Flood detection method based on BeiDou MP and CGMC dual combined observations

Abstract

The present invention discloses a flood detection method based on Beidou MP and CGMC dual combined observation values, including the following steps: S1, obtaining satellite observation data; S2, preprocessing the observation data; S3, establishing a flood detection model; S4, detecting the flood twice based on the flood detection values ​​generated based on the pseudo-range MP combined observation value and the phase CGMC combined observation value, and outputting the detection results; S5, combining the detection results of the flood detection with the two flood detection values, and marking the detection results according to the detection threshold; S6, issuing flood warnings according to the detection results and updating the detection model with the detection results. The above technical solution can be used to make full use of the existing Beidou navigation system satellites and low-cost measurement receivers for flood detection, which can effectively improve the detection time and space resolution and detection accuracy.

Description

Technical Field

[0001] The invention relates to the technical field of navigation satellite inversion, and in particular to a flood detection method based on Beidou MP (Multipath-Pseudorange combination) and CGMC (Carrier-phase Geometry-free Multipath Combination) dual combination observation values. Background Art

[0002] Since the current weather forecast accuracy cannot effectively prevent extreme heavy rainfall, and the rainfall intensity forecast system cannot be effectively connected with the urban drainage system and other functions, urban waterlogging or local flooding caused by extreme heavy rainfall often occurs, which has a great impact on the safety of people's lives and property. Even in the current technological development, major floods still occur, which bring great danger and loss to people's lives and property. Therefore, it is urgent to propose a low-cost, high-efficiency, fast-response and high-precision flood warning method to provide reliable and timely information for flood disaster prevention departments, protect people's lives and safety, and reduce property losses caused by disasters.

[0003] At present, flood detection and early warning methods can be roughly divided into three categories. The first category is flood detection and early warning based on the traditional ground rainfall collection station method, but this method has a small observation range, low temporal resolution, and requires a wide range of collection points to conduct large-scale ultra-wide monitoring, which increases the hardware cost. The second category is a method based on professional remote sensing satellites and ground radar stations for rainfall detection. Although this method can meet the temporal resolution requirements of flood detection, it requires the launch of professional remote sensing satellites and the establishment of professional ground radar receiving stations. At the same time, the construction of professional radar receiving stations in cities has high requirements for geographical conditions and is likely to affect people's lives. The construction of observation stations in suburban or rural open areas requires the construction of other supporting facilities, such as corresponding power supply facilities, which leads to problems such as high construction difficulty and long construction period, and the hardware cost is also high. The third category is based on the GNSS (Global Navigation Satellite System) navigation system satellite and geodetic receivers to perform ground surface water inversion for flood detection and early warning. The existing navigation system satellite-based flood detection method is mainly based on the carrier-to-noise ratio method for detection, but some low-end receivers cannot output carrier-to-noise ratio observations, which makes this method invalid. At the same time, the types of carrier-to-noise ratio outputs of different receivers vary greatly, and the carrier-to-noise ratios of different satellites and frequencies also vary significantly, resulting in low detection accuracy based on the carrier-to-noise ratio method, which cannot meet the needs of high-precision flood detection. Therefore, it is particularly urgent to study the use of low-cost navigation satellite receivers to output only conventional observations (including only pseudorange and phase observations) for high-precision flood detection. This can not only meet the requirements of flood detection time and space resolution, but also meet the requirements of detection accuracy and application universality. In addition, low-cost receivers can further reduce hardware costs and provide strong support for flood disaster prevention. Summary of the invention

[0004] In view of the shortcomings of the prior art, the present invention proposes a flood detection method based on Beidou MP and CGMC dual combined observation values, which can realize a low-cost navigation satellite receiver and perform high-precision flood detection by only outputting conventional observation values. It can not only meet the requirements of spatiotemporal resolution of flood detection, but also meet the requirements of detection accuracy and application universality. The low-cost receiver can further reduce the hardware cost, providing strong support for flood disaster prevention.

