variable long signal integration

Abstract

A vehicle, system, and method of determining a velocity of an object. The system includes a radar system and a processor. The radar system is configured to obtain, over an integration interval, a radar signal regarding the object, the radar signal including a plurality of velocity samples. The processor is configured to divide the integration interval into a plurality of time segments, each time segment including a subset of the velocity samples, perform a first integration of the subset of the velocity samples over a selected time segment using a first set of velocity hypotheses to obtain a first stage integration value for the time segment, perform a second integration using the first stage integration value using a second set of velocity hypotheses to obtain a second stage integration value over the integration interval, and determine a velocity of the object from the second stage integration value.

Description

Technical Field

[0001] This subject matter discloses radar systems and methods for determining the velocity of an object, and more specifically, systems and methods for increasing the signal-to-noise ratio of a signal obtained over an integral interval of a radar system with significant velocity shifts. Background Technology

[0002] A radar system used on a vehicle transmits a source signal into the vehicle's environment and receives reflections of the source signal from objects in the environment. The source and reflected signals can then be compared to determine the object's velocity relative to the vehicle. The reflected signal is sampled over a time interval called the integration interval, and a transformation is performed to generate a frequency distribution in a frequency space with a grid defining multiple velocity assumptions. The vehicle's velocity can be determined by finding the peak location of the frequency distribution in the frequency space. As an object moves through space and / or changes its velocity during the integration interval, the peak of the frequency distribution broadens, reducing velocity resolution. Resolution can be improved by increasing the number of radar samples by sampling the reflected signal over longer integration intervals. However, increasing the duration of the integration interval increases the required processing power and computation time. Therefore, it is desirable to provide a method for determining object velocity that reduces the computational complexity inherent in long integration intervals. Summary of the Invention

[0003] In one exemplary embodiment, a method for determining the velocity of an object is disclosed. A radar signal about the object is obtained over an integration interval. The radar signal includes multiple velocity samples. The integration interval is divided into multiple time periods. Each time period includes a subset of the velocity samples. A first integration of the subset of velocity samples is performed over the selected time period using a first set of velocity assumptions to obtain a first-stage integration value for the selected time period. A second integration is performed using the first-stage integration value and a second set of velocity assumptions to obtain a second-stage integration value over the integration interval. The velocity of the object is determined by the second-stage integration value.

[0004] In addition to one or more features described herein, the first resolution of the first set of velocity assumptions is smaller than the second resolution of the second set of velocity assumptions, and the number of velocity samples used as input for the first integration is greater than the number of first integration values ​​used as input for the second integration. The first-stage integration value is obtained by performing a first integration on the product of velocity samples within a selected time period and a first phase term corresponding to a velocity assumption selected from the first set of velocity assumptions. The second-stage integration value is obtained by performing a second integration on the product of the first-stage integration value and a second phase term corresponding to a velocity assumption from the second set of velocity assumptions. Performing the second integration includes determining an interpolation function for the first-stage integration value and performing coherent integration on the product of the interpolation function and the second phase term. Performing the first integration includes obtaining the first-stage integration value for multiple time periods and each of the first set of velocity assumptions. The vehicle is navigated based on the velocity of the object relative to the object.

[0005] In another exemplary embodiment, a system for determining the velocity of an object is disclosed. The system includes a radar system and a processor. The radar system is configured to acquire radar signals about the object over an integration interval, the radar signals including multiple velocity samples. The processor is configured to divide the integration interval into multiple time periods, each time period including a subset of velocity samples; to perform a first integration of the subset of velocity samples within the selected time period using a first set of velocity assumptions to obtain a first-stage integration value for that time period; to perform a second integration using the first-stage integration value using a second set of velocity assumptions to obtain a second-stage integration value over the integration interval; and to determine the velocity of the object based on the second-stage integration value.

[0006] In addition to one or more features described herein, the first resolution of the first set of velocity assumptions is smaller than the second resolution of the second set of velocity assumptions, and the number of velocity samples used as input for the first integration is greater than the number of first-stage integration values ​​used as input for the second integration. The processor is also configured to perform a first integration on the product of velocity samples within a selected time period and a first phase term corresponding to a velocity assumption selected from the first set of velocity assumptions. The processor is also configured to perform a second integration on the product of the first-stage integration value and a second phase term corresponding to the second set of velocity assumptions. The processor is further configured to perform the second integration by determining an interpolation function for the first-stage integration value and performing a coherent integration on the product of the interpolation function and the second phase term. The processor is also configured to obtain first-stage integration values ​​for each of multiple time periods and for each of the first set of velocity assumptions. The processor is also configured to navigate the vehicle based on the velocity of the object relative to the object.

