Scanning type millimeter wave and sea wave measuring device and measuring method

Abstract

The embodiment of the invention provides a scanning type millimeter wave and sea wave measuring device and method. The system is applied to the technical field of marine environment monitoring and comprises a remote display control unit for data interaction with a remote communication unit; the remote communication unit is used for issuing the instruction information transmitted by the remote display control unit; the instruction information comprises a working parameter and an irradiation angle; the control unit is used for receiving the instruction information sent by the remote communication unit and controlling the radar unit to operate; the radar unit is used for transmitting corresponding radar transmitting signals according to the working parameters, rotating the two-axis turntable according to the irradiation angle so as to drive the transmitting antenna and the receiving antenna to face the sea surface, and calculating sea wave information according to the received sea surface echo signals; the clock synchronization unit is used for synchronizing timestamps into the control unit and the radar unit; the method has the advantages of reliable monitoring result, unattended operation and the like, and is of great significance to construction of a marine monitoring system.

Description

Technical Field

[0001] The present invention relates to the field of marine environment monitoring, and particularly to a scanning millimeter-wave sea wave measuring device and a measuring method. Background Art

[0002] According to the current technical regulations for wave observation and analysis in water transportation engineering, the main elements of sea wave observation are variables that characterize the sea wave conditions in a certain water area and a specific time period, such as wave height, wave direction, wave period, and wavelength. The acquisition of marine information relies on marine monitoring, and there are many methods for sea wave monitoring, such as manual monitoring, satellite observation, buoy wave measurement, radar detection, etc.

[0003] However, the use of manual monitoring has a high cost and is not easy to achieve real-time and continuous monitoring. In the case of strong winds and waves, it may even affect the safety of personnel. Satellite monitoring is relatively convenient, but the monitoring accuracy is not high. The buoy system has a high accuracy, but it needs to be deployed in a certain sea area for a long time, the data cannot be transmitted back in real time, and the implementation cost is high. Using radar wave measurement has the advantages of high accuracy, simple installation, all-weather, maintenance-free, and no need for on-site calibration, and is suitable for use in all weather conditions and is widely used in offshore wind power.

[0004] Existing radar wave measurement devices are vulnerable to the influence of salt spray corrosion, temperature and humidity changes, mechanical vibration interference, etc. in a complex marine environment, resulting in a decline in equipment stability, and lack an automated maintenance mechanism. It is difficult to detect and repair faults of offshore equipment in a timely manner, and relying on manual intervention leads to high operation and maintenance costs. In addition, the traditional array scanning method relies on multiple radars to work synchronously, its spatial resolution is limited by the physical element spacing, and multi-channel calibration is easily affected by the environment and generates different performance drifts, resulting in large errors in the final measurement results.

[0005] Therefore, there is an urgent need to develop a scanning millimeter-wave sea wave measuring device and a measuring method to solve the above problems. Summary of the Invention

[0006] The present invention provides a scanning millimeter-wave sea wave measuring device and a measuring method, which solve the technical problems that the existing device relies on multiple radars to work synchronously, the spatial resolution is limited by the physical element spacing, and the measurement result error is large.

[0007] According to the first aspect of the present invention, a scanning millimeter-wave sea wave measuring device is provided. The measuring device includes: a remote display and control unit for data interaction with a remote communication unit; a remote communication unit for sending down the instruction information transmitted by the remote display and control unit; the instruction information includes working parameters and irradiation angles; a control unit for receiving the instruction information sent by the remote communication unit and controlling the operation of the radar unit; A radar unit, configured to transmit corresponding radar transmission signals according to operating parameters, and at the same time rotate a two-axis turntable according to the illumination angle to drive a transmitting antenna and a receiving antenna to face the sea surface, and calculate sea wave information based on the received sea surface echo signals; A clock synchronization unit, configured to synchronize timestamps to the control unit and the radar unit.

[0008] In the aspect and any possible implementation manner as described above, a further implementation manner is provided. The radar unit includes a transmitting part, a transmitting antenna, a two-axis turntable, a receiving antenna, and a receiving part; an output end of the transmitting part is connected to the transmitting antenna, an output end of the receiving antenna is connected to an input end of the receiving part, and both the transmitting antenna and the receiving antenna are arranged on the two-axis turntable.

[0009] In the aspect and any possible implementation manner as described above, a further implementation manner is provided. The transmitting part is sequentially provided with a frequency control unit, a DDS, a DA, a first up-conversion unit, a first band-pass filter, a second up-conversion unit, a second band-pass filter, and a power amplifier according to the signal transmission path.

[0010] In the aspect and any possible implementation manner as described above, a further implementation manner is provided. An output end of the DDS is further connected to an input end of a frequency synthesizer module, and an output end of the frequency synthesizer module is connected to input ends of the first up-conversion unit and the second up-conversion unit.

[0011] In the aspect and any possible implementation manner as described above, a further implementation manner is provided. The receiving part is sequentially provided with an LNA, a third band-pass filter, a de-slope mixing unit, a low-pass filter, an AD, a radar signal processing unit, and a radar data processing module according to the signal transmission path.

[0012] In the aspect and any possible implementation manner as described above, a further implementation manner is provided. An output end of the second band-pass filter is connected to an input end of the de-slope mixing unit, and an output end of the two-axis turntable is connected to an input end of the radar signal processing unit.

[0013] According to a second aspect of the present invention, a scanning millimeter-wave sea wave measurement method is provided. The measurement method includes the following steps: S1. A remote display and control unit transmits operating parameters to a control unit through a remote communication unit. The control unit receives the timestamp of the clock synchronization unit to mark the operating parameters, and issues the marked operating parameters to the radar unit; S2. In the radar unit, a corresponding radar transmission signal is transmitted according to the marked operating parameters, and the two-axis turntable is controlled to rotate according to the marked irradiation angle to irradiate the target sea area. The wave information of the target sea area is calculated based on the received sea surface echo signal. The radar unit receives the time stamp from the clock synchronization unit to mark the wave information; S3. The marked wave information is sent to the remote communication unit, and the remote communication unit exchanges information with the remote display and control unit; S4. The remote display and control unit visually displays the marked wave information and the information output by the status monitoring and maintenance unit.

