River state detection method and river state detection device

Abstract

The present application relates to a river state detection method and a river state detection device, wherein the river state detection method comprises: obtaining measurement data of a target river based on a millimeter wave radar; performing feature conversion processing on the measurement data to obtain a corresponding feature vector; processing the feature vector based on a pre-built flow estimation model to obtain a flow estimation value of the target river; and processing the feature vector based on a pre-built flow velocity estimation model to obtain a flow velocity estimation value of the target river. It can eliminate the dependence on lighting conditions based on millimeter wave radar, thereby realizing accurate river state monitoring around the clock.

Description

Technical Field

[0001] The present application relates to the field of radar data processing, and in particular to a river state detection method and a river state detection device. Background Art

[0002] Real-time monitoring of river conditions plays an important role in preventing some geological disasters. For example, in the rainy season, continuous heavy rainfall is prone to occur, which in turn triggers geological disasters such as floods, mountain torrents, and mudslides. In the early stages of floods or mountain torrents, the river flow rate often increases and the flow rate surges.

[0003] Currently, visible light cameras are usually used as surveillance equipment to monitor target scenes. Heavy rainfall can occur at any time and in any weather, and visible light cameras rely on lighting conditions and cannot work effectively in dim or no lighting conditions. Therefore, there is still a problem that it is impossible to achieve river monitoring around the clock.

[0004] There is no effective solution to the problem that related technologies cannot achieve river monitoring around the clock. Summary of the invention

[0005] In this embodiment, a river state detection method and a river state detection device are provided to solve the problem in the related art that it is impossible to implement river monitoring around the clock.

[0006] In a first aspect, a river state detection method is provided in this embodiment, including:

[0007] Based on millimeter wave radar, obtain measurement data of target rivers;

[0008] Performing feature conversion processing on the measurement data to obtain a corresponding feature vector;

[0009] Based on a pre-constructed flow estimation model, the feature vector is processed to obtain a flow estimation value of the target river; and based on a pre-constructed flow velocity estimation model, the feature vector is processed to obtain a flow velocity estimation value of the target river.

[0010] In some embodiments, after obtaining the estimated value of the flow rate of the target river and the estimated value of the flow velocity of the target river, the method further includes:

[0011] Determine flow change information in the time dimension according to the estimated flow values ​​at different times; determine whether a preset warning event occurs in the target river according to the flow change information;

[0012] and / or,

[0013] According to the estimated flow velocity values ​​at different times, the flow velocity change information in the time dimension is determined; according to the flow velocity change information, it is determined whether a preset warning event occurs in the target river.

[0014] In some embodiments, the measurement data is subjected to feature conversion processing to obtain a corresponding feature vector, including:

[0015] Performing clustering processing on the measurement data to obtain cluster data;

[0016] According to the clustering data, the speed feature, the intensity feature and the clustering convex hull feature of the target river are determined to form a corresponding feature vector.

[0017] In some embodiments, clustering the measurement data to obtain cluster data includes:

[0018] According to the direction represented by the measurement data, the measurement data is divided into outgoing measurement data and incoming measurement data;

[0019] According to the river flow direction corresponding to the target river, retaining the measurement data consistent with the river flow direction from the outgoing measurement data and the incoming measurement data as the target measurement data;

[0020] Clustering is performed on the target measurement data to obtain clustered data.

[0021] In some of the embodiments, the characteristic vector is processed based on a pre-built flow estimation model to obtain the flow estimation value of the target river, including:

[0022] According to the system estimation of the pre-built flow estimation model, the feature vector is expanded to obtain a target vector;

[0023] The flow weight vector corresponding to the target vector in the flow estimation model is multiplied by the target vector to obtain the flow estimation value of the target river.

[0024] In some of the embodiments, the feature vector is processed based on a pre-built velocity estimation model to obtain the velocity estimation value of the target river, including:

[0025] According to the system estimation of the pre-built velocity estimation model, the feature vector is expanded to obtain a target vector;

[0026] The velocity weight vector corresponding to the target vector in the velocity estimation model is multiplied by the target vector to obtain an estimated value of the velocity of the target river.

[0027] In some of the embodiments, before obtaining the measurement data of the target river based on the millimeter wave radar, the method further includes:

[0028] Initialize and set various radar detection parameters; the radar detection parameters are at least used to estimate the flow velocity and flow rate of the target river.

[0029] In some embodiments, the method further comprises:

[0030] Acquiring image information of the target river;

[0031] According to the image information, the various radar detection parameters are corrected.

[0032] In some embodiments, the method further comprises:

[0033] Acquire measured and marked data of the target river;

[0034] According to the measured annotation data, the various radar detection parameters are corrected.

[0035] In a second aspect, a river state detection device is provided in this embodiment, including: an acquisition module, a feature conversion module and a state estimation module; wherein:

[0036] The acquisition module is used to acquire measurement data of the target river based on the millimeter wave radar;

[0037] The feature conversion module is used to perform feature conversion processing on the measurement data to obtain a corresponding feature vector;

[0038] The state estimation module is used to process the feature vector based on a pre-constructed flow estimation model to obtain a flow estimation value of the target river; and to process the feature vector based on a pre-constructed flow velocity estimation model to obtain a flow velocity estimation value of the target river.