[0007] In yet another exemplary embodiment, a vehicle is disclosed. The vehicle includes a radar system and a processor. The radar system is configured to acquire radar signals from an object over an integration interval, the radar signals including multiple velocity samples. The processor is configured to divide the integration interval into multiple time periods, each time period including a subset of velocity samples; perform a first integration of the subset of velocity samples within the selected time period using a first set of velocity assumptions to obtain a first-stage integration value for that time period; perform a second integration using the first-stage integration value using a second set of velocity assumptions to obtain a second-stage integration value over the integration interval; and determine the velocity of the object based on the second-stage integration value.

[0008] In addition to one or more features described herein, the first resolution of the first set of velocity assumptions is smaller than the second resolution of the second set of velocity assumptions, and the number of velocity samples used as input to the first integration is greater than the number of first-stage velocity values ​​used as input to the second integration stage. The processor is also configured to perform a first integration on the product of velocity samples within a selected time period and a first phase term corresponding to a velocity assumption selected from the first set of velocity assumptions. The processor is also configured to perform a second integration on the product of the first-stage integration value and a second phase term corresponding to the second set of velocity assumptions. The processor is further configured to perform the second integration by determining an interpolation function for the first-stage integration value and performing a coherent integration on the product of the interpolation function and the second phase term. The processor is also configured to obtain the first-stage integration value for each of a plurality of time periods and for each of the first set of velocity assumptions.

[0009] The above-described features and advantages, as well as other features and advantages, of this disclosure will become apparent when taken in conjunction with the accompanying drawings and the following detailed description. Attached Figure Description

[0010] Other features, advantages, and details appear by way of example only in the following detailed description, which refers to the accompanying drawings, wherein:

[0011] Figure 1 An autonomous vehicle is shown in an exemplary embodiment;

[0012] Figure 2 A schematic diagram illustrating the effect of the duration of the integration interval on velocity resolution is shown;

[0013] Figure 3 The signal integration process suitable for determining the velocity of an object from radar signals is illustrated in one embodiment.

[0014] Figure 4 A diagram illustrating the hierarchical stages of the signal processing method disclosed herein is shown;

[0015] Figure 5A diagram illustrating the complexity of a traditional single-stage integration process using time intervals without time segments or hierarchical integration stages is shown.

[0016] Figure 6 A diagram illustrating the complexity of each stage of the multi-stage integration process disclosed herein is shown;

[0017] Figure 7 A flowchart of the signal integration process disclosed herein is shown;

[0018] Figure 8 The image shows a distance Doppler plot obtained from an object approaching the autonomous vehicle; and

[0019] Figure 9 The distance Doppler map of the object is shown, obtained using a long-duration integral interval and the processing method disclosed herein. Detailed Implementation

[0020] The following description is merely exemplary in nature and is not intended to limit this disclosure, its application, or use. It should be understood that in all the drawings, corresponding reference numerals denote similar or corresponding parts and features.

[0021] According to an exemplary embodiment, Figure 1 An autonomous vehicle 10 is illustrated. In an exemplary embodiment, the autonomous vehicle 10 is a so-called Level 4 or Level 5 automation system. A Level 4 system signifies "high automation," referring to the performance of the automated driving system in specific driving modes across all aspects of a dynamic driving task, even if the human driver does not appropriately respond to intervention requests. A Level 5 system signifies "full automation," referring to the full-time performance of the automated driving system across all aspects of a dynamic driving task under all road and environmental conditions manageable by a human driver. It should be understood that the systems and methods disclosed herein can also be used in autonomous vehicles operating at any of Levels 1 through 5.

[0022] An autonomous vehicle 10 typically includes at least a navigation system 20, a propulsion system 22, a transmission system 24, a steering system 26, a braking system 28, a sensor system 30, an actuator system 32, and a controller 34. The navigation system 20 determines a road-level route plan for the autonomous vehicle 10. The propulsion system 22 provides power for generating motion for the autonomous vehicle 10 and, in various embodiments, may include an internal combustion engine, an electric motor such as a traction motor, and / or a fuel cell propulsion system. The transmission system 24 is configured to transmit power from the propulsion system 22 to two or more wheels 16 of the autonomous vehicle 10 according to a selectable speed ratio. The steering system 26 affects the position of two or more wheels 16. Although described for illustrative purposes as including a steering wheel 27, in some embodiments contemplated within the scope of this disclosure, the steering system 26 may not include a steering wheel 27. The braking system 28 is configured to provide braking torque to two or more wheels 16.

