequency band is: in passage 1, the position matrix A _CW1 =[ a ₁ ,a ₂ ,…a _n1 ], calculate the Doppler frequency value at the corresponding point according to the following rules, and obtain the Doppler frequency matrix as FA _CW1 =[fd _a1 ,fd _a2 ,…fd _an1 ]; for channel 1 Among them, the position matrix C _CW2 =[c ₁ ,c ₂ ,…c _n3 ] of the third section of constant frequency wave CW2 passing the threshold point, calculate the Doppler frequency value at the corresponding point according to the following rules, and obtain the Doppler frequency matrix For FC _CW1 = [fd _c1 , fd _c2 ,... fd _cn3 ]; The rule is, if the number of points is 1≤x _i ≤128 (1≤i≤n), it is judged that the target is approaching, and the Doppler frequency at the corresponding point (1≤i≤n); if the number of points is 128< _xi ≤256 (1≤i≤n), it is judged that the target is far away, and the Doppler frequency on the corresponding point (1<i≤n). 10. The method according to claim 3, characterized in that, the method for calculating the beat frequency value of the sawtooth band is: in the passage 1, the position matrix B _FMCW1 =[b ₁ of the second section sawtooth wave FMCW1 crossing the threshold point , _{b 2} ,... _{b n2} ], calculate the beat frequency matrix FB _FMCW1 at the corresponding point according to the following rules _: The position matrix D _FMCW2 =[d ₁ ,d ₂ ,...d _n4 ], and the difference frequency matrix FD _FMCW1 =[fo _d1 ,fo _d2 ,...fo _dn4 ] at the corresponding point is calculated according to the following rules; The rule is, that is, if the number of points is 1≤y _j ≤128 (1≤j≤n), the difference frequency value on the corresponding point (1≤j≤n); if the number of points is 128<y _j ≤256(1≤j≤n), the difference frequency value on the corresponding point ${\mathrm{fo}}_j=-\frac{(256-({\mathrm{the\; y}}_j-1\left)\right)*f_{\mathrm{the\; s}}}{256}$ (1<j≤n). 11. The method according to claim 5, characterized in that the method for calculating the relative distance matrix is: the calculation distance formula is Among them, T is the action time of each waveform, B is the frequency modulation bandwidth, fd is the Doppler frequency value, fo is the difference frequency value; According to channel 1, the Doppler frequency matrix FA _CW1 =[fd _a1 , fd _a2 ,... fd _an1 ] obtained by the first constant frequency wave CW1 and the difference frequency matrix FB FMCW1 obtained by the second sawtooth wave _FMCW1 =[ fo _b1 ,fo _b2 ,…fo _bn2 ], all the elements in the Doppler frequency matrix and all the elements in the beat frequency matrix are paired one by one to calculate the relative distance matrix, and the calculated relative distance matrix is Where r _aibj (1≤i≤n1,1≤j≤n2) represents the difference frequency between the i-th element in the Doppler frequency matrix obtained by the first constant frequency wave CW1 and the second sawtooth wave FMCW1 The distance value obtained by calculating the jth element in the frequency matrix; the Doppler frequency matrix FC _CW1 =[fd _c1 , fd _c2 ,... fd _cn3 ] obtained for the second segment of constant frequency wave CW2 and the fourth segment of sawtooth wave FMCW2 The obtained difference frequency matrix FD _FMCW2 =[fo _d1 ,fo _d2 ,…fo _dn4 ], the above-mentioned processing is also carried out, and finally the relative distance matrix is obtained as Among them, r _cicj (1≤i≤n3,1≤j≤n4) represents the difference frequency between the i-th element in the Doppler frequency matrix obtained by the third segment of constant frequency wave CW2 and the fourth segment of sawtooth wave FMCW2 The distance value obtained by calculating the jth element in the frequency matrix. 12. the method for claim 6 is characterized in that, the calculation method of the azimuth angle of multi-target is: (1) First calculate the phase value corresponding to each threshold point In channel 1, the position matrix A _CW1 of the first section of constant frequency wave CW1 passing the threshold point =[a ₁ ,a ₂ ,…a _n1 ], calculate the phase value at the corresponding point according to the following calculation method, and obtain the phase matrix as ψA _CW1 =[ψ _a1 ,ψ _a2 ,...ψ _an1 ]; In channel 1, the position matrix B _FMCW1 of the second sawtooth wave FMCW1 passing the threshold point =[b ₁ ,b ₂ ,…b _n2 ], calculate the difference frequency value ψB at the corresponding point according to the following calculation method _FMCW1 =[ψ _b1 ,ψ _b2 ,...ψ _bn2 ]; In channel 2, the position matrix E _CW1 of the first section of constant frequency wave CW2 passing the threshold point =[e ₁ ,e ₂ ,…e _n5 ], calculate the phase value at the corresponding point according to the following calculation method, and obtain the phase matrix as ψE _CW1 =[ψ _e1 ,ψ _e2 ,...ψ _en5 ]; In channel 2, the position matrix G _FMCW1 =[g ₁ ,g ₂ ,…g _n6 ] of the second segment of the sawtooth wave FMCW1 passing the threshold point, calculate the difference frequency value at the corresponding point according to the following calculation method ψG _FMCW1 =[ψ _g1 ,ψ _g2 ,...