[0039] Compared with the related art, a river state detection method and a river state detection device are provided in this embodiment. The river state detection method obtains the measurement data of the target river based on the millimeter wave radar; performs feature conversion processing on the measurement data to obtain the corresponding feature vector; processes the feature vector based on the pre-built flow estimation model to obtain the flow estimation value of the target river; and processes the feature vector based on the pre-built flow velocity estimation model to obtain the flow velocity estimation value of the target river. It can eliminate the dependence on lighting conditions based on the millimeter wave radar, thereby realizing accurate river state monitoring all day and all night.

[0040] Details of one or more embodiments of the present application are set forth in the following drawings and description to make other features, objects, and advantages of the present application more readily apparent. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0042] Figure 1 is a hardware structure block diagram of a terminal of the river state detection method of this embodiment;

[0043] Figure 2 is a flow chart of the river state detection method of this embodiment;

[0044] Figure 3 This is a schematic diagram of a river flow surge phenomenon in this embodiment;

[0045] Figure 4 is a schematic diagram of the data structure of measurement data collected by a millimeter wave radar in this embodiment;

[0046] Figure 5 is a schematic diagram of drawing measurement data of this embodiment;

[0047] Figure 6 is a river flow diagram that changes with time in this embodiment;

[0048] Figure 7 is a schematic diagram of a clustering convex hull of this embodiment;

[0049] Figure 8a This is a schematic diagram of drawing vector measurement data in this embodiment;

[0050] Figure 8b This is a schematic diagram of drawing dequantized measurement data in this embodiment;

[0051] Fig. 9 is a flow chart of a river state detection method of some of the embodiments;

[0052] Fig.10 A method for detecting flow rate and velocity of a river in some of the embodiments;

[0053] Fig.11 It is a structural block diagram of the river state detection device of this embodiment. DETAILED DESCRIPTION

[0054] In order to more clearly understand the purpose, technical solutions and advantages of the present application, the present application is described and illustrated below in conjunction with the accompanying drawings and embodiments.

[0055] Unless otherwise defined, the technical terms or scientific terms involved in this application shall have the general meaning understood by people with general skills in the technical field to which this application belongs. The words "one", "a", "the", "these" and the like in this application do not indicate a quantitative limitation, and they may be singular or plural. The terms "include", "comprise", "have" and any variants thereof involved in this application are intended to cover non-exclusive inclusions; for example, a process, method and system, product or device comprising a series of steps or modules (units) is not limited to the listed steps or modules (units), but may include unlisted steps or modules (units), or may include other steps or modules (units) inherent to these processes, methods, products or devices. The words "connect", "connected", "coupled" and the like involved in this application are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. The "multiple" involved in this application refers to two or more. "And / or" describes the association relationship of associated objects, indicating that there may be three relationships, for example, "A and / or B" may mean: A exists alone, A and B exist at the same time, and B exists alone. Generally, the character " / " indicates that the objects associated with each other are in an "or" relationship. The terms "first", "second", "third", etc. in this application are only used to distinguish similar objects and do not represent a specific ordering of the objects.

[0056] The method embodiment provided in this embodiment can be executed in a terminal, a computer or a similar computing device. For example, running on a terminal, Figure 1 FIG. 1 is a hardware structure diagram of a terminal of the river state detection method of this embodiment. Figure 1 As shown, the terminal may include one or more ( Figure 1 Only one is shown in the figure) a processor 102 and a memory 104 for storing data, wherein the processor 102 may include but is not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA. The above terminal may also include a transmission device 106 and an input and output device 108 for communication functions. It can be understood by those skilled in the art that Figure 1 The structure shown is only for illustration and does not limit the structure of the above terminal. Figure 1 More or fewer components as shown, or with Figure 1 Different configurations shown.

[0057] The memory 104 can be used to store computer programs, for example, software programs and modules of application software, such as the computer program corresponding to the river state detection method in this embodiment. The processor 102 executes various functional applications and data processing by running the computer program stored in the memory 104, that is, to implement the above method. The memory 104 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 104 may further include a memory remotely arranged relative to the processor 102, and these remote memories may be connected to the terminal via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0058] The transmission device 106 is used to receive or send data via a network. The above network includes a wireless network provided by a communication provider of the terminal. In one example, the transmission device 106 includes a network adapter (Network Interface Controller, referred to as NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device 106 can be a radio frequency (Radio Frequency, referred to as RF) module, which is used to communicate with the Internet wirelessly.

[0059] In this embodiment, a method for detecting river status is provided. Figure 2 is a flow chart of the river state detection method of this embodiment. Figure 2 As shown, the process includes the following steps:

[0060] Step S210, obtaining measurement data of the target river based on the millimeter wave radar.