[0014] In the aspect and any possible implementation manner as described above, a further implementation manner is provided. In the radar unit, S2 specifically includes the following steps: S21. The frequency control unit of the radar unit controls the DDS to generate a corresponding baseband radar transmission signal according to the marked operating parameters, and then generates an intermediate frequency analog baseband signal through the DA and sends it to the first up-conversion. The radar unit sends the irradiation angle to the two-axis turntable to control the two-axis turntable to point to the target sea area corresponding to the irradiation angle; S22. The frequency synthesizer module of the radar unit generates a 12 GHz mixing reference point frequency signal and sends it to the first up-conversion and the second up-conversion of the radar unit; S23. The first up-conversion of the radar unit performs mixing processing on the baseband radar transmission signal, so that the center frequency of the baseband radar transmission signal is shifted to 12 GHz, and the harmonic components are filtered out through the first band-pass filter; S24. The second up-conversion of the radar unit performs mixing processing on the 12 GHz baseband radar transmission signal after passing through the first band-pass filter, shifting it to 24 GHz to generate a millimeter-wave radar transmission signal; The millimeter-wave radar transmission signal is filtered through the second band-pass filter to remove the harmonic components, and one path is sent to the power amplifier for amplification and then transmitted by the transmitting antenna; one path is sent to the de-slope mixing unit to provide a reference signal for the de-slope mixing unit; S26. The ocean echo signal received by the receiving antenna is amplified by the LNA and sent to the third band-pass filter to filter out the out-of-band interference components; S27. The amplified and filtered signal is sent to the de-slope mixing unit to complete the matched filtering processing using the reference signal, and then after passing through the low-pass filter, the spectrum of the sea surface echo signal is shifted to zero intermediate frequency to generate the one-dimensional range image feature of the sea surface echo; S28. After the one-dimensional range image feature of the sea surface echo is processed by the AD, it is sent to the radar signal processing unit for feature extraction to generate feature data; S29. The radar data processing module integrates the feature data transmitted by the radar signal processing unit and the illumination angle of the two-axis turntable at the moment corresponding to the timestamp sent by the clock synchronization unit, maps them to the reference coordinate system, and obtains the wave surface elevation data of the target sea area; S210. Fuse and process the wave surface elevation data at different illumination angles, calculate the sea wave information of the target sea area using a deep learning algorithm, and store the data according to the timestamp given by the clock synchronization unit. The sea wave information includes wave height, wave period, wave direction, one-dimensional energy spectrum, and two-dimensional azimuth spectrum.

[0015] In the above-described aspects and any possible implementation manners, a further implementation manner is provided. Calculating the sea wave information of the target sea area using a deep learning algorithm includes: inputting the fused wave surface elevation data into a multi-modal spatio-temporal fusion deep learning model; the multi-modal spatio-temporal fusion deep learning model includes a radar feature extraction module and a transformation processing module; wherein, the radar feature extraction module is used to input the fused wave surface elevation data and output a time series feature vector; the transformation processing module is used to input the time series feature vector and output the sea wave information.

[0016] In the above-described aspects and any possible implementation manners, a further implementation manner is provided. The radar feature extraction module includes a convolutional neural network and a bidirectional long short-term memory network; the transformation processing module is a fully connected decoder.

[0017] Compared with the prior art, the present invention has the following beneficial effects: This application uses a scanning millimeter-wave radar device, controls the angles of the transmitting antenna and the receiving antenna with a two-axis turntable to achieve time-sharing observation with angle stepping, and at the same time performs strict time synchronization by setting a clock synchronization unit, so as to map the sea surface feature data at different times to the corresponding spatial positions, realize scanning observation on a finer spatial scale, accumulate multi-angle data using the slow time dimension, combine with a deep learning algorithm to achieve fusion, improve the calculation accuracy, have greater advantages in the calculation of sea wave velocity and direction, and can realize high-resolution sea wave direction spectrum inversion; this application can complete sea surface monitoring in real time and effectively, has the advantages of high real-time performance, reliable monitoring results, and unattended operation, and has important significance for the construction of a new generation of ocean monitoring systems.

[0018] It should be understood that the content described in the invention content part is not intended to limit the key or important features of the embodiments of the present invention, nor is it used to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In conjunction with the accompanying drawings and with reference to the following detailed description, the above and other features, advantages, and aspects of the embodiments of the present invention will become more apparent. The accompanying drawings are used to better understand the solution and do not limit the present invention. In the drawings, the same or similar reference numerals denote the same or similar elements, where: Figure 1 Shows a schematic diagram of a scanning millimeter-wave sea wave measurement device and a radar unit; Figure 2 Shows a schematic diagram of DA; Figure 3 Shows a schematic diagram of a frequency synthesizer module; Figure 4 Shows a schematic diagram of the relationship between frequency and gain; Figure 5 Shows a schematic diagram of the relationship between frequency and noise figure; Figure 6 Shows a schematic diagram of AD; Figure 7 Shows a schematic diagram of the relationship between the frequency and the energy spectral density observed by the measurement device provided in the third embodiment; Figure 8 Shows a schematic diagram of the two-dimensional energy spectrum observed by the measurement device provided in the third embodiment. Detailed implementation manners

[0020] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts fall within the scope of protection of the present invention.

[0021] In addition, the term "and / or" in this document is merely a description of the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B may represent: A exists alone, A and B exist simultaneously, and B exists alone. In addition, the character " / " in this document generally represents an "or" relationship between the associated objects before and after.

[0022] This application provides a scanning millimeter-wave sea wave measurement device. Refer to Figure 1 , specifically including: A remote display and control unit for data interaction with a remote communication unit; A remote communication unit for sending down the instruction information transmitted by the remote display and control unit; the instruction information includes working parameters and irradiation angles; A control unit for receiving the instruction information sent by the remote communication unit and controlling the operation of the radar unit; A radar unit, which is configured to transmit corresponding radar transmission signals according to operating parameters, and at the same time rotate a two-axis turntable that supports a transmitting antenna and a receiving antenna according to an illumination angle for illumination, and calculate sea wave information based on the received sea surface echo signals; A clock synchronization unit, which is configured to synchronize timestamps into the control unit and the radar unit.

[0023] The remote display and control unit is used to issue the operating parameters of the radar unit, control the operating process of the radar unit, detect the operating status, and display the data obtained from measurement and analysis. The hardware of the remote display and control unit can be any model of computer. Just install the display and control software of the measurement device of this application in the control computer and configure the corresponding network information according to the parameter requirements.

[0024] In a specific embodiment, the remote communication unit adopts a USR-G806w industrial router, which supports 5G / 4G networks and WIFI functions, greatly improves the anti-interference ability and connection stability, and has three modes: WIFI hotspot, WIFI client, and WIFI relay; in addition, it integrates technologies such as 4G-LTE, Wi-Fi, built-in patch cards, Ethernet ports (2 LANs and 1 WAN / LAN), and VPN to provide advanced Internet connections and high-speed data access for the measurement device of this application, and quickly build its own application network.

[0025] Among them, the radar unit includes a transmitting part, a transmitting antenna, a two-axis turntable, a receiving antenna, and a receiving part; the output end of the transmitting part is connected to the transmitting antenna, the output end of the receiving antenna is connected to the input end of the receiving part, and both the transmitting antenna and the receiving antenna are arranged on the two-axis turntable; The transmitting part is sequentially provided with a frequency control unit, a DDS, a DA, a first up-conversion, a first band-pass filter, a second up-conversion, a second band-pass filter, and a power amplifier according to the signal transmission path; Among them, the output end of the DDS is also connected to the input end of the frequency synthesizer module, and the output end of the frequency synthesizer module is connected to the input ends of the first up-conversion and the second up-conversion; The receiving part is sequentially provided with an LNA, a third band-pass filter, a de-slope mixing unit, a low-pass filter, an AD, a radar signal processing unit, and a radar data processing module according to the signal transmission path; Among them, the output end of the second band-pass filter is connected to the input end of the de-slope mixing unit, and the output end of the two-axis turntable is connected to the input end of the radar signal processing unit.

[0026] The following explanations are made for each part of the radar unit respectively: (1) Two-axis turntable The two-axis turntable can adopt a miniaturized electric pan-tilt head. The control azimuth of the two-axis turntable is 0 - 360°, the pitch angle is -45° to 45°, and the control speed is 1° / s. By receiving the control instructions from the control unit, it can realize the scanning of the illumination radar in the corresponding target sea area and transmit the illumination angle to the radar signal processing unit.

