A CFA steady-state filtering method for velocity and direction detection

Abstract

This invention discloses a CFAR steady-state filtering method for velocity and direction detection, belonging to the field of target detection technology. The method includes: acquiring and processing real-time fluid radar signal data to determine target data; analyzing and identifying the target data to determine the CFAR steady-state filtering velocity and direction detection results; visually displaying the CFAR steady-state filtering velocity and direction detection results; and implementing timely fluid control based on these results. This invention solves the problem of existing methods failing to effectively detect fluid velocity and direction using CFAR steady-state filtering, resulting in low accuracy and poor performance. This invention can effectively detect fluid velocity and direction using CFAR steady-state filtering, providing reliable results under various environmental conditions, thus improving both accuracy and effectiveness.

Description

Technical Field

[0001] This invention relates to the field of target detection technology, specifically to a CFAR steady-state filtered flow velocity and direction detection method. Background Technology

[0002] Groundwater is one of the most important freshwater resources on Earth. Accurately measuring groundwater flow velocity and direction is of great significance for water resource management, pollution control, and agricultural irrigation.

[0003] Chinese patent CN105652034B discloses an intelligent sensor for detecting groundwater flow velocity and direction in landslide monitoring. It is suitable for measuring groundwater flow velocity and direction, a parameter used in landslide deformation calculations. It can be connected to an inclinometer, is easy to operate, facilitates data processing, and offers high measurement accuracy. It uses a nozzle-extruded tracer, which is pollution-free, has a good aqueous solution, a specific gravity similar to water, and good fluidity. However, this patent has the following drawbacks:

[0004] Existing technologies cannot effectively detect the velocity and direction of fluids based on CFAR steady-state filtering, thus failing to provide reliable velocity and direction detection results under various environmental conditions, resulting in low accuracy and poor performance in velocity and direction detection. Summary of the Invention

[0005] The purpose of this invention is to provide a CFAR steady-state filtering method for detecting flow velocity and direction. This method can effectively detect the flow velocity and direction of fluids based on CFAR steady-state filtering, providing reliable flow velocity and direction detection results under various environmental conditions. It can improve the accuracy and effectiveness of flow velocity and direction detection, and solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] A CFA steady-state filtered flow velocity and direction detection method includes:

[0008] Collect and process real-time data of fluid radar signals to determine target data of fluid radar signals;

[0009] A steady-state filtered flow velocity and direction detection model for CFAR is established to analyze and identify target data of fluid radar signals, predict the flow velocity and direction of the fluid, and determine the CFAR steady-state filtered flow velocity and direction detection results.

[0010] The system visualizes the flow velocity and direction detection results of CFAR steady-state filtering and enables timely control of the fluid based on these results.

[0011] Preferably, real-time data of fluid radar signals is acquired, including:

[0012] The radar sensor is installed next to the location of the fluid to be detected, in an open and unobstructed location.

[0013] Real-time monitoring of the fluid under test is based on radar sensors. The radar sensor emits microwave or millimeter wave signals that pass through the fluid and interact with scattering objects in the fluid. The radar sensor receives the radar waves reflected back by the scattering objects in the fluid and performs frequency analysis on the received radar signals to extract the frequency information of the scattering objects and determine the real-time data of the fluid radar signals.

[0014] Preferably, the real-time data processing of the fluid radar signal includes:

[0015] Based on Kalman filtering, steady-state filtering is performed on real-time fluid radar signal data to filter out specific frequency bands in the real-time fluid radar signal data that are not valuable for the detection of flow velocity and direction in CFA steady-state filtering, thereby eliminating noise and interference in the real-time fluid radar signal data.

[0016] Preferably, processing the real-time data of the fluid radar signal further includes:

[0017] Target detection is performed on real-time data of fluid radar signals based on constant false alarm rate (CFAR).

[0018] Specifically, the window is selected based on the real-time data of the fluid radar signal. A detection area is selected in the real-time data of the fluid radar signal to determine whether there is a target signal. Multiple reference areas are selected around the detection area to estimate the statistical characteristics of the background noise. A protection area is set between the detection area and the reference areas to prevent the target signal energy from leaking into the reference area.

[0019] Based on the Gaussian distribution, the statistical characteristics of background noise are estimated according to the signal data in the reference area, and the noise estimation is dynamically adjusted according to real-time environmental changes to adapt to different noise levels.

