Direct detection lidar system and method using synthetic Doppler processing

Abstract

The present invention discloses a detection system, which includes a signal transmitter for transmitting a transmission signal into an area and a receiver, the receiver is used to receive a reflection signal generated by reflection of the transmission signal, and generate a reception signal indicating the reflection signal. A processor is coupled to the receiver to receive the reception signal and process the reception signal to generate a detection of one or more objects in the area. The processing includes changing a phase shift to generate a phase modulation signal from the reception signal, and generating a detection from the phase modulation signal.

Description

Background Art 1. Technical Field

[0002] The present disclosure relates to lidar systems, and in particular to lidar systems using synthetic Doppler processing, which may be used in automotive or other applications.

[0003] 2. Related fields

[0004] LiDAR is often an acronym for light detection and ranging, and in this sense LiDAR is often viewed as the optical analog of radar. Generally speaking, there are two types of LiDAR systems, incoherent LiDAR and coherent LiDAR. Incoherent LiDAR (also often called direct detection or direct energy detection LiDAR) primarily uses amplitude measurements of light echoes, while coherent LiDAR is more suited to phase-sensitive measurements or other more complex transmitter waveform modulation techniques. Coherent systems generally use optical heterodyne detection, which is more sensitive than direct detection, allowing them to operate at much lower power and provide greater measurement accuracy and resolution, but at the expense of more complex transceiver requirements and cost. Summary of the invention

[0005] According to a first aspect, a detection system is provided. The detection system includes a signal transmitter for transmitting a transmission signal into an area and a receiver for receiving a reflection signal generated by reflection of the transmission signal and generating a reception signal indicative of the reflection signal. A processor is coupled to the receiver to receive the reception signal and process the reception signal to generate a detection of one or more objects in the area. The processing includes changing a phase shift to generate a phase modulation signal from the reception signal, and generating a detection from the phase modulation signal.

[0006] The phase shift may be varied by a phase modulator that varies the phase shift in a local oscillator (LO) signal. Alternatively, the phase shift may be varied by varying the output of a direct digital synthesizer (DDS) used in generating the transmit and receive signals. The phase shift may be varied by accumulating the phase shift value to the in-phase and quadrature phase detector signals generated by the receive signal. The phase shift value may be digitally indexed.

[0007] The frequency of the transmission signal may be controlled to vary according to a ramp between a first frequency and a second frequency. The ramp may be a linear ramp. The ramp may have a slope such that the first frequency is lower than the second frequency or the first frequency is higher than the second frequency.

[0008] The transmit signal may be a frequency modulated continuous wave (FMCW) signal. Alternatively, the transmit signal may be a pulse signal.

[0009] The signal transmitter can scan the transmission signal into the area at an angle at a varying angle. In some embodiments, the signal transmitter includes a scanning mirror for scanning the transmission signal into the area at a varying angle. The scanning mirror can be a micro-electromechanical (MEMS) scanning mirror.

[0010] The detection system may be an automobile detection system.

[0011] In some exemplary embodiments, the detection system is a radar system. In these embodiments, the signal transmitter is a radar signal transmitter, the transmitted signal is a transmitted radar signal, and the reflected signal is a reflected radar signal.

[0012] In some exemplary embodiments, the detection system is a radar system. In these embodiments, the signal transmitter is a radar signal transmitter, the transmitted signal is a transmitted radar signal, and the reflected signal is a reflected radar signal.

[0013] At least one of the one or more objects may be stationary relative to the detection system.Alternatively or additionally, at least one of the one or more objects is in motion relative to the detection system. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] The present disclosure is further described in the following detailed description, by way of non-limiting examples of embodiments of the disclosure, with reference to the indicated various figures wherein like reference numerals refer to like parts throughout the several views of the drawings.

[0015] Figure 1 Included are three curves illustrating a transmitter envelope modulation technique using a basic sinusoidal modulation envelope as applied to direct detection lidar according to an exemplary embodiment.

[0016] Figure 2 Included are three curves according to other exemplary embodiments, which illustrate transmitter envelope modulation techniques using a pulse burst modulation envelope as applied to direct detection lidar.

[0017] Figure 3 Included are three curves illustrating transmitter envelope modulation techniques using linear frequency envelope modulation as applied to direct detection lidar according to other exemplary embodiments.

[0018] Figure 4 Included are schematic functional block diagrams of quadrature demodulation systems and techniques according to some exemplary embodiments.

[0019] Figure 5 Included is a schematic functional block diagram of a conventional direct detection lidar system.

[0020] Fig. 6AIncludes a schematic functional block diagram according to some exemplary embodiments, which shows the use of Figure 2 The laser radar system shown has pulse burst transmission envelope modulation and orthogonal demodulation.

[0021] Figure 6B is a schematic diagram showing a schematic diagram according to some exemplary embodiments Fig. 6A Details of the pulse burst modulation signal used in the system.

[0022] Figure 7 is a graph including three curves showing experimental data associated with the pulse burst envelope modulated lidar system described in detail herein, according to an exemplary embodiment.

[0023] Figure 8 Included are schematic diagrams illustrating pulse burst data acquisition for an I-channel or a Q-channel according to some exemplary embodiments.

[0024] Fig. 9 is a schematic diagram showing a Fast Fourier Transform (FFT) implementation of a filter bank of N filters.

[0025] Fig.10 Included is a graph of complex FFT processing gain simulation results according to an exemplary embodiment.

[0026] Fig.11 Includes a schematic functional block diagram according to some exemplary embodiments, which shows the use of Figure 3 The laser radar system with FMCW transmit envelope modulation and orthogonal demodulation is shown.

[0027] Fig. 12A is a schematic timing diagram of a frequency modulated signal having a frequency modulated substantially sinusoidal configuration according to an exemplary embodiment.

[0028] Fig. 12B is a schematic timing diagram of a frequency modulated signal having a frequency modulated basic square wave configuration according to an exemplary embodiment.

[0029] Fig. 12C is used according to an exemplary embodiment Fig. 12A Schematic timing diagram of an envelope modulated optical signal generated from a basic sinusoidal FM signal.

[0030] Fig.12D According to an exemplary embodiment Fig. 12A and Fig. 12B Schematic timing diagram of frequency deviation of an FM signal versus time.

[0031] Fig.13A schematic diagram including a frequency versus time graph to show the intermediate frequency (IF) output frequency from a target object with a distance-dependent delay t d .

[0032] Fig.14 A schematic timing diagram is included for ADC sampling of I-channel and Q-channel IF signals according to an exemplary embodiment.

[0033] Fig.15 Included is a graph with two curves showing experimental data for a FMCW envelope modulated lidar system according to an exemplary embodiment.

[0034] Fig.16 Included is a schematic timing diagram of multi-frequency ramp data acquisition for two-dimensional FFT processing according to an exemplary embodiment.

[0035] Fig.17 Included is a schematic diagram of an M x N data matrix populated with samples from each frequency ramp for a two-dimensional FFT method according to an exemplary embodiment.

[0036] Fig.18 A table is included that lists exemplary parameterized data values ​​for an exemplary exemplary operating configuration for a FMCW lidar system according to an exemplary embodiment.

[0037] Fig.19 is a schematic timing diagram of frequency deviation versus time for a linear frequency ramp envelope modulation waveform having a combination of positive and negative frequency / time ramp envelope modulation slopes according to an exemplary embodiment.

[0038] Fig. 20 is a schematic timing diagram of frequency deviation versus time for a stepped frequency ramp envelope modulation waveform according to an exemplary embodiment.

[0039] Fig.21 is a schematic timing diagram of frequency deviation versus time for a pseudo-random stepped frequency envelope modulation waveform according to an exemplary embodiment.

[0040] Fig. 22 Included are schematic functional block diagrams illustrating a lidar system using stepped frequency modulated (FM) pulse burst transmit envelope modulation and quadrature demodulation and a MEMS scanning mirror according to some exemplary embodiments.

[0041] Fig.23 Includes detailed schematic diagrams according to some exemplary embodiments, which show Fig. 22 The MEMS scanning mirror used in the lidar system transmits the beam pattern.

[0042] Fig.24Includes a schematic timing diagram according to some exemplary embodiments, which shows Fig. 22 and Fig.23 The LiDAR system utilizes a MEMS scanning mirror data acquisition process with a single pulse burst frequency at each scanning mirror position.

[0043] Fig.25 Including according to some exemplary embodiments Fig. 22 Schematic detailed functional block diagram of the MEMS mirror controller / actuator shown.

[0044] Fig.26 A schematic timing diagram showing the step FM modulation pulse and the Fig. 22 and Fig.25 The MEMS mirror controller / actuator provides synchronization of the scanning mirror position.

[0045] FIG. 27A to FIG. 27D Included is a series of four exemplary data matrices for scanning mirror data acquisition according to some exemplary embodiments.

[0046] Fig.28 Included are tables of parameterized data in typical automotive operating scenarios for a LiDAR system using a MEMS scanning mirror for data acquisition according to certain exemplary embodiments.

[0047] Fig.29 Includes a schematic functional block diagram according to some exemplary embodiments, which shows the use of Figure 3 The FMCW transmission envelope modulation and orthogonal demodulation laser radar system shown in FIG. Fig.11 The system shown has been varied.

[0048] FIG. 30A to FIG. 30C Included are data matrices showing FMCW data acquisition for 2-D FFT processing of a single scan angle position according to some exemplary embodiments.

[0049] Fig.31 A schematic logic flow diagram of synthetic Doppler simulation using an FMCW envelope modulated waveform according to an exemplary embodiment is included.

[0050] Fig.32 Included is a table of FMCW synthetic Doppler parameters for simulations with a single beam angle.

[0051] Fig.33 Includes the results of the synthetic Doppler range FFT calculation for a single scan angle.

[0052] Fig.34 Included are the results of the synthetic Doppler FFT for a single scan angle.

[0053] Fig.35 Supplementary data including simulation results of the Doppler FFT after the CFFT is performed.

[0054] Fig.36 Additional supplementary data including simulation results.

[0055] Fig.37 Included are detailed schematic diagrams illustrating transmit beam patterns employed in a lidar system as disclosed herein, according to some exemplary embodiments.

[0056] Fig.38 A table listing parameters associated with data for k=50 beam angle positions is included.

[0057] Fig.39 Includes a schematic functional block diagram according to some exemplary embodiments, which shows the use of Figure 2 The laser radar system with pulsed transmission envelope modulation and orthogonal demodulation shown in FIG. Fig. 6A The system shown has been varied.

[0058] Fig.40A is a timing diagram illustrating data acquisition timing in pulsed lidar processing according to some exemplary embodiments.

[0059] Fig.40B Included are schematic diagrams showing pulsed data acquisition for a single scan angle position according to some exemplary embodiments.

[0060] Fig.40C Included is a raw data matrix generated during pulse data acquisition according to some exemplary embodiments.

[0061] Fig.40D A repopulated data matrix generated during pulse data acquisition according to some exemplary embodiments is included.

[0062] Fig.41 Included is a schematic logic flow diagram of synthetic Doppler simulation using a pulsed envelope modulated waveform according to an exemplary embodiment.

[0063] Fig.42 Contains a table of pulsed synthetic Doppler parameters for simulations of individual beam angles.

[0064] Fig.43 A method for generating noise according to an exemplary embodiment is included Program scripts, signals for pulsed synthetic Doppler simulation and CFFT Program script and simulation results for the CFFT of n data samples from a single range bin.

[0065] Fig.44 A schematic diagram is included showing data acquisition for the I-channel and the Q-channel for a single scan according to an exemplary embodiment.