[0061] Millimeter wave radar is a radar system that uses radio waves in the millimeter wave frequency band to observe and detect objects. Millimeter wave radar has a strong ability to detect moving objects and can be used to detect slightly moving targets in real time, obtaining measurement information of moving targets with high detection probability and high frame rate. In addition, millimeter wave radar supports all-weather and all-day monitoring. Therefore, in this step, millimeter wave radar is used to monitor the target river.

[0062] Figure 3 FIG. 1 is a schematic diagram of a sudden increase in river flow in this embodiment. Figure 3 As shown, the sudden increase in river flow will cause the water level to rise, the river to widen, and the river's scouring effect on the river bank to increase, increasing the risk of river bank collapse.

[0063] It is worth noting that millimeter-wave radar has high sensitivity to moving targets and is suitable for detecting slightly moving targets. For rivers in mountainous areas and other areas, there are splashes and waves due to the influence of rocks, but the statistical characteristics of the measured data are stable and suitable for clustering. In addition, the river water is mixed with objects such as mud, branches, etc., and the movement trend of all impurities in the process of floods from far to near or from near to far is consistent, which is also suitable for clustering. When the river flow rate surges, the characteristic curve and the target intensity information of the corresponding area in a short period of time will change significantly. When the river flow rate increases slowly, the phenomenon of slow increase in flow rate can also be captured based on the detection of river flow, but in this case there is enough time to take emergency measures, so this embodiment pays more attention to river surge events. Based on the above content, this embodiment needs to capture the situation where the state of the target river changes sharply in a short period of time based on the millimeter-wave radar, and then determine whether a river surge event occurs.

[0064] Among them, the millimeter-wave radar can be fixed on a stable bracket so that the detection range of the millimeter-wave radar covers the water surface area of ​​the target river that needs to be monitored. The millimeter-wave radar can obtain the measurement information of each detected target, which may specifically include: distance, angle, radial speed (RadialSpeed), radar reflection area (RCS). The measurement frequency or measurement period of the millimeter-wave radar can also be set according to the needs of the actual application scenario. For example, the measurement period of the millimeter-wave radar can be set to 0.05 seconds, that is, the operating frequency is 20 Hz, which is not specifically limited in this embodiment.

[0065] Figure 4 FIG. 1 is a schematic diagram of the data structure of the measurement data collected by a millimeter wave radar in this embodiment. Figure 4 As shown, "Time" represents the frame number, and the number following "Time" represents the specific value of the actual frame number; "Num" means the number of measurements of the frame, and the number following "Num" represents the actual value of the measured number; time format strings such as "2022-03-31 11:24:48", "2022-03-31 11:24:49", and "2022-03-31 11:24:49" represent the measurement time of the corresponding frame. Figure 4 In the rest of the data (such as rows 2 to 7), the first column such as "0", "1", "2", "3", "4", "5" is the ID of the measurement data, the 2nd to 3rd columns are the X and Y coordinates (in meters), the 4th column is the RCS in db, which is the intensity unit, and the 5th column is the radial velocity in meters per second. Each value in the measurement data has a clear physical meaning, but it is not intuitive enough for the representation of the river state. The radar measurement data can be plotted. Figure 5 FIG. 1 is a schematic diagram of drawing measurement data of this embodiment. Figure 5 As shown in the figure, the horizontal axis is the relative speed between the target and the radar, and the vertical axis is the relative distance between the target and the radar. The solid points represent the target points close to the millimeter-wave radar (the corresponding speed is less than zero), the hollow points represent the target points far away from the millimeter-wave radar (the corresponding speed is greater than zero), and the larger black circular points represent the stationary target points. Figure 5 It can be seen that by plotting the measured data, a more intuitive representation of the river status can be formed.

[0066] Step S220, performing feature conversion processing on the measurement data to obtain a corresponding feature vector.

[0067] After the measurement data of the target river is obtained based on step S210, feature conversion processing can be performed on the measurement data to obtain a corresponding feature vector. For example, the feature vector can be constructed from dimensions such as intensity, speed, and clustering characteristics of the measurement data.

[0068] Step S230, based on the pre-constructed flow estimation model, the feature vector is processed to obtain the flow estimation value of the target river; and based on the pre-constructed flow velocity estimation model, the feature vector is processed to obtain the flow velocity estimation value of the target river.

[0069] Among them, the flow estimation model can be a pre-established water flow estimation function. The flow velocity estimation model can be a pre-established water flow velocity estimation function. That is, the flow estimation model can be an expression that characterizes the mathematical relationship between the above-mentioned characteristic vector and the water flow, and the flow velocity estimation model can be an expression that characterizes the mathematical relationship between the above-mentioned characteristic vector and the water flow velocity. The estimation parameters involved in the flow estimation model and the estimation parameters involved in the flow velocity estimation model can be obtained based on the measured data collected in advance, and continuously corrected during the river state detection process. In this way, the measured data can be converted into flow estimation values ​​and flow velocity estimation values ​​respectively, so as to achieve a clearer and more accurate representation of the river state.

[0070] It will be understood by those skilled in the art that after obtaining the estimated flow and velocity of the target river, an accurate characterization of the river state can be achieved, thereby enabling real-time monitoring of the river state based on the real-time estimated flow and velocity of the river, and timely detecting abnormalities in the river flow and velocity.

