Accelerating radar object detection and parameter estimation

Abstract

A system for estimating object parameters, comprising: a receiver configured to detect a return signal of a radar signal; and a processing device configured to sample the return signal to generate a series of signal samples, divide a time frame into a plurality of consecutive segments k, and for each segment k, apply a Doppler Fourier transform and calculate as the Doppler frequency f D The complex value y of the function k The processing device is further configured to calculate an index based on the acceleration hypothesis and the velocity hypothesis in the set of hypotheses, and for each segment, select one or more Doppler frequency bins based on the index, and extract the complex value y associated with each selected Doppler frequency bin. k (f D The processing device is further configured to calculate a velocity and acceleration spectrum and estimate object parameters based on the spectrum.

Description

Technical Field

[0001] The subject disclosure relates to estimating object position and motion using radar. Background Art

[0002] Vehicles (e.g., cars, trucks, aircraft, construction equipment, agricultural equipment, automated factory equipment) are increasingly being equipped with detection systems for monitoring their surroundings. Radar systems can be used to detect and track objects, for example, to avoid obstacles. Radar systems can be used in vehicles to warn the driver or user and / or take evasive action. Detection and tracking systems are also useful in autonomously operating vehicles. The position of a moving object may not be accurately detected using conventional radar processing within typical integration times, especially if the object is accelerating. Therefore, it would be desirable to provide a system that uses radar to accurately estimate the position of an accelerating object. Summary of the Invention

[0003] In one exemplary embodiment, a system for estimating parameters of an object includes a receiver configured to detect a return signal comprising reflections of a radar signal, the radar signal comprising a series of transmit pulses transmitted within a selected time frame. The system also includes a processing device configured to sample the return signal within the time frame to generate a series of signal samples, divide the time frame into a plurality of consecutive segments k such that an assumption of linear phase variation within each segment k is valid, and for each segment k, apply a Doppler Fourier transform and calculate a plurality of Doppler frequencies f as D The complex value of the function y k The processing device is further configured to select a set of hypotheses including an acceleration hypothesis and a velocity hypothesis, and for the set of hypotheses, calculate an index based on the acceleration hypothesis and the velocity hypothesis, and for each segment, select one or more Doppler frequency bins based on the index, and extract a complex value y associated with each selected Doppler frequency bin. k (f D The processing device is further configured to combine the extracted components to calculate a velocity and acceleration spectrum of the time frame, and detect the object and estimate object parameters based on the velocity and acceleration spectrum.

[0004] In addition to one or more features described herein, the object parameters include at least one of an object position, an object velocity, and an object acceleration.

[0005] In addition to one or more features described herein, the processing device is configured to detect the object based on a strength of the velocity and acceleration spectra exceeding a selected threshold.

[0006] In addition to one or more features described herein, the processing device is configured to select a plurality of sets of hypotheses, each set of hypotheses having a respective velocity hypothesis and a respective acceleration hypothesis.

[0007] In addition to one or more features described herein, the processing device is configured to calculate a corresponding index for each set of hypotheses, generate a velocity and acceleration spectrum for each set of hypotheses based on the corresponding index, estimate the velocity and acceleration of the object by comparing the strength of each corresponding velocity and acceleration spectrum, and select at least one of the velocity and acceleration spectra based on the comparison.

[0008] In addition to one or more of the features described herein, the set of assumptions also includes the initial velocity assumption and acceleration assumptions And calculating the index includes calculating a velocity value for each segment k based on:

[0009]

[0010] in is the Doppler index, M is the number of samples in each segment k, and T is the sampling interval. The velocity value for each segment k is converted to a Doppler frequency based on:

[0011]

[0012] where λ is the wavelength, and at least one Doppler frequency point is selected for each segment k corresponding to the Doppler frequency.

[0013] In addition to one or more features described herein, the processing device is further configured to determine a phase correction for the extracted component, the phase correction being determined based on:

[0014]

[0015] In addition to one or more features described herein, combining the extracted components includes correlating the selected one or more points with a matched filter having a set of composite points, the composite points corresponding to the set of hypotheses.

[0016] In addition to one or more functions described herein, velocity and acceleration spectra are calculated based on:

[0017]

[0018] in, are the calculated velocity and acceleration spectra, λ is the wavelength, N is the total number of samples in the time frame, and M is the number of samples in each segment k.

[0019] In addition to one or more of the features described herein, the extracted components are represented by:

[0020]

[0021] Among them, x kM+m is the value of the signal sample corresponding to the extracted component, and fD is the Doppler frequency of the extracted component.

[0022] In one exemplary embodiment, a method of estimating a parameter of an object includes detecting a return signal comprising reflections of a radar signal, the radar signal comprising a series of transmit pulses transmitted within a selected time frame, sampling the return signal within the time frame to generate a series of signal samples, dividing the time frame into a plurality of consecutive segments k such that an assumption of linear phase variation within each segment k is valid, and for each segment k, applying a Doppler Fourier transform and calculating as a plurality of Doppler frequencies f D The complex value y of the function k The method further includes selecting a set of hypotheses including an acceleration hypothesis and a velocity hypothesis, and for the set of hypotheses, calculating an index based on the acceleration hypothesis and the velocity hypothesis, and for each segment, selecting one or more Doppler frequency points based on the index, and extracting a complex value y associated with each selected Doppler frequency k The method further includes combining the extracted components to calculate velocity and acceleration spectra within the time frame, and detecting the object and estimating object parameters based on the velocity and acceleration spectra.