[0023] Sensor system 30 includes radar system 40, which senses objects in the external environment of autonomous vehicle 10 and determines various parameters of the objects. These parameters help to locate the position and relative speed of various remote vehicles in the autonomous vehicle environment. These parameters can be provided to controller 34. In operation, radar system 40 transmits radio frequency (RF) reference signals 48, which are reflected by one or more objects 50 in the field of view of autonomous vehicle 10 as one or more reflected echo signals or reflected signals 52, which are received back at radar system 40. The one or more reflected signals 52 can be used to determine various parameters of one or more objects 50, such as the distance of the object, the Doppler frequency or relative radial velocity of the object, and azimuth. Sensor system 30 includes additional sensors, such as digital cameras, for identifying road features, etc.

[0024] The controller 34 includes a processor 36 and a computer-readable storage device or storage medium 38. The storage medium 38 includes a program or instructions 39 that, when executed by the processor 36, operate the autonomous vehicle 10 based on outputs from the sensor system 30. The controller 34 can establish a trajectory for the autonomous vehicle 10 based on the outputs of the sensor system 30 and provide this trajectory to the actuator system 32 to control the propulsion system 22, the transmission system 24, the steering system 26, and / or the braking system 28, thereby navigating the autonomous vehicle 10 relative to one or more objects 50 based on parameters of one or more objects 50, such as speed, distance, etc.

[0025] Figure 2A schematic diagram 200 illustrates the effect of the duration of the integration interval on velocity resolution. Both the short integration interval 206 and the long integration interval 208 of the radar signal are shown in the time domain 202. A Fast Fourier Transform (FFT) is performed over the integration interval to generate a frequency distribution in the frequency space. A Fourier Transform is performed on the short integration interval 206 to generate a first frequency distribution with a first peak 210 in the frequency domain 204. Performing a Fourier Transform on the short integration interval 206 results in a first frequency grid with first grid markings 212, where each first grid marking provides a first set of velocity assumptions that can be used during computation to determine the location of the first peak 210. The first grid markings 212 are... Figure 2 The large grid markers in the diagram represent this. In various embodiments, the position of the first peak 210 is related to the velocity of the object, but it can also be used to determine the distance to the object.

[0026] A Fourier transform is performed over the long integration interval 208 to generate a second peak 214 in the frequency domain 204. Performing the Fourier transform over the long integration interval 208 produces a second frequency grid with second grid markings 216. The second grid markings 216 include large grid markings and small grid markings. Each second grid marking 216 provides a second set of velocity assumptions that can be used during computation to determine the location of the second peak 214. The first resolution of the first peak 210 is less than the second resolution of the second peak 214. Therefore, the velocity resolution determined from the first peak 210 is less than the velocity resolution determined from the second peak 214.

[0027] like Figure 2 As shown, the number of first grid markers 212 is less than the number of second grid markers 216. In various embodiments, the first set of first grid markers 212 is a suitable subset of the second set of second grid markers 216. Figure 2 In the illustrative example, every fifth grid marker in the second group is also a grid marker in the first group. Therefore, the second frequency grid has five times the velocity assumption of the first frequency grid, allowing the second frequency grid to have higher resolution for the resulting peaks. This is evident, as shown by the fact that the second peak 214 is narrower than the first peak 210.

[0028] In one embodiment, Figure 3A signal integration process 300, applicable to determining the velocity of an object based on radar signals, is schematically illustrated. The signal integration process comprises a series of integration stages, in which the input or samples of each integration stage are integrated with a velocity hypothesis. In each successive integration stage, the resolution of the velocity hypothesis increases while the number of input samples decreases. The signal integration process 300 includes dividing the integration interval into multiple time periods; in a first integration stage, low-resolution integration is performed over each time period using a velocity hypothesis corresponding to a short duration; and in a second integration stage, high-resolution integration is performed using the result of the first integration stage and a velocity hypothesis corresponding to a long duration time interval. For long integration intervals, the signal integration process provides high velocity resolution while reducing computational complexity.

[0029] The signal integration process 300 begins with a long integration interval 302. The number of samples obtained over the long integration interval 302 is N. During the preprocessing stage, the long integration interval 302 is divided into P time periods 304. Figure 3 The symbols are labeled TS1, TS2, ..., TS P Each time interval 304 is equal in duration and includes at least one sample from the long integral interval 302. The number of samples in each time interval is given by N1 = N / P. As an example, the long integral interval 302 may include 30 samples and may be divided into 5 time intervals 304, each time interval 304 having 6 samples.