[0071] In the related art, the state of the river is monitored by visible light cameras. Since visible light cameras cannot work effectively in dim or no light conditions, it is impossible to achieve all-weather and all-day monitoring. Due to the randomness of heavy rainfall, it is impossible to conduct environmental monitoring or disaster monitoring all-weather and all-day, and it will be impossible to detect early warning events before the disaster occurs, which is not conducive to timely taking risk avoidance measures.

[0072] In this regard, this embodiment takes into account the sensitivity of millimeter-wave radar for detecting moving targets, as well as the characteristics of millimeter-wave radar that can adapt to all-weather and all-day monitoring, and monitors the state of the target river based on the millimeter-wave radar. Specifically, based on the measurement data collected by the millimeter-wave radar, the estimated flow rate and flow velocity of the target river are ultimately obtained. The river state detection implemented in this embodiment has a high degree of intelligence, does not require manual intervention, has strong scene adaptability, high accuracy, and can work around the clock, thereby achieving accurate and real-time monitoring of the state of the target river, and can promptly detect river warning events based on changes in flow rate and flow velocity, so as to take timely risk avoidance measures.

[0073] In the above steps S210 to S230, based on the millimeter wave radar, the measurement data of the target river is obtained; the measurement data is subjected to feature conversion processing to obtain the corresponding feature vector; based on the pre-built flow estimation model, the feature vector is processed to obtain the flow estimation value of the target river; and based on the pre-built flow velocity estimation model, the feature vector is processed to obtain the flow velocity estimation value of the target river. It can eliminate the dependence on lighting conditions based on the millimeter wave radar, thereby realizing accurate river status monitoring all day and all night.

[0074] In one embodiment, after obtaining the estimated flow rate of the target river and the estimated flow velocity of the target river, the above-mentioned river state detection method may further include:

[0075] Based on the estimated flow values ​​at different times, determine the flow change information in the time dimension; based on the flow change information, judge whether a preset warning event occurs in the target river; and / or, based on the estimated flow velocity values ​​at different times, determine the flow velocity change information in the time dimension; based on the flow velocity change information, judge whether a preset warning event occurs in the target river.

[0076] The warning event may be an event in which the flow rate or velocity changes drastically within a preset period of time, i.e., a river flood event. The flow change can be curve-fitted based on the two dimensions of time and flow, and an S-shaped curve based on time and flow can be obtained. The corresponding S-shaped curve function can be estimated to represent the flow change information. Figure 6 is a river flow diagram that changes with time in this embodiment. Figure 6 The horizontal axis is time; the vertical axis is the estimated flow value. Figure 6 The points in the graph represent the estimated flow values ​​at different times, and the curve represents the S-shaped curve of flow changing with time. The corresponding S-shaped curve function can be expressed as follows:

[0077]

[0078] Among them, exp represents exponential operation; a, b, c, d, and e represent the parameters to be estimated in the S-curve function respectively, and the parameters to be estimated can be estimated by nonlinear optimization method based on the estimated flow values ​​at different times. After obtaining the above-mentioned S-curve, the difference in river flow and the rate of change during the change period can be determined based on the S-curve to determine whether a river flood event has occurred. For example, based on the above-mentioned S-curve function of flow, it is determined that the flow difference within a preset time period exceeds a preset difference threshold, or the rate of change within a preset time period exceeds a preset rate threshold, then it is determined that a river flood event has occurred. Similarly, after obtaining the estimated flow rate, the S-curve function corresponding to the flow rate can also be obtained by referring to the determination process of the above-mentioned flow change information, and the river flood event detection can be performed based on this, which will not be repeated in this embodiment. In actual application scenarios, the S-curve function of flow can be mainly referred to to realize the early warning event monitoring of rivers.

[0079] This embodiment can realize the timely detection of river warning events based on flow change information and flow velocity change information. In practical applications, it is necessary to determine in real time whether a river flood event has occurred. If so, it is necessary to report the time of occurrence, the real-time flow and velocity of the river, and upload the reported data to the PTZ. The PTZ integrates more information (multiple locations, multiple millimeter-wave radars, and information generated by other monitoring equipment working simultaneously) to comprehensively determine whether a geological disaster event has occurred, determine the corresponding disaster level, and determine the downstream warning time, thereby helping relevant personnel make timely risk avoidance decisions.

[0080] In one embodiment, based on the above step S220, feature conversion processing is performed on the measurement data to obtain a corresponding feature vector, which may specifically include:

[0081] The measured data are clustered to obtain cluster data; based on the cluster data, the velocity characteristics, intensity characteristics and cluster convex hull characteristics of the target river are determined to form a corresponding feature vector.

[0082] In this embodiment, the composition expression of the characteristic vector can be pre-established, and the confidence interval can be constructed by realizing the statistics of the fluctuation range of the water flow characteristic vector based on the measurement data. The measurement data in the adjacent preset historical period is collected and substituted into the expression of the characteristic vector established above to calculate the corresponding characteristic value.