[0023] In addition to one or more features described herein, the object parameters include at least one of an object position, an object velocity, and an object acceleration.

[0024] In addition to one or more features described herein, an object is detected based on a strength of the velocity and acceleration spectra exceeding a selected threshold.

[0025] In addition to one or more features described herein, the method also includes selecting a plurality of sets of hypotheses, each set of hypotheses having a corresponding velocity hypothesis and a corresponding acceleration hypothesis.

[0026] In addition to one or more features described herein, the method also includes calculating a corresponding index for each set of hypotheses, generating a velocity and acceleration spectrum for each set of hypotheses based on the corresponding index, and estimating the velocity and acceleration of the object by comparing the strength of each corresponding velocity and acceleration spectrum, and selecting at least one of the velocity and acceleration spectra based on the comparison.

[0027] In addition to one or more of the features described herein, the set of assumptions also includes the initial velocity assumption and acceleration assumptions And calculating the index includes calculating the velocity value for each segment k based on:

[0028]

[0029] in, is the Doppler index, M is the number of samples in each segment k, T is the sampling interval, and the velocity value for each segment k is converted to Doppler frequency based on:

[0030]

[0031] where λ is the wavelength, and at least one Doppler frequency point is selected for each segment k corresponding to the Doppler frequency.

[0032] In addition to one or more features described herein, the method also includes determining a phase correction for the extracted component, the phase correction being determined based on:

[0033]

[0034] In addition to one or more features described herein, combining the extracted components includes correlating the selected one or more points with a matched filter having a set of composite points corresponding to the set of hypotheses.

[0035] In addition to one or more features described herein, velocity and acceleration spectra are calculated based on:

[0036]

[0037] in, are the calculated velocity and acceleration spectra, λ is the wavelength, N is the total number of samples in the time frame, and M is the number of samples in each segment k.

[0038] In addition to one or more of the features described herein, the extracted components are represented by:

[0039]

[0040] Among them, x kM+m is the value of the signal sample corresponding to the extracted component, and fD is the Doppler frequency of the extracted component.

[0041] The above features and advantages and other features and advantages of the present disclosure will be apparent from the following detailed description when considered in conjunction with the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Additional features, advantages and details appear by way of example only in the following detailed description, which refers to the accompanying drawings, in which:

[0043] Figure 1 is a top view of a motor vehicle including a radar system;

[0044] Figure 2 depicts a radar system according to an exemplary embodiment;

[0045] Figure 3 is a flow chart depicting a method of estimating one or more parameters of an object using a radar system according to an exemplary embodiment;

[0046] Figure 4 depicts examples of radar transmission and detection time frames, and examples of data structures for detecting and estimating parameters of objects, according to an exemplary embodiment;

[0047] Figure 5 is a diagram illustrating an exemplary embodiment of the present invention. Figure 3 a flowchart of an example of a method;

[0048] Figure 6 depicts an example of a Doppler frequency signal generated by conventional radar processing techniques;

[0049] Figure 7 depicts an example of a Doppler frequency signal generated according to an exemplary embodiment; and

[0050] Figure 8 Depicted are examples of a Doppler range-frequency map generated in accordance with an exemplary embodiment and a Doppler range-frequency map produced by conventional radar processing techniques. DETAILED DESCRIPTION

[0051] The following description is merely exemplary in nature and is in no way intended to limit the present disclosure, its application, or uses. It should be understood that throughout the drawings, corresponding reference numerals indicate like or corresponding parts and features.

[0052] According to one or more exemplary embodiments, methods and systems for radar detection and object parameter estimation are described herein. Embodiments of the radar system are configured to estimate one or more parameters of an object, such as position, acceleration, and / or velocity. An object can be any feature or condition (e.g., a vehicle, a person, a ship, a weather phenomenon, etc.) that reflects an emitted radar signal. The radar system can be included in or connected to a vehicle for detecting objects such as road features, road obstacles, other vehicles, trees, people, and others. The radar system is not limited to use with a vehicle and can be used in any environment (e.g., weather, aviation, and others).

[0053] A radar system includes or is in communication with a processor configured to perform a method for detecting a dynamic vehicle or other object and accurately estimating object parameters (e.g., velocity and acceleration) during acceleration of the object. The method includes dividing a frame (e.g., a coherent processing frame) into a plurality of consecutive time segments and performing a Doppler signal processing technique on return signal components (e.g., samples) in each time segment. Each segment is selected under an approximation or assumption of constant velocity within the segment (i.e., such that an approximation of linear phase variation within the segment is valid).