[0027] (2)Frequency control unit The frequency control unit generates corresponding phase control words and frequency control words according to the waveform design requirements of the measuring device in this application, and is used to control the DDS to generate corresponding radar transmission signal waveform data.

[0028] (3)DDS The DDS can generate arbitrary waveforms, meet the radar waveform design requirements of the measuring device in this application, and provide a general architecture for the subsequent expansion of the radar unit.

[0029] Compared with traditional frequency synthesizers, the DDS has the advantages of low cost, low power consumption, high resolution, and fast conversion time, and can adapt to the complex transmission waveform design for sea wave observation.

[0030] (4)DA The DA is used to convert the digital intermediate frequency transmission signal encoded according to the corresponding frequency in the DDS into an analog signal within the 300MHz bandwidth range.

[0031] In a specific embodiment, for the convenience of subsequent expansion, the DA needs to be able to adapt to various signal bandwidth requirements and meet the expansion requirements of the subsequent radar unit. See Figure 2 . The DA is a single-channel module with a sampling accuracy of 16 bits and a maximum adaptable working signal bandwidth of 120MHz.

[0032] First, it receives the clock and trigger signals from the previous module, inputs them into the clock management unit for time synchronization, and obtains the control instructions (i.e., configuration parameters) and status signals (i.e., feedback system status) of the previous module. The clock management unit drives the DA chip with a 300MHz clock, and realizes high-speed digital-to-analog conversion through Σ-Δ modulation or pipelined architecture to ensure that the sampling rate meets the Nyquist criterion. Before output, the high-frequency components are filtered by the anti-aliasing filter integrated in the radio frequency transformer to ensure the quality of the analog signal. After output from the radio frequency transformer, the conversion from the digital domain to the analog radio frequency domain is completed.

[0033] It adopts a multi-channel interleaved architecture and cooperates with a low-jitter clock circuit to ensure the spurious-free dynamic range performance of this module at 2.5 times the bandwidth, that is, 750Msps.

[0034] (5)Frequency synthesizer module The frequency synthesizer module uses principles such as frequency division, frequency multiplication, and frequency mixing to generate the local oscillator signal required for the frequency conversion of the measurement device in this application and the clock reference signal required for the digital circuit.

[0035] The frequency synthesizer module adopts a fully coherent design to ensure the performance of the device. Structurally, the frequency synthesizer module is placed inside the radar unit to provide the frequency synthesis signal for the entire device.

[0036] See Figure 3 , the phase-locked loop is a key part for frequency synthesis, which locks the output frequency through feedback control. The phase detector can divide the input signal into two paths and process them separately in different phase-locked loops. During operation, the oven-controlled crystal oscillator is used to provide a reference signal with a high stability of 100 MHz and a power of +10 dBm. The internal clock and the external clock are both input into the phase detector for power distribution or signal mixing, and finally the two signals are respectively input into two phase detectors to generate the first frequency synthesis signal of 3.5 GHz and the second frequency synthesis signal in the range of 10.75 - 14.75 GHz.

[0037] (6) The first up-conversion and the second up-conversion The first up-conversion is used to up-convert the broadband intermediate-frequency baseband signal generated by the intermediate-frequency baseband frequency source to 12 GHz, and then the second up-conversion up-converts the 12 GHz signal to the 24 GHz frequency band.

[0038] (7) The first band-pass filter and the second band-pass filter The first band-pass filter is a small-size high-pass filter with a working frequency covering 10 - 14 GHz, an insertion loss of less than 1.5 dB, a voltage standing wave ratio of 1.5, and a rejection rate exceeding 50 dB.

[0039] The second band-pass filter is cascaded after the frequency conversion unit to filter out the frequency multiplication components, interference, and noise signal components outside the receiver band, and improve the signal-to-noise ratio of the signal.

[0040] In a specific embodiment, the radio frequency filters (the first band-pass filter and the second band-pass filter) are of models TLLF-DC-30G-L-1 and TLLF-DC-30G-L-2; (8) The power amplifier The power amplifier is used to amplify the signal output by the second band-pass filter to meet the corresponding radiation power requirements.

[0041] In a specific embodiment, the HGC367 amplifier is used to achieve power amplification. From Figure 4 and Figure 5 it can be seen that the signal has a stable gain in a very wide frequency band range, meeting the usage requirements of this application.

[0042] (9) The transmitting antenna and the receiving antenna In this application, both the transmitting antenna and the receiving antenna adopt PCB antennas. Multiple transceiver sub-arrays form an antenna transceiver array, which meets the corresponding technical index requirements and forms a narrow spatial transceiver beam.

[0043] (10)LNA The LNA (Low-Noise Amplifier) is located at the receiving end and is used to amplify the received sea echo signal to meet the dynamic range requirements of the AD conversion. Generally, the intensity of the sea echo signal is lower than the order of -50 dBm, while the input requirement for the intermediate-frequency AD signal acquisition is -10 dBm to 0 dBm.

[0044] In a specific embodiment, the signal is cascaded and amplified by using a cascaded method of three-stage HGC367 amplifiers to meet the input requirements of the AD.

[0045] (11)Third band-pass filter The third band-pass filter is cascaded after the LNA and is used to filter out the interference and noise signal components outside the band of the receiving antenna and improve the signal-to-noise ratio of the received signal.

[0046] In a specific embodiment, the radio frequency filter model is TLLF-DC-300M-L, which is a small-size high-pass filter with a working frequency covering DC - 0.3 GHz, an insertion loss of less than 1.5 dB, a voltage standing wave ratio of 1.5, and a rejection rate exceeding 50 dB.

[0047] (12)De-slope mixing unit The de-slope mixing unit completes the matched filtering process of the sea echo signal, mixes the received sea echo signal with another path of radar transmitting signal coupled over, transforms it to zero intermediate frequency, and obtains the one-dimensional range image feature of the sea echo corresponding to the angle.

[0048] The de-slope mixing unit is a technology that converts a broadband linear frequency modulation signal into a fixed intermediate frequency signal by eliminating the frequency modulation (frequency modulation slope) of the signal. De-slope mixing converts the broadband linear frequency modulation signal into a narrowband fixed intermediate frequency signal by matching the local oscillator signal with the frequency modulation slope, significantly reducing the complexity of subsequent processing. The functions of the de-slope mixing unit mainly include the following two aspects: 1. The local oscillator signal is strictly synchronized with the frequency modulation slope of the received signal; 2. Low-pass filtering accurately extracts the fixed frequency component and removes residual frequency components such as harmonics.

[0049] Through the de-slope mixing process, the one-dimensional range image feature of the sea echo in the target echo can be extracted.

[0050] (13)AD AD samples and converts the sea surface echo signal received and amplified by the receiving antenna into a digital signal for subsequent radar data processing module to process.

[0051] See also Figure 6 AD achieves efficient digitization of RF signals by integrating clock management unit, control logic and high-speed ADC chip, ensures timing accuracy through time management unit, ensures scalability through FMC interface, and ensures system coordination through linkage between time management unit and RF transformer.