[0030] The first integration phase is represented by grid 306, where multiple first-stage integration values ​​z are determined for each time interval 304. The first-stage integration values ​​z are generated by integrating samples within a selected time interval with velocity hypotheses selected from a set of low-resolution velocity hypotheses (i.e., the first set of velocity hypotheses). For P time intervals and Q hypotheses, the number of first-stage integration values ​​is P*Q. As schematically shown, the first-stage integration generates P*Q blocks arranged in grid 306, where the grid length indicates the number of time intervals and the grid width indicates the number of velocity hypotheses. Each block in the grid represents a first-stage integration value obtained by coherently integrating the product of a velocity sample within a time interval and a phase term corresponding to a velocity hypothesis. The velocity hypothesis includes a velocity term v and related terms, such as an acceleration term. (i.e., the first derivative of velocity with respect to time) and the jerk term (i.e., the second derivative of velocity with respect to time). In one embodiment, the first-stage integral value of the selected block is given by equation (1):

[0031]

[0032] Where i is the index of the i-th time interval indicating the integration interval, N1 is the number of samples in each time interval, and n is the index of the samples in that time interval. The parameter T is the duration of the time interval, and λ is the frequency of the source signal. It is the nth sample in the i-th time period. It is an assumption about the i-th time period and a specific speed. The first stage integral value.

[0033] The second integration stage is represented by grid 308. In the second integration stage, integration is performed using the product of the first-stage integration value z and the phase term corresponding to a set of high-resolution velocity assumptions (i.e., the second set of velocity assumptions). Because the second set of velocity assumptions has a higher resolution than the first set of velocity assumptions, the second-stage integration produces a higher resolution for the frequency peak. The second integration stage includes a coherent integration process to produce the second-stage integration value u using equation (2):

[0034]

[0035] Where N2 is the number of time periods, and It is an interpolation function relative to the selected first integration stage value. The interpolation function can be obtained by using nearest neighbor values ​​or... The interpolation is used to determine this.

[0036] Figure 4 Figure 400 illustrates the hierarchical stages of the signal processing method disclosed herein. (See Figure 400.) Figure 4 As shown, the signal processing method may include more than two integration stages. Results of multiple integrations are displayed. The first integration stage 404 is shown as a 5×5 block grid. The number of columns in the grid is equal to P (i.e., the number of time intervals in the integration interval 402), and the number of rows is proportional to N / P (i.e., the number of samples in each time interval). The number of velocity assumptions in the second integration stage 406 is greater than the number of velocity assumptions in the first integration stage 404, while the number of inputs to the second integration stage 406 is less than the number of inputs to the first integration stage 404. The number of velocity assumptions in the third integration stage 408 is greater than the number of velocity assumptions in the second integration stage 406, while the number of inputs to the third integration stage 408 is less than the number of inputs to the second integration stage 406. For each successive stage, the number of velocity assumptions increases, while the number of inputs from the previous integration stage decreases.

[0037] Figure 5 Figure 500 illustrates the complexity of a conventional single-stage integration process using time intervals without time segmentation or hierarchical integration stages. Rectangle 502 represents the complexity of the conventional process. The complexity of this processing operation is on the order of N^2, where n is the number of samples in the long integration interval.

[0038] Figure 6 Figure 600 illustrates the complexity of each stage of the multi-stage integration process disclosed herein. The complexity of each integration stage is represented by a rectangular region. Stage complexity refers to the amount of computational processing required in that stage. The length of each rectangle represents the number of inputs, and the width of each rectangle represents the number of velocity assumptions used in the corresponding stage. Figure 6 As shown, the complexity of the first stage (represented by the area of ​​rectangle 602) is roughly the same as the complexity of the second stage (represented by the area of ​​rectangle 604). Similarly, the complexity of the third stage (represented by the area of ​​rectangle 606) is roughly the same as the complexity of the second stage (represented by the area of ​​rectangle 604). Furthermore, the complexity of the fourth stage (represented by the area of ​​rectangle 608) is roughly the same as the complexity of the third stage (represented by the area of ​​rectangle 606). The fourth stage has approximately the same complexity (the area of ​​rectangle 608), and is also quite complex. The general consistency of complexity is due to the decrease in the number of inputs in each stage, while the number of velocity assumptions increases in each stage.

[0039] Will Figure 5 The area of ​​rectangle 502 is compared with the sum of the areas of rectangles 602, 604, 606, and 608. The process disclosed here is less complex than that required by traditional methods. Specifically, Figure 6 The complexity of the first integration stage 602 shown is on the order of N*Q, where N is the number of samples in the long integration interval and Q is the number of samples in the time period of the first integration stage. The complexity of the second integration stage 604 is on the order of N*P, where P is the number of input samples in the second integration stage. Therefore, the complexity of the two-stage integration process is on the order of N*(P+Q). Thus, the complexity of the two-stage integration process is reduced by a factor P, i.e., N*N, compared to the complexity of the traditional process.