[0054] Doppler processing is performed on the samples in each segment to transform the time domain return signal into the frequency domain. For each segment, a Doppler processed output is generated, which includes the amplitude for each of the multiple Doppler frequencies. For example, a discrete Fourier transform (DFT) is performed on the samples in the initial segment, and a discrete DFT is performed separately for each segment. For each segment, the components of the DFT output are stored in or associated with one or more Doppler frequency bins. In one embodiment, a "bin" represents a frequency value or range within a frequency spectrum. As described below, a matrix including columns corresponding to time frame segments and rows corresponding to frequency bins can be generated.

[0055] The Doppler processing outputs are then effectively combined to estimate parameters including the velocity and acceleration of the object.Other parameters may be estimated such as position, distance and / or orientation.

[0056] In one embodiment, one or more sets of velocity and acceleration hypotheses are selected. For each set of hypotheses, a velocity or frequency index (also known as a Doppler index) is calculated for each segment based on the velocity and acceleration selected for that set. The index instructs the processor which point or points in each segment to select when combining the DFT outputs. For a given segment, a single point or a group of multiple points is selected (for example, if the Doppler index falls between adjacent points). Based on the index, for a given set of hypotheses, DFT components for each segment are extracted from the points, and the selected components are integrated or otherwise combined to generate a combined velocity and acceleration spectrum. If the combined spectrum has a peak that exceeds a certain threshold, an object is detected, and the object velocity and acceleration are determined.

[0057] In one embodiment, multiple sets of hypotheses are selected, each set having a different combination of velocity and acceleration hypotheses. For example, each set includes an initial velocity hypothesis (the velocity of the earliest segment) and a different acceleration hypothesis. For each set of hypotheses, a Doppler index is calculated, and a DFT output component or point is selected based on the Doppler index. The Doppler index can be expressed as one or more values, equations, slopes, or any other suitable representation.

[0058] The selected components are summed, integrated, or otherwise combined to generate a combined velocity and acceleration spectrum for each set of hypotheses. The combined spectrum includes intensity values ​​for each Doppler frequency. This process is repeated to generate a velocity and acceleration spectrum for each set of hypotheses, and peaks in the spectrum are identified to determine which set of hypotheses produces a peak with sufficient intensity to identify the object. This process thus provides an indication of the actual object velocity and acceleration.

[0059] The embodiments described herein present numerous advantages. For example, radar systems configured according to the embodiments described herein can accurately estimate parameters (e.g., position, velocity, and / or acceleration) used to accelerate and decelerate objects. Furthermore, because the complexity and required processing power are significantly reduced compared to conventional techniques and systems, the embodiments provide a more efficient and faster process for detecting objects. The embodiments described herein also provide a higher strength combined signal and a lower probability of missed detection than conventional systems.

[0060] Figure 1 An embodiment of a motor vehicle 10 is shown that includes a body 12 that at least partially defines a passenger compartment 14. The body 12 also supports various vehicle subsystems, including an engine assembly 16 and other subsystems to support the functions of the engine assembly 16 and other vehicle components, such as a braking subsystem, a steering subsystem, a fuel injection subsystem, an exhaust subsystem, etc.

[0061] Vehicle 10 includes aspects of a radar system 20 for detecting and tracking objects, which can be used to warn a user, perform evasive maneuvers, assist a user, and / or autonomously control vehicle 10. Radar system 20 includes one or more radar sensing assemblies 22, each of which can include one or more transmitting elements and / or one or more receiving elements. Vehicle 10 can include multiple radar sensing assemblies positioned at different locations and with different angular orientations.

[0062] For example, each radar sensing assembly 22 includes a transmit portion and a receive portion. The transmit and receive portions may include separate transmit and receive antennas, or may share an antenna in a transceiver configuration. Each radar sensing assembly 22 may include additional components, such as a low-pass filter (LPF) and / or a controller or other processing device. The radar sensing assembly and / or radar system 20 may be configured as a coherent radar.

[0063] Radar sensing assembly 22 communicates with one or more processing devices, such as a processing device in each assembly and / or a remote processing device, such as an onboard processor 24 and / or a remote processor 26. Remote processor 26 may be, for example, part of a mapping system or a vehicle diagnostic system. Vehicle 10 may also include a user interaction system 28 and other components, such as a GPS device.

[0064] Radar system 20 is typically configured to acquire radar signals and analyze the radar signals to estimate parameters of an object, such as the object's position, acceleration, and / or velocity. Such parameters are typically estimated by processing and integrating the acquired signals over a selected time frame. The length of the time frame is selected to provide a desired resolution.

[0065] Radar system 20 is configured to transmit radar signals from one or more transmitters, each transmitter comprising a series of continuous pulses transmitted within a selected time frame. In one embodiment, the selected time frame is selected as a coherent processing interval associated with coherent radar technology. Reflections of the transmitted pulses are detected by a receiver and multiplied or mixed with a reference signal (e.g., a waveform corresponding to the transmitted radar signal) to produce a return signal.