[0066] Fig.45 A schematic diagram is included showing data acquisition for I-channel and Q-channel at beam position 3 for multiple scans according to an exemplary embodiment.

[0067] Fig.46 A table listing parameters associated with data for a plurality of beam angle positions is included.

[0068] Fig.47 Includes a schematic functional block diagram according to some exemplary embodiments, which shows the use of Figure 3 The FMCW transmission envelope modulation and orthogonal demodulation laser radar system shown in FIG. Fig.29 The system shown has been varied.

[0069] Fig.48 Included is a schematic perspective view of a vehicle equipped with one or more lidar systems described in detail herein according to an exemplary embodiment.

[0070] Fig.49 Included is a schematic top view of a car equipped with two lidar systems as described in detail herein according to an exemplary embodiment. DETAILED DESCRIPTION

[0071] Direct detection lidar systems are characterized by simplicity of construction and function, and unlike more complex homodyne or heterodyne lidar systems, do not utilize frequency translation or down-conversion stages that would promote signal detection and processing gain advantages. The signal detection and processing gain advantages of homodyne / heterodyne lidar systems are achieved by a combination of advanced modulation and coding of the transmitted signal and sophisticated correlation processing techniques within the lidar receiver.

[0072] Transmit signal modulation and coding and advanced related processing techniques have been utilized within radar systems, from complex military object imaging systems to commercial automotive autonomous cruise control applications. Except for very advanced measurement requirements (e.g., NASA's measurement of CO2 emissions), lidar systems have not yet utilized these techniques. However, in accordance with the present disclosure, the development of laser transmit signal envelope modulation and quadrature demodulation of the recovered envelope modulated signal has demonstrated advantages similar to those associated and achieved via radar science. Laser transmitter envelope modulation and quadrature demodulation represent a modest increase in complexity for direct detection lidar systems, and significant benefits in measurement capability and lower operating power are achieved by realizing the signal processing gains of direct detection lidar.

[0073] According to the exemplary embodiments detailed herein, laser transmitter envelope modulation and receiver quadrature demodulation techniques are applied to direct detection lidar systems. Specific transmitter modulation envelope waveforms, such as pulse bursts and frequency modulated continuous waves (FMCW), are described in detail. Data acquisition techniques and processing gains are also described in detail herein. Specific measurement enhancements and parameters associated with each envelope modulation waveform are also described in detail.

[0074] Figure 1 Included are three plots showing a general example of a transmitter envelope modulation technique using a basic sinusoidal modulation envelope as applied to direct detection lidar according to an exemplary embodiment. Figure 1 , showing the modulation envelope signal, the sinusoidal carrier signal and the envelope modulation carrier waveform. The mathematical definitions associated with the envelope modulation waveform, the carrier and the transmission envelope modulation waveform are according to the following equation (1):

[0075] Mod(t)=sin(2πf m t)→Modulation waveform

[0076] Car(t)=sin(2πf c t)→Carrier

[0077] T x (t) = Mod(t)·Car(t) → Envelope modulated carrier (1)

[0078] It should be noted that envelope modulated carrier means the modulating waveform and frequency multiplication of the carrier signal. Direct detection lidar systems perform frequency multiplication within the laser modulator element, as detailed below. Unlike other systems that use modulated carriers, envelope modulation techniques result in the transmission of two sidebands.

[0079] Figure 2 Included are three curves illustrating a transmitter envelope modulation technique using a pulse burst modulation envelope as applied to a direct detection lidar according to other exemplary embodiments. Figure 2 , shows the modulated envelope signal, the sinusoidal carrier signal and the envelope modulated carrier waveform. Figure 2 In pulse burst envelope modulation as shown, a repetitive pulse waveform modulates a carrier. Laser modulators are capable of pulse modulation at very high repetition rates (eg, hundreds of megahertz), which facilitates coherent detection of the recovered modulation waveform with attendant signal processing benefits, as will be described in detail herein.

[0080] It should be noted that according to an exemplary embodiment, the temporal position of the modulation pulses may be variable, which allows for pulse position modulation (PPM) encoding.

[0081] Figure 3Included are three curves illustrating a transmitter envelope modulation technique using linear frequency envelope modulation as applied to direct detection lidar according to other exemplary embodiments. Figure 3 , shows a modulated envelope signal with linear frequency variation, a sinusoidal carrier signal, and an envelope modulated carrier waveform. Figure 3 Linear frequency envelope modulation is shown, where in this particular exemplary embodiment, the modulation waveform frequency changes linearly from f1 to f2 (ΔF) within a specific time interval (ΔT). The linear frequency modulation envelope is beneficial for the implementation of FMCW lidar due to the ability to provide high range resolution based on frequency deviation (ΔF), lower detection bandwidth, and the unique spectrum resolution characteristics of the Fast Fourier Transform (FFT) calculation technique.

[0082] One principle of transmitter envelope modulation is that during transmission, the modulated envelope is phase delayed according to the envelope frequency. The total transmitted phase shift from the LiDAR system to the two-way distance of the object is described by the following equation (2):

[0083]

[0084] Transmission Phase Shift

[0085] R: distance from the object

[0086] λ mod : Envelope modulation wavelength

[0087] In an exemplary embodiment, as described in detail below, the amplitude and transmission phase of the modulation envelope are demodulated in an orthogonal demodulator after envelope recovery in a photodiode or photodetector or photodetector.

[0088] Figure 4 Included are schematic functional block diagrams of quadrature demodulation systems and techniques according to some exemplary embodiments. Quadrature demodulation is a technique that takes advantage of coherent signals to provide efficient detection of orthogonal or vector signal components of a modulated signal. Quadrature demodulation is general in the sense that it has the ability to recover the amplitude modulated (AM), frequency modulated (FM), and phase modulated (PM) components of a modulated signal. See Figure 4 , receiving a modulated input signal to be demodulated having a certain frequency and applying it to in-phase and quadrature phase mixers. The quadrature demodulator according to an exemplary embodiment includes a coherent continuous wave (CW) local oscillator (LO) signal at a modulated carrier input frequency f0, a 0° / 90° power divider, in-phase and quadrature phase mixers, and a low pass filter to eliminate the LO signal and other spurious signals from the demodulated output, as shown in the figure, which is provided at the I-channel output and the Q-channel output. In order to make the LO coherent with the received envelope modulation signal, according to an exemplary embodiment, a single frequency source is used for both the envelope modulation and the quadrature demodulator LO.

[0089] This is quite different from the envelope modulation / quadrature demodulation direct detection lidar method of the exemplary embodiment. Figure 5 A schematic functional block diagram of a conventional direct detection laser radar system 50 is included. Figure 5 , a typical operating configuration involves the transmission of a short duration (typically 2.0 to 20 nanoseconds) high power laser transmit pulse at transmit optics 52 via optical transmitter 66, which is modulated by laser modulator 64 under the control of digital signal processor and control (DSPC) 68. A portion of the transmit pulse is reflected from an object within the transmitter beam width and subsequently captured by receive optics 54 and photodetector 56 after a two-way flight time to the object. The received signal is then amplified by transimpedance amplifier (TIA) 58 and filtered by low pass filter (LPF) 60. Analog to digital converter (ADC) 62 samples range bins commensurate with the pulse width. Under the control of DSPC 68, a target is declared if the signal is determined to exceed a specific threshold level within a specific range bin. Other processing strategies may be employed to improve the signal to noise ratio, such as range bin sampling after multiple transmitter pulses and integration of the signal energy received from each transmit pulse (also known as incoherent detection); however, the basic operation is limited to high power pulse transmission and receive signal detection and amplification.

[0090] In addition to the above-described direct detection LiDAR system 50, a time of flight (TOF) system transmits multiple pulses in the form of a square wave and measures the two-way time of flight using a phase detector upon reception. A time of flight system must limit the square wave modulation frequency to avoid phase ambiguity.

[0091] Fig. 6A Includes a schematic functional block diagram according to some exemplary embodiments, which shows the use of Figure 2 The laser radar system 100 with pulse burst transmission envelope modulation and orthogonal demodulation is shown. Figure 5 The basic direct detection lidar system 50 shown is quite different from the conventional pulse modulation. Fig. 6A In the system 100, bursts of repetitive pulses are used instead of single repetitive pulses. Figure 6B is a schematic diagram showing a schematic diagram according to some exemplary embodiments Fig. 6A The pulse burst modulation signal 115 used in the system 100 is detailed. Fig. 6AIn the system 100 of FIG. 6 , the burst frequency (which is 200 MHz in some specific exemplary embodiments) produces the transmission of two sidebands, each of which is offset from the operating laser wavelength (frequency) by the modulation frequency. In addition, in the system 100 of FIG. 6 , the modulation frequency is coherent with the envelope recovered upon reception, thereby providing an efficient means for modulated signal detection. In addition, a bandpass filter centered at the modulation burst frequency attenuates the broadband noise of the TIA and also attenuates the l / f noise associated with the photodetector and the TIA. In addition, an orthogonal demodulator is employed to recover the bidirectional transmission phase shift of the pulse burst envelope and the accompanying modulation envelope.

[0092] See also Fig. 6A and Figure 6B , the laser radar system 100 according to an exemplary embodiment includes a receiving optical device 154, at which optical energy is received, including optical echoes from one or more target objects. The optical energy from the receiving optical device 154 is received at a photodetector 156, which converts the received optical energy into one or more electrical signals. These electrical signals are amplified by a TIA 158 and filtered by a BPF 160, which has a center frequency at a burst modulation frequency f0. In some specific exemplary embodiments, the center frequency f0 of the BPF 160 is 200 MHz, but other frequencies may be used. The resulting amplified and filtered signal is applied to the first input of the I / Q mixer 162, 164 at a node 161.

[0093] An optical signal for illuminating the one or more target objects is generated by a signal generator 182. The output signal of the signal generator 182, which is a continuous substantially sinusoidal signal in an exemplary embodiment, is applied to a power divider 184, which divides the signal and provides the divided signal at two outputs. The first output 185 is routed to a distribution and phase shift circuit or a 90-degree power divider 186, which divides the signal, applies a phase shift to one of the resulting divided signals, and generates a pair of output signals with phase shifts. In an exemplary embodiment, a 90-degree phase shift is applied to one of the signals, so that the distribution and phase shift circuit or the 90-degree power divider 186 generates a first "in-phase" local oscillator (LO) signal 189 and a second "quadrature phase" or "quadrature" LO signal 191, which is phase-shifted 90 degrees relative to the in-phase LO signal 189. The in-phase and quadrature phase LO signals 189, 191 are applied to the second inputs of the VQ mixers 162, 164, respectively. The I / Q mixers 162, 164 mix the amplified and filtered input signals with the in-phase and quadrature phase LO signals 189, 191, respectively, at node 161 to generate output signals 193, 195, respectively, which are low pass filtered by low pass filters (LPF) 166 and LPF 168, respectively. The resulting filtered analog signals are converted to digital signals by analog-to-digital converters (ADCs) 170, 172, respectively, and sampled under the control of a sample control signal 197 generated by a digital signal processor and control (DSPC) 168. The resulting sampled digital VQ (quadrature) signals (i.e., I-channel and Q-channel signals) 105, 107 are processed by the DSPC 168 to determine the distance and / or velocity of the one or more target objects. The results of this detection processing performed by DSPC 168 may be forwarded via system interface 109 to, for example, a user interface, as desired.