[0040] Figure 7 A flowchart 700 of the signal integration process disclosed herein is shown. In block 702, a radar signal or reflected signal is received at a radar system sampling point on a first integration interval. In block 704, the first integration interval is divided into a plurality of second time intervals. In block 706, a first coherent integration is performed on short time intervals using a set of low-resolution velocity assumptions to obtain a first-stage integration value. In block 708, a second coherent integration is performed on the first-stage integration value using a set of high-resolution velocity assumptions to obtain a second-stage integration value. In block 710, the velocity of the object is determined using the second-stage integration value.

[0041] Figure 8The range Doppler image 800 obtained from an object approaching the autonomous vehicle 10 is shown. The radar system was operated using conventional processing with a long-duration integration interval of approximately 200 microseconds (msec). Due to the object's motion, the target's range and velocity shift during the integration interval. As a result, the peak frequency 802 in the range Doppler image 800 is relatively wide and undefined, and the signal-to-noise ratio is low. This leads to low-resolution velocity values.

[0042] Figure 9 The distance Doppler image 900 of the object is shown, obtained using a long-duration integration interval and the processing method disclosed herein. Although target motion exists within the long-duration integration interval, the peak frequency 902 is well-defined and has a high signal-to-noise ratio. This allows for higher resolution values ​​of the object's velocity.

[0043] While the foregoing disclosure has been described with reference to exemplary embodiments, those skilled in the art will understand that various changes can be made without departing from its scope, and equivalents can replace its elements. Furthermore, many modifications can be made to adapt particular situations or materials to the teachings of this disclosure without departing from the basic scope of this disclosure. Therefore, it is intended that this disclosure be limited to the specific embodiments disclosed, but will include all embodiments falling within its scope.

Claims

1. A method for determining the velocity of an object, comprising: A radar signal about an object is obtained within an integration interval, the radar signal comprising multiple velocity samples; The integration interval is divided into multiple time periods, each of which includes a subset of velocity samples; A first integration is performed on a subset of velocity samples within a selected time period using a first set of velocity assumptions, which are a set of low-resolution velocity assumptions, to obtain the first-stage integration value for the selected time period. A second integration is performed using a second set of velocity assumptions, which is a set of high-resolution velocity assumptions, with the first-stage integral value to obtain the second-stage integral value over the integration interval; and The velocity of the object is determined from the integral value in the second stage. The first stage integral value is obtained by performing a first integration on the product of a velocity sample within a selected time period and a first phase term corresponding to a velocity hypothesis selected from the first set of velocity hypotheses, and the second stage integral value is obtained by determining an interpolation function for the first stage integral value and performing coherent integration on the product of the interpolation function and a second phase term corresponding to the second set of velocity hypotheses.

2. The method according to claim 1, wherein, The first resolution of the first set of velocity assumptions is smaller than the second resolution of the second set of velocity assumptions, and the number of velocity samples used as input to the first integral is greater than the number of first integral values ​​used as input to the second integral.

3. The method according to claim 1, wherein, Performing the first integration also includes obtaining a first-stage integration value for each of the plurality of time periods and the first set of velocity assumptions.

4. A system for determining the velocity of an object, comprising: A radar system is configured to acquire radar signals about an object within an integration interval, the radar signals comprising multiple velocity samples; and The processor is configured as follows: The integration interval is divided into multiple time periods, each of which includes a subset of velocity samples; The first set of velocity assumptions, which is a set of low-resolution velocity assumptions, is used to perform the first integration on a subset of velocity samples within a selected time period to obtain the first stage integration value for that time period. A second set of velocity assumptions, which is a set of high-resolution velocity assumptions, is used to perform a second integration using the first-stage integral value to obtain the second-stage integral value over the integration interval. and The velocity of the object is determined from the integral value in the second stage. The processor is further configured to perform a first integration on the product of a velocity sample within a selected time period and a first phase term corresponding to a velocity hypothesis selected from the first set of velocity hypotheses, and the processor is further configured to perform a second integration by determining an interpolation function for the first stage integration value and performing a coherent integration on the product of the interpolation function and a second phase term corresponding to the second set of velocity hypotheses.

5. The system according to claim 4, wherein, The first resolution of the first set of velocity assumptions is smaller than the second resolution of the second set of velocity assumptions, and the number of velocity samples used as input to the first integral is greater than the number of first-stage integral values ​​used as input to the second integral.

6. The system according to claim 4, wherein, The processor is also configured to obtain a first-stage integral value for each of the plurality of time periods and each of the first set of velocity assumptions.

Images