[0020] Based on the noise estimate and the predetermined false alarm rate, the detection threshold is calculated, and the range of the protection zone is dynamically adjusted according to the signal strength in the detection area to avoid interference with the target signal.

[0021] The signal in the detection area is compared with the calculated threshold. If the signal strength in the detection area is higher than the threshold, it is determined to be the target signal; otherwise, it is determined to be noise or clutter.

[0022] The detection results are fed back to the constant false alarm rate (CFAR) in real time to update noise estimation and threshold calculation. The detection threshold is dynamically adjusted according to real-time data and environmental changes to ensure stable detection performance under different clutter environments, thereby determining the target data of the fluid radar signal.

[0023] Preferably, the data acquisition window is compensated and adjusted based on real-time data from the fluid radar signal, including:

[0024] Extract real-time data of fluid radar signals after steady-state filtering;

[0025] By comparing the real-time data of the fluid radar signal after steady-state filtering with the real-time data of the fluid radar signal before steady-state filtering, the signal-to-noise ratio corresponding to the real-time data of the fluid radar signal is obtained.

[0026] The signal-to-noise ratio corresponding to the real-time data of the fluid radar signal is compared with a preset signal-to-noise ratio threshold.

[0027] When the signal-to-noise ratio (SNR) of the real-time fluid radar signal data is lower than a preset SNR threshold, the real-time fluid radar signal data with an SNR lower than the preset SNR threshold is retrieved as the target real-time data.

[0028] For each target's real-time data, obtain the signal vector norm corresponding to the target's real-time data;

[0029] The standard deviation of the signal vector norm is obtained by using the signal vector norm corresponding to the real-time data of each target.

[0030] The standard deviation of the signal vector norm is compared with a preset standard deviation threshold.

[0031] When the standard deviation of the signal vector norm exceeds the preset standard deviation threshold, the data acquisition window is compensated and adjusted according to the real-time data of the current fluid radar signal.

[0032] Preferably, when the standard deviation of the signal vector norm exceeds a preset standard deviation threshold, the data acquisition window is compensated and adjusted based on the current real-time data of the fluid radar signal, including:

[0033] When the standard deviation of the signal vector norm exceeds a preset standard deviation threshold, the percentage of the signal vector norm standard deviation exceeding the preset standard deviation threshold is obtained.

[0034] Retrieve the signal-to-noise ratio corresponding to the real-time data of each target;

[0035] The percentage of targets whose real-time data signal-to-noise ratio is lower than the preset signal-to-noise ratio threshold is obtained by comparing the signal-to-noise ratio of each target with the preset signal-to-noise ratio threshold.

[0036] The average ratio of the signal-to-noise ratio (SNR) of all real-time data corresponding to the targets that is lower than the preset SNR threshold is obtained by using the ratio of the SNR of each target real-time data corresponding to the targets that is lower than the preset SNR threshold.

[0037] The data acquisition window is adjusted by combining the excess ratio with the average ratio to obtain the adjusted data acquisition window.

[0038] Preferably, a steady-state filtered flow velocity and direction detection model for CFAR is established, including:

[0039] Historical data of fluid radar signals were collected and divided into training and testing sets.

[0040] Based on deep learning technology, a training set is used to train the deep learning model, enabling the deep learning model to autonomously learn the CFAR steady-state filtering velocity and direction detection behavior from the training set, and predict the velocity and direction of the fluid, thus determining the deep learning-based CFAR steady-state filtering velocity and direction detection model.

[0041] The deep learning-based CFAR steady-state filtering flow velocity and direction detection model was tested using a test set to evaluate its performance and determine the optimal CFAR steady-state filtering flow velocity and direction detection model.

[0042] Preferably, the optimal CFA steady-state filtered velocity and direction detection model includes:

[0043] The test set is input into the deep learning-based CFAR steady-state filtering velocity and direction detection model;

[0044] The deep learning-based CFAR steady-state filtering velocity and direction detection model was tested on the test set to evaluate its performance and determine whether it could achieve the expected effect of predicting the velocity and direction of fluids.

[0045] When the deep learning-based CFAR steady-state filtering velocity and direction detection model fails to achieve the expected effect of predicting the velocity and direction of the fluid, the parameters of the deep learning-based CFAR steady-state filtering velocity and direction detection model are continuously adjusted and optimized until the deep learning-based CFAR steady-state filtering velocity and direction detection model can achieve the expected effect of predicting the velocity and direction of the fluid, thereby determining the optimal CFAR steady-state filtering velocity and direction detection model.