ψ _gn6 ]; Among them, the method of calculating the phase value is to obtain the matrix of the respective complex values after calculating the FFT of each segment of the waveform, and the method of calculating the phase value according to the characteristics of the complex number is assuming that the complex number is c=a+j*b=cosθ+j *sinθ, then $ta\mathrm{no}\mathrm{\&psi;}=\frac{\sin \mathrm{\&theta;}}{co\mathrm{the\; s}\mathrm{\&theta;}}=\frac{b}{a},$ get the phase value of the complex number $\mathrm{\&psi;}=arcta\mathrm{no}\left(\frac{\sin \mathrm{\&theta;}}{co\mathrm{the\; s}\mathrm{\&theta;}}\right)=ta\mathrm{no}\left(\frac{b}{a}\right);$ (2) Calculate the phase difference Calculate the phase difference between the first constant frequency wave CW1 of channel 1 and the first constant frequency wave CW2 of channel 2, and the phase difference matrix is obtained as Calculate the phase difference between the second sawtooth wave FMCW1 of channel 1 and the second sawtooth wave FMCW1 of channel 2, and the phase difference matrix is obtained as The formula for calculating the specific phase difference is: $ta\mathrm{no}\left({\mathrm{\&Delta;\&psi;}}_{aiej}\right)=ta\mathrm{no}({\mathrm{\&psi;}}_{ej}-{\mathrm{\&psi;}}_{ai})=\frac{ta\mathrm{no}\left({\mathrm{\&psi;}}_{ej}\right)-ta\mathrm{no}\left({\mathrm{\&psi;}}_{ai}\right)}{1+ta\mathrm{no}\left({\mathrm{\&psi;}}_{ej}\right)ta\mathrm{no}\left({\mathrm{\&psi;}}_{ai}\right)},1\&le;i\&le;\mathrm{no}1,1\&le;j\&le;5,$ ${\mathrm{\&Delta;\&psi;}}_{aiej}=\arctan \left({\mathrm{\&psi;}}_{ej}-{\mathrm{\&psi;}}_{ai}\right)=\arctan \left(\frac{\mathrm{the\; tan}\left({\mathrm{\&psi;}}_{ej}\right)-\mathrm{the\; tan}\left({\mathrm{\&psi;}}_{ai}\right)}{1+\mathrm{the\; tan}\left({\mathrm{\&psi;}}_{ej}\right)\mathrm{the\; tan}\left({\mathrm{\&psi;}}_{ai}\right)}\right),\left(1\&le;i\&le;\mathrm{no}1,1\&le;j\&le;\mathrm{no}5\right);$ (3) Calculate the azimuth After obtaining the phase difference matrix, according to the formula azimuth angle formula, Among them, d is the antenna spacing, and λ is the wavelength; Calculate the azimuth angle between the first constant frequency wave CW1 of channel 1 and the first constant frequency wave CW2 of channel 2, and the azimuth matrix is obtained as Calculate the azimuth angle between the second sawtooth wave FMCW1 of channel 1 and the second sawtooth wave FMCW1 of channel 2, and obtain the phase difference matrix as 13. The method according to claim 7, wherein the method for finding the true distance of multiple targets is: Based on the matrix whose sum of the rows and columns of the relative distance matrix R _AB and the relative distance matrix R _CD is the smallest, search for the true value of the distance; find a distance in the relative distance matrix R _AB and the relative distance matrix R _CD When the true value is found, record the row value and column value of the matrix where the true value is located. At the same time, remove all the row and column values in the matrix where the obtained true value is located, and search all the true values in turn until the benchmark until the matrix dimension is 0. 14. The method of claim 7, wherein, The method of speed value matching is: use the row value of the matrix where the found real distance value is located, as the relative speed matrix VA _CW1 =[v _a1 ,v _a2 ,...v _an1 ] or VC _CW1 =[v _c1 ,v _c2 ,... v _cn3 ] column value of the matrix, find the speed value corresponding to the real target distance value, so that the matching of the distance and speed of the real target is completed; The method of azimuth matching is: when finding the true value of the distance in the relative distance matrix R _AB , record the row value and column value of the relative distance matrix where the true value of the target is located at the same time, and when finding the true value of the speed in the relative speed matrix VA _CW1 , record the column values of the relative velocity matrix, use the row values where the found real distance values are located, all the azimuth angle values in the same row of the azimuth angle matrix θ _AE and the column values where the real velocity values are located are the same for the azimuth angle matrix θ _BG For all the azimuth values in the row of numerical values, find the value of the same azimuth angle. After finding the azimuth angle corresponding to the target, record the row value and column value of the matrix where the azimuth angle is located. At the same time, the θ _AE and θ _BG obtained In the matrix where the direction angle is located, all the values of the corresponding rows and columns are removed, and according to the number of rows where the real target is located in the relative distance matrix R _AB , the corresponding direction angle of the target is searched in turn until the direction corresponding to all the real targets is found horn. 15. An automobile lane-changing assisting system, characterized in that it executes the signal processing method of the multi-target detection combined waveform automobile lane-changing assisting system described in any one of claims 1-14.