[0083] The measurement data may be clustered based on a preset clustering method. For example, a variable threshold clustering method may be used to cluster the measurement data to obtain clustered data. In addition, those skilled in the art may also select other clustering methods according to the needs of actual application scenarios, and this embodiment does not specifically limit this.

[0084] Taking the clustering method with variable threshold as an example, the clustering operation process is as follows:

[0085]

[0086]

[0087]

[0088]

[0089]

[0090] in, e i Indicates i The corresponding spatial position is ( x i , y i ), the speed is v i , rcs yes Rcs i ; e j represents the first j goals; dis ( e i , e j ) is the elliptical distance between the two targets; is the distance threshold, and Represent the speed difference threshold and rcs Difference threshold, and e i It is related to the spatial location or wind speed in the environment. N ( C ( k )) represents clustering C ( k ), N t Indicates the target number threshold for clustering.

[0091] Afterwards, the average speed, average intensity, intensity variance and cluster convex area of ​​the target river can be determined according to the clustering data to form the corresponding feature vector. Specifically:

[0092] m _ v = mean ( C ( k ) v )

[0093] m _ R = mean ( C ( k ) Rcs )

[0094] v _ R =var( C ( k ) Rcs )

[0095] Dx = D ( C ( k ))

[0096] in, m _ v Representation clustering C ( k ), m _ R Representation clustering C ( k ), v _ R Representation clustering C ( k ), Dx Representation clustering C ( k ) is the area of ​​the convex hull in . Figure 7 is a schematic diagram of a clustering convex hull of this embodiment, wherein: Figure 7 The outer circles in the figure represent the cluster convex hull vertices, the lines represent the cluster boundaries, and the inner circles represent the cluster centers.

[0097] Finally, the constructed feature vector is: V(k)=( m _ v , m _ R,v _ R, Dxy ).

[0098] This embodiment performs clustering processing on the measurement data and forms a feature vector based on the clustered data, thereby improving the accuracy and efficiency of feature conversion.

[0099] In one embodiment, clustering the measurement data to obtain cluster data may include:

[0100] According to the direction represented by the measurement data, the measurement data is divided into de-vectored measurement data and incoming measurement data; according to the river flow direction corresponding to the target river, the measurement data consistent with the river flow direction is retained from the de-vectored measurement data and the incoming measurement data as the target measurement data; the target measurement data is clustered to obtain clustered data.

[0101] Among them, the outgoing vector measurement data and the incoming vector measurement data relative to the millimeter wave radar can be distinguished according to the speed in the measurement data, and the measurement data can be grouped based on this. Figure 8a A schematic diagram of drawing vector measurement data according to this embodiment; Figure 8b FIG. 1 is a schematic diagram of drawing a method for dequantizing measurement data in this embodiment. Figure 8a and Figure 8b The horizontal axis is the relative speed between the target and the radar, and the vertical axis is the relative distance between the target and the radar.

[0102] Since the direction of the river is fixed, based on the predetermined flow direction of the target river, the measurement data that is different from the flow direction of the river can be filtered out from the de-vectored measurement data and the incoming measurement data, and only the measurement data that is consistent with the flow direction of the river is retained. Afterwards, the above clustering process is performed on the retained target measurement data to obtain the above clustered data.

[0103] This embodiment filters the measurement data according to the direction of the measurement data based on the river flow direction, and can filter out the measurement data representing weeds and trees on the waterside and ships moving in the opposite direction, thereby reducing the interference and influence of interference objects in the environmental area corresponding to the target river on the measurement results and improving the accuracy of subsequent river state estimation.

[0104] Additionally, in one embodiment, based on the above step S230, based on the pre-built flow estimation model, processing the feature vector to obtain the flow estimation value of the target river may include:

[0105] According to the system estimation of the pre-built flow estimation model, the characteristic vector is expanded to obtain the target vector; the flow weight vector corresponding to the target vector in the flow estimation model is multiplied by the target vector to obtain the flow estimation value of the target river.

[0106] The flow and velocity of a river are affected by the topographical factors of the area in which it is located. The system estimate corresponding to the river can be determined in advance based on the topographical conditions of the rivers in different areas and the corresponding measured data. The system estimate can be set manually based on the measured data of the target river, or the system estimate can be determined and corrected by taking a photo of the target river and analyzing the river state through the image. Afterwards, the feature vector is expanded according to the system estimate to obtain the target vector, which can be shown as follows:

[0107]

[0108] in S is the system estimate, m _ v Represents the above clustering C ( k ), m _ R Represents the above clustering C ( k ), v _ R Represents the above clustering C ( k ), Dx Represents the above clustering C ( k ) is the area of ​​the convex hull in .

[0109] Alternatively, the flow weight vector composed of the flow weights corresponding to the elements of the above target vectors, which are pre-set for the target river, can be obtained. β c , the traffic weight vector can also be set according to the measured data. Based on this, the above target vector and the traffic weight vector β c , determine the estimated flow of the target river Y c :

[0110]

[0111] in:

[0112]

[0113] In this embodiment, the feature vector is expanded and weighted according to the pre-built flow estimation model, and finally the accurate flow estimation of the target river is realized.