[0005] In order to solve the above technical problems, the technical solution of the present invention is:

[0006] A flood detection method based on Beidou MP and CGMC dual combined observation values ​​includes the following steps:

[0007] S1. Obtain satellite observation data

[0008] S1-1, read the original satellite observation values, wherein the original observation values ​​include pseudorange observation values ​​and carrier phase observation values;

[0009] S1-2, respectively forming a three-frequency pseudorange MP combined observation value and a phase CGMC combined observation value according to the pseudorange observation value and the carrier phase observation value;

[0010] S1-3, calculating the altitude angle and azimuth angle at the corresponding epoch time according to the acquired pseudorange observation value and carrier phase observation value;

[0011] S2, preprocessing the observation data;

[0012] S3, establish a flood detection model, that is, write the processed original observations into the database in chronological order;

[0013] S4, performing two flood detections on the flood based on the flood detection values ​​generated based on the pseudorange MP combined observation value and the phase CGMC combined observation value, and outputting the detection results;

[0014] S5, combining the two flood detection methods to detect floods, and marking the detection results according to the detection threshold;

[0015] S6. Issue flood warnings based on the detection results and update the detection model based on the detection results.

[0016] Preferably, the pseudorange MP combined observation value is generated as follows: the three frequency bands of the Beidou navigation system satellite are marked as B1 band, B2 band and B3 band respectively, and the original observation values ​​of any two frequency bands are selected to obtain the three-frequency pseudorange MP combined observation value through calculation.

[0017] Preferably, the calculation method is as follows:

[0018] (1)

[0019] In the formula, MP 12 represents the pseudorange MP combined observation value composed of B1 and B2 frequency bands, P and L represent the satellite original pseudorange and carrier phase observation value respectively, and the subscript represents the frequency band, that is, P 1 and L 1 represents the original pseudorange and carrier phase observation values ​​of the B1 frequency band, for ,in is the frequency of the signal, N 1 It represents the fixed deviation component of the first MP combination observation value. For the same observation segment data, the fixed deviation component remains unchanged. m i represents the impact of multipath on pseudorange, n iis the carrier phase ambiguity parameter, Indicates the wavelength. Similarly, other second and third MP combination observations can be generated by the following formula:

[0020] (2)

[0021] The corresponding parameters have the same meaning as above. 21 MP represents the second pseudo-range MP combined observation value composed of B1 and B2 frequency bands, MP 31 It represents the third pseudo-range MP combination observation value composed of the B1 and B3 frequency bands, and the triple-frequency pseudo-range MP combination observation value can be obtained.

[0022] (3)

[0023] Where: IF represents the ionosphere-free combined observation value, F represents the raw carrier phase observation, f is the frequency of the signal, and the subscript indicates the frequency band. F 1 represents the raw carrier phase observation value in the B1 frequency band, and CGMC represents the geometry-free and ionosphere-free combined observation value.

[0024] Preferably, the step S2 comprises:

[0025] S2-1, based on the segmented model, eliminate the influence of fixed constants and remove gross error observations;

[0026] S2-2. Perform noise reduction processing based on the singular value eigendecomposition denoising algorithm with variable threshold strategy.

[0027] Preferably, the step S2-1 is specifically performed as follows:

[0028] 1) Arrange all pseudorange MP combined observations and phase CGMC combined observations from low to high according to the altitude information, and divide them into segments according to the fixed altitude intervals;

[0029] 2) Perform sliding average processing on each segment of data separately, set a sliding window, and set the sliding direction of the window as follows: the front data adopts forward sliding, that is, taking data forward; the middle data adopts forward and backward sliding, that is, taking half of the front and half of the back respectively; the rear data adopts backward sliding, that is, taking data backward;

[0030] 3) The observation values ​​with the fixed constant bias removed are stored to obtain the pseudorange MP combination residual and phase CGMC combination observation values ​​that only retain the multipath signal and random noise error.