[0046] Preferably, the flow velocity and direction of the fluid are predicted to determine the CFAR steady-state filtered flow velocity and direction detection results, including:

[0047] Obtain the optimal CFAR steady-state filtered flow velocity and direction detection model, and deploy the optimal CFAR steady-state filtered flow velocity and direction detection model in the actual CFAR steady-state filtered flow velocity and direction detection environment;

[0048] The fluid radar signal target data is input into the CFAR steady-state filtered velocity and direction detection model. Based on the CFAR steady-state filtered velocity and direction detection model, the fluid radar signal target data is analyzed and identified, and the fluid velocity and direction are predicted to determine the CFAR steady-state filtered velocity and direction detection results.

[0049] Preferably, the steady-state filtered flow velocity and direction detection results of CFAR are visualized, and the fluid is controlled in a timely manner based on these results, including:

[0050] The system acquires the steady-state filtered flow velocity and direction detection results of CFAR and displays them to users in real time in a visual format. When there are abnormalities in the flow velocity and direction of the fluid, it provides timely warnings, enabling users to manage the fluid in a timely manner based on the CFAR steady-state filtered flow velocity and direction detection results.

[0051] Compared with the prior art, the beneficial effects of the present invention are:

[0052] This invention acquires real-time fluid radar signal data using a radar sensor, performs steady-state filtering on the real-time fluid radar signal data using Kalman filtering, and performs target detection on the real-time fluid radar signal data using a constant false alarm rate (CFAR) to determine the target data of the fluid radar signal. It then analyzes and identifies the fluid radar signal target data by establishing a CFAR steady-state filtering velocity and direction detection model, predicts the fluid velocity and direction, determines the CFAR steady-state filtering velocity and direction detection results, and visualizes the CFAR steady-state filtering velocity and direction detection results. When abnormalities in the fluid velocity and direction are detected, timely warnings are issued, allowing users to manage the fluid in a timely manner based on the CFAR steady-state filtering velocity and direction detection results. This invention enables effective detection of fluid velocity and direction using CFAR steady-state filtering, providing reliable velocity and direction detection results under various environmental conditions, thus improving the accuracy and effectiveness of velocity and direction detection. Attached Figure Description

[0053] Figure 1 This is a flowchart of the CFA steady-state filtering flow velocity and direction detection method of the present invention;

[0054] Figure 2 This is a flowchart for evaluating the performance of the deep learning-based CFAR steady-state filtering flow velocity and direction detection model of the present invention. Detailed Implementation

[0055] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0056] To address the problem that existing methods cannot effectively detect fluid velocity and direction using CFA steady-state filtering, resulting in unreliable velocity and direction detection under various environmental conditions and consequently low accuracy and poor performance, please refer to [the relevant documentation / reference]. Figures 1-2 This embodiment provides the following technical solution:

[0057] Example 1

[0058] A CFA steady-state filtered flow velocity and direction detection method includes:

[0059] Collect and process real-time data of fluid radar signals to determine target data of fluid radar signals;

[0060] In this embodiment, real-time data acquisition of fluid radar signals includes:

[0061] The radar sensor is installed next to the location of the fluid to be detected, in an open and unobstructed location.

[0062] Real-time monitoring of the fluid under test is based on radar sensors. The radar sensor emits microwave or millimeter wave signals that pass through the fluid and interact with scattering objects in the fluid. The radar sensor receives the radar waves reflected back by the scattering objects in the fluid and performs frequency analysis on the received radar signals to extract the frequency information of the scattering objects and determine the real-time data of the fluid radar signals.

[0063] In this embodiment, the real-time data of the fluid radar signal is processed, including:

[0064] Based on Kalman filtering, steady-state filtering is performed on real-time fluid radar signal data to filter out specific frequency bands in the real-time fluid radar signal data that are not valuable for the detection of flow velocity and direction in CFA steady-state filtering, thereby eliminating noise and interference in the real-time fluid radar signal data.

[0065] In this embodiment, the real-time processing of fluid radar signal data further includes:

[0066] Target detection is performed on real-time data of fluid radar signals based on constant false alarm rate (CFAR).

[0067] Specifically, the window is selected based on the real-time data of the fluid radar signal. A detection area is selected in the real-time data of the fluid radar signal to determine whether there is a target signal. Multiple reference areas are selected around the detection area to estimate the statistical characteristics of the background noise. A protection area is set between the detection area and the reference areas to prevent the target signal energy from leaking into the reference area.