[0114] Accordingly, in one embodiment, based on the above step S230, based on the pre-built velocity estimation model, the feature vector is processed to obtain the velocity estimation value of the target river, which may specifically include:

[0115] According to the system estimation of the pre-built velocity estimation model, the characteristic vector is expanded to obtain the target vector; the velocity weight vector corresponding to the target vector in the velocity estimation model is multiplied by the target vector to obtain the velocity estimation value of the target river.

[0116] Among them, the velocity weight vectorβ v It can be expressed as:

[0117]

[0118] Among them, the velocity weight vector β v The elements in correspond to the flow velocity weights of the elements of the above target vector.

[0119] Afterwards, the estimated flow velocity can be solved Y v for:

[0120]

[0121] Based on this, this embodiment can achieve accurate estimation of the flow velocity of the target river through measurement data combined with a pre-built flow velocity estimation model.

[0122] In addition, in one embodiment, before obtaining the measurement data of the target river based on the millimeter wave radar, the above-mentioned river state detection method may further include:

[0123] Initialize and set various radar detection parameters; the radar detection parameters are at least used to estimate the flow velocity and flow rate of the target river.

[0124] Specifically, before the detection, the working time of the river state detection can be set (specifically, it can be based on Beijing time), the river direction of the target river can be set, the sensitivity of the millimeter wave radar can be set, etc. It is also possible to set the initialization settings for the various estimation parameters (such as the system estimation S mentioned in the above embodiment) in the estimation model required for the calculation process from the measured data to the flow estimation value and the flow velocity estimation value. Optionally, the above-mentioned estimation parameters can be initialized based on preset default values, and the various estimation parameters can be corrected according to the river scene image. Alternatively, the above-mentioned estimation parameters can be determined or adjusted according to the measured annotation of the river on site (for example, the annotation of the river surface width, the river flow rate, etc.). Based on this, the initialization setting of the radar detection parameters is realized. The specific adjustment or determination process can be adjusted and tested based on a number of measured data, and this embodiment does not specifically describe the adjustment process.

[0125] In one embodiment, the above-mentioned river state detection method may further include:

[0126] Obtain image information of the target river; based on the image information, correct various radar detection parameters.

[0127] For example, taking one year as the observation span, the velocity and flow state of the river are steady state for about 99% of the time. Therefore, most of the time, the measurement data detected by the millimeter-wave radar for the target river are stable data with small fluctuations. The factor that causes misjudgment is most likely the correction to the estimation result. This is because the millimeter-wave radar's measurement of the river is an indirect measurement rather than a direct measurement, so there are linear or nonlinear system deviations that need to be corrected. In this embodiment, the correction process can be regarded as a linear correction.

[0128] The target river can be photographed based on a visual camera or a mobile device with a shooting function to obtain the corresponding river photo. The riverbed depth, river width and flow rate corresponding to the target river can be identified and classified by image analysis of the river photo (for example, with the help of image analysis algorithms such as neural networks, or manual analysis), and the scene to which the target river belongs can be determined (such as a river with a fast flow rate, a deep riverbed and a narrow river surface in a mountainous area, or a river with a slow flow rate and a wide river surface in a plain area), and the estimation parameters (such as system estimation) in the above-mentioned related estimation models involving flow estimation and flow rate estimation can be corrected accordingly.

[0129] The above correction process can be carried out throughout the entire monitoring process of the target river by the millimeter wave radar. For example, the above parameters can be corrected based on the image information at each preset time interval. The correction can also be carried out based on the image information before detection.

[0130] In this embodiment, by correcting various radar detection parameters based on the image information of the target river, the accuracy of the flow estimation and flow velocity estimation finally realized based on the measurement data can be improved, thereby achieving accurate detection of the state of the target river.

[0131] Optionally, in one embodiment, the above-mentioned river state detection method may further include:

[0132] Obtain the measured and marked data of the target river; modify various radar detection parameters based on the measured and marked data.

[0133] For example, various radar detection parameters can be corrected based on the measured annotation data obtained in advance from the target river, or the measured annotation data obtained by manually observing the target river for a preset period of time during the process of the millimeter wave radar detecting the state of the target river. Specifically, the estimation parameters in the above-mentioned estimation models related to flow estimation and flow velocity estimation can be corrected. Based on this, this embodiment can improve the accuracy of the final flow estimation and flow velocity estimation based on manual annotation.

[0134] Fig. 9 FIG. 4 is a flow chart of a river state detection method in some embodiments. Fig. 9 As shown, the river state detection method includes the following steps:

[0135] Step S901, installing the millimeter wave radar device. The millimeter wave radar device can be installed on a stable bracket near the target river, and the detection area of ​​the millimeter wave radar covers the area of ​​the target river where the state detection needs to be performed.

[0136] Step S902, adjusting the radar detection parameters for river detection, wherein the radar detection parameters include the flow direction of the target river, the sensitivity of the millimeter wave radar, the working time of the detection, and other parameters; on the other hand, they may also include the estimation parameters involved in the process of realizing flow estimation and flow velocity estimation based on the radar measurement data.