[0066] Figure 2 Aspects of an embodiment of a computer system 30 are shown that is in communication with or part of the radar system 20 and can perform various aspects of the embodiments described herein. The computer system 30 includes at least one processing device 32, which generally includes one or more processors for performing aspects of the radar detection and analysis methods described herein. The processing device 32 can be integrated into the vehicle 10, such as as an onboard processor 24, or can be a processing device separate from the vehicle 10, such as a server, a personal computer, or a mobile device (e.g., a smartphone or tablet). For example, the processing device 32 can be part of or in communication with one or more engine control units (ECUs), one or more vehicle control modules, a cloud computing device, a vehicle satellite communication system, and / or the like. The processing device 32 can be configured to perform the radar detection and analysis methods described herein and can also perform functions related to the control of various vehicle subsystems.

[0067] Components of computer system 30 include a processing device 32 (e.g., one or more processors or processing units) and a system memory 34. System memory 34 may include various computer system-readable media. Such media may be any available media that can be accessed by processing device 32 and includes volatile and non-volatile media, removable and non-removable media.

[0068] For example, system memory 34 includes nonvolatile memory 36, such as a hard drive, and may also include volatile memory 38, such as random access memory (RAM) and / or cache memory. Computer system 30 may also include other removable or non-removable and volatile or nonvolatile computer system storage media.

[0069] The system memory 34 may include at least one program product having a set (e.g., at least one) of program modules configured to perform the functions of the embodiments described herein. For example, the system memory 34 stores various program modules 40 that generally perform the functions and / or methods of the embodiments described herein. For example, a receiver module 42 may be included to perform functions related to acquiring and processing received signals (e.g., radar return signals), and an analysis module 44 may be included to perform functions related to object detection and estimation of object parameters (e.g., velocity, acceleration, and / or position). The system memory 34 may also store various data structures 46, such as data files or other structures storing data related to radar detection and analysis. Examples of such data include sampled return signals, frequency data, range-Doppler maps and spectra, and object position, velocity, and / or orientation data. As used herein, the term "module" refers to processing circuitry, which may include application-specific integrated circuits (ASICs), electronic circuits, processors (shared, dedicated, or grouped), and memory that executes one or more software or firmware programs, combinational logic circuits, and / or other suitable components that provide the described functionality.

[0070] Processing device 32 can communicate with radar sensing assembly 22 via, for example, input / output (I / O) interface 55. Processing device 32 can also communicate with one or more external devices 48, such as a keyboard, a pointing device, and / or any device (e.g., a network card, a modem, etc.) that enables processing device 32 to communicate with one or more other computing devices. In addition, processing device 32 can communicate with one or more devices that can be used in conjunction with radar system 20, such as a global positioning system (GPS) device 50 and a camera 52.

[0071] The GPS device 50 and the camera 52 may be used, for example, in conjunction with the radar system 20 for autonomous and / or semi-autonomous control of the vehicle 10. Communications with various devices may occur via the input / output interface 54.

[0072] The processing device 32 may also communicate with one or more networks 56, such as a local area network (LAN), a general wide area network (WAN), and / or a public network (e.g., the Internet), via a network adapter 58. It should be understood that, although not shown, other hardware and / or software components may be used in conjunction with the computer system 30. Examples include, but are not limited to, microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, and data archival storage systems.

[0073] Figure 3Aspects of an embodiment of a computer-implemented method 70 for radar detection and analysis are shown, which includes estimating parameters of an object, such as location or position, acceleration, and / or velocity. The method 70 may be performed by one or more processors disposed in a vehicle (e.g., a processing device 32 as an ECU or onboard computer) and / or disposed in a device such as a smartphone, tablet, or smartwatch. For illustrative purposes, the method 70 is described in conjunction with FIG. Figure 1 Radar system 20 and Figure 2 The method 70 is discussed with reference to the components shown. Note that aspects of the method 70 may be performed by any suitable processing device or system, either exclusively or in conjunction with a human operator.

[0074] Method 70 includes a plurality of stages or steps represented by blocks 71-75, all of which may be performed sequentially. However, in some embodiments, one or more stages may be performed in a different order than shown, or fewer stages than shown may be performed.

[0075] At block 71, a radar signal is transmitted by one or more transmitting elements in a radar system (e.g., radar system 20). Each transmitting element transmits a radar signal comprising a series of pulses. In one embodiment, each transmitting element transmits a linear frequency modulated continuous wave (LFM-CW) signal. This signal may be referred to as a "chirp signal," and each pulse may be referred to as a "chirp."

[0076] Each radar signal is transmitted within a selected time frame. The time frame can be of any suitable length. In one embodiment, the time frame is selected to be equal to or related to the coherent processing interval (CPI). Although this document discusses only a single transmitting element and a single radar signal (whose duration is equal to the selected time frame), it should be understood that method 70 is applicable to multiple radar signals from a single transmitting element or multiple transmitting elements.

[0077] For example, method 70 may be performed as part of a multiple-input multiple-output (MIMO) and / or phased array radar system including multiple transmitters. Execution of method 70 in this example may include individually processing and analyzing each received return signal associated with each transmitter.