[0068] Based on the Gaussian distribution, the statistical characteristics of background noise are estimated according to the signal data in the reference area, and the noise estimation is dynamically adjusted according to real-time environmental changes to adapt to different noise levels.

[0069] Based on the noise estimate and the predetermined false alarm rate, the detection threshold is calculated, and the range of the protection zone is dynamically adjusted according to the signal strength in the detection area to avoid interference with the target signal.

[0070] The signal in the detection area is compared with the calculated threshold. If the signal strength in the detection area is higher than the threshold, it is determined to be the target signal; otherwise, it is determined to be noise or clutter.

[0071] The detection results are fed back to the constant false alarm rate (CFAR) in real time to update noise estimation and threshold calculation. The detection threshold is dynamically adjusted according to real-time data and environmental changes to ensure stable detection performance under different clutter environments, thereby determining the target data of the fluid radar signal.

[0072] Specifically, constant false alarm rate (CFAR) is an adaptive detection technique that maximizes the target detection probability while maintaining a constant false alarm rate by dynamically adjusting the detection threshold. It utilizes the characteristics of background noise and clutter to adaptively calculate the detection threshold, thereby achieving robust signal detection in complex environments.

[0073] It should be noted that by using a constant false alarm rate (CFAR) to perform target detection on real-time fluid radar signal data, and by analyzing background clutter, the detection threshold is dynamically adjusted according to the environmental noise level to ensure accurate target detection even in complex environments. It can accurately detect the speed and direction of fluids while maintaining a constant false alarm rate, and can be used for monitoring flow velocity and direction in complex water bodies such as rivers and oceans.

[0074] A steady-state filtered flow velocity and direction detection model for CFAR is established to analyze and identify target data of fluid radar signals, predict the flow velocity and direction of the fluid, and determine the CFAR steady-state filtered flow velocity and direction detection results.

[0075] In this embodiment, a CFAR steady-state filtered velocity and direction detection model is established, including:

[0076] Historical data of fluid radar signals were collected and divided into training and testing sets.

[0077] Based on deep learning technology, a training set is used to train the deep learning model, enabling the deep learning model to autonomously learn the CFAR steady-state filtering velocity and direction detection behavior from the training set, and predict the velocity and direction of the fluid, thus determining the deep learning-based CFAR steady-state filtering velocity and direction detection model.

[0078] The deep learning-based CFAR steady-state filtering flow velocity and direction detection model was tested using a test set to evaluate its performance and determine the optimal CFAR steady-state filtering flow velocity and direction detection model.

[0079] In this embodiment, determining the optimal CFAR steady-state filtered velocity and direction detection model includes:

[0080] The test set is input into the deep learning-based CFAR steady-state filtering velocity and direction detection model;

[0081] The deep learning-based CFAR steady-state filtering velocity and direction detection model was tested on the test set to evaluate its performance and determine whether it could achieve the expected effect of predicting the velocity and direction of fluids.

[0082] When the deep learning-based CFAR steady-state filtering velocity and direction detection model fails to achieve the expected effect of predicting the velocity and direction of the fluid, the parameters of the deep learning-based CFAR steady-state filtering velocity and direction detection model are continuously adjusted and optimized until the deep learning-based CFAR steady-state filtering velocity and direction detection model can achieve the expected effect of predicting the velocity and direction of the fluid, thereby determining the optimal CFAR steady-state filtering velocity and direction detection model.

[0083] In this embodiment, the flow velocity and direction of the fluid are predicted to determine the CFAR steady-state filtered flow velocity and direction detection results, including:

[0084] Obtain the optimal CFAR steady-state filtered flow velocity and direction detection model, and deploy the optimal CFAR steady-state filtered flow velocity and direction detection model in the actual CFAR steady-state filtered flow velocity and direction detection environment;

[0085] The fluid radar signal target data is input into the CFAR steady-state filtered velocity and direction detection model. Based on the CFAR steady-state filtered velocity and direction detection model, the fluid radar signal target data is analyzed and identified, and the fluid velocity and direction are predicted to determine the CFAR steady-state filtered velocity and direction detection results.

[0086] It should be noted that by establishing a CFAR steady-state filtered flow velocity and direction detection model to analyze and identify fluid radar signal target data, and predicting the flow velocity and direction of the fluid, the CFAR steady-state filtered flow velocity and direction detection results can be determined, which can facilitate users to manage the fluid in a timely manner based on the CFAR steady-state filtered flow velocity and direction detection results.