[0031] Preferably, the step S2-2 is specifically performed as follows:

[0032] 1) Perform singular value decomposition on the pseudorange MP combined observation value and the phase CGMC combined observation value;

[0033] 2) Using variable threshold strategy to denoise its singular values;

[0034] 3) Use the denoised singular values ​​to perform inverse transformation on the signal, complete the signal reconstruction, and obtain the denoised signal.

[0035] Preferably, the singular value decomposition method is as follows:

[0036] According to matrix decomposition theory, for an m-row n-column matrix A∈R m*n Assuming the rank of the matrix is ​​r, there exists an orthogonal matrix U∈R m*m and the orthogonal matrix V∈R n*n , so that it satisfies:

[0037] (4)

[0038] In the formula, the singular value l i satisfy l 1 ≥ l 2 ≥ ... ≥ l r >0, the column vectors of U and V are AA T and A T The characteristic matrix of A, T is the transpose of the matrix, and min(˙) means taking the minimum value. From linear algebra theory, we can get:

[0039] (5)

[0040] In the formula u i is the column vector of U, v i is the column vector of V. Through the above decomposition, the signal can be decomposed into mutually unrelated sub-matrices and the corresponding singular values.

[0041] Preferably, the variable threshold strategy model is expressed as:

[0042] (6)

[0043] In the formula, β = n / m For the matrix A The ratio limit of dimensions, are the singular values ​​after denoising.

[0044] Preferably, the signal reconstruction model is expressed as:

[0045] (7)

[0046] This is the reconstructed signal after denoising.

[0047] Preferably, the method for establishing the flood detection model in step S3 is as follows:

[0048] Continuously collect the original pseudorange observations and carrier phase observations of the satellite on rainless days, process the original data according to steps S1 and S2, and finally obtain accurate pseudorange MP and phase CGMC combined observations, and model the above two detection quantities according to the altitude angle and azimuth angle information according to the orbit type and satellite, and finally obtain an accurate flood detection model of pseudorange MP and phase CGMC combined observations.

[0049] Preferably, step S4 includes

[0050] S4-1, flood detection based on pseudo-range MP combined observations;

[0051] S4-2. Flood detection is performed using flood detection measurements based on phase CGMC combined observations.

[0052] Preferably, the flood detection method in step S4-1 is as follows:

[0053] 1) Calculate the residual of the pseudo-range MP combined observation value of the current detection day, and the calculation process is as in step S1 and step S2, to obtain the satellite pseudo-range MP combined observation value detection quantity and the corresponding altitude angle azimuth information at the current epoch;

[0054] 2) Compare the pseudo-range MP combined observation value at this moment with the pseudo-range MP combined observation value flood detection model established in step S3. The comparison reference is mainly the altitude angle and azimuth angle of the satellite. If the difference is greater than the set threshold, a flood mark is performed. If not, no mark is performed.

[0055] 3) Output the detection results,

[0056] The flood detection method in step S4-2 is the same as that in step S4-1.

[0057] Preferably, the threshold is set such that, for the pseudorange MP combined observation value, the average pseudorange MP combined observation value is taken as the center value, and a downward or upward deviation of 0.2m indicates that there is a strong flood correlation. For the phase CGMC combined observation value, the average pseudorange CGMC combined observation value is taken as the center value, and a downward or upward deviation of 0.2cm indicates that there is a strong flood correlation.