[0094] Continue to see Fig. 6A and Figure 6B, the second output 187 of the power divider 184 is routed to the pulse modulator 174, which converts the continuous basic sinusoidal signal 187 from the power divider 184 into a pulsed basic sinusoidal signal 111. The timing of the pulses in the pulsed basic sinusoidal signal 111 is controlled by the pulse burst modulation signal 115 on the output signal line 113 from the DSPC 168. That is, the continuous basic sinusoidal signal 187 is modulated by the pulse modulator 174 using the pulse burst modulation signal 115 to generate the pulsed basic sinusoidal signal 111. The pulsed amplitude modulated signal 111 obtained from the pulse modulator 174 is applied as a modulation signal to the laser modulator 176 to generate a control / modulation signal, which is applied to the optical transmitter 178 to generate a pulsed amplitude modulated light signal, which is transmitted to the transmitting optical device 180, and the pulse modulated light signal is transmitted to the one or more target objects through the transmitting optical device.

[0095] Therefore, according to an exemplary embodiment, quadrature detection precedes analog-to-digital conversion. The quadrature detector recovers the pulse modulation envelope associated with the low-frequency pulse modulation. The data samples are then processed via spectral parsing of each range bin data set or otherwise. The spectral parsing method used reduces the detection bandwidth and effectively integrates the energy of the range bin sample set.

[0096] According to an exemplary embodiment, Doppler frequency detection of moving objects within the field of view of the laser radar system 100 is achieved by coherent detection of phase changes caused by distance changes caused by the relative velocity between the laser radar system 100 and the target object. Doppler detection is important because the reduction in detection bandwidth associated with Doppler processing can increase the signal-to-noise ratio proportional to the bandwidth ratio. Therefore, according to the present disclosure as described in detail below, the synthetic Doppler technique of the present disclosure is a way to synthetically induce "motion" of stationary objects, thereby facilitating 2-D FFT processing.

[0097] According to some exemplary embodiments, in addition to Doppler processing, a coherent pulse integration method may be performed using multi-frame range bin sampling, which also increases the signal-to-noise ratio proportional to the number of pulses integrated.

[0098] Figure 7 is a graph including three curves showing experimental data associated with the pulse burst envelope modulation lidar system 100 described in detail herein, according to an exemplary embodiment. Figure 7, the upper and middle curves are the I-channel and Q-channel video outputs, indicating the detection of the modulation envelope after coherent detection. If the detected object exhibits motion, the I-channel and Q-channel signal amplitudes will change according to the transmission phase changes associated with the object motion. This phenomenon gives rise to the ability to detect Doppler frequency using pulse burst envelope modulation and orthogonal demodulation. Figure 7 The trace at the bottom of the graph is the photodiode signal at the output of the TIA 158, which shows a 200 MHz envelope modulated signal before coherent detection. The time scale is 50 nanoseconds per division.

[0099] A detailed description of data acquisition and signal processing techniques that may be used in conjunction with a pulse burst envelope modulation lidar system 100 employing quadrature demodulation is now provided. The first step in the pulse burst envelope modulation lidar system signal processing is to acquire a data set representing the signal level for each range bin from the successive transmission pulses at the output of each channel of the quadrature demodulator.

[0100] Figure 8 Included are schematic diagrams illustrating pulse burst data acquisition for an I-channel or a Q-channel according to some exemplary embodiments. Figure 8 represents the filling of the data matrix (data set) after consecutive transmission pulses. According to an exemplary embodiment, a frame is defined as a single transmission pulse burst followed by a receive interval during which the ADC 170, 172 collects samples from each channel of the orthogonal demodulator output at each range bin. The range bins are defined according to the pulse burst width which sets the range measurement resolution. The range resolution can be written as equation (3):

[0101]

[0102] δR: distance resolution

[0103] t W : Pulse burst width

[0104] c: speed of light

[0105] Refer to the I-channel or Q-channel signal. Figure 8 , the arrows designate the sample points of the ADCs 170, 172. A variable amplitude echo signal is indicated in range bin 12. The variable amplitude is a result of phase changes of moving objects within range bin 12. Stationary objects produce fixed amplitude signal levels.

[0106] After the range bin samples in each of the N frames are completed, an M×N data matrix is ​​populated for each channel of the quadrature demodulator. It should be noted that the variable amplitude range bin pulses are generated by moving objects that are coherently detected by the quadrature demodulator, and Doppler frequency measurement is achieved via spectral resolution of the range bin samples.

[0107] There are two processing methods that can be used to enhance signal detection via an increase in signal-to-noise ratio: coherent pulse integration and spectral resolution of each column of the data matrix. Each column of the data matrix represents a range bin sample acquired at a discrete time point after the pulse burst transmission. The method related to coherent pulse integration is captured in the following equation (4):

[0108]

[0109] I n : The nth I-channel value of the distance bin column

[0110] Q n : The nth Q-channel value of the distance bin column

[0111] In the presence of zero-mean, Gaussian noise, ideal coherent pulse integration improves the signal-to-noise ratio by N, the number of pulses integrated.

[0112] The spectrum analysis is performed according to the discrete Fourier transform and the following equation (5):

[0113]

[0114] DFT k : The kth element of the discrete Fourier transform

[0115] I n and Q n : As defined previously

[0116] The FFT is a computationally efficient technique for the calculation of the discrete Fourier transform, which implements a set of identical filters or filter banks uniformly spaced in the frequency domain at intervals of 1 / NT, where T is the time interval of N samples of the acquired waveform. Fig. 9 , which is a schematic diagram showing an FFT implementation of a filter bank of N filters. FFT is particularly suitable for IF spectrum resolution in FMCW radars, because the narrow information bandwidth requires a filter bank that can be digitally implemented using a moderately capable digital signal processor.

[0117] It can be used to quantify the signal processing gain of the pulse burst lidar envelope modulation waveform. According to an exemplary embodiment, Fig.10 The simulation results are recorded in the curve graph of , which is a curve graph of the simulation results of complex FFT processing gain. Fig.10 , the results show a processing gain of about 25.5 dB, as estimated from the graph. The pre-processing signal-to-noise ratio was set to 0 dB. The theoretical processing gain is calculated by PG = 10·log(N = 512) = 27.1 dB. A signal of -1.5 dB in bin k = 93 was recorded; a noise level of -27.0.

[0118] According to an exemplary embodiment, the direct detection lidar described in detail herein may also use FMCW modulation (instead of pulse burst modulation) as well as orthogonal demodulation. Fig.11 Includes a schematic functional block diagram according to some exemplary embodiments, which shows the use of Figure 3 A laser radar system 200 with FMCW transmit envelope modulation and orthogonal demodulation is shown. Fig.11 2 shows a system block diagram of a transmitter envelope modulated laser radar system 200, which uses a linear deviation of the modulation frequency from a first frequency f1 to a second frequency f2, so that the linear deviation ΔF of the modulation frequency is given by ΔF=f2-f1 within a ΔT time interval, as shown in Fig.11 It should be noted that all of the above detailed descriptions for pulse burst modulation in the laser radar system 100 are applicable to FMCW modulation in the laser radar system 200 unless the context dictates otherwise.

[0119] See also Fig.11 , the laser radar system 200 according to an exemplary embodiment includes a receiving optical device 254 at which optical energy is received, including optical echoes from one or more target objects. The optical energy from the receiving optical device 254 is received at a light detector 256, which converts the received optical energy into one or more electrical signals. According to an exemplary embodiment, these electrical signals are amplified by a TIA 258 and filtered by a BPF 260, which has a low cutoff frequency f1 and a high cutoff frequency f2. h , these frequencies define the nominal endpoints of the FM frequency in each linear sweep used to modulate the optical carrier signal. In some specific exemplary embodiments, f1=100 MHz and f2=200 MHz, but other frequencies may be utilized. The resulting amplified and filtered signal is applied to the first input of the I / Q mixer 262, 264 at node 261.

[0120] The modulated FM signal is generated by a voltage controlled oscillator (VCO) 282 under the control of a control signal from a phase locked loop (PLL) control circuit 283, which in turn is controlled by the DSPC 268. Fig.11As shown, the frequency of the VCO 282 output is controlled to take a linear ramp configuration from a first frequency f1 to a second frequency f2. The output signal of the VCO 282 is applied to a power divider 284, which divides the signal and provides the divided signal at two outputs. The first output 285 is routed to a distribution and phase shift circuit or a 90-degree power divider 286, which distributes the signal, applies a phase shift to one of the resulting distributed signals, and generates a pair of output signals with phase shifts. In an exemplary embodiment, a 90-degree phase shift is applied to one of these signals, so that the distribution and phase shift circuit or the 90-degree power divider 286 generates a first "in-phase" local oscillator (LO) signal 289 and a second "quadrature phase" or "quadrature" LO signal 291, which is phase-shifted 90 degrees relative to the in-phase LO signal 289. The in-phase and quadrature-phase LO signals 289, 291 are applied to the I / Q mixers 262, 264, f if_max The second input of . I / Q mixers 262, 264 mix the amplified and filtered input signals with in-phase and quadrature phase LO signals 289, 291, respectively, at node 261 to generate output signals 293, 295, respectively, which are low pass filtered by low pass filters (LPF) 266 and LPF 268, respectively, which have cutoff frequencies at the maximum IF frequency if max of the mixers 262, 264. The resulting filtered analog signals are converted to digital signals by analog-to-digital converters (ADCs) 270, 272, respectively, and sampled under the control of a sample control signal 297 generated by the DSPC 268. The resulting sampled digital I / Q (quadrature) signals (i.e., I-channel and Q-channel signals) 205, 207 are processed by the DSPC 268 to determine the distance and / or speed of the one or more target objects. The results of the detection processing performed by the DSPC 268 can be forwarded to, for example, a user interface via the system interface 209 as needed.

[0121] Continue to see Fig.11 , the second output 287 of the power divider 284 is routed to the laser modulator 274, which modulates the optical output signal at the optical transmitter 278 using the linear FM signal 217, which has a frequency that varies linearly from f1 to f2. The resulting frequency modulated signal is applied to the optical transmitter 278 to generate an envelope modulated optical signal 219 (see Fig. 12C ), wherein the envelope is a frequency modulated signal, the envelope modulated light signal is transmitted to the transmitting optical device 280, and the envelope frequency modulated light signal 219 is transmitted to the one or more target objects through the transmitting optical device.

[0122] Fig. 12Ais a schematic timing diagram of a frequency modulated signal 217 having a frequency modulated substantially sinusoidal configuration according to an exemplary embodiment. Fig. 12B is a schematic timing diagram of a frequency modulated signal 217A having a frequency modulated basic square wave configuration according to an exemplary embodiment. Fig. 12C is used according to an exemplary embodiment Fig. 12A Schematic timing diagram of an envelope modulated optical signal 219 generated by a basic sinusoidal frequency modulated signal 217. Fig.12D According to an exemplary embodiment Fig. 12A and Fig. 12B Schematic timing diagram of frequency deviation of FM signals 217 and 217A versus time. Fig.11 and FIG. 12A to FIG. 12D , the lidar antenna or transmitting optical device 280 transmits an envelope modulated light signal 219 toward an object at a distance R. The frequency variation of the envelope modulation waveform is linear within a finite duration ΔT and a frequency deviation ΔF. It should be noted that according to exemplary embodiments, other envelope frequency modulation waveforms may be utilized, such as FSK (frequency shift keying), stepped FSK, PRFM (pseudo-random frequency modulation), etc.

[0123] Upon impacting an object within the beam width of the transmit antenna 280, the transmit signal is scattered or reflected depending on the geometry and other physical properties of the object. A portion of the scattered signal is received by the lidar photodetector, where the FM modulated envelope is recovered and then amplified by the transimpedance amplifier 258.