[0087] Example 2

[0088] The data acquisition window is compensated and adjusted based on real-time fluid radar signal data, including:

[0089] Extract real-time data of fluid radar signals after steady-state filtering;

[0090] By comparing the real-time data of the fluid radar signal after steady-state filtering with the real-time data of the fluid radar signal before steady-state filtering, the signal-to-noise ratio corresponding to the real-time data of the fluid radar signal is obtained.

[0091] The signal-to-noise ratio corresponding to the real-time data of the fluid radar signal is compared with a preset signal-to-noise ratio threshold.

[0092] When the signal-to-noise ratio (SNR) of the real-time fluid radar signal data is lower than a preset SNR threshold, the real-time fluid radar signal data with an SNR lower than the preset SNR threshold is retrieved as the target real-time data.

[0093] For each target's real-time data, obtain the signal vector norm corresponding to the target's real-time data;

[0094] The standard deviation of the signal vector norm is obtained by using the signal vector norm corresponding to the real-time data of each target.

[0095] The standard deviation of the signal vector norm is compared with a preset standard deviation threshold.

[0096] When the standard deviation of the signal vector norm exceeds the preset standard deviation threshold, the data acquisition window is compensated and adjusted according to the real-time data of the current fluid radar signal.

[0097] The technical effects of the above solution are as follows: By extracting and comparing real-time fluid radar signal data before and after steady-state filtering, the signal-to-noise ratio (SNR) can be effectively evaluated. Comparing the SNR with a preset threshold filters out target real-time data with low SNR, avoiding measurement errors caused by poor signal quality and improving data reliability and accuracy. Calculating the signal vector norm and standard deviation of the target's real-time data further characterizes the signal. When the standard deviation of the signal vector norm exceeds a preset threshold, compensation adjustments are made to identify and handle abnormal signal fluctuations, making the acquired data more stable and accurate, and reducing the impact of noise interference. Based on the SNR and the standard deviation of the signal vector norm, this solution compensates and adjusts the data acquisition window to adapt to different working scenarios and signal environment changes. In complex situations such as signal interference, the acquisition window can be adjusted promptly to ensure the system continuously and stably acquires high-quality data, enhancing the system's adaptability to complex working conditions. Through the above evaluation and processing of signal quality, measurement errors caused by signal problems can be specifically addressed, making fluid radar signal measurements more accurate, ensuring the accuracy of flow velocity, flow rate, and other measurement results, and improving the overall performance of the fluid radar measurement system.

[0098] Specifically, when the standard deviation of the signal vector norm exceeds a preset standard deviation threshold, the data acquisition window is compensated and adjusted based on the current real-time data of the fluid radar signal, including:

[0099] When the standard deviation of the signal vector norm exceeds a preset standard deviation threshold, the percentage of the signal vector norm standard deviation exceeding the preset standard deviation threshold is obtained.

[0100] Retrieve the signal-to-noise ratio corresponding to the real-time data of each target;

[0101] The percentage of targets whose real-time data signal-to-noise ratio is lower than the preset signal-to-noise ratio threshold is obtained by comparing the signal-to-noise ratio of each target with the preset signal-to-noise ratio threshold.

[0102] The average ratio of the signal-to-noise ratio (SNR) of all real-time data corresponding to the targets that is lower than the preset SNR threshold is obtained by using the ratio of the SNR of each target real-time data corresponding to the targets that is lower than the preset SNR threshold.

[0103] The data acquisition window is adjusted by combining the excess ratio with the average ratio to obtain the adjusted data acquisition window.

[0104] The adjusted data acquisition window is obtained using the following formula:

[0105]