[0052] The high-speed ADC chip is located at the key node of the system signal processing link, connecting the upstream RF module and the downstream data processing unit through the FMC interface. In this application, in order to match the transmitted signal, a 300MHz pipeline or successive approximation (SAR) ADC chip is used, and the digital quantization bit is 16 bits to adapt to the high dynamic range of the sea surface echo signal.

[0053] AD uses digital error correction logic to improve reliability. It coordinates the timing relationship between clock, control signal and data output through the synchronization signal generated by the clock management unit to avoid timing conflicts or data loss.

[0054] (14) Radar signal processing unit The radar signal processing unit is used to complete the feature extraction of the sea surface echo signal and realize the analysis of the wave characteristics.

[0055] In a specific embodiment, the radar signal processing unit uses Intel's second-generation I7 processor, which includes a set of 256-bit AVX (Advanced Vector Extensions) vector floating-point computing instruction sets. The AVX vector operation unit can greatly improve the floating-point computing capability of traditional X86 instructions. At a main frequency of 2.1GHz, the peak computing capability of the i7-2715QE processor can reach 135 GFLOPS.

[0056] (15) Radar data processing module The radar data processing module is used to analyze the information output by the radar signal processing unit, that is, to estimate the parameters of the sea surface characteristics and extract the wave characteristics.

[0057] The radar data processing module completes the one-dimensional range image data sorting, correlation processing and characteristic parameter estimation of the range Doppler processing results of different channels. While considering the computing power, it is also necessary to consider characteristics such as operating power consumption.

[0058] In a specific embodiment, the radar data processing module uses an embedded industrial control computer, and develops corresponding data processing software on the platform to analyze the measured radar data and estimate parameters, and can have low power consumption. Specifically, an EPC-S202 embedded industrial control computer is used, and corresponding data processing software is developed to implement the full correlation processing and feature extraction of data.

[0059] The control unit can use an EPC-S202 embedded industrial control computer as the hardware architecture, and develop corresponding embedded control software to forward the working parameters and issue the control parameters of the two-axis turntable, and implement the scheduling control of the working process and the control of the state pointing during the observation process. Exemplarily, the two-axis turntable can be controlled to scan a 180° sector within 30 minutes, and the time-sharing observation method with angle stepping (such as 0.1°) is adopted to obtain 3,600 angle samples, so as to accumulate multi-angle data in the slow time dimension.

[0060] The clock synchronization unit provides a clock reference signal for different modules of the measurement device of the present application, is used to control the radar pointing time of the two-axis turntable and the working time of the millimeter-wave radar irradiated, realizes the synchronization of the working time of the radar unit and the irradiated area, and provides timestamp information for subsequent sea wave feature analysis, that is, fuses the scanning sea area information corresponding to different moments through strict time synchronization.

[0061] In a specific embodiment, the clock synchronization unit can use a GPS Beidou synchronous clock, adopt high-precision timing GPS and Beidou modules and precise time synchronization technology to provide accurate clock synchronization signals.

[0062] In addition, the measurement device of the present application further includes: A state monitoring and maintenance unit, which is used to collect environmental data and the working state data of other units in real time, determine the dynamic relationship between the environment and the working state, and then judge the occurrence of faults according to the dynamic relationship. If a fault occurs, a maintenance strategy is generated and processed, and data interaction is carried out with the remote display and control unit.

[0063] Considering the influence of the marine environment on the aforementioned device, such as salt spray will accelerate the oxidation and corrosion of metal parts, especially the oxidation of the antenna support, two-axis turntable, etc., resulting in a decrease in mechanical strength and jamming of moving parts; the marine environment has high humidity, which is easy to cause internal condensation of the equipment, causing the circuit board, etc. to get damp... which is not conducive to the long-term use of the measurement device. Therefore, a state monitoring and maintenance unit is added.

[0064] Specifically, the state monitoring and maintenance unit includes: (1) An acquisition unit, which is used to collect the working state data of other units except the state monitoring and maintenance unit in real time.

[0065] In a specific embodiment, the acquisition unit includes: An environmental sensor module, including a wind speed sensor and a precipitation measurement sensor provided on the radar unit, a fog and salt sensor provided on the receiving antenna and the transmitting antenna, a humidity sensor and a temperature sensor provided in the radar unit, and a GPS signal strength sensor provided on the clock synchronization unit, which are respectively used to collect wind speed data, rainfall data, salt fog concentration, humidity data, temperature data, and radar signal strength data in the environment where the radar unit is located; Exemplarily, the model of the wind speed sensor is Gill WindSonic4, the model of the precipitation measurement sensor is Texas Electronics TE525MM, the model of the fog and salt sensor is Halymor-EC5 electrochemical sensor, the model of the humidity sensor is Sensirion SHT45, the model of the temperature sensor is PT100 RTD, and the model of the GPS signal strength sensor is u-blox NEO-M9N. The GPS signal strength monitor monitors the carrier-to-noise ratio. When the threshold is less than 35 dB / Hz, it indicates that the clock synchronization unit may have clock synchronization drift or loss of lock, affecting the accuracy of the sea wave parameter measurement. Therefore, maintenance and calibration should be carried out in a timely manner.

[0066] A working state data acquisition module, including a bus monitoring module, a transmitted signal acquisition module, a radar unit detection module, a device sensor module, and a data verification module, which is used to output working state data.

[0067] Among them, the bus monitoring module is provided between the control unit and the radar unit and between the control unit and the radar unit, and is used to output the abnormal transmission state of the working parameters; the FPGA is used to realize real-time data capture. For example, the number of CRC check errors within every 10 seconds is counted. When the threshold is greater than 5 times / second, it is determined that a fault has occurred. After the control unit issues an instruction, if the response time of the radar unit is greater than 200 ms, it is determined as a transmission anomaly.

[0068] The transmitted signal acquisition module is used to output the quality indicators of the radar transmitted signal; the center of the transmitted signal can be monitored by integrating a spectrum analyzer in the radar unit. For example, when the deviation is greater than 1 MHz, calibration is required, and fields such as output signal power and bandwidth are output.

[0069] The radar unit detection module is used to output the detection results of software crashes and configuration errors of the radar unit; the detection of software crashes can be realized by setting a watchdog counter, and whether a configuration error occurs can be judged by comparing whether the current configuration of the radar unit is consistent with the remotely issued configuration file.

[0070] The device sensor module includes a vibration sensor and a rotation angle sensor arranged on a two-axis turntable, which are respectively used to output the vibration data and rotation angle data of the two-axis turntable; Exemplarily, the model of the vibration sensor is ADI ADXL357 three-axis MEMS accelerometer, and the model of the rotation angle sensor is ALPS RDC503013A.

[0071] The data verification module is used to verify the output data at the output end of the remote communication unit and output the verification result; CRC-32 verification can be performed on the output data of the remote communication unit.

[0072] (2) The acquisition time synchronization unit is used to synchronize the timestamp to the remote protection unit and mark the working state data; The acquisition time synchronization unit is dedicated to the status monitoring and maintenance unit and provides a timestamp reference for fault records. Its accuracy requirement is lower than that of the clock synchronization unit. Therefore, the acquisition time synchronization unit can be connected to the clock synchronization unit through the PTP protocol or RS422 interface to synchronize the time of the clock synchronization unit to the acquisition time synchronization unit, ensuring the overall timing consistency of the device. And the acquisition time synchronization unit acts as a slave clock. Even if the clock synchronization unit fails briefly, the acquisition time synchronization unit can still maintain short-term synchronization.