[0078] The return signal is detected or measured by one or more receiving elements as a measurement signal. For example, the analog signal detected by the receiving element is sampled and converted into a digital signal, referred to herein as a sample. In one embodiment, the return signal is sampled within a selected time frame (whose duration is equal to the selected time frame of the transmitted radar signal) according to a selected sampling frequency, and a detection signal x is generated for each of a plurality of samples n. n The total number of samples n within a given time frame is denoted as N.

[0079] At block 72, the time frame is divided into a plurality of consecutive time periods k. In one embodiment, the time periods k are of equal time length (duration) and are selected so that the phase change in the return signal within the time period k can be effectively approximated as linear and the speed can be approximately constant.

[0080] At block 73, the Doppler Fourier transform is applied to the return signal x in each segment k. n The time domain signal in the segment is converted to the Doppler frequency domain by applying a Doppler Fourier transform to the signal. In one embodiment, a discrete Fourier transform (DFT) is performed on the samples in the segment k, producing a DFT output. The execution of the DFT includes computing a complex value including the Doppler frequency for a plurality of Doppler shifts (referred to herein as the Doppler frequency f D ) in each of the amplitude (A) and phase (φ). As discussed further below, the Doppler frequency can be expressed as a Doppler frequency point or f D For example, for a given segment k, multiple f D Point, each f D The points correspond to different Doppler frequencies f D .

[0081] For example, the processor performs a DFT on the samples in each segment k. The DFT output for a given time segment can be represented by a complex signal y containing amplitude and phase information. k (f D ) indicates that it is the Doppler frequency f D (a single frequency value or a range of frequency values).

[0082] At block 74, one or more sets of hypotheses are selected, where each set of hypotheses includes a velocity hypothesis and acceleration assumptions Multiple sets of assumptions can be selected. For example, the processor selects or receives at least one initial velocity assumption and multiple acceleration assumptions For each acceleration assumption You can define a set of values Each value has an initial velocity assumption and a different acceleration assumption.

[0083] The processor assumes that Calculate the Doppler index based on a set of velocity assumptions and acceleration assumptions The Doppler index indicates which components of the DFT output are selected for a given segment k. A separate Doppler index is calculated for each set of hypotheses. The Doppler index provides a separate set of indices for each time frame segment. Each individual index indicates which f is selected for the segment associated with the individual index. D point.

[0084] The processor then extracts the DFT output y for each segment k based on the Doppler index k (f D ) for further processing or analysis. Unlike prior art techniques that assume constant velocity or linear phase variation along the frame (and thus do not assume acceleration during the frame), the embodiments described herein do not assume constant velocity, but rather consider that the object may accelerate. Additionally, by selecting only a subset of the DFT outputs (f D points), which can greatly reduce the number of samples that need to be processed.

[0085] The Doppler index can be represented as a series or vector of velocity or Doppler frequency values ​​(e.g., Doppler frequency amplitudes), where each velocity or Doppler frequency value corresponds to a segment k (also referred to as a segment index). For example, the Doppler index can be configured to indicate a specific Doppler frequency or frequencies for each segment, or to indicate a specific Doppler frequency bin or bins for each segment. The index instructs the processor which Doppler frequency bin or bins to extract for each segment. In another example, the index is represented as a slope or other information that indicates which Doppler frequency bins to select for each segment k when combining or otherwise processing the DFT outputs in the frame.

[0086] At block 75, one or more parameters (e.g., velocity and acceleration) of the object are estimated based on the DFT output and the Doppler index. For a given set of hypotheses, a subset of the DFT output is extracted based on the Doppler index, and the subset is processed by combining the DFT outputs to estimate the velocity, acceleration, and / or other characteristics of the object during the time frame. For example, the subsets (DFT output components extracted according to the Doppler index) are combined to generate a velocity and acceleration spectrum, such as a range-frequency plot. In one embodiment, a velocity and acceleration spectrum is generated for each set of hypotheses, and the spectrum with the highest peak (e.g., exceeding a detection threshold) is selected. The hypothesized velocity and acceleration associated with the selected Doppler frequency are considered to be correlated with the actual velocity and acceleration of the object.

[0087] Figure 4 Depicted are examples of various data structures that can be used or generated as part of method 70. In this example, a radar signal is transmitted from one or more objects and reflected as a return signal. The radar signal is transmitted and the return signal is sampled within a time frame selected as a coherent processing time frame (e.g., approximately 50 milliseconds). The time frame is represented by the horizontal time (t) axis of graph 80. If the object is accelerating, the phase of the sampled signal is not linear, as shown by phase curve 82.

[0088] The time frame is divided into k segments, each of which has a duration T selected so that the phase can be approximately linear within the segment. In this example, a number of equal-length segments k are selected, and the return signal samples n are subdivided into consecutive subsets.

[0089] In each segment k, the samples are converted to the frequency domain using DFT and a DFT output is generated which includes the frequency domain at various Doppler frequencies f D The amplitude and phase strength at . In this example, the DFT output of segment k is the vector y k , which includes each Doppler frequency f D The complex value of .