[0106] Where E represents the adjusted data acquisition window; E0 represents the original data acquisition window; ||S|| b η represents the percentage of the signal vector norm standard deviation that exceeds the standard deviation threshold; η represents the spread gain coefficient, ranging from 0.12 to 0.83; b represents the transition smoothing factor, ranging from 0.5 to 1.2; P represents the average percentage of all target real-time data whose signal-to-noise ratio (SNR) is lower than a preset SNR threshold. It is an exponentially decaying function, as... It decreases as it increases. This transforms the signal into a function ranging from 0 to 1, used to quantify the adjustment ratio of the acquisition window under the combined influence of signal fluctuation and signal quality. The closer the value is to 1, the larger the window adjustment range; the closer it is to 0, the smaller the adjustment range. This formula can adaptively adjust the data acquisition window size based on two key indicators: the ratio of the standard deviation of the signal vector norm exceeding the threshold and the average ratio of the signal-to-noise ratio falling below the threshold. When the signal fluctuation is large (||S|| b When the signal strength (P) is high and the signal-to-noise ratio (SNR) is low, the acquisition window is increased to obtain more data, thereby improving signal quality and measurement accuracy. Conversely, when the signal condition is good, the window size is appropriately reduced or maintained to avoid resource waste and enable the system to adapt to different signal environments. The smoothing factor b in the transition region ensures the smoothness of the data acquisition window adjustment process. This avoids excessively large or frequent window adjustments due to small signal fluctuations or slight changes in SNR, enhancing system stability and reducing measurement errors or system instability caused by sudden window changes. By combining the extended gain coefficient η with a calculation method that comprehensively considers signal fluctuations and SNR, precise control of the data acquisition window can be achieved. Under different signal conditions, a suitable window adjustment range can be determined based on quantitative indicators, improving the quality of acquired data and thus enhancing the accuracy and reliability of subsequent analysis and measurement based on this data.

[0107] The technical effects of the above solution are as follows: By calculating the average ratio of signal vector norm standard deviation exceeding the threshold and the average ratio of signal-to-noise ratio below the threshold, quantification is performed from two dimensions: signal fluctuation and signal quality. Based on these two quantified indicators, the data acquisition window is compensated and adjusted, enabling more precise optimization of signal problems, reducing errors caused by noise interference and signal fluctuation, thereby improving the accuracy of the acquired data. The data acquisition window is dynamically adjusted according to the actual signal conditions under different operating conditions (such as the degree of signal fluctuation and signal-to-noise ratio), allowing the system to adapt to signal changes. Compared to a fixed-parameter acquisition window, this dynamic adjustment method better matches actual signal requirements, ensuring the consistency between the acquired data and the real signal, further improving data accuracy. This technical solution establishes an adaptive feedback mechanism; when the signal vector norm standard deviation and signal-to-noise ratio are abnormal, the system can automatically adjust the data acquisition window based on the quantified indicators. This mechanism can effectively cope with sudden signal changes or environmental interference, reducing system measurement fluctuations or errors caused by signal problems and enhancing system stability. By comprehensively considering signal fluctuation and signal-to-noise ratio, it can more comprehensively address various complex signal scenarios. Whether the signal-to-noise ratio decreases due to external interference or the signal itself fluctuates abnormally, the system can compensate by adjusting the acquisition window, improving its anti-interference capability and robustness in complex environments. Evaluating from two important dimensions—signal fluctuation and signal quality—and adjusting the acquisition window accordingly avoids the limitations of single-index evaluation. This multi-dimensional comprehensive evaluation and optimization more reliably ensures the validity of the acquired data, thereby improving the reliability of subsequent measurements and analyses based on this data, making the measurement results more trustworthy.

[0108] Example 3

[0109] The system visualizes the flow velocity and direction detection results of CFAR steady-state filtering and enables timely control of the fluid based on these results.

[0110] In this embodiment, the steady-state filtered flow velocity and direction detection results of CFAR are visualized, and the fluid is controlled in a timely manner based on the CFAR steady-state filtered flow velocity and direction detection results, including:

[0111] The system acquires the steady-state filtered flow velocity and direction detection results of CFAR and displays them to users in real time in a visual format. When there are abnormalities in the flow velocity and direction of the fluid, it provides timely warnings, enabling users to manage the fluid in a timely manner based on the CFAR steady-state filtered flow velocity and direction detection results.

[0112] In summary, this method involves acquiring real-time fluid radar signal data via radar sensors, performing steady-state filtering on the data using Kalman filtering, and then using a constant false alarm rate (CFAR) model to detect targets within the data. A CFAR steady-state filtering velocity and direction detection model is established to analyze and identify these targets, predicting the fluid's velocity and direction. The CFAR steady-state filtering velocity and direction detection results are then visualized and displayed. When abnormalities in the fluid's velocity and direction are detected, timely warnings are provided, allowing users to manage the fluid effectively based on the CFAR steady-state filtering results. This method enables effective detection of fluid velocity and direction under various environmental conditions, improving both the accuracy and effectiveness of velocity and direction detection.