[0124] The recovered / amplified FM modulated waveform envelope is further processed using a bandpass filter 260 centered at the arithmetic mean (f2-f1) / 2 of the frequency extremes and a bandwidth (f2-f1) commensurate with the envelope modulation frequency extremes. The bandpass filter 260 rejects extraneous signals as well as broadband noise from the TIA 258 and 1 / f noise of the detector 256 and TIA 258. The received signal propagates to the input of the quadrature demodulator where the difference frequency is detected and applied to the ADC 270, 272. The frequency difference is caused by the two-way range delay and the delay difference between the coherent local oscillators at the quadrature demodulator inputs.

[0125] According to an exemplary embodiment, coherent detection of the linear FM modulation envelope is based on Fig.13 The distance information is provided by a graph of frequency versus time, which includes a schematic diagram of an intermediate frequency (IF) output frequency from a target object and a distance-dependent delay t d . Fig.13 The relationship between the transmitted signal, the received (delayed) signal and the resulting difference frequency is shown. Fig.13 , the signal frequency at the output of the quadrature demodulator can be written according to equation (6):

[0126]

[0127] t d : Two-way delay

[0128] R: object distance,

[0129] ΔF: Ramp frequency deviation

[0130] ΔT: Ramp deviation interval

[0131] c: speed of light

[0132] ADC 270, 272 collects samples of the quadrature demodulator output during the linear frequency ramp interval ΔT; the sample sequence is then subjected to spectral analysis. The spectral analysis method is generally performed based on a Fast Fourier Transform (FFT), which implements a filter bank of discrete range bins. Each range bin is checked to determine whether a signal is present at a specific threshold level.

[0133] Fig.14 A schematic timing diagram including ADC sampling of I-channel and Q-channel IF signals according to an exemplary embodiment. Fig.14 During the transmit interval, the I-channel IF signal and the Q-channel IF signal are continuously sampled by ADCs 270, 272 and the samples are stored in the DSPC 268 memory. Fig.14 In the example, ADC samples are represented by vertical arrows. The sampling rate f s is the number of samples N divided by the frequency ramp time ΔT, and must be greater than the highest IF frequency (for I / Q sampling) in order to comply with the Nyquist sampling theorem. Therefore:

[0134]

[0135] Data acquisition and signal processing techniques that can be used in conjunction with FMCW envelope modulated laser radar systems and orthogonal demodulation according to some exemplary embodiments will now be described in detail. In addition to the one-dimensional FFT processing described above, a two-dimensional FFT can also be performed, which further enhances the detection process and provides Doppler frequency detection. The two-dimensional FFT is also detailed herein, where a processing gain of 24.1 dB is calculated for a 256-point range FFT and a processing gain of 21.1 dB is calculated for a 128-point Doppler FFT.

[0136] Fig.15 Included is a graph with two curves showing experimental data of FMCW envelope modulated lidar system 200 according to an exemplary embodiment. Fig.15, the curve marked by a "diamond" or a square rotated 45 degrees is the I-channel signal, and the curve marked by an upright square is the Q-channel signal. Fig.15 The experiment shown, ie, the frequency was manually changed from 100 MHz to 200 MHz in 1.0 MHz increments, while the I-channel and Q-channel amplitudes were recorded at each frequency step. Fig.15 The graph of shows quadrature signals as expected and an illustrative exemplary object distance of approximately 3.0 meters. The distance is calculated by recording the frequency change required for a full cycle (2π) of the I-channel or Q-channel and executing in the following equation (8):

[0137]

[0138] Solution

[0139] In addition to a single object within the transmit optics beamwidth, additional objects induce additional IF frequencies that are proportional to the distance of the individual objects. A useful technique for detecting multiple objects uses spectral resolution of the aggregated sampled data. The FFT is a computationally efficient procedure for the calculation of the Discrete Fourier Transform (DFT) that implements a set of identical filters or filter banks uniformly distributed in the frequency domain at intervals of 1 / NT, where T is the time interval of N samples of the acquired waveform (also ΔT in this case). The FFT is particularly suitable for IF spectral resolution in FMCW radars because the narrow information bandwidth requires filter banks that can be implemented digitally using moderately capable digital signal processors. As described above, Fig. 9 is a schematic diagram showing the filter bank properties of an FFT according to an exemplary embodiment. The effect of the filter bank is important because the signal energy from each detected object is concentrated within a single filter of the filter bank to the extent of range resolution; and more importantly, the noise detection bandwidth is reduced, which improves the signal-to-noise ratio. The initial noise detection bandwidth is a low pass filter 266, 268 following the I-channel and Q-channel mixers with a cutoff frequency established by the IF signal frequency at the maximum distance defined previously.

[0140] The processing gain of an N-point FFT is given by

[0141]

[0142] After acquiring the I-channel and Q-channel data, a signal processing step (e.g., FFT) is performed and the signal level of each filter is tested and compared to a previously established threshold. Threshold testing is used to determine the presence or absence of an object and can trigger additional techniques for object recognition.

[0143] In addition to FFT processing of singular frequency ramp data, known as a one-dimensional FFT, according to an exemplary embodiment, additional signal processing gain is achieved via a two-dimensional FFT process, where sample data is collected from multiple frequency ramps to extract object relative velocity, such as related to Doppler frequency, and further reduce noise detection bandwidth, thereby providing additional processing gain. According to an exemplary embodiment, the two-dimensional FFT uses a two-dimensional FFT process such as Fig.16 The figure shows sampled data from multiple frequency ramps, and includes a schematic timing diagram for multi-frequency ramp data acquisition for two-dimensional FFT processing.

[0144] Fig.17 The two-dimensional FFT process sequence is further illustrated in FIG, which includes a schematic diagram of an M×N data matrix for a two-dimensional FFT method according to an exemplary embodiment, the data matrix being populated with samples from each frequency ramp. Fig.17 , a range FFT may be performed on each row of the data matrix, while a Doppler FFT may be performed on each column of the data matrix. Additional processing is used to associate or identify specific objects with the appropriate Doppler frequency (velocity).

[0145] Continue to see Fig.17 , the data matrix is ​​populated in a manner similar to the pulse burst lidar data matrix detailed above, according to the two-dimensional FMCW signal processing for the FMCW lidar system 200. However, in the case of the FMCW lidar system 200, the rows of the data matrix are populated with samples collected during the linear frequency ramp. The columns of the data matrix are populated by samples at discrete frequencies of the linear ramp at a fixed time after the linear frequency ramp is initiated. In this particular exemplary embodiment, the data matrix is ​​generated by N samples of each of the M frequency ramps within a frequency difference of ΔF=f2-f1, where f2=200MHz and f1=100MHz. The sampling frequency or sampling rate f in samples per second (SPS) within the time interval ΔT sR Get N samples.

[0146] Continue to see Fig.17, in the data matrix for the FMCW lidar system 200, the row data represents the received signal samples from each channel of the orthogonal demodulator during the frequency ramp time interval; the subsequent FFT processing is called the range-FFT. The column data represents the received signal samples from each channel of the orthogonal demodulator for the corresponding discrete time points of each frequency ramp, and is called the Doppler-FFT. Two-dimensional FMCW signal processing achieves a basic signal processing gain advantage. Unlike the range-Doppler processing of the pulse burst envelope modulated lidar 100 detailed above, in the FMCW lidar system 200, the range samples require a less restricted sampling rate, that is, the pulse burst envelope modulated lidar 100 requires at least one sample at each range bin, which converts to 1 / t W The FMCW LiDAR system 200 requires a minimum distance sampling rate based on the IF frequency at the maximum operating distance. This is important because high-speed ADCs have a higher cost. Mathematically stated as follows:

[0147]

[0148] f s : Sampling rate

[0149] R max : Maximum operating distance

[0150] c: speed of light

[0151] ΔF: Ramp frequency deviation

[0152] ΔT: frequency ramp time

[0153] Fig.18 A table is included that lists exemplary parameterized data values ​​for an exemplary exemplary operating configuration of the FMCW lidar system 200 according to an exemplary embodiment. Fig.18 The parameterized data of a typical road vehicle application are given in this article as a basis for performance calculations. It should be emphasized that the parameterized data, especially speed, range, range resolution, data cycle time and object detection geometry are application-specific parameters and therefore differences are expected for other operating scenarios. Fig.18 The exemplary values ​​in vary widely.

[0154] See also Fig.17 and Fig.18 In order to achieve the maximum processing gain provided by two-dimensional FFT calculation, the detection object is detected during the data acquisition time T acq The duration of the range resolution unit must remain within a single range resolution unit. For FMCW modulation waveforms, the range resolution unit is determined according to the following formula:

[0155]

[0156] δR: distance resolution unit (meter)

[0157] c: speed of light (m / s)

[0158] ΔF: Frequency deviation (Hz)

[0159] At maximum approach speed, the object distance unit dwell time can be written as:

[0160]

[0161] T dw : Distance unit time (seconds)

[0162] v max :70(m / s)

[0163] The data acquisition time is the time required to fill the data matrix and can be written as:

[0164] T acq =MΔT=0.0168; (13)

[0165] T acq : Data collection time (seconds)

[0166] M: frequency ramp number

[0167] ΔT: Frequency ramp time (seconds)

[0168] A comparison of the range cell object dwell time to the data acquisition time indicates that the conditions for optimum processing gain are met.

[0169] The minimum number of range samples affects the range signal processing gain and is limited by the IF frequency at the maximum operating range according to the following equation:

[0170]

[0171] IF frequency at maximum distance (Hz)

[0172] R max : Maximum operating distance (m)

[0173] Therefore, the minimum number of samples can be found through the following equation (15):

[0174] N min =f if_max ΔT=128 (15)

[0175] It should be noted that for additional processing gain, the sampling rate can be increased. In this case, the ADC 270, 272 will be able to use a higher sampling rate without affecting performance. For the illustrated exemplary embodiment, the number of samples has been increased to provide greater processing gain, especially for conditions where the object spans adjacent range bins. The number of samples and the range sampling rate for the illustrative example can now be written as:

[0176] N = 256 samples per frequency ramp

[0177] F sR =2.0×10 per second 6 Samples(16)

[0178] Continuing with the parameterization and numerical analysis, the maximum Doppler frequency can be calculated by knowing the nominal envelope modulation wavelength / frequency and the maximum velocity:

[0179]

[0180] f D_max : Maximum Doppler frequency (Hz)

[0181] λ m : Modulation wavelength (m)

[0182] f m : Modulation frequency (Hz)

[0183] In some exemplary embodiments, the modulation frequency is generally limited by the laser modulator and is typically less than 2 GHz for low-cost modulators; however, significantly higher frequency laser modulators have been reported.

[0184] It is advantageous to calculate the noise detection bandwidth for both the range and Doppler parameters as indicated below in equation (18). The range and Doppler bandwidth are important in the detection process because the noise level is determined by the values ​​of the range and Doppler bandwidth.

[0185]

[0186] B wR : FFT distance bandwidth (Hz)

[0187] f sR : Distance sampling rate (samples / second)

[0188] N: number of distance samples; (18)

[0189]

[0190] f sD : Doppler sampling rate (samples / second)

[0191] M: number of Doppler samples

[0192] It should be noted that the ratio of the range sampling rate to the Doppler sampling rate provides an estimate of the Doppler processing gain and serves as a complement to the range signal processing gain.

[0193] The processing gain of the range and Doppler FFT can be estimated using the following equation (19):

[0194] PG R_dB =10·log(N)=24.1dB

[0195] PG D_dB =10·log(M)=21.1dB (19)

[0196] Important elements of the exemplary embodiment include transmission phase shift variation of the envelope modulation waveform within bidirectional distance from the object, and coherent detection of the envelope modulation waveform within the quadrature demodulator.The local oscillator for the quadrature demodulator is also the source of the envelope modulation signal.