[0058] The present invention has the following characteristics and beneficial effects:

[0059] By adopting the above technical scheme, the existing Beidou navigation system satellites and low-cost measurement receivers can be fully utilized for flood detection, which can effectively improve the detection time and space resolution and detection accuracy; without the investment and construction of third-party equipment, it can avoid the problems of high construction cost and long construction period; it can solve the problems of method failure and low accuracy caused by the type of receiver, and at the same time can greatly save hardware costs, and realize economic and universal high-precision flood detection. Considering the influence of fixed constants on the detection value observation, the segmented sliding model is used to eliminate the fixed constants in the combined observation value of pseudorange MP and phase CGMC, filter out the coarse values ​​in the observation value, and only retain the residual of the combined observation value containing multipath signals and random noise. At the same time, the singular value eigendecomposition denoising algorithm based on the variable threshold strategy is used to denoise the residual of the combined observation value of pseudorange MP and phase CGMC, filter out the high-frequency random noise, retain the residual containing only multipath signals, improve the detection accuracy, and improve the detection success rate. Taking full account of the correlation between the pseudo-range MP combination and the phase CGMC combination, the two combined detection values ​​are used to detect floods together, avoiding the low precision of single detection and false detection, ensuring detection accuracy and improving detection success rate. At the same time, the detection model is updated and iterated using new data to ensure the long-term detection accuracy and stability of the model. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0061] Figure 1 Flowchart of flood detection method based on dual combined observations of BeiDou MP and CGMC.

[0062] Figure 2 Algorithm flow chart based on Beidou pseudorange MP combined observations.

[0063] Figure 3 Flowchart of the Beidou phase CGMC combined observation algorithm. DETAILED DESCRIPTION

[0064] It should be noted that, in the absence of conflict, the embodiments of the present invention and the features in the embodiments may be combined with each other.

[0065] In the description of the present invention, it should be understood that the terms "center", "longitudinal", "lateral", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, the terms "first", "second", and the like are only used for descriptive purposes, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, features defined as "first", "second", and the like may explicitly or implicitly include one or more of the features. In the description of the present invention, unless otherwise specified, "multiple" means two or more.

[0066] In the description of the present invention, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood by specific circumstances.

[0067] The present invention provides a flood detection method based on Beidou MP and CGMC dual combined observation values, such as Figure 1-Figure 3 As shown, the following steps are included:

[0068] S1. Obtain satellite observation data, where the satellite is a Beidou navigation system satellite;

[0069] S1-1, read the original satellite observation values, wherein the original observation values ​​include pseudorange observation values ​​and carrier phase observation values;

[0070] S1-2, respectively forming a three-frequency pseudorange MP combined observation value and a phase CGMC combined observation value according to the pseudorange observation value and the carrier phase observation value;

[0071] S1-3, calculating the altitude angle and azimuth angle at the corresponding epoch time according to the acquired pseudorange observation value and carrier phase observation value;

[0072] S2, preprocessing the observation data;

[0073] S3, establish a flood detection model, that is, write the processed original observations into the database in chronological order;

[0074] S4, performing two flood detections on the flood based on the flood detection values ​​generated based on the pseudorange MP combined observation value and the phase CGMC combined observation value, and outputting the detection results;

[0075] S5, combining the two flood detection methods to detect floods, and marking the detection results according to the detection threshold;

[0076] S6. Issue flood warnings based on the detection results and update the detection model based on the detection results.

[0077] Specifically, step 1-2: use the pseudorange and carrier phase original observation values ​​to generate the three-frequency pseudorange MP and phase CGMC combined observation values ​​respectively. Among them, the pseudorange MP combined observation value is generated by the following formula:

[0078] (1)

[0079] In the formula, MP 12 Represents the pseudorange MP combined observation value composed of B1 and B2 frequency bands. P and L represent the Beidou satellite original pseudorange and carrier phase observation values ​​respectively. The subscript represents the frequency band, such as P 1 and L 1 Represents the raw pseudorange and carrier phase observations in the B1 band. for ,in f is the frequency of the signal. 1 It represents the fixed deviation component of the first MP combination observation value. For the same observation segment data, the fixed deviation component remains unchanged. m i represents the impact of multipath on pseudorange, n i It is the carrier phase ambiguity parameter. Indicates the wavelength. Similarly, other second and third MP combination observations can be generated by the following formula:

[0080] (2)

[0081] The corresponding parameters have the same meaning as above. 21 MP represents the second pseudo-range MP combined observation value composed of B1 and B2 frequency bands, MP 31 It represents the third pseudo-range MP combined observation value composed of B1 and B3 frequency bands.