[0073] The acquisition time synchronization unit can adopt a temperature-compensated crystal oscillator (TCXO), and also adds a clock offset compensation algorithm to predict the crystal oscillator offset according to historical synchronization data to achieve dynamic correction.

[0074] (3) The coupling unit is used to process the working state data and environmental data to obtain the dynamic relationship between the environment and the working state, including: The feature module: used to obtain historical working state data and historical environmental data for preprocessing and extracting features; The features include physical features and device response features; The preprocessed historical working state data and features are divided into a training set and a validation set; The dynamic relationship construction model; used to construct a dynamic relationship model, and the dynamic relationship model includes a physical drive model and a data drive model; The baseline compensation amount output by the physical drive model is input into the data drive model; The training module: used to train the data drive model using the training set, update the parameters of the data drive model using the loss function, and verify the performance of the data drive model using the validation set; The data drive model is used to input the to-be-tested working state data and to-be-tested environmental data and output the precise compensation amount.

[0075] (4) The remote protection unit is used to judge whether a fault occurs according to the output result of the coupling unit, the working state data and the environmental data, generate a maintenance strategy and perform data interaction with the remote display and control unit, including: A fault diagnosis module, which is used to identify faults in the output results, working status data, and environmental data of the coupling unit, and classify the identified faults according to their levels; A protection module, which is used to generate maintenance decisions and warnings according to the classification results; A transmission module, which is used to perform data interaction with the remote display and control unit.

[0076] Among them, the fault diagnosis module includes: A preprocessing module, which is used to obtain multiple groups of historical fault data of the measuring device and the corresponding output results of the coupling unit for each group and perform preprocessing; the historical fault data includes fault working status data and fault environmental data; A fusion module, which is used to fuse the preprocessed historical fault data and the corresponding output results of the coupling unit for each group, and divide them into a training set and a verification set; A model construction module, which is used to construct a lightweight fault diagnosis model, train the model using the training set, and verify the model performance using the verification set. The model is used to output the fault judgment result of the working status data to be detected; A judgment module, which is used to input the working status data to be detected and the environmental data to be detected into the trained model, and output the fault judgment result.

[0077] Exemplarily, according to the impact of the fault on the usage of the measuring device of the present application, the fault levels can be artificially divided as follows: Level 1 fault: Hardware damage, mechanical jamming, such as the two-axis turntable cannot rotate, etc.; Level 2 fault: Parameter overrun (frequency deviation > 10 MHz, humidity > 90% RH), etc.; Level 3 fault: Communication delay (RTT > 500 ms), salt spray amount > 150 μg / cm², etc.; Among them, the level 1 fault is the most urgent situation, which may cause the complete paralysis of the device. For example, a short circuit occurs in the radar unit, resulting in the interruption of radar signal transmission and the inability to detect, or a hardware short circuit causes local overheating or fire risk, etc., and emergency repair is required; the level 2 fault is the second most urgent situation, which will greatly affect the measurement accuracy and requires dehumidification, calibration, etc.; the level 3 fault has the least impact on the device and can be postponed for processing.

[0078] The urgency levels of each level are different. For the most urgent level 1 fault, an early warning needs to be sent by text message or other means for emergency repair; for level 2 and level 3 faults, remote maintenance can be carried out through the fault processing unit.

[0079] (5)Fault handling unit, which is used to handle faulty devices according to the maintenance strategies generated by the remote protection unit, mainly for automatic maintenance of secondary and tertiary faults, reducing the complexity of manual operations and being beneficial to the long-term use of the measuring device, including: Hardware control module, including salt spray prevention spraying device, intelligent temperature control device, mechanical emergency braking device and GPS signal enhancement device; Software control module, including parameter adjustment unit, transmission management unit, communication guarantee unit and recovery module; Among them, the parameter adjustment unit is used to adjust the radar transmission signal output by the radar unit; the transmission management unit is used to prevent interruption of work parameter transmission; the communication guarantee unit is used to guarantee the communication quality of the remote communication unit, and the recovery module is used to recover radar unit faults.

[0080] Among them, in the parameter adjustment unit, the output frequency of the DDS is adjusted and the transmission power is adjusted in real time according to the signal-to-noise ratio of the sea surface echo signal. When the quality of the radar transmission signal does not meet the requirements, the parameter adjustment unit is triggered.

[0081] In the transmission management unit, the multi-path TCP protocol can be used to achieve link aggregation and dynamic routing selection, and the secondary link is switched when the main link data transmission is interrupted; when the bus monitoring module detects data anomalies, the transmission management unit is triggered.

[0082] The communication guarantee unit can dynamically select a suitable modulation method by adopting adaptive modulation and coding technology, or first perform local storage, temporarily store data during communication interruption, and re-upload after recovery; when communication delay occurs, the transmission management unit is triggered.

[0083] In the recovery module, relevant recovery strategies can be defined, such as solving software crashes by restarting and solving parameter configuration errors by restoring to the latest version.

[0084] This application provides a scanning millimeter-wave sea wave measurement method as follows: S1. The remote display and control unit transmits work parameters to the control unit through the remote communication unit. The control unit receives the time stamp of the clock synchronization unit to mark the work parameters, and issues the marked work parameters to the radar unit; S2. In the radar unit, corresponding radar transmission signals are emitted according to the marked work parameters, and the two-axis turntable is controlled to rotate according to the marked irradiation angle to irradiate the target sea area, and the sea wave information of the target sea area is calculated according to the received sea surface echo signal. The radar unit receives the time stamp of the clock synchronization unit to mark the sea wave information; S3. The marked sea wave information is sent to the remote communication unit, and the remote communication unit and the remote display and control unit perform information interaction; S4. The remote display and control unit visually displays the marked sea wave information and the information output by the status monitoring and maintenance unit.

[0085] Among them, in the radar unit, S2 specifically includes the following steps: S21. The frequency control unit of the radar unit controls the DDS to generate a corresponding baseband radar transmission signal according to the marked operating parameters, and then generates an intermediate-frequency analog baseband signal through the DA and sends it to the first up-conversion; the radar unit sends the irradiation angle to the two-axis turntable to control the two-axis turntable to point to the target sea area corresponding to the irradiation angle; S22. The frequency synthesizer module of the radar unit generates a mixing reference point frequency signal of 12 GHz and sends it to the first up-conversion and the second up-conversion of the radar unit; S23. The first up-conversion of the radar unit performs mixing processing on the baseband radar transmission signal, so that the center frequency of the baseband radar transmission signal is shifted to 12 GHz, and the harmonic components are filtered out through the first band-pass filter; S24. The second up-conversion of the radar unit performs mixing processing on the 12 GHz baseband radar transmission signal after passing through the first band-pass filter, shifts it to 24 GHz, and generates a millimeter-wave radar transmission signal; The millimeter-wave radar transmission signal is filtered through the second band-pass filter to remove the harmonic components, and one path is sent to the power amplifier for amplification and then transmitted by the transmitting antenna; one path is sent to the de-slope mixing unit to provide a reference signal for the de-slope mixing unit; The ocean echo signal received by the receiving antenna is amplified by the LNA and sent into the third band-pass filter to filter out the out-of-band interference components; The amplified and filtered signal is sent to the de-slope mixing unit, and the matched filtering process is completed using the reference signal, and the spectrum of the sea surface echo signal is shifted to zero intermediate frequency to generate the one-dimensional range image feature of the sea surface echo; After the one-dimensional range image feature of the sea surface echo is processed by the AD, it is sent to the radar signal processing unit for feature extraction to generate feature data; The radar data processing module integrates the feature data transmitted by the radar signal processing unit and the irradiation angle of the two-axis turntable at the time corresponding to the timestamp sent by the clock synchronization unit, maps it to the reference coordinate system, and obtains the wave surface elevation data of the target sea area; The wave surface elevation data at different irradiation angles are fused and processed, and the sea wave information of the target sea area is calculated using a deep learning algorithm and stored according to the timestamp given by the clock synchronization unit. The sea wave information includes wave height, wave period, wave direction, one-dimensional energy spectrum, and two-dimensional azimuth spectrum.