[0137] Step S903, collecting images of the target river to obtain image information of the target river.

[0138] Step S904: Manual annotation is performed based on the measured data of the target river, wherein information such as river width, river flow rate, and riverbed depth may be manually annotated.

[0139] Step S905, based on the data obtained in step S903 and step S904, the radar detection parameters are corrected. For example, based on the collected image information (panoramic photo) of the target river, the result of manual annotation, and the sensitivity of the millimeter wave radar, the estimation parameters used to achieve flow estimation and flow velocity estimation in the millimeter wave radar can be corrected.

[0140] Step S906, based on the measurement data of the target river obtained by the millimeter wave radar, the river state is estimated, wherein the flow rate and flow velocity of the target river can be detected in real time to obtain the estimated flow rate and flow velocity.

[0141] Step S907, judging whether a river warning event occurs according to the detection result of step S906; if so, executing step S908 or step S909; otherwise, executing step S910. For example, judging whether a river flood event occurs according to an S-curve of flow rate variation over time.

[0142] Step S908: Report the warning event, for example, through the Internet or other channels, to notify downstream residents and tourists in real time to avoid risks.

[0143] Step S909: Manual review. The on-duty personnel can be notified in real time through the Internet or other channels to review whether the millimeter wave radar has any false alarms or missed alarms, and if so, mark them.

[0144] Step S910, uploading the currently detected measurement data, flow estimation value, flow velocity estimation value and other information. Specifically, it can be uploaded to the cloud and shared with radar equipment systems in other regions or upstream and downstream of the river.

[0145] Fig.10 For some of the embodiments of the present invention, the flow rate detection method of the river is as follows: Fig.10 As shown, the flow rate and velocity detection method of the river may include:

[0146] Step S1001, detection initialization. Specifically, information such as target clustering parameters, accumulated data frames, and river direction may be initialized.

[0147] Step S1002, collecting image information of the target river.

[0148] Step S1003: Upload the image information to the millimeter wave radar device or the data processing device.

[0149] Step S1004, setting radar detection parameters. Specifically, the radar detection parameters are set based on the data collected in step S1003 and the image analysis result.

[0150] Step S1005, collecting measurement data: The millimeter wave radar device is operated to collect measurement data within a period of time.

[0151] Step S1006: annotate the measurement data collected in step S1005, wherein the annotated content may include the actual flow rate or actual flow rate of the data in the corresponding time period.

[0152] Step S1007: parameter optimization: radar detection parameters are optimized according to the labeling result of S1006 and the image analysis result of S1003.

[0153] Step S1008, based on the millimeter wave radar sensing environment, collecting measurement data. The environment is measured using the target detection algorithm built into the millimeter wave radar, and point cloud data of the environment is output in real time as measurement data.

[0154] Step S1009: Filter the measurement data to obtain target measurement data, wherein the measurement data is divided into the de-measured measurement data and the received measurement data, and only the measurement data within the river area and consistent with the river flow direction is retained.

[0155] Step S1010: clustering the target measurement data to obtain clustered data.

[0156] Step S1011, solving the characteristic vector for the cluster data.

[0157] Step S1012, determining a flow estimation value and a flow velocity estimation value according to the feature vector. Specifically, referring to the above embodiment, the flow estimation value is determined based on a pre-built flow estimation model, and the flow velocity estimation value is determined based on a pre-built flow velocity estimation model.

[0158] Step S1013, judging the warning event according to the result of step S1012, and obtaining the judgment result.

[0159] Step S1014: if a warning event occurs, it is uploaded to the cloud and shared with other monitoring systems or radar equipment systems upstream and downstream of the target river.

[0160] In this embodiment, a river state detection device is also provided, which is used to implement the above-mentioned embodiments and preferred implementation modes, and the descriptions that have been made will not be repeated. The terms "module", "unit", "subunit", etc. used below can implement a combination of software and / or hardware for a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceivable.

[0161] Fig.11 : is a structural block diagram of the river state detection device 11 of this embodiment, as shown in Fig.11 As shown, the river state detection device 11 includes: an acquisition module 112, a feature conversion module 114 and a state estimation module 116; wherein:

[0162] The acquisition module 112 is used to acquire the measurement data of the target river based on the millimeter wave radar; the feature conversion module 114 is used to perform feature conversion processing on the measurement data to obtain the corresponding feature vector; the state estimation module 116 is used to process the feature vector based on a pre-built flow estimation model to obtain the flow estimation value of the target river; and, based on a pre-built flow velocity estimation model, process the feature vector to obtain the flow velocity estimation value of the target river.

[0163] It should be noted that the above modules can be functional modules or program modules, and can be implemented by software or hardware. For modules implemented by hardware, the above modules can be located in the same processor; or the above modules can be located in different processors in any combination.

[0164] It should be noted that the specific examples in this embodiment can refer to the examples described in the above embodiments and optional implementation modes, and will not be repeated in this embodiment.

[0165] It should be understood that the specific embodiments described herein are only used to explain the application, rather than to limit it. Based on the embodiments provided in this application, all other embodiments obtained by ordinary technicians in this field without creative work are within the protection scope of this application.