[0197] The basic characteristic of the transmitter envelope modulation according to the exemplary embodiment is that when transmitted, the modulation envelope is phase delayed according to the envelope modulation frequency. After recovering the modulation envelope in the photodetector diode, the amplitude and transmission phase of the modulation envelope are detected in the orthogonal demodulator.

[0198] According to an exemplary embodiment, the total transmission phase shift from the lidar to the two-way distance of the object is described by the following equation (20):

[0199]

[0200] Transmission Phase Shift

[0201] R: distance from the object

[0202] λ: envelope modulation wavelength

[0203] According to an exemplary embodiment, the mathematical establishment of the Doppler frequency is as follows:

[0204] R=2(R 0+ vt)

[0205] R0 is a fixed distance from the object

[0206] v·t is the change in distance, that is, the speed-time product after substitution

[0207]

[0208] Define the Doppler frequency:

[0209] Perform a difference:

[0210] Throughout this detailed description, embodiments have been described in which the envelope modulation waveform varies linearly with a positive frequency / ramp slope, i.e., the frequency of the envelope modulation follows a linear ramp function (such as Fig.12D It should be understood that the present disclosure is also applicable to linear ramp envelope modulation waveforms that also have a negative slope, that is, envelope modulation in which the frequency of the envelope decreases rather than increases over time.

[0211] Furthermore, in addition to the linear ramp FMCW waveform detailed herein, the present disclosure is also applicable to other alternative envelope modulation waveforms that provide some flexibility and, in some cases, unique operating advantages. Such alternative envelope modulation waveforms may include positive / negative linear frequency ramp envelope modulation waveforms, stepped frequency ramp envelope modulation waveforms, and pseudo-random envelope frequency modulation (FM) waveforms. Examples of these waveforms are described in Fig.19 , Fig. 20 and Fig.21 Shown in.

[0212] Fig.19 is a schematic timing diagram of frequency deviation versus time for a linear frequency ramp envelope modulation waveform with a combination of positive and negative frequency / time ramp envelope modulation slopes according to an exemplary embodiment. Fig.19 , positive and negative frequency ramps are used together with equal or unequal durations. When positive and negative ramps are deployed together, Doppler frequency measurement can be achieved according to the following equation. Fig.11 The equations for the IF frequencies at the I-Ch and Q-Ch outputs 205, 207 of the laser radar system 200 shown in the block diagram are:

[0213]

[0214]

[0215] After substituting the variable distance, the IF equation can be written as:

[0216]

[0217] Therefore, the IF frequency consists of two components: the component at the beginning of the ramp due to the initial distance (R0); and the component due to the change in distance (vt) due to the relative velocity. The positive or negative slope of the ramp imparts a positive or negative offset to the IF frequency, which provides the Doppler frequency after spectral resolution of the acquired data set during each frequency ramp. The Doppler frequency has a positive / negative (±) effect, depending on the increasing (-) or decreasing (+) distance due to velocity.

[0218] Fig. 20is a schematic timing diagram of frequency deviation versus time for a stepped frequency ramp envelope modulation waveform according to an exemplary embodiment. The difference between the linear frequency ramp and the stepped frequency ramp is the time dwell at the discrete frequency points.

[0219] According to exemplary embodiments, parameterized operational considerations for the stepped frequency ramp are made. For example, in some exemplary embodiments, the frequency step (Δf) does not exceed a value that causes a bidirectional phase shift greater than 2π at the maximum operating distance. This condition can be mathematically shown by the following equation (25):

[0220]

[0221] or

[0222] Δf: Maximum frequency step (Hz)

[0223] cSpeed ​​of light (m / s)

[0224] R max : Maximum operating distance (m),

[0225] For example, if R max =150 meters, then Δf≤1.0×10 6 Hz

[0226] In addition, in addition to the maximum frequency step increment, the frequency also stays at a fixed value for a time increment greater than the two-way flight time from the target at the maximum operating distance. This is mathematically expressed in equation (26) as follows:

[0227]

[0228] For example, if R max =150 meters, then Δt>1.0×10 -6

[0229] Fig.21 is a schematic timing diagram of frequency deviation versus time for a pseudo-random stepped frequency envelope modulation waveform according to an exemplary embodiment. The pseudo-random stepped frequency envelope modulation waveform is similar in some respects to Fig. 20 The stepped frequency waveform shown has similar restrictions on eligibility. Fig.21 The pseudo-random stepped frequency envelope modulation waveform can be used, for example, to mitigate interference generated by multiple lidar sensors or systems operating in a common environment.

[0230] See also Fig.21, the frequency is pseudo-randomly stepped over time intervals in a manner that assigns the individual frequencies of the composite linear ramp to known time positions. Additional processing is performed to aggregate the acquired data points in a manner that reconstructs the linear frequency ramp. The pseudo-random pattern may be fixed, or the pattern may change in each time interval. It should be noted that the same restrictions on the magnitude of the frequency and time increments apply to the pseudo-random frequency step waveform. The pseudo-random frequency step waveform is particularly useful in environments where multiple lidar systems may be operational and interference must be mitigated.

[0231] According to the exemplary embodiments detailed herein, transmitter envelope modulation and receiver quadrature demodulation techniques are applied to a direct detection lidar system. It has been demonstrated that the technique of transmit envelope modulation combined with receive quadrature demodulation as applied to a direct detection lidar system provides a signal processing gain, as determined by an increase in the signal-to-noise ratio at the detection stage of the system. Important operating factors associated with the exemplary embodiments include the transmission phase shift variation of the envelope modulation waveform over a bidirectional distance from an object, and the coherent detection of the envelope modulation waveform within the quadrature demodulator. Additionally, in the exemplary embodiment, an envelope modulation waveform is derived from a quadrature demodulation local oscillator, thereby establishing a coherent signal for detection.

[0232] The signal processing gains achieved in direct detection lidar systems according to the present disclosure far outweigh the modest increase in hardware complexity. The availability of integrated circuit phase-locked loop and quadrature demodulation functions provides ease of implementation with minimal impact on system capacity, operating power and cost. In addition, the lidar architecture detailed herein provides a system with lower transmit power, longer measurement range, reduced power consumption and better performance in multiple system deployment conditions.

[0233] According to an exemplary embodiment, synthetic Doppler technology is used in a scanning lidar detection system to achieve additional signal processing gain. The scanning lidar implementation may be of the type described in co-pending U.S. patent application Ser. No. 15 / 410,158, filed Jan. 19, 2017, docket No. AAM-0031US, which is incorporated herein by reference in its entirety. In the co-pending application, the invention is combined with Figures 14 to 20 The scanning lidar system is described in detail. The figures are Figure 22 to Figure 28 Included in this article. Figure 22 to Figure 28 The following detailed description thereof is extracted from and substantially copied from the co-pending application, but appropriate record changes have been made for clarity and to avoid the inclusion of irrelevant material and excessive redundancy. It should be noted that the scanning and synchronization techniques described in the co-pending application are described in conjunction with the step FM pulse burst transmission optical signal. It should be understood that the detailed description is applicable to the waveforms and embodiments detailed herein.

[0234] In some exemplary embodiments, the laser radar system may include a micro-electromechanical system (MEMS) scanning mirror for enhancing the processing of the optical signal. The MEMS scanning mirror is one of the technologies used to achieve laser beam scanning. The MEMS mirror is manufactured using semiconductor technology, which promotes high-yield manufacturing, repeatable performance and low cost. Additional attributes of the MEMS scanning mirror technology are high tolerance to vibration and operating environment, accurate / fast scanning, electronic control of the scanning mirror position, and small capacity.

[0235] Fig. 22 Included is a schematic functional block diagram illustrating a LiDAR system 200A using stepped frequency modulated (FM) pulse burst transmit envelope modulation and quadrature demodulation and a MEMS scanning mirror, according to some exemplary embodiments. Fig. 22 , the laser radar system 200A according to an exemplary embodiment includes a receiving optical device 154A, at which optical energy is received, including optical echoes from one or more target objects. The optical energy from the receiving optical device 154A is received at a photodetector 156A, which converts the received optical energy into one or more electrical signals, such as the square wave signal shown. These electrical signals are amplified by TIA 158A and filtered by BPF 160A, which has a center frequency at a burst modulation frequency fi to fn. The resulting amplified and filtered signal (at node 161A) is Fig. 22 A basic sinusoidal signal (shown in FIG. 1 ) is applied to the first input of the I / Q mixers 162A, 164A.

[0236] A modulated stepped FM signal is generated by a stepped FM source including a voltage controlled oscillator (VCO) 182A under the control of a control signal from a phase locked loop (PLL) control circuit 183A, which in turn is controlled by a DSPC 168A via a control signal on line 181A. The output signal of the VCO 182A is applied to a power divider 184A, which divides the signal and provides the divided signal at two outputs. A first output 185A is routed to a divider and phase shift circuit or 90 degree power divider 186A, which divides the signal, applies a phase shift to one of the resulting divided signals, and generates a pair of output signals that are phase offset. In an exemplary embodiment, a 90 degree phase shift is applied to one of these signals such that a splitting and phase shifting circuit or 90 degree power divider 186A generates a first "in-phase" local oscillator (LO) signal 189A and a second "quadrature phase" or "quadrature" LO signal 191A that is phase shifted 90 degrees relative to the in-phase LO signal 189A. The in-phase and quadrature phase LO signals 189A, 191A are applied to the second "L" inputs of the I / Q mixers 162A, 164A, respectively. The I / Q mixers 162A, 164A mix the amplified and filtered input signals applied at the first "R" inputs of the I / Q mixers 162A, 164A at node 161A with in-phase and quadrature phase LO signals 189A, 191A, respectively, to generate output signals 193A, 195A, respectively, which are low pass filtered by low pass filters (LPF) 166A and LPF 168A, respectively. The resulting filtered analog signals are converted to digital signals by analog-to-digital converters (ADCs) 170A, 172A, respectively, and sampled under the control of a sample control signal 197A generated by the DSPC 168A. The resulting sampled digital I / Q (quadrature) signals (i.e., I-channel and Q-channel signals) 105A, 107A are processed by the DSPC 168A to determine the distance and / or velocity of the one or more target objects. The results of this detection processing performed by DSPC 168A may be forwarded via system interface 109A to, for example, a user interface, as desired.

[0237] Continue to see Fig. 22, the second output 187A of the power divider 184A is routed to the pulse modulator 174A, which converts the basic sinusoidal signal 187A from the power divider 184A into a pulsed basic sinusoidal signal 111A. The timing of the pulses in the pulsed sinusoidal signal 111A is controlled by the stepped FM pulse burst modulation signal 115A on the output signal line 113A from the DSPC 168A. That is, the basic sinusoidal signal 187A is modulated by the pulse modulator 174A using the stepped FM pulse burst modulation signal 115A to generate the pulsed basic sinusoidal signal 111A. The pulse modulated signal 111A obtained from the pulse modulator 174A is applied as a modulation signal to the laser modulator 176A to generate a control / modulation signal 117A, which is applied to the optical transmitter 178A to generate a stepped FM pulse burst modulated light signal, which is transmitted to the transmitting optical device 180A, and the stepped FM pulse burst modulated light signal is transmitted to the one or more target objects through the transmitting optical device.