[0082] The carrier phase geometry-free observations can be formed by the following model:

[0083] (3)

[0084] Where: IF represents the ionosphere-free combined observation value, Frepresents the raw carrier phase observation, f is the frequency of the signal. The subscript indicates the frequency band, such as F 1 Represents the raw carrier phase observation value in the B1 band. CGMC represents the geometry-free and ionosphere-free combined observation value, which further eliminates the influence of geometric distance.

[0085] It can be understood that at this time the CGMC combined observation value still contains fixed ambiguity constants, but they are hidden in the carrier phase.

[0086] Step 2-1: Use the segmented sliding model to eliminate the fixed constants in the combined observations of pseudorange MP and phase CGMC, filter out the coarse values ​​in the observations, and only retain the combined observation residuals containing multipath signals and random noise. This step of the algorithm mainly consists of three parts:

[0087] 1) Arrange all pseudorange MP and phase CGMC combined observation data from low to high according to the altitude information, and divide them into segments with an altitude interval of 10 degrees. A total of 9 segments are: 0-10, 11-20, 21-30, 31-40, 41-50, 51-60, 61-70, 71-80, 81-90.

[0088] 2) Perform sliding average processing on each data segment separately, and set the sliding window to 50 epochs. This setting needs to be determined according to the data sampling rate and data density. Here, 50 is set only for reference. The window sliding direction setting mode is: the front data adopts forward sliding (i.e., taking data forward), the middle data adopts forward and backward sliding (i.e., taking half of the front and half of the back), and the rear data adopts backward sliding (i.e., taking data backward).

[0089] 3) The observation values ​​with the fixed constant bias removed are stored to obtain the pseudorange MP combination residual and phase CGMC combination observation values ​​that only retain the multipath signal and random noise error.

[0090] Step 2-2: Use the singular value eigendecomposition denoising algorithm based on the variable threshold strategy to denoise the pseudorange MP and phase CGMC combined observations processed in step 3, and obtain the residual of the pseudorange MP and phase CGMC combined observations that only retain the multipath signal. This step of the algorithm mainly consists of three parts:

[0091] 1) Perform singular value decomposition on the pseudorange MP combined observation value and the phase CGMC combined observation value. According to the matrix decomposition theory, for a matrix A∈R with m rows and n columns, m*n Assuming the rank of the matrix is ​​r, there exists an orthogonal matrix U∈R m*m and the orthogonal matrix V∈R n*n , so that it satisfies:

[0092] (4)

[0093] In the formula, the singular value l i satisfy l 1 ≥ l 2 ≥ ... ≥ l r > 0. The column vectors of U and V are AA T and A T The characteristic matrix of A. T is the transpose of the matrix, and min(˙) means taking the minimum value. From linear algebra theory, we can get:

[0094] (5)

[0095] In the formula u i is the column vector of U, v i is the column vector of V. Through the above decomposition, the signal can be decomposed into mutually unrelated sub-matrices and the corresponding singular values ​​are multiplied.

[0096] 2) Use variable threshold strategy to denoise its singular values. Conventional threshold strategies are mainly divided into hard threshold and soft threshold, but both threshold strategies set the threshold to 0, resulting in loss of effective information and failure to obtain optimal denoising. Therefore, the present invention adopts variable threshold strategy to improve this problem. Its variable threshold strategy is expressed as:

[0097] (6)

[0098] In the formula, β = n / m For the matrix A After the above threshold strategy denoising, the singular values ​​after the noise can be effectively filtered out , we can know from formula (6) It is just an excessive value.