[0090] The components of the DFT output are stored in or assigned to a matrix 90 comprising rows 92 and K number of columns 94, wherein each column 94 corresponds to a time frame segment k. Rows 92 represent successive Doppler frequencies f D point.

[0091] For a given segment, the DFT output y is scanned along column 94. k (complex value), and the components of the DFT output corresponding to different Doppler frequencies are assigned to the corresponding f D For example, each DFT output includes or represents one or more peaks representing reflections from one object. If a DFT output has multiple peaks (e.g., due to multiple objects), they can be assigned to multiple f D point.

[0092] Figure 4 An example of the Doppler index calculated for a given set of assumptions is shown. This includes the initial velocity assumption of 10 m / s and 0.1m / s 2 The acceleration assumption In this example, the Doppler index is given by It can be calculated according to the following formula:

[0093]

[0094] Where M is the number of samples in the segment (i.e., the number of DFT input samples in the segment). T is the sampling interval (the time between two samples), and k is the segment index. Although the index is defined above as a function of velocity, the index can also be expressed in terms of Doppler frequency or the number of Doppler frequency bins.

[0095] The Doppler index calculated within the time frame Figure 4 is shown as a linear function represented by the diagonal line 96, which is related to multiple f DBased on this index, select one or more f D For example, for a given segment, the processor uses the index to identify which f point or points indicated by the index (ie, intersecting the diagonal line 96) D Point. Extract the identified f D points and combine the DFT output components stored therein to estimate the object parameters.

[0096] In some cases, the Doppler index falls within f D Between points or between adjacent f D In some cases, the value of the Doppler index at segment k may be rounded up or down (ie, moved to the next point above or below), or adjusted in any other suitable manner, such as by linear interpolation.

[0097] For example, choose a set of hypotheses The Doppler index is calculated for the initial segment (k=0) using equation (1). Thus, the Doppler index corresponds to the initial velocity. The initial velocity can be associated with a frequency (f) and the f corresponding to that frequency can be selected. D For the next consecutive segment (k=1), calculate the Doppler index according to equation (1) with k=1 to obtain the radial velocity and select the corresponding f D point. Repeat this process for each consecutive segment.

[0098] For each set of hypotheses, the DFT points identified by the index are extracted and the signal components are combined from them to derive velocity and acceleration information. Due to the Doppler index, the number of calculations is reduced compared to conventional techniques that utilize all samples in each segment, which makes the process faster and less complex than conventional techniques. For example, for a given velocity and acceleration hypothesis Selecting a point from each of the K segments corresponds to a complexity of K. In contrast, conventional techniques involve calculating all input samples for each hypothesis of velocity and acceleration (M samples per segment). Such conventional techniques have a complexity of K*M, which is M times more complex than the method described herein.

[0099] Once the selected points are extracted, for each set of hypotheses Combine the DFT output components to generate a coherently combined magnitude and phase. For example, combine the DFT output components using the following summation:

[0100]

[0101] where K is the number of segments k, is a function representing the Doppler index, and is the phase correction. v,a Represents the selected fD The sum of the points, where f D The points are chosen based on the Doppler index. The output of this summation is a relatively high-intensity signal for the case where the velocity and acceleration are assumed to be a real object, and low-intensity (i.e., the intensity of the various frequencies is higher than that of a single sample or block of samples) when the hypothesis corresponds to a non-existent object.

[0102] In one embodiment, the extracted DFT points are input to a matched filter, which outputs strength values ​​for each velocity and acceleration hypothesis. The outputs are distributed or stored in a matched filter output matrix 100, which includes rows 102 representing velocity hypotheses and columns 104 representing acceleration hypotheses.

[0103] Still refer to Figure 4 The Doppler index calculation can be repeated for multiple acceleration and velocity hypotheses. Another example of a set of hypotheses is shown by line 95, which includes an initial velocity hypothesis of 10 m / s and 0 m / s. 2 acceleration assumption.

[0104] Figure 5 It shows that for a given set of assumptions A flowchart of an example of a method 110 for estimating object parameters, which may be used Figure 4 The method 110 may be performed by, but is not limited to, a matrix of blocks 111-116. The method 110 includes a plurality of stages or steps represented by blocks 111-116, all of which may be performed sequentially. However, in some embodiments, one or more stages may be performed in a different order than shown, or fewer stages may be performed than shown.

[0105] Initially, a radar signal is transmitted within a selected time frame, and the return signal is sampled to generate a number of samples n according to a desired sampling frequency and sampling interval T. The time frame is divided into segments k.

[0106] At block 111, each set of M samples is transformed using a DFT, thereby producing a DFT output. The components of the DFT output (amplitude and phase at each Doppler frequency) are assigned to the corresponding f D points (e.g., assigning a DFT component with a given frequency to a point corresponding to that frequency).