[0238] Fig.23 Includes detailed schematic diagrams according to some exemplary embodiments, which show Fig. 22 The MEMS scanning mirror used in the lidar system 200A emits a beam pattern. Fig.23 Schematically illustrates a transmit beam pattern 282A of a MEMS scanning mirror 210A as processed via transmit optics according to an exemplary embodiment. Fig. 22 and Fig.23 In system 200A, a MEMS mirror controller / driver 216A provides a mirror drive signal 212A to a MEMS mirror 210A on line 214A, which causes the MEMS mirror 210A to rotate about an axis 211A, thereby being oriented to provide an azimuth rotation or an altitude rotation of the MEMS mirror 210A. As shown, the signal mirror drive signal 212A may be substantially sinusoidal. The MEMS scanning mirror 210A is tilted to allow high-speed, controlled beam steering in LiDAR range and imaging applications and a variety of other optical systems. Fig.23 The narrow beam width shown in FIG. 1 and the fast azimuth scanning or altitude scanning of the MEMS scanning mirror 210A are suitable for high azimuth resolution scanning requirements. The stepped FM pulse burst envelope modulation waveform of the present disclosure is well suited to provide complementary high range resolution and is compatible with the scanning rate of the MEMS mirror 210A.

[0239] The timing of the pulses in the pulsed sinusoidal signal 111A is controlled by the stepped FM pulse burst modulation signal 115A on the output signal line 113A from the DSPC 268A. That is, the basic sinusoidal signal 187A is modulated by the pulse modulator 174A using the stepped FM pulse burst modulation signal 115A to generate the pulsed basic sinusoidal signal 111A. The pulsed modulation signal 111A obtained from the pulse modulator 174A is applied as a modulation signal to the laser modulator 176A to generate the control / modulation signal 117A, which is applied to the optical transmitter 178A to generate the stepped FM pulse burst modulated optical signal. In system 200A, a stepped FM pulse burst modulated optical signal is transmitted along an optical path 240 to a MEMS mirror 210, where the signal is reflected by the MEMS mirror 210 along an optical path 242A to a transmitting optical device 280A, through which the stepped FM pulse burst modulated optical signal is transmitted to the one or more target objects in a transmitting beam pattern 282A of the MEMS scanning mirror 210A.

[0240] Fig.24 A schematic timing diagram is included that illustrates the data acquisition process of the lidar system 200A using the MEMS scanning mirror 210A with a single pulse burst frequency at each scanning mirror position according to an exemplary embodiment. Fig.24 The top schematic diagram in shows the scan angle (ie, angular position) of the MEMS mirror 210A over time during data acquisition, and the bottom schematic diagram shows the stepped FM pulse burst frequency over time during data acquisition. Figure 22 to Figure 24 According to an exemplary embodiment, a single fixed frequency pulse burst cluster is used at each scan increment for the duration of each scan. After successive scans, the frequency of the fixed frequency pulse burst cluster is increased by a step frequency Δf. Thus, n scans are used to complete the data acquisition process, where n is the number of frequency steps. Fig.23 The synchronization of the pulse burst frequency and the angular position of the MEMS scanning mirror 210A as the scanning mirror 210A beam is scanned is shown. According to an exemplary embodiment, data is collected for each beam position (i.e., scanning increment) at each discrete frequency step. It should be noted that, in general, the MEMS mirror position is not linear with time. However, in an exemplary embodiment, through the process of synchronization of the MEMS mirror position with the pulse burst waveform, this effect results in a linear or synchronized pulse burst relationship with the mirror position.

[0241] Continue to see Figure 22 to Figure 24 , the MEMS mirror controller / driver 216A provides synchronization of the pulses of the step-FM pulse burst modulated signal 115A, and hence the optical illumination pulses of the step-FM pulse burst modulated light signal, with the angular position of the MEMS scanning mirror 210A. Fig.25Included is a schematic detailed functional block diagram of a MEMS mirror controller / driver 216A according to some exemplary embodiments. Fig.26 Included are schematic timing diagrams illustrating synchronization of stepped FM modulation pulses with scanning mirror positions provided by MEMS mirror controller / driver 216A, according to some exemplary embodiments.

[0242] See also Figure 22 to Figure 26 , digital signals MirrorCycle X2 and MirrorCycle are provided in the MEMS mirror controller / driver 216A. MirrorCycle X2 goes high at the beginning of a scan when the MEMS mirror 210A is at its maximum angle, which is shown as ±12 degrees for exemplary purposes only. It switches low again when the MEMS mirror 210A is in its neutral or zero degree position. MirrorCycle can be used to determine whether the mirror 210A is moving left or right (in the case of an azimuth scan) or up or down (in the case of an altitude scan). The rising and falling edges of the MirrorCycle signal coincide with the zero crossings of the MEMS mirror 210A.

[0243] As described above, according to some exemplary embodiments, in order to accurately collect data at a repeatable scanning mirror position, synchronization between the scanning mirror position and the pulse burst transmission time is achieved. One synchronization technique is to divide the time between the start and stop position scanning signals into many smaller isochronous increments, which are close to the angular position of the scanning mirror 210A. This division can be completed using a phase-locked loop (PLL) 252A configured as a frequency multiplier. The output of the PLL frequency multiplier 252A is applied to a counter 254A, which acts as a frequency divider. That is, the counter 254A output value represents the time of the scan, and the angular position of the mirror 210A can be calculated using a cosine function or determined by a lookup table (LUT) according to the time of the scan, as shown in 256A. A D flip-flop 258A is used to determine the mirror direction, and thus generate a synchronous transmission pulse burst. Therefore, PLL 252A generates a clock from the mirror cycle X2 signal. For each scan (which can be forward or reverse), a single pulse is generated in the mirror cycle X2 signal. PLL 252A is configured to divide the single pulse into, for example, 1024 shorter pulses evenly spaced in time. These pulses are routed to counter 254A, the current value of which corresponds to the time of the scan. The angular position of the scanning mirror 210A can be calculated using a cosine function or determined by a lookup table (LUT), as shown in 256A. When combined with a single D flip-flop 258A to monitor the direction of the mirror motion, a synchronized pulse train 215A is generated by DSPC 268A. To this end, the output of the D flip-flop is applied to DSPC 268A on line 218A, and the output of the LUT / cosine function 256A indicating the mirror position is also applied to DSPC 268A on line 218A. The mirror drive signal 212A, which is also output from the LUT / cosine function 256A, is applied to the MEMS scanning mirror 210A on line 214A to control its rotation.

[0244] With respect to samples collected during the negative slope of the scan cycle, there is a potential anomaly. For example, in order to reconstruct samples from the relevant scan increments, the FFT sampling rate can be adjusted. For example, in some specific exemplary embodiments, the FFT sampling rate can be adjusted to twice the scan time.

[0245] FIG. 27A to FIG. 27D Included is a series of four exemplary data matrices for scanning mirror data acquisition according to some exemplary embodiments. Fig.28 Tables containing parameterized data for typical automotive operating scenarios for a LiDAR system 100A using a MEMS scanning mirror for data acquisition according to certain exemplary embodiments. FIG. 27A to FIG. 27D and Fig.28, the data acquisition process of the MEMS scanning mirror includes filling or populating n data matrices of dimension k×N; where n is the number of scans (ie, the number of frequency steps), k is the number of scan increments, and N is the number of range bin samples.

[0246] According to the present disclosure, laser transmitter stepped FM pulse burst envelope modulation and receiver quadrature demodulation techniques consistent with direct detection laser radar systems have been described in detail. Data acquisition techniques and signal processing gains have also been described in detail. According to the present disclosure, it has been demonstrated that the techniques of transmit envelope modulation combined with receive quadrature demodulation as applied to direct detection laser radar systems provide signal processing gain, as determined by an increase in the signal-to-noise ratio at the detection stage of the system. Important operating factors include the transmission phase shift variation of the envelope modulation waveform over a bidirectional distance from an object, and the coherent detection of the envelope modulation waveform within the quadrature demodulator. In addition, the envelope modulation waveform is derived from the quadrature demodulation local oscillator to establish the coherent signal required for detection.

[0247] It has been demonstrated that the disclosed stepped FM pulse burst envelope modulation waveform is compatible with MEMS fast scanning mirrors.

[0248] The signal processing gains achieved in direct detection lidar systems far outweigh the modest increase in hardware complexity. The availability of integrated circuit phase-locked loop and quadrature demodulation functions ensures ease of implementation with minimal impact on system capacity, operating power, and cost. It should be noted that the lidar architecture detailed herein facilitates systems with lower transmit power, longer measurement range, reduced power consumption, and potentially better performance in multi-system deployment conditions. In addition, according to an exemplary embodiment, the range measurement error or variance is reduced due to the increase in signal-to-noise ratio.

[0249] In some other exemplary embodiments, repetition of the data acquisition process to populate additional data matrices may provide simultaneous high-resolution range and Doppler measurements.

[0250] According to an exemplary embodiment, synthetic Doppler techniques are used to increase processing gain in radar and lidar measurement systems using linear FM ramps (i.e., FMCW) and pulse transmission modulation waveforms as detailed above. The frequency ramp deviation ΔF and duration ΔT are determined for the linear FMCW modulation waveform. The pulse width and pulse repetition frequency are determined for the pulse modulation waveform. According to the present disclosure, synthetic Doppler is digitally applied to each waveform for the purpose of increasing the effective processing gain. Two-dimensional FFT and conventional pulse Doppler processing techniques are utilized using digital incremental increases in phase, which are indexed with the frequency ramp or pulse repetition frequency (for single beam systems) or antenna scanning (for multi-beam systems). The synthetic Doppler techniques of the present disclosure are particularly effective in detecting stationary objects, which are generally not suitable for two-dimensional FFT processing.

[0251] Lidar systems (e.g., automotive Lidar systems) have challenging and conflicting operational objectives. Accurate, high-resolution range measurements require wide transmission bandwidths, which conflict with the cost of high-sampling-rate analog-to-digital converters (ADCs) with wide dynamic range. High-resolution azimuth and elevation beam scanning requires narrow beamwidths and beam position dwell times, which impact data acquisition and system response time. Range and position measurement accuracy requires high signal-to-noise ratios (SNRs), which are limited by transmit power and available signal processing gain techniques. The transmit envelope modulation techniques described previously coupled with quadrature demodulation have proven to improve direct detection Lidar architectures to meet automotive deployment challenges. However, due to Doppler resolution and detection limitations, the SNR enhancement provided by two-dimensional (2-D) processing at the signal detection stage cannot be achieved, even at high relative speeds. To facilitate 2-D signal processing and thereby improve detection SNR, the techniques described herein employ digital incremental phase addition to the demodulated signal. In addition, because the technique implements a known "synthesized" Doppler frequency, spectrally resolved analysis is required to determine individual elements of the signal spectrum and thereby reduce computation time.

[0252] As used herein, the term "synthetic Doppler" refers to the accumulation of incremental phases of a signal in a manner similar to the Doppler shift frequency produced by changes in signal transmission phase caused by changes in range due to relative velocity. This synthetic Doppler is not used to measure relative velocity; however, the techniques described herein provide signal processing options similar to those provided by true Doppler frequencies, such as 2-D FFT (for FMCW waveforms) and pulsed Doppler processing (for pulsed waveforms), and are applicable to operating scenarios where true Doppler is limited due to stationary or small relative motion or operating frequencies that limit the magnitude of the Doppler frequency.