[0086] Compared with traditional array scanning radars, the present invention uses a simultaneous irradiation method to complete sampling at different angles. Due to the limited number of arrays, the angle sampling rate is low, usually only having data at three angles, and the angle resolution is extremely low. In the present invention, a time-division scanning method is adopted, which greatly improves the angle sampling accuracy. Through data fusion processing of time-division accumulation, the angle resolution is effectively improved, thereby having extremely high two-dimensional spectral resolution and wave direction measurement accuracy.

[0087] Exemplarily, the scanning millimeter-wave sea wave measurement device deployed at a certain ocean observation station controls the radar antenna to scan a 180° sea area in 0.1° steps through a two-axis turntable, and collects multi-angle sea surface echo data.

[0088] The time-division accumulated data strictly aligns the timestamps of the echo data at each irradiation angle through a clock synchronization unit, maps the wave surface elevation data collected at different times to a unified spatial coordinate system, forms a multi-angle data set with spatio-temporal correlation; uses an interpolation algorithm to fill the scanning gaps, and reconstructs the discrete angle data into a continuous spatial distribution matrix based on the azimuth-pitch angle grid, obtaining a fused time-frequency spectrum diagram, thus realizing data fusion.

[0089] The fused data is input into a multi-modal spatio-temporal fusion deep learning model. The multi-modal spatio-temporal fusion deep learning model includes a radar feature extraction module and a transformation processing module; among them, the radar feature extraction module includes a convolutional neural network and a bidirectional long short-term memory network. The fused time-frequency spectrum diagram is input into the convolutional neural network for feature extraction, and a feature map is output. The feature map is transformed into a sequence varying with time and input into the bidirectional long short-term memory network, and a time series feature vector is output; the time series feature vector is input into the transformation processing module, and the transformation processing module is a fully connected decoder. After being processed by the fully connected decoder, a fused feature is output, namely wave height, wave period, and two-dimensional direction spectrum.

[0090] The multi-modal spatio-temporal fusion deep learning model can be represented by the following formula:

[0091] Among them, y represents the overall output of the model, W represents the weights of the fully connected layer in the transformation processing module, is the second bias term, and F represents the time series feature vector obtained after being processed by the bidirectional long short-term memory network.

[0092]

[0093] Among them, BiLSTM(·) represents the processing by the bidirectional long short-term memory network, represents the feature map obtained after being processed by the convolutional neural network.

[0094]

[0095] Among them, It means that it has been processed by convolutional neural network, w is the convolution kernel weight, X is the fused time-frequency spectrum, Indicates a convolution operation, and b is the first bias term.

[0096] The convolutional network is combined to extract time-frequency features, and the bidirectional long short-term memory network is combined to capture time series associations, and finally the wave height, wave period, wave direction, one-dimensional energy spectrum and two-dimensional directional spectrum are output. The training data contains 100,000 sets of samples from three months of measured storms and calm periods. After pruning and quantization, the multimodal spatiotemporal fusion deep learning model is embedded in the radar data processing module. The actual measurement shows that the mean square error of wave height prediction is reduced to 0.08m, which is 33% higher than the traditional buoy method, and the directional spectrum resolution reaches 1°.

[0097] When a condition monitoring and maintenance unit is added, S1 also includes: The status monitoring and maintenance unit starts to monitor the environmental data and working status data in real time, determines whether a fault occurs and generates a maintenance strategy to handle the faulty equipment; the status monitoring and maintenance unit exchanges data with the remote display and control unit in real time; the working parameters include radar modulation period, signal bandwidth, working frequency, and illumination angle; For other contents, please refer to the contents of the aforementioned measuring device, and this application will not elaborate on them.

[0098] Determining whether a fault has occurred includes the following steps: S111, obtaining historical working status data and historical environmental data for preprocessing and extracting features; the features include physical features and device response features; dividing the preprocessed historical working status data and features into a training set and a verification set; In this application, first of all, outliers must be eliminated and data that is beyond a reasonable range must be deleted, such as temperatures greater than 100°C; secondly, various types of data must be aligned with their time axes to form a hierarchical cache queue; finally, missing data must be filled in. For example, interpolation is used to reconstruct missing values based on data from adjacent times.

[0099] Physical characteristics include environmental factors and mechanical coefficients, among which the environmental factor E s Calculated by the following formula:

[0100] Among them, S M is the salt spray concentration, H is the humidity data, and R is the rainfall data; the environmental factors quantify the effects of rainfall, salt spray and humidity.

[0101] Mechanical coefficient M l Calculated by the following formula:

[0102] Among them, V E (t) represents the variation of the vibration data of the two-axis turntable with time t, and W s (t) is the variation of the wind speed with time t, which reflects the cumulative effect of the wind on the two-axis turntable.

[0103] S112. Construct a dynamic relationship model; The dynamic relationship model includes a physical-driven model and a data-driven model. The physical-driven model is used to calculate the baseline compensation amount according to the physical model, including a rainfall attenuation model and a turntable limit model; The rainfall attenuation model is expressed by the following formula:

[0104]

[0105] Among them, represents the baseline compensation amount of the transmission power of the radar transmission signal. K represents the attenuation coefficient, and γ represents the rainfall attenuation index, which is used to express the non-linear relationship between the rainfall amount data and the attenuation of the radar transmission signal.

[0106] The turntable limit model is expressed by the following formula:

[0107]

[0108] Among them, represents the angular baseline compensation amount of the two-axis turntable, represents the wind resistance coefficient. The physical-driven model provides the baseline compensation amount based on known physical laws to ensure compliance with the engineering safety boundary. The baseline compensation amount output by the physical-driven model is input into the data-driven model, and the data-driven model accurately captures complex internal effects, improving the overall adaptability of the dynamic relationship model.

[0109] The data-driven model is a deep learning network, which successively includes an input layer, a 3-layer fully connected network, an activation function connected after each layer of the fully connected network, and an output layer; the output layer is provided with a mathematical clipping function; the mathematical-driven model is used to predict the difference between the baseline compensation amount and the actual compensation amount, that is, the prediction difference amount, which covers complex relationships not included in the physical model, such as non-linear effects of equipment aging and salt spray corrosion, etc., to obtain an accurate compensation amount; among them, the actual compensation amount is the value issued by the control unit minus the value actually measured after being disturbed by wind and rain.