[0107] At block 112, a set of hypotheses is selected And the Doppler index is calculated for each of the k segments using, for example, equation (1). The Doppler index can be expressed as velocity, frequency, f D point or allows the processor to determine which f should be selected for each segment k D Other values ​​of the point.

[0108] At block 113, one or more f are selected for each segment based on the Doppler index. D point, and from the selected f D Extract the DFT components from the points. For example, refer to Figure 4 , extracting at least one f from each column 94 of the matrix 90 based on the Doppler index D Point, that is, select the point f that intersects the diagonal line 96 D points and extract the DFT output components from the selected points.

[0109] At block 114, a phase correction is calculated for each set of samples based on the phase at the beginning of the set of samples. The phase correction may be calculated as:

[0110]

[0111] In equation (3), M is the number of samples in a segment, k is the segment index, and T is the sampling interval. As described at block 115, phase correction is incorporated into the matched filter.

[0112] At block 115, phase correction is applied and the sets of samples are integrated using, for example, coherent integration. In one embodiment, the sets of samples are input to a matched filter. The matched filter provides a set of synthetic points that are the values ​​that would be expected to be calculated if the set of hypotheses corresponded to the velocity and acceleration of the actual object. The synthetic points can be represented by the exponential terms of equation (8). The matched filter outputs some values ​​or peaks that are compared to a threshold to determine whether the set of hypotheses corresponds to the velocity and acceleration of the actual object.

[0113] The DFT process (e.g., matched filtering) produces a complex vector z for each point (i.e., each combination of velocity and acceleration hypotheses). The amplitude (real part) is taken from vector z and entered or assigned to the corresponding velocity and acceleration (v, a) point, for example, in matrix 100.

[0114] An example of a suitable matched filter is represented, for example, by:

[0115]

[0116] where x n is the signal amplitude at sample n. Conventionally, the use of such a matched filter requires processing all samples n in a frame. Using this index, a subset of samples is selected based on a given set of assumed Doppler indices and processed to derive velocity and acceleration spectra as follows:

[0117]

[0118] Where M is the number of samples in each of the K segments. Equation (5) can be rewritten as:

[0119]

[0120] Since the phase is assumed to be linear in each segment k, the term representing the nonlinear phase change is assumed to be small and can be neglected. This simplifies equation (6) to:

[0121]

[0122] The above equation can be further simplified by replacing the sum of M samples with a function of the Doppler index in equation (1), yielding:

[0123]

[0124] in:

[0125]

[0126] The above expression is based on the Doppler index The Doppler index is defined in terms of velocity using equation (1) for the selected DFT components. Convert it to Doppler frequency f D , we get:

[0127]

[0128] where X kM+m is the DFT output of the kth group of samples. k It can be expressed as:

[0129] X k =DFT{x kM+1 …x kM+m}

[0130] The output z is the value of each chosen f D A coherent combination of the intensity values ​​at a point and is represented by:

[0131]

[0132] At block 116, the matched filter output is analyzed to identify peaks therein. Peaks exceeding a selected detection threshold (e.g., an amplitude or intensity threshold) can be identified based on the output to determine estimated velocity and acceleration. Other parameters (e.g., distance and position) can also be derived from the matched filter output.

[0133] The complexity of the solution provided by the above embodiment is significantly less than that of conventional techniques. For example, the complexity O of the conventional solution is expressed as O(N*Na*Nv), where N is the number of samples in a frame, Nv is the number of velocity (e.g., initial velocity) hypotheses, and Na is the number of acceleration hypotheses.

[0134] The complexity of the methods described herein is significantly less than conventional techniques. For example, the complexity of methods 70 and 110 can be expressed as O(N*Nv+M*Na*Nv)=O(M*Na*Nv), where M is the number of samples in each segment. The complexity of the above methods is significantly increased by a factor of N / M, which is the number of segments. For example, for a 50ms frame and a signal with a frequency of 40MHz and a chirp duration of 25us, the complexity is reduced by a factor of 200. This reduced complexity provides advantages in the form of, for example, reduced processing power and faster performance.

[0135] As described above, the methods described herein produce stronger peaks, which improve object detection. The following examples illustrate the advantages of Doppler processing according to embodiments described herein. In these examples, there is an accelerating object.

[0136] Figure 6 The result of Doppler processing using conventional techniques is shown, where velocity is assumed to be constant and phase changes are assumed to be linear (as shown in phase-time plot 200). In this example, the result of this processing is a velocity and acceleration spectrum 202, with a first portion 204 (the return signal component) representing the return signal and a second portion 206 representing noise and clutter. As can be seen, the intensity of the return signal component is below the detection threshold 208, and no object is detected.

[0137] Figure 7 The result of Doppler processing using method 70 or 110 is shown, where the velocity is not assumed to be constant (the phase change curve 300 is assumed to be nonlinear). The result of this processing is a velocity and acceleration spectrum 302, which has a return signal component 304 and a noise and clutter component 306. As shown, the method described herein causes the return signal component peak to be above the detection threshold 208. As a result, the object is detected.