[0113] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0114] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A CFAR steady-state filtering flow rate flow direction detection method, characterized by, The method comprises the following steps: Collecting fluid radar signal real-time data and processing the same to determine fluid radar signal target data; Processing fluid radar signal real-time data, comprising: Performing target detection on fluid radar signal real-time data based on constant false alarm rate (CFAR); Wherein, window selection is performed according to fluid radar signal real-time data, a detection region is selected in the fluid radar signal real-time data for judging whether a target signal exists, a plurality of reference regions are selected around the detection region for estimating statistical characteristics of background noise, and a guard region is set between the detection region and the reference regions for avoiding target signal energy leakage into the reference regions; Based on Gaussian distribution, the statistical characteristics of the background noise are estimated according to the signal data in the reference regions, and the noise estimation is dynamically adjusted according to real-time environmental changes for adapting to different noise levels; According to the noise estimation value and a predetermined false alarm rate, a detection threshold is calculated, and the range of the guard region is dynamically adjusted according to the signal strength of the detection region for avoiding interference of the target signal; The signal in the detection region is compared with the calculated threshold, and if the signal strength of the detection region is higher than the threshold, it is determined as a target signal; otherwise, it is determined as noise or clutter; The detection result is fed back to the constant false alarm rate (CFAR) in real time for updating noise estimation and threshold calculation, and the detection threshold is dynamically adjusted according to real-time data and environmental changes to ensure stable detection performance in different clutter environments, thereby determining fluid radar signal target data; Wherein, when the signal vector norm standard deviation exceeds a preset standard deviation threshold, the data collection window is compensated and adjusted according to the current fluid radar signal real-time data; A CFAR steady-state filtering flow rate and flow direction detection model is established to analyze and identify the fluid radar signal target data and predict the flow rate and flow direction of the fluid, and a CFAR steady-state filtering flow rate and flow direction detection result is determined. The CFAR steady-state filtering flow rate and flow direction detection result is visualized and displayed, and the fluid is timely controlled according to the CFAR steady-state filtering flow rate and flow direction detection result.

2. A CAF steady state filtering flow rate flow direction detection method as claimed in claim 1, wherein, Collecting fluid radar signal real-time data, comprising: The radar sensor is installed beside the position of the fluid to be detected, and the installation position is open and unobstructed around; The radar sensor transmits microwave or millimeter wave signals through the fluid and interacts with the scattering objects in the fluid, and the radar waves reflected by the scattering objects in the fluid are received by the radar sensor, and the frequency of the received radar signals is analyzed to extract the frequency information of the scattering objects, and the fluid radar signal real-time data is determined.

3. A CAF steady state filtering flow rate flow direction detection method as claimed in claim 1, wherein, Processing fluid radar signal real-time data, comprising: Based on Kalman filtering, the fluid radar signal real-time data is steady-state filtered to filter out specific band frequencies in the fluid radar signal real-time data that are worthless for CFAR steady-state filtering flow rate and flow direction detection, and to eliminate noise and interference in the fluid radar signal real-time data.

4. A CAF steady state filtering flow rate flow direction detection method as claimed in claim 3, wherein, According to the fluid radar signal real-time data, the data collection window is compensated and adjusted, comprising: Extracting the fluid radar signal real-time data after steady-state filtering processing; The fluid radar signal real-time data after the steady-state filtering processing is compared with the fluid radar signal real-time data before the steady-state filtering processing, and a signal-to-noise ratio corresponding to the fluid radar signal real-time data is obtained; The signal-to-noise ratio corresponding to the fluid radar signal real-time data is compared with a preset signal-to-noise ratio threshold value; When the signal-to-noise ratio corresponding to the fluid radar signal real-time data is lower than the preset signal-to-noise ratio threshold value, fluid radar signal real-time data corresponding to the signal-to-noise ratio lower than the preset signal-to-noise ratio threshold value is called as target real-time data; A signal vector norm corresponding to each target real-time data is obtained; A signal vector norm standard deviation is obtained by using the signal vector norm corresponding to each target real-time data; The signal vector norm standard deviation is compared with a preset standard deviation threshold value; When the signal vector norm standard deviation exceeds the preset standard deviation threshold value, a data acquisition window is compensated and adjusted according to the current fluid radar signal real-time data.