[0166] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties.

[0167] Obviously, the drawings are only some examples or embodiments of the present application. For ordinary technicians in the field, the present application can also be applied to other similar situations based on these drawings without creative work. In addition, it is understandable that although the work done in this development process may be complicated and lengthy, for ordinary technicians in the field, certain changes in design, manufacturing or production based on the technical content disclosed in this application are only conventional technical means and should not be regarded as insufficient content disclosed in this application.

[0168] The term "embodiment" in this application refers to a specific feature, structure or characteristic described in conjunction with the embodiment that can be included in at least one embodiment of the present application. The appearance of this phrase in various locations in the specification does not necessarily mean the same embodiment, nor does it mean that it is mutually exclusive with other embodiments and is independent or optional. It is clearly or implicitly understood by those of ordinary skill in the art that the embodiments described in this application can be combined with other embodiments without conflict.

[0169] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of patent protection. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the scope of protection of the present application. Therefore, the scope of protection of the present application shall be subject to the attached claims.

Claims

1. A river state detection method, characterized in that: include: Based on millimeter wave radar, obtain measurement data of the target river; Performing feature conversion processing on the measurement data to obtain a corresponding feature vector; Taking the system estimation quantity of the pre-built flow estimation model as a vector element, expanding it into the feature vector, and obtaining a target vector for flow estimation; multiplying the flow weight vector corresponding to the target vector in the flow estimation model by the target vector to obtain a flow estimation value of the target river; and Taking a system estimation quantity of a pre-built velocity estimation model as a vector element, expanding it into the feature vector, and obtaining a target vector for velocity estimation; Multiplying the flow velocity weight vector corresponding to the target vector for flow velocity estimation in the flow velocity estimation model by the target vector for flow velocity estimation to obtain a flow velocity estimation value of the target river; Among them, the flow weight vector is composed of flow weights pre-set for the target river and corresponding to each element in the target vector for flow estimation; the flow velocity weight vector is composed of flow velocity weights pre-set for the target river and corresponding to each element in the target vector for flow velocity estimation.

2. The river state detection method according to claim 1, characterized in that: After obtaining the estimated value of the flow rate of the target river and the estimated value of the flow velocity of the target river, the method further includes: Determine flow change information in the time dimension according to the estimated flow values ​​at different times; determine whether a preset warning event occurs in the target river according to the flow change information; and / or, According to the estimated flow velocity values ​​at different times, the flow velocity change information in the time dimension is determined; according to the flow velocity change information, it is determined whether a preset warning event occurs in the target river.

3. The river state detection method according to claim 1, characterized in that: The measured data is subjected to feature conversion processing to obtain a corresponding feature vector, including: Performing clustering processing on the measurement data to obtain cluster data; According to the clustering data, the speed feature, the intensity feature and the clustering convex hull feature of the target river are determined to form a corresponding feature vector.

4. The river state detection method according to claim 3, characterized in that: The measurement data is clustered to obtain cluster data, including: According to the direction represented by the measurement data, the measurement data is divided into outgoing measurement data and incoming measurement data; According to the river flow direction corresponding to the target river, retaining the measurement data consistent with the river flow direction from the outgoing measurement data and the incoming measurement data as the target measurement data; Clustering is performed on the target measurement data to obtain clustered data.

5. The river state detection method according to any one of claims 1 to 4, characterized in that: Before obtaining the measurement data of the target river based on the millimeter wave radar, the method further includes: Initialize and set various radar detection parameters; the radar detection parameters are at least used to estimate the flow velocity and flow rate of the target river.

6. The river state detection method according to claim 5, characterized in that: The method further comprises: Acquiring image information of the target river; According to the image information, the various radar detection parameters are corrected.

7. The river state detection method according to claim 5, characterized in that: The method further comprises: Acquire measured and marked data of the target river; According to the measured annotation data, the various radar detection parameters are corrected.

8. A river state detection device, characterized in that: include: Acquisition module, feature conversion module and state estimation module; wherein: The acquisition module is used to acquire measurement data of the target river based on the millimeter wave radar; The feature conversion module is used to perform feature conversion processing on the measurement data to obtain a corresponding feature vector; The state estimation module is used to expand the system estimation quantity of the pre-built flow estimation model as a vector element into the feature vector to obtain a target vector for flow estimation; multiply the flow weight vector corresponding to the target vector in the flow estimation model by the target vector to obtain the flow estimation value of the target river; and The system estimation quantity of the pre-constructed velocity estimation model is used as a vector element and expanded into the feature vector to obtain a target vector for velocity estimation; the velocity weight vector corresponding to the target vector for velocity estimation in the velocity estimation model is multiplied by the target vector for velocity estimation to obtain the velocity estimation value of the target river; Among them, the flow weight vector is composed of flow weights pre-set for the target river and corresponding to each element in the target vector for flow estimation; the flow velocity weight vector is composed of flow velocity weights pre-set for the target river and corresponding to each element in the target vector for flow velocity estimation.

Images