[0138] Figure 8 A comparison is shown between data obtained from the method described herein and data obtained from a conventional method in which object velocity is assumed to be constant during a time frame. In this example, radar detection is performed according to the method described herein, and a range map 400 is generated. The range map 400 is color-coded or shaded according to a legend 402, showing the output of the detection method as a function of range and Doppler frequency. Figure 8 Also shown is a distance map 500 generated according to conventional procedures.

[0139] As shown, the range map 400 produced by the method described herein results in a high-intensity signal that provides sharp peaks in the velocity and acceleration spectra and provides higher gain on the target compared to the velocity and acceleration spectra shown by the conventionally generated range map 500. The Doppler spectrum peak in the conventionally generated range map 500 has a significantly higher spread, resulting in a higher uncertainty about the object's velocity compared to the output of the method described herein.

[0140] Although the above disclosure has been described with reference to exemplary embodiments, it will be understood by those skilled in the art that various changes may be made and equivalents may be substituted for elements thereof without departing from the scope of the present invention. In addition, many modifications may be made to adapt a particular situation or material to the teachings of the present disclosure without departing from the basic scope of the present disclosure. Therefore, it is intended that the present disclosure is not limited to the specific embodiments disclosed, but rather includes all embodiments falling within its scope.

Claims

1. A system for estimating parameters of an object, the system comprising: a receiver configured to detect a return signal comprising a reflection of a radar signal comprising a series of transmit pulses transmitted within a selected time frame; as well as A processing device configured to perform: sampling the return signal within the time frame to generate a series of signal samples; Divide the time frame into a number of consecutive segments k such that the assumption of linear phase variation within each segment k is valid; For each segment k, apply the Doppler Fourier transform and calculate as a multiple of the Doppler frequencies f D The complex value y of the function k ; Select a set of assumptions that includes an acceleration assumption and a velocity assumption; as well as For this set of assumptions, execute: Calculate the index based on the acceleration assumption and the velocity assumption; For each segment, one or more Doppler frequency bins are selected based on the index, and the complex value y associated with each selected Doppler frequency bin is extracted. k (f D ) as well as The extracted components are combined to calculate velocity and acceleration spectra for the time frame, and the object is detected and object parameters are estimated based on the velocity and acceleration spectra.

2. The system according to claim 1, wherein: The processing device is configured to select a plurality of groups of hypotheses, each group of hypotheses having a respective velocity hypothesis and a respective acceleration hypothesis, and the processing device is configured to: Calculate the corresponding index for each set of hypotheses; generating velocity and acceleration spectra for each set of hypotheses based on the corresponding indices; as well as The velocity and acceleration of the object are estimated by comparing the strength of each respective velocity and acceleration spectrum, and at least one of the velocity and acceleration spectrum is selected based on the comparison.

3. The system according to claim 1, wherein: This set of assumptions includes the initial velocity assumption and acceleration assumptions And the calculation index includes: The velocity value for each segment k is calculated based on the following equation: in is the Doppler index, M is the number of samples in each segment k, and T is the sampling interval; The velocity value of each segment k is converted to Doppler frequency based on the following equation: where λ is the wavelength; and At least one Doppler frequency point is selected for each segment k corresponding to a Doppler frequency.

4. The system according to claim 3, wherein: The processing device is further configured to perform a determination of a phase correction of the extracted component, the phase correction being determined based on the following equation:

5. The system according to claim 4, wherein: Combining the extracted components includes correlating the selected one or more points with a matched filter having a set of synthetic points corresponding to the set of hypotheses.

6. The system according to claim 5, wherein: The velocity and acceleration spectra are calculated based on the following equations: in, are the calculated velocity and acceleration spectra, λ is the wavelength, N is the total number of samples in the time frame, and M is the number of samples in each segment k.

7. The system according to claim 6, wherein: The extracted components are expressed as follows: Among them, x kM+m is the value of the signal sample corresponding to the extracted component, and f D is the Doppler frequency of the extracted component.

8. A method for estimating a parameter of an object, the method comprising: detecting a return signal comprising a reflection of a radar signal comprising a series of transmit pulses transmitted within a selected time frame; sampling the return signal within the time frame to generate a series of signal samples; dividing the time frame into a plurality of consecutive segments k such that the assumption of linear phase variation within each segment k is valid; For each segment k, the Doppler Fourier transform is applied and the Doppler frequencies f are calculated as D The complex value y of the function k ; Select a set of assumptions that includes an acceleration assumption and a velocity assumption; as well as For this set of assumptions, execute: Calculate the index based on the acceleration assumption and the velocity assumption; For each segment, one or more Doppler frequency points are selected based on the index, and the complex value y associated with each selected Doppler frequency is extracted. k The weight; as well as The extracted components are combined to calculate velocity and acceleration spectra within the time frame, and the object is detected and object parameters are estimated based on the velocity and acceleration spectra.

9. The method according to claim 8, wherein The object parameter includes at least one of an object position, an object velocity, and an object acceleration.

10. The method according to claim 8, wherein The object is detected based on the strength of the velocity and acceleration spectrum exceeding a selected threshold.

Images