[0253] According to exemplary embodiments, the synthetic Doppler of the present disclosure can be implemented using a variety of techniques. For example, in a first embodiment, a phase modulator can be inserted directly in the local oscillator (LO) path leading to the orthogonal demodulator. In a second embodiment, a direct digital synthesizer (DDS) for both the transmit signal and the receive signal can be phase modulated during the receive interval (i.e., after the transmit pulse). Alternatively, a second DDS for the LO can be added to achieve phase modulation after a continuous frequency ramp while operating from the same clock. In a third embodiment, the incremental indexed phase shift can be digitally accumulated to the I-channel and Q-channel signal components after the ADC is acquired or within the complex FFT (CFFT) signal component of the FMCW waveform.

[0254] The synthetic Doppler techniques of the present disclosure detailed herein include FMCW and / or pulsed LiDAR data acquisition and two-dimensional (2-D) Fast Fourier Transform (FFT) processing. These techniques facilitate 2-D processing of operating scenarios where Doppler frequency resolution is limited, or enhance detection of stationary objects via signal processing gain (as measured by an increase in signal-to-noise ratio (SNR)). It should be noted that the techniques of the present disclosure achieve increased signal processing gain without an increase in ADC sampling rate. An estimate of the signal processing gain for one-dimensional processing is 10·log(N). An estimate of the signal processing gain for two-dimensional processing is 10·log(n×N).

[0255] Fig.29 Includes a schematic functional block diagram according to some exemplary embodiments, which shows the use of Figure 3 The FMCW transmission envelope modulation and orthogonal demodulation laser radar system shown in FIG. Fig.11 The system shown has been varied. Fig.29 The block diagram and Fig.11 The difference between the block diagram is Fig.29 The direct digital synthesizer (DDS) 238A described above for generating the envelope modulated signal is shown. Fig.29 The remaining components and Fig.11 are identical to those of the present invention and are identified by the same reference numerals. It should also be noted that although not in Fig.29 shown in Fig.29 A LiDAR system may include Fig. 22 Shown in and combined with this article Fig. 22 Scanning capability described in detail. As detailed above, the FMCW sensor operation provides a mapping of IF frequency to distance according to the above equation (6). According to the method, a linear frequency ramp is applied to modulate the envelope signal, and the modulated signal is transmitted. The echo signal from the target object at a distance R is mixed with the transmitted signal. The mixing process generates an IF frequency proportional to the object distance R. N samples of the IF signal are collected from the I-channel and Q-channel by ADC 270, 272. The distance is generated by spectral analysis of the IF signal via a complex Fourier transform.

[0256] FIG. 30A to FIG. 30C Included are data matrices showing FMCW data acquisition for 2-D FFT processing of a single scan angle position according to some exemplary embodiments. Specifically, Fig. 30A Includes raw FMCW data matrix; Fig. 30B including the repopulated data matrix; and Fig. 30C Includes the second repopulated data matrix. See Fig.14 , Fig.29 and FIG. 30A to FIG. 30C, generate n frequency ramps, and collect N complex (I-channel and Q-channel) data points at each frequency ramp in the I-channel and Q-channel within the time ΔT. After n ramps, construct Fig. 30A The single scan angle (n×N) data matrix shown in FIG. Fig. 30A In the raw data matrix shown, the acquired I-channel and Q-channel data samples are used to fill the complex valued n×N matrix. Each matrix position has a structured complex form: (I-Ch i,k +j·Q-Ch i,k ), where i∈1,...,n, and k∈1,...,N. Fig. 30A A complex FFT (range FFT) of length N is performed on each row of the original data matrix. The complex FFT elements from each row are then used to refill Fig. 30A The original matrix is ​​created Fig. 30B The repopulated data matrix. Fig. 30B Each row of the refilled data matrix represents the resolved IF spectrum of the corresponding frequency ramp. Fig. 30B Each row element of the refilled matrix represents the adjacent distance bin according to the above frequency-to-distance mapping equation (ie, equation (6)). Fig. 30B Each column of the refilled matrix represents the range bin data from the continuous frequency ramp. It should be noted that for a stationary target object within that range bin, Fig. 30B The column data in the refilled matrix is ​​constant.

[0257] Use the incremented (ie, indexed) phase term value or values ​​added to the number of each element The phase shift is used to enhance Fig. 30B Each column of the refilled data matrix is ​​filled with , where i∈1,…,n. This action implements the synthetic Doppler frequency for the column range bin data. Refilling with numerically increasing phase elements Fig. 30B The repopulated data matrix is ​​generated Fig. 30C The second repopulated data matrix. Fig. 30C A complex FFT of length n is performed on the columns of the second refilled data matrix and the resulting spectrum is resolved with unique synthetically induced Doppler frequencies. Fig. 30B The phase angle associated with each element of the first refilled data matrix is ​​defined by equation (27):

[0258]

[0259] It should be noted that Fig. 30CEach column of the second refilled data matrix represents a discrete range bin according to the CFFT index. A spectral analysis is performed on each column of the data matrix. Since the artificial "Doppler" frequency is known, the spectral search can be limited to the known artificial "Doppler" frequency.

[0260] Figure 31 to Figure 35 Included are schematic diagrams showing the FMCW synthetic Doppler technique of the present disclosure according to an exemplary embodiment and simulation results of the technique. Specifically, Fig.31 A schematic logic flow diagram of synthetic Doppler simulation using an FMCW envelope modulated waveform according to an exemplary embodiment is included. Fig.32 Included is a table of FMCW synthetic Doppler parameters for simulations with a single beam angle. Fig.33 Includes the results of the synthetic Doppler range FFT calculation for a single scan angle. Fig.34 Included are the results of the synthetic Doppler FFT for a single scan angle. Fig.35 Supplementary data including simulation results of the Doppler FFT after the CFFT is performed. Fig.36 Additional supplementary data of the simulation results are included. Specifically, Fig.36 shows the distance CFFT Program scripts and Doppler CFFT Program scripts. Fig.36 As shown, the range CFFT is performed on a single row of data from the FMCW lidar using the following parameters: R = 75 meters, ΔF = 100 MHz, ΔT = 10 μs, amplitude A0 = 1.0, noise = 1.0. Similarly, the Doppler CFFT fills a single column of data at range bin 320 from the FMCW lidar using the following parameters: R = 75 meters, ΔF = 100 MHz, ΔT = 10 μs, amplitude A0 = 1.0, noise = 1.0. For the Doppler CFFT, each data point is incremented as the distance from the FMCW lidar increases. The differential phase shift increases. Fig.35 The scripts shown are commented to clarify the definitions of parameters and variables.

[0261] See Fig.31 , in step S301 of the 2-D FFT of the FMCW waveform, the phase increment is defined Next, in step S303, the length n of the Doppler FFT is defined. Next, in step S305, the length N of the range FFT is defined. Next, in step S307, the IF equation for the range FFT is defined. Next, in step S309, an N-point CFFT is performed and the result FFTR is stored in a matrix. Next, in step S311, the result of the CFFT (bin 320) is stored in the column matrix M. r,c(Bin 320 is set by R=75 meters). Next, in step S313, a single frequency search is performed for the known synthetic Doppler signal level. It should be noted that efficient digital processing techniques (such as the Goertzel algorithm) can be used to speed up the single frequency spectrum search.

[0262] According to an exemplary embodiment, a processing gain is achieved according to 10×Log(n×N). The data acquisition time is defined as T acq = n × ΔT. An N-point CFFT is performed for each frequency ramp. Since the "Doppler" frequency is known, a single-frequency spectrum analysis is performed on the synthetic "Doppler". An additional processing gain of 10 × Log (n) is achieved at the expense of the computation of the n-point CFFT and spectrum search execution time.

[0263] In terms of antenna beam scanning, Fig.37 Included are detailed schematic diagrams according to some exemplary embodiments, which illustrate the transmit beam pattern employed in a laser radar system as disclosed herein. In some exemplary embodiments, the laser radar system is generated by using a MEMS scanning mirror as described in detail herein. Fig.37 The transmit beam pattern is shown. Fig.37 and Fig.23 Similar, but with some differences as detailed below. Fig.37 , which shows k beam positions covering a total scan range of -25° to +25°. The time required for a single scan is T 扫描 , which may be, for example, about 500 μs. A single scan is used to acquire data for one-dimensional signal processing with a single frequency ramp ΔF at each of the k beam positions. Multiple scans are used to acquire data for 2-D signal processing. Fig.38 A table listing the parameters associated with data for k=50 beam angle positions is included. As shown, the FMCW data acquisition time T acq is calculated as:

[0264] T acq =n×T 扫描 =0.064 seconds.

[0265] The above detailed description relates to the synthetic Doppler technique of the exemplary embodiment using an FMCW envelope modulation waveform. As mentioned above, the synthetic Doppler technique of the present disclosure may also be performed using a pulsed envelope modulation waveform. As described in detail below, the implementation of the synthetic Doppler technique using a pulsed envelope modulation waveform is more straightforward and requires only a single complex FFT after data acquisition.

[0266] Fig.39 Includes a schematic functional block diagram according to some exemplary embodiments, which shows the use of Figure 2 The laser radar system with pulsed transmission envelope modulation and orthogonal demodulation shown in FIG. Fig. 6A The system shown has been varied. Fig.39 The block diagram and Fig. 6A The difference between the block diagram is Fig.39 The above-mentioned direct digital synthesizer (DDS) 238A is shown. Fig.39 The remaining components and Fig. 6A are identical to those of the present invention and are identified by the same reference numerals. It should also be noted that although not in Fig.39 shown in Fig.39 A LiDAR system may include Fig. 22 Shown in and combined with this article Fig. 22 As detailed above, pulsed sensor operation involves a time delay (based on t d =2R / c) to provide distance measurement. In an exemplary embodiment, Fig.40A is a timing diagram illustrating data acquisition timing in pulsed lidar processing according to some exemplary embodiments. Fig.40A , a pulse is transmitted at the beginning of each frame from which data is to be acquired. N data samples from consecutive range bins are acquired from each receive channel (I-Ch and Q-Ch) during the receive interval after the transmit pulse (indicated by arrows). The data samples from n frames form an n×N matrix, from which additional processing is performed.

[0267] Fig.40B Included are schematic diagrams showing pulsed data acquisition for a single scan angle position according to some exemplary embodiments. Fig.40B Represents the I-channel and Q-channel receiver outputs. Fig.40B The arrows in the Figure 8 Sample data points in similar distance bins. Fig.40C Included is a raw data matrix generated during pulse data acquisition according to some exemplary embodiments. Fig.40D Includes a repopulated data matrix generated during pulse data acquisition according to some exemplary embodiments. Fig.39 and FIG. 40A to FIG. 40D , in the time interval T prf Complex data points from I-channel and Q-channel data samples are collected at each range bin. After n frames at a single scan angle, these data samples are used to fill the complex-valued n×N raw data matrix, as Fig.40C Each matrix position has a structured complex form: (I-Ch i,k +j·Q-Ch i,k), where i∈1,…,n, and k∈1,…,N. It should be noted that the column data represents the complex signal from each consecutive range bin. This is different from the FMCW example described above, where a range FFT is performed to structure the columns into range bins.

[0268] Use an incremental (ie, indexed) phase term or value added to each element The phase shift is used to enhance Fig.40C Each column of the original data matrix of , where i∈1,…,n. This action implements the synthetic Doppler frequency for the column range bin data. Refilling with numerically increasing phase elements Fig.40C The original data matrix is ​​generated Fig.40D The repopulated data matrix of Fig.40D A complex FFT of length n is performed on the column data of the refilled matrix, where each column represents a discrete range bin, and the resulting spectrum is resolved with a unique synthetic Doppler frequency. A spectral analysis based on the known synthetic Doppler frequency is performed on each range bin to assess the presence or absence of an object. It should be noted that the two-dimensional signal processing for pulsed processing does not require a second FFT. Fig.40C The raw data matrix of is collected and filled in range bins, so no frequency-to-range transform or FFT process is required. The pulse envelope modulated waveform requires only a single CFFT at each range bin (i.e., column). It should be noted that FIG. 40A to FIG. 40D In the exemplary illustration of , an object is identified in a range bin 12 .