[0110] The mathematical-driven model finally outputs the sum of the baseline compensation amount and the prediction difference amount, that is, the accurate compensation amount.

[0111] S113. Train the data-driven model using the training set, update the model parameters using the loss function, and verify the performance of the data-driven model using the validation set. The data-driven model is used to input the data of the to-be-detected working state, the data of the to-be-detected environment, and the corresponding baseline compensation amount, and output the precise compensation amount, such as the adjustment amount of the transmission power of the radar transmission signal and the adjustment angle of the two-axis turntable. The loss function L can be expressed by the following formula:

[0112] where, represents the actual compensation amount of the transmission power of the radar transmission signal, represents the precise compensation amount of the transmission power of the radar transmission signal; θ represents the actual compensation amount of the angle of the two-axis turntable, and θ’ represents the precise compensation amount of the angle of the two-axis turntable; α and β are weight coefficients.

[0113] S114. Compress the trained dynamic relationship model.

[0114] The above process establishes the interaction between the environment and the device, emphasizes the physical effect of the environment on the device, and provides a stable input for the remote protection unit. If a fault judgment model is directly constructed based on the environment data and the working state data, early warning can only be carried out a few hours before the fault occurs, and the fault risk caused by the real-time environment cannot be avoided.

[0115] S121. Obtain multiple groups of historical fault data of the measuring device and the output results of the corresponding coupling units for each group and perform preprocessing. The historical fault data includes fault working state data and fault environment data. In this application, the types of historical fault data are diverse, including bus transmission status data (digital signals), radar transmission signal quality (time-frequency domain waveform), environmental sensor data (physical quantities such as salt spray / humidity / vibration), GPS signal strength (electromagnetic field parameters), etc. from different dimensions. First, various types of data are respectively aligned with their time axes to form a hierarchical cache queue. Secondly, missing data is filled. Finally, feature extraction is respectively performed on different types of data. For example, short-time Fourier transform is performed on the time-frequency domain waveform to generate a spectrogram, such as calculating the accumulation amount of fog and salt in a day.

[0116] S122. Fuse the preprocessed historical fault data and the output results of the corresponding coupling units for each group, and divide them into a training set and a validation set.

[0117] In this application, the data features of various types after feature extraction are merged, spliced, and added to the queue after filling in the missing data according to the feature extraction time to form a mapping relationship between time and features.

[0118] S123. Build a lightweight fault diagnosis model, train the model using the training set, and verify the model performance using the validation set. The model is used to output the fault judgment results of the data of the working state to be detected and the data of the environment to be detected.

[0119] The model is a structure combining depthwise separable convolution and bidirectional gated recurrent network. In the depthwise separable convolution, the quality of the radar transmission signal in the input training set, that is, the radar spectrogram after time-domain analysis, including data in two dimensions of frequency and amplitude, is input; in the bidirectional gated recurrent network, a one-dimensional sequence is input for processing.

[0120] Among them, the depthwise separable convolution includes two parts: depth convolution and pointwise convolution. The pointwise convolution extracts the temporal evolution law of various types of data with a 1×8 convolution kernel. The 1×8 depth convolution kernel only needs to process the spatial features of a single channel, which is much smaller than the standard 3×3 convolution, reducing the model complexity, and the 1×8 convolution kernel improves the accuracy of anomaly detection of radar spectral data.

[0121] Concatenate the output features of the depthwise separable convolution and the bidirectional gated recurrent network. The output layer of the model uses a tree-structured classifier. Its first level uses a support vector machine to distinguish major categories of hardware, software, or environmental faults; if it is a hardware fault, then at the second level, for hardware faults, the random forest algorithm is used for further subdivision; if it is an environmental fault, at the third level, for environmental interference, the fuzzy logic is used to evaluate the influence degree.

[0122] Update the model parameters using the hybrid loss. The model is compressed by pruning and quantization-aware training to generate an inference engine with small memory.

[0123] Step 124. Input the data of the working state to be detected into the trained model, and output the fault judgment results, such as hardware damage, communication delay, etc.

[0124] This application provides a specific embodiment. Install the measuring device of this application at an observation station.

[0125] The central frequency of the radar unit of the measuring device is 24 GHz, the bandwidth is 1.5 GHz, the modulation period is 100 us, the output power is 0.1 W, the radiation antenna system is a millimeter-wave array antenna, and the rest of the settings are not elaborated here and can be referred to the foregoing content.

[0126] After the sea surface echo signal is isolated and amplified and down-converted to zero intermediate frequency, it is input to the AD. The AD uses 12 bits, and the radar signal processing unit is a 32-bit floating-point DSP to complete the subsequent radar signal processing and data processing work.

[0127] Use the measuring device of this application to scan the area to be measured. Through inversion calculation, obtain the one-dimensional energy spectrum as shown in Figure 7 andFigure 8 The two-dimensional energy spectrum shown

[0128] Feature extraction is performed, and the obtained feature data can be seen in Table 1.

[0129] Table 1 Calculation results of sea wave feature data

[0130] According to the embodiments of the present invention, the following technical effects are achieved: The measurement results of the measurement device are basically equivalent to the accuracy of the test results of the high-cost wave cage, and it can effectively solve the problems of high cost, high difficulty, poor real-time performance, high danger, etc. in sea wave monitoring.

[0131] Based on the above technical solution, the present application adopts a scanning millimeter-wave radar device, uses a two-axis turntable to control the angles of the transmitting antenna and the receiving antenna, realizes time-sharing observation with angle stepping, and at the same time sets a clock synchronization unit for strict time synchronization, so as to map the sea surface feature data at different times to the corresponding spatial positions, realizes scanning observation on a finer spatial scale, accumulates multi-angle data using the slow time dimension, combines with deep learning algorithms to achieve fusion, improves the calculation accuracy, has greater advantages for the calculation of sea wave flow velocity and direction, and can realize high-resolution sea wave direction spectrum inversion; The present application sets up a status monitoring and maintenance unit, monitors the working status data of each unit in real time, processes the monitored data to judge the type of fault, constructs a dynamic relationship model, calculates the baseline compensation amount that needs to be provided for the device parameters based on the physical model, and captures the complex internal effects between the device and the environment by constructing a data-driven model, improving the overall adaptability of the dynamic relationship model; The dynamic relationship model analyzes the interaction between the environment and the device, emphasizes the physical effect of the environment on the device, and provides a stable input for the remote protection unit; A lightweight fault diagnosis model is constructed in the remote protection unit for fault classification, and active protection or emergency warning processing is carried out accordingly according to the level of the fault, significantly improving the reliability and stability of the device in a harsh marine environment, and still being able to maintain the operation of the core function in extreme weather. The system will have stronger environmental adaptability and self-maintenance ability, further meeting the design goal of "unattended"; among them, the time alignment of the acquisition time synchronization unit and the clock synchronization unit realizes the full-link time consistency from acquisition and processing to fault recovery; The present application can complete sea surface monitoring in real time and effectively, has the advantages of high real-time performance, reliable monitoring results, unattended, etc., and has important significance for the construction of a new generation of marine monitoring systems.