5. A CFAR steady state filtering flow rate flow direction detection method as claimed in claim 4, wherein, When the signal vector norm standard deviation exceeds the preset standard deviation threshold value, a data acquisition window is compensated and adjusted according to the current fluid radar signal real-time data, including: When the signal vector norm standard deviation exceeds the preset standard deviation threshold value, an exceeding ratio of the signal vector norm standard deviation exceeding the preset standard deviation threshold value is obtained by using the signal vector norm standard deviation exceeding the preset standard deviation threshold value; The signal-to-noise ratio corresponding to each target real-time data is called; A ratio of the signal-to-noise ratio lower than the preset signal-to-noise ratio threshold value corresponding to each target real-time data is obtained by using the signal-to-noise ratio corresponding to each target real-time data and the preset signal-to-noise ratio threshold value; A ratio average value of the signal-to-noise ratio lower than the preset signal-to-noise ratio threshold value corresponding to all target real-time data is obtained by using the ratio of the signal-to-noise ratio lower than the preset signal-to-noise ratio threshold value corresponding to each target real-time data; The data acquisition window is compensated and adjusted by using the exceeding ratio and the ratio average value, and an adjusted data acquisition window is obtained.

6. A CFAR steady state filtering flow rate flow direction detection method as claimed in claim 1, wherein, A cfar steady-state filtering flow rate and flow direction detection model is established, including: Fluid radar signal historical data is collected, and the collected fluid radar signal historical data is divided into a training set and a test set; Based on a deep learning technology, the training set is used to train a deep learning model, so that the deep learning model learns a cfar steady-state filtering flow rate and flow direction detection behavior from the training set and predicts the flow rate and flow direction of the fluid, to determine a cfar steady-state filtering flow rate and flow direction detection model based on deep learning; The cfar steady-state filtering flow rate and flow direction detection model based on deep learning is tested by using the test set, the performance of the cfar steady-state filtering flow rate and flow direction detection model based on deep learning is evaluated, and thus an optimal cfar steady-state filtering flow rate and flow direction detection model is determined.

7. A CFAR steady state filtering flow rate flow direction detection method as claimed in claim 6, wherein, An optimal cfar steady-state filtering flow rate and flow direction detection model is determined, including: The test set is input into the cfar steady-state filtering flow rate and flow direction detection model based on deep learning; The cfar steady-state filtering flow rate and flow direction detection model based on deep learning is tested based on the test set, the performance of the cfar steady-state filtering flow rate and flow direction detection model based on deep learning is evaluated, and whether the cfar steady-state filtering flow rate and flow direction detection model based on deep learning can achieve the expected effect of predicting the flow rate and flow direction of the fluid is determined. When the cfar steady-state filtering flow rate and flow direction detection model based on deep learning cannot achieve the expected effect of predicting the flow rate and flow direction of the fluid, the parameters of the cfar steady-state filtering flow rate and flow direction detection model based on deep learning are continuously adjusted and optimized until the cfar steady-state filtering flow rate and flow direction detection model based on deep learning can achieve the expected effect of predicting the flow rate and flow direction of the fluid, so as to determine the optimal cfar steady-state filtering flow rate and flow direction detection model.

8. A CFAR steady state filtering flow rate flow direction detection method as claimed in claim 7, wherein, The flow rate and flow direction of the fluid are predicted, and the cfar steady-state filtering flow rate and flow direction detection result is determined, including: The optimal cfar steady-state filtering flow rate and flow direction detection model is obtained, and the optimal cfar steady-state filtering flow rate and flow direction detection model is deployed in the actual cfar steady-state filtering flow rate and flow direction detection environment; The fluid radar signal target data is input into the cfar steady-state filtering flow rate and flow direction detection model, the fluid radar signal target data is analyzed and recognized according to the cfar steady-state filtering flow rate and flow direction detection model, and the flow rate and flow direction of the fluid are predicted to determine the cfar steady-state filtering flow rate and flow direction detection result.

9. The CFAR steady-state filtered flow rate flow direction detection method of claim 1 wherein, The cfar steady-state filtering flow rate and flow direction detection result is visualized and displayed, and the fluid is timely controlled according to the cfar steady-state filtering flow rate and flow direction detection result, including: The cfar steady-state filtering flow rate and flow direction detection result is obtained, and the cfar steady-state filtering flow rate and flow direction detection result is displayed to the user in a visual form in real time, when the flow rate and flow direction of the fluid exist abnormal conditions, timely warning is carried out, so that the user can timely control the fluid according to the cfar steady-state filtering flow rate and flow direction detection result.

Images