[0269] Figure 41 to Figure 46 Included are schematic diagrams showing the synthetic Doppler technique of the present disclosure according to an exemplary embodiment and simulation results of the technique when a pulsed envelope modulation waveform is used. Specifically, Fig.41 A schematic logic flow diagram of synthetic Doppler simulation using a pulsed envelope modulated waveform according to an exemplary embodiment is included. Fig.41 In step S331, the noise parameters are defined. Next, in step S333, the phase increment is defined. Next, in step S335, the length n of the Doppler FFT is defined. Next, in step S337, the signal equation SSIG is defined. Next, in step S339, an n-point CFFT is performed and the result FFTR is stored in a matrix. Next, in step S341, a single frequency search is performed for the known synthetic Doppler signal level. It should be noted that effective digital processing techniques (such as the Goertzel algorithm) can be used to speed up the single frequency spectrum search.

[0270] Fig.42 Contains a table of pulsed synthetic Doppler parameters for simulations of individual beam angles. Fig.43 A method for generating noise according to an exemplary embodiment is included Program scripts, signals for pulsed synthetic Doppler simulation and CFFT Program script and simulation results for the CFFT of n data samples from a single range bin. Fig.43 Also included are simulation results for SNRIN, SNRouT, and processing gain PGDB. According to an exemplary embodiment, processing gain is achieved according to 10×Log(n). The data acquisition time is defined as T acq =n×T prf A single frequency spectrum analysis is performed on the synthetic Doppler and the Doppler frequency is known.

[0271] Fig.44 A schematic diagram is included showing data acquisition for the I-channel and the Q-channel for a single scan according to an exemplary embodiment. Fig.44 Represents the I-channel and Q-channel receiver outputs. Fig.44 The arrows in the Figure 8 Similar range bin sample data points as in Figure 40. Note that the object appears at beam position 3 in range bin 12. Fig.44 , in the time interval T prf N data points are collected at each beam scan position (k). A single scan is required to fill the k×N data matrix at each beam scan position. At each beam scan position, the range bin signal level can be evaluated for object detection. If a single scan is used, processing gain is not available. The beam scan position dwell time must be increased to accommodate multiple PRFs to provide processing gain.

[0272] Fig.45 A schematic diagram is included that shows data acquisition for the I-channel and Q-channel at beam position 3 for multiple scans according to an exemplary embodiment. It should be noted that the object appears in range bin 12. See Fig.45 , repeat the single scan data acquisition for n scans. A k×N data matrix is ​​formed for each scan beam position. Therefore, the complete data matrix is ​​3-dimensional and indexed by beam scan position k. Signal processing requires incremental phase accumulation Then, an n-point spectral analysis is performed for each range bin (N) at each beam position (k), where i∈1,…,n.

[0273] Fig.46 A table is included that lists parameters associated with data at multiple beam angle positions. As shown in the figure, the pulse data acquisition time T acq is calculated as:

[0274] T acq =n×T 扫描 =0.0512 seconds.

[0275] As described above, according to exemplary embodiments, the synthetic Doppler of the present disclosure may be implemented using a variety of techniques. For example, in another embodiment, a phase modulator may be inserted directly in the local oscillator (LO) path leading to the quadrature demodulator. Fig.47 Includes a schematic functional block diagram according to some exemplary embodiments, which shows the use of Figure 3 The FMCW transmission envelope modulation and orthogonal demodulation laser radar system shown in FIG. Fig.29 The system shown has been varied. Fig.47 The block diagram and Fig.29 The difference between the block diagram is Fig.47 A phase modulator 293A is shown inserted between the output 285 of the power divider 284 and the input of the 90 degree power divider 286. The phase shift introduced to produce the synthetic "Doppler" of the exemplary embodiment is controlled by the DSPC 268 via control line / interface 295A. Fig.47 The remaining components and Fig.29 are identical to those of the present invention and are identified by the same reference numerals. It should also be noted that although not in Fig.47 shown in Fig.47 A LiDAR system may include Fig. 22 Shown in and combined with this article Fig. 22 A detailed description of the scanning capabilities.

[0276] The application of the synthetic Doppler technique of the present disclosure detailed herein provides advantageous operating characteristics in environments including relative motion. The technique enables the distinction between stationary and moving target objects and target objects with decreasing and increasing distances. In accordance with these principles, according to the present disclosure, the synthetic Doppler delta phase is accumulated to all range bin columns of the first FFT. Since the synthetic Doppler frequency is known, all range bins of stationary objects will be located at the known synthetic Doppler frequency. In addition, all approaching objects will be located in frequency bins above the synthetic Doppler frequency, while objects with increasing distances will be located below the synthetic Doppler frequency. The same signal processing gain can be used for stationary and moving objects.

[0277] Fig.48 A schematic perspective view of a vehicle 500 is included that is equipped with one or more lidar systems 100 as described in detail herein, according to an exemplary embodiment. Fig.48, it should be noted that while only a single lidar system 100 is shown, it should be understood that multiple lidar systems 100 according to exemplary embodiments may be used in the automobile 500. In addition, to simplify the description, the lidar system 100 is shown as being mounted on or on the front portion of the automobile 500. It should also be understood that one or more lidar systems 100 may be mounted at various locations on the automobile 500. In addition, while only system 100 is shown, it should be understood that any embodiment of the lidar system described herein may be used.

[0278] Fig.49 Schematic top view of a car 500 equipped with two lidar systems 100 as described in detail above according to an exemplary embodiment. Fig.49 In the particular embodiment shown, the first lidar system 100 is connected to a first CAN bus electronic control unit (ECU) 558A via bus 560 (which in some embodiments may be a standard automotive controller area network (CAN) bus). Detections generated by the lidar processing detailed herein in the lidar system 100 may be reported to ECU 558A, which processes these detections and may provide detection alerts via the CAN bus 560. Similarly, in some exemplary embodiments, the second lidar system 100 is connected to a second CAN bus electronic control unit (ECU) 558B via CAN bus 560. Detections generated by the lidar processing detailed herein in the lidar system 100 may be reported to ECU 558B, which processes these detections and may provide detection alerts via the CAN bus 560. It should be noted that this configuration is merely exemplary, and many other automotive lidar configurations within the automobile 500 may be implemented. For example, a single ECU may be used in place of multiple ECUs. In addition, these individual ECUs may be omitted entirely. In addition, although only system 100 is shown, it should be understood that any embodiment of the lidar system described herein may be used.

[0279] Experimental data indicate further opportunities to exploit signal processing gain. Additional modulation and coding waveforms are considered based on system performance goals of increased processing gain, measurement accuracy, and spatial resolution.

[0280] It should be noted that the present disclosure describes a lidar system installed in a car. It should be understood that the system of the present disclosure is applicable to any type of vehicle, such as a bus, a train, etc., or the lidar system of the present disclosure does not have to be associated with any type of vehicle.

[0281] It should also be noted that the above detailed description mainly relates to a laser radar detection system. It should be understood that the present disclosure is also applicable to other forms of detection systems, such as radar systems.

[0282] Although many changes and modifications of the present disclosure will be apparent to those of ordinary skill in the art after reading the above description, it should be understood that the specific embodiments shown and described in an illustrative manner are not intended to be considered limiting. In addition, the subject matter has been described with reference to specific embodiments, but variations within the spirit and scope of the present disclosure will occur to those skilled in the art. It should be noted that the above examples are provided only for explanation and should in no way be construed as limiting the present disclosure.

[0283] While the inventive concept has been particularly shown and described with reference to exemplary embodiments thereof, it will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the inventive concept as defined by the appended claims.

Claims

1. A detection system, comprising: a signal transmitter for transmitting a frequency modulated continuous wave (FMCW) transmission signal into an area, the transmission signal comprising a series of signal pulses, each of the signal pulses comprising a signal having a varying frequency, the frequency within each of the signal pulses varying according to a ramp between a first frequency and a second frequency over time; a receiver configured to receive a reflection signal generated by reflection of the transmission signal and generate a reception signal indicative of the reflection signal; and a processor coupled to the receiver for receiving the received signal and processing the received signal to generate a detection of one or more objects in the area, the processing comprising: generating a first two-dimensional data matrix comprising n rows of N complex data values, wherein n is the number of signal pulses of the transmit signal, each row of the two-dimensional data matrix corresponding to one of the signal pulses of the transmit signal, and wherein N is the number of complex data values ​​generated for each signal pulse, each column of the two-dimensional data matrix corresponding to n complex data values ​​for all signal pulses of the transmit signal, performing a complex Fast Fourier Transform (FFT) of length N on each row of the two-dimensional data matrix, generating a second refilled two-dimensional data matrix comprising n rows of N complex FFT elements, each row of the two-dimensional data matrix corresponding to one of the signal pulses of the transmitted signal, each row comprising complex FFT elements obtained by performing the complex FFT on each row of the first two-dimensional data matrix, each column of the second refilled two-dimensional data matrix representing range bin data of a signal pulse from the transmitted signal, and generating a third refilled two-dimensional data matrix comprising n rows of N complex data values, each row of the two-dimensional data matrix corresponding to one of the signal pulses of the transmit signal, each column of the second refilled two-dimensional data matrix representing range bin data of the signal pulse from the transmit signal, each column of the third refilled two-dimensional data matrix being augmented with an adjusted phase term so that a synthetic Doppler frequency is performed in each column range bin, the adjusted phase term being used in generating the detection, performing a complex FFT of length n on the column data of the second refilled two-dimensional data matrix and resolving the resulting spectrum at a synthetic Doppler frequency; The synthetic Doppler refers to the accumulation of incremental phase of the signal in a certain manner, which is similar to the Doppler shift frequency generated by the change in signal transmission phase caused by the distance change caused by the relative speed.

2. The detection system of claim 1, wherein the adjusted phase term is generated using a phase modulator that changes a phase shift in a local oscillator (LO) signal.

3. The detection system of claim 1, wherein the adjusted phase term is generated by changing the output of a direct digital synthesizer (DDS) used in generating the transmit signal and the receive signal.

4. The detection system of claim 1, wherein the adjusted phase term is generated by accumulating phase shift values ​​to in-phase and quadrature-phase detector signals generated from the received signal.

5. The detection system of claim 4, wherein the phase shift values ​​are digitally indexed. The detection system of claim 1 , wherein the ramp is a linear ramp. The detection system of claim 1 , wherein the first frequency is lower than the second frequency. The detection system of claim 1 , wherein the first frequency is higher than the second frequency.

9. The detection system of claim 1, wherein the signal transmitter angularly scans the transmission signal into the area at varying angles.

10. The detection system of claim 1, wherein the signal transmitter comprises a scanning mirror for scanning the transmission signal into the area at a varying angle.

11. The detection system of claim 10, wherein the scanning mirror is a micro-electromechanical (MEMS) scanning mirror.

12. The detection system of claim 1, wherein the detection system is an automotive detection system.

13. The detection system according to claim 1, wherein: The detection system is a radar system; The signal transmitter is a radar signal transmitter; The transmitted signal is a transmitted radar signal; and The reflected signal is a reflected radar signal.

14. The detection system according to claim 1, wherein: The detection system is a laser radar system; The signal transmitter is a laser radar signal transmitter; The transmission signal is a laser radar signal; and The reflected signal is a reflected lidar signal.

Images