[0132] It should be noted that, for the foregoing method embodiments, for the sake of simple description, they are all expressed as a series of action combinations. However, those skilled in the art should know that the present invention is not limited by the described action sequence, because according to the present invention, certain steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all optional embodiments, and the actions and modules involved are not necessarily essential to the present invention.

[0133] The various embodiments of the systems and techniques described above in this document can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGA), application specific integrated circuits (ASIC), application specific standard products (ASSP), system on a chip (SOC), complex programmable logic devices (CPLD), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a dedicated or general-purpose programmable processor, and can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit the data and instructions to the storage system, the at least one input device, and the at least one output device.

[0134] The program code for implementing the method of the present invention can be written in any combination of one or more programming languages. These program codes can be provided to the processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing devices, such that when the program codes are executed by the processor or controller, the functions / operations specified in the flowcharts and / or block diagrams are implemented. The program codes can be executed entirely on the machine, partially on the machine, executed partially on the machine as an independent software package and partially on a remote machine, or executed entirely on a remote machine or server.

[0135] It should be understood that various forms of the flow shown above can be used, reordering, adding, or deleting steps. For example, the steps described in the present invention can be executed in parallel, sequentially, or in different orders, as long as the desired results of the technical solutions disclosed in the present invention can be achieved. No limitations are imposed herein.

[0136] The above specific embodiments do not constitute a limitation to the protection scope of the present invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention shall be included within the protection scope of the present invention.

Claims

1. A scanning millimeter-wave sea wave measuring device, wherein, Comprising: A remote display and control unit for data interaction with a remote communication unit; A remote communication unit for transmitting the instruction information sent by the remote display and control unit; the instruction information includes working parameters and irradiation angles; A control unit for receiving the instruction information sent by the remote communication unit and controlling the operation of the radar unit; A radar unit for transmitting corresponding radar emission signals according to the working parameters, and at the same time rotating a two-axis turntable according to the irradiation angle to drive the transmitting antenna and the receiving antenna to face the sea surface, and calculating sea wave information according to the received sea surface echo signals; A clock synchronization unit for synchronizing time stamps into the control unit and the radar unit.

2. The measuring device according to claim 1, wherein, The radar unit includes a transmitting section, a transmitting antenna, a two-axis turntable, a receiving antenna and a receiving section; the output end of the transmitting section is connected to the transmitting antenna, the output end of the receiving antenna is connected to the input end of the receiving section, and both the transmitting antenna and the receiving antenna are arranged on the two-axis turntable.

3. The measuring device according to claim 2, wherein, The transmitting section is sequentially provided with a frequency control unit, a DDS, a DA, a first up-conversion, a first band-pass filter, a second up-conversion, a second band-pass filter, and a power amplifier according to the signal transmission path.

4. The measuring device according to claim 3, wherein, The output end of the DDS is also connected to the input end of a frequency synthesizer module, and the output end of the frequency synthesizer module is connected to the input ends of the first up-conversion and the second up-conversion.

5. The measuring device according to claim 3, wherein, The receiving section is sequentially provided with an LNA, a third band-pass filter, a de-slope mixing unit, a low-pass filter, an AD, a radar signal processing unit, and a radar data processing module according to the signal transmission path.

6. The measuring device according to claim 5, wherein, The output end of the second band-pass filter is connected to the input end of the de-slope mixing unit, and the output end of the two-axis turntable is connected to the input end of the radar signal processing unit.

7. A scanning millimeter-wave sea wave measurement method, based on the scanning millimeter-wave sea wave measurement device according to any one of claims 1 to 6, wherein, Including the following steps: S1. The remote display and control unit transmits the working parameters to the control unit through the remote communication unit. The control unit receives the time stamp of the clock synchronization unit to mark the working parameters, and then issues the marked working parameters to the radar unit; S2. In the radar unit, corresponding radar emission signals are emitted according to the marked working parameters, and the two-axis turntable is controlled to rotate according to the marked irradiation angle to irradiate the target sea area. The sea wave information of the target sea area is calculated according to the received sea surface echo signals. The radar unit receives the time stamp of the clock synchronization unit to mark the sea wave information; S3. The marked sea wave information is sent to the remote communication unit, and the remote communication unit and the remote display and control unit perform information interaction; S4. The remote display and control unit visually displays the marked sea wave information and the information output by the status monitoring and maintenance unit.

8. The measuring method according to claim 7, wherein, In the radar unit, S2 specifically includes the following steps: S21. The frequency control unit of the radar unit controls the DDS to generate corresponding baseband radar emission signals according to the marked working parameters, and then generates intermediate-frequency analog baseband signals through the DA and sends them to the first up-conversion. The radar unit sends the irradiation angle to the two-axis turntable to control the two-axis turntable to point to the target sea area corresponding to the irradiation angle; S22. The frequency synthesizer module of the radar unit generates a 12 GHz mixing reference point frequency signal and sends it to the first up-conversion and the second up-conversion of the radar unit; S23. The first up-conversion of the radar unit performs mixing processing on the baseband radar transmission signal, shifting the center frequency of the baseband radar transmission signal to 12 GHz, and filtering out the harmonic components through the first band-pass filter; S24. The second up-conversion of the radar unit performs mixing processing on the 12 GHz baseband radar transmission signal after the first band-pass filter, shifting it to 24 GHz to generate a millimeter-wave radar transmission signal; S25. Filter out the harmonic components of the millimeter-wave radar transmission signal through the second band-pass filter, send one path to the power amplifier for amplification and then transmit it through the transmitting antenna; send one path to the de-slope mixing unit to provide a reference signal for the de-slope mixing unit; S26. The ocean echo signal received by the receiving antenna is amplified by the LNA and sent into the third band-pass filter to filter out out-of-band interference components; S27. Send the amplified and filtered signal to the de-slope mixing unit, complete the matched filtering processing using the reference signal, and then shift the spectrum of the sea surface echo signal to zero intermediate frequency through low-pass filter processing to generate the one-dimensional range image feature of the sea surface echo; S28. After the one-dimensional range image feature of the sea surface echo is processed by AD, it is sent to the radar signal processing unit for feature extraction to generate feature data; S29. The radar data processing module integrates the feature data transmitted by the radar signal processing unit and the illumination angle of the two-axis turntable at the corresponding moment of the timestamp sent by the clock synchronization unit, maps it to the reference coordinate system, and obtains the wave surface elevation data of the target sea area; S210. Perform fusion processing on the wave surface elevation data at different illumination angles, calculate the sea wave information of the target sea area using a deep learning algorithm, and store the data according to the timestamp given by the clock synchronization unit. The sea wave information includes wave height, wave period, wave direction, one-dimensional energy spectrum, and two-dimensional azimuth spectrum.

9. The measuring method according to claim 8, wherein, Calculating the sea wave information of the target sea area using a deep learning algorithm includes: Inputting the fused wave surface elevation data into a multi-modal spatio-temporal fusion deep learning model; The multi-modal spatio-temporal fusion deep learning model includes a radar feature extraction module and a transformation processing module; among them, The radar feature extraction module is used to input the fused wave surface elevation data and output a time series feature vector; The transformation processing module is used to input the time series feature vector and output the sea wave information.

10. The measuring method according to claim 9, wherein, The radar feature extraction module includes a convolutional neural network and a bidirectional long short-term memory network; the transformation processing module is a fully connected decoder.

Images