[0099] 3) Use the denoised singular values ​​to perform inverse transformation on the signal to complete the signal reconstruction and obtain the denoised signal. The signal reconstruction model is expressed as:

[0100] (7)

[0101] In the formula It is the singular value after denoising. This is the reconstructed signal after denoising. The meanings of other parameters are the same as above.

[0102] Through the above process, high-frequency random noise can be effectively filtered out, and the pseudorange MP and phase CGMC combined observation value residuals that only retain the multipath signal are obtained.

[0103] Step 3: Establish a detection model based on pseudorange MP and phase CGMC combined observations. The modeling process is: continuously collect the original pseudorange and carrier phase observations of Beidou satellites on rainless days, process the original data according to steps S1 and S2, and finally obtain accurate pseudorange MP and phase CGMC combined observations, and model the above two detection quantities according to the altitude and azimuth information according to the orbit type and satellite, and finally obtain an accurate flood detection model of pseudorange MP and phase CGMC combined observations.

[0104] It should be noted that the flood detection model is actually a database that stores preprocessed observation values, or it can be a text file that writes preprocessed observations, which is used to classify and store the processed pseudorange MP and phase CGMC dual combined observation values, and the above two detection quantities according to the altitude angle and azimuth angle information according to the orbit type and satellite, and can realize data update and iteration, which is convenient for later data reading and lays the foundation for flood detection accuracy.

[0105] Step 4-1: Detect floods using pseudo-range MP combined observations. This step of the algorithm consists of three main parts:

[0106] 1) Calculate the residual of the pseudo-range MP combined observation value of the current detection day. The calculation process is the same as described in step S1 and step S2, and obtain the satellite pseudo-range MP combined observation value detection quantity and the corresponding altitude angle azimuth information at the current epoch.

[0107] 2) Compare the pseudo-range MP combination observation value at this moment with the pseudo-range MP combination observation value flood detection model established in step S3. The comparison reference is mainly the satellite altitude angle and azimuth angle. If the difference is greater than the set threshold, the flood is marked, otherwise it is not marked. The threshold is obtained according to the empirical model. It is generally believed that the average pseudo-range MP combination observation value decreases or increases by 0.2m, indicating that there is a strong correlation with the flood. The empirical threshold is calculated from the local observation station data when the flood occurs.

[0108] It can be understood that the empirical model specifically refers to a database storing historical pseudorange MP combination observation values.

[0109] 3) Output the detection results.

[0110] Step 4-2: Detect floods using phase-based CGMC combined observations. This step of the algorithm consists of three main parts:

[0111] 1) Calculate the residual of the phase CGMC combined observation value of the current detection day. The calculation process is the same as that described in steps S1 and S2, and obtain the satellite phase CGMC combined observation value detection quantity and the corresponding altitude angle azimuth information at the current epoch.

[0112] 2) Compare the current phase CGMC combined observation value with the pseudo-phase CGMC combined observation value flood detection model established in step S3. The comparison reference is mainly the satellite altitude and azimuth. If the difference is greater than the set threshold, the flood is marked, otherwise it is not marked. The threshold is obtained based on historical data. It is generally believed that a decrease or increase deviation of 0.2 cm in the average phase CGMC combined observation value indicates a strong correlation with the flood. The empirical threshold is calculated from the local observation station data when the flood occurs.

[0113] It can be understood that the empirical model specifically refers to a database storing historical phase CGMC combination observations.

[0114] 3) Output the detection results.

[0115] Steps S5 and S6: The detection results of the two detection algorithms are jointly analyzed. If both detection results are marked, it is considered that there is a flood risk and a flood warning is required. If only one detection result is marked or neither is marked, no warning is issued, and the original detection model is updated to improve the detection precision and accuracy of the original model and ensure the long-term stability of the detection model.

[0116] The embodiments of the present invention are described in detail above with reference to the accompanying drawings, but the present invention is not limited to the described embodiments. For those skilled in the art, various changes, modifications, substitutions and variations of these embodiments including components are made without departing from the principles and spirit of the present invention, and still fall within the scope of protection of the present invention.

Claims

1. A flood detection method based on BeiDou MP and CGMC dual combined observation values, characterized in that: The steps include: S1. Obtain satellite observation data S1-1, read the original satellite observation values, wherein the original observation values ​​include pseudorange observation values ​​and carrier phase observation values; S1-2, forming a three-frequency pseudorange MP combined observation value and a phase CGMC combined observation value according to the pseudorange observation value and the carrier phase observation value, wherein CGMC represents a geometry-free ionosphere-free combined observation value; S1-3, calculating the altitude angle and azimuth angle at the corresponding epoch time according to the acquired pseudorange observation value and carrier phase observation value; S2, preprocessing the observation data; S3, establish a flood detection model, that is, write the processed original observations into the database in chronological order; S4, performing two flood detections on the flood based on the flood detection values ​​generated based on the pseudorange MP combined observation value and the phase CGMC combined observation value, and outputting the detection results; S5, combining the two flood detection methods to detect floods, and marking the detection results according to the detection threshold; S6. Issue flood warnings based on the detection results and update the detection model based on the detection results.

2. The flood detection method based on Beidou MP and CGMC dual combined observation values ​​according to claim 1 is characterized in that: Generation of the pseudo-range MP combined observation value: The three frequency bands of the Beidou navigation system satellite are marked as B1 band, B2 band and B3 band respectively, and the original observation values ​​of any two frequency bands are selected to obtain the three-frequency pseudo-range MP combined observation value through calculation.

3. The flood detection method based on Beidou MP and CGMC dual combined observation values ​​according to claim 2 is characterized in that: The calculation method is as follows: In the formula, MP 12 represents the pseudorange MP combined observation value composed of B1 and B2 frequency bands, P and L represent the original pseudorange and carrier phase observation values ​​of the satellite respectively, and the subscript represents the frequency band, that is, P1 and L1 represent the original pseudorange and carrier phase observation values ​​of the B1 frequency band, and α is Where f is the frequency of the signal, N1 represents the fixed deviation component of the first MP combined observation value. For the same observation segment data, the fixed deviation component remains unchanged, m1 represents the impact of multipath on pseudorange, n1 is the carrier phase ambiguity parameter, and λ represents the wavelength. Similarly, the other second and third MP combined observation values ​​can be generated by the following formula: The corresponding parameters have the same meaning as above. 21 MP represents the second pseudo-range MP combined observation value composed of B1 and B2 frequency bands, MP 31 It represents the third pseudo-range MP combination observation value composed of the B1 and B3 frequency bands, and the three-frequency pseudo-range MP combination observation value can be obtained.

4. The flood detection method based on BeiDou MP and CGMC dual combined observation values ​​according to claim 1 is characterized in that: The phase CGMC combined observations are formed by the following model: Where: IF represents the ionosphere-free combined observation value, φ represents the original carrier phase observation value, f is the frequency of the signal, the subscript represents the frequency band, φ1 represents the original carrier phase observation value of the B1 frequency band, and CGMC represents the geometry-free ionosphere-free combined observation value.

5. The flood detection method based on Beidou MP and CGMC dual combined observation values ​​according to claim 1 is characterized in that: The step S2 comprises: S2-1, based on the segmented model, eliminate the influence of fixed constants and remove gross error observations; S2-2. Perform noise reduction processing based on the singular value eigendecomposition denoising algorithm with variable threshold strategy.

6. The flood detection method based on Beidou MP and CGMC dual combined observation values ​​according to claim 5 is characterized in that: The specific method of step S2-1 is as follows: 1) Arrange all pseudorange MP combined observation values ​​and phase CGMC combined observation values ​​from low to high according to the altitude information, and divide them into segments according to the fixed altitude intervals; 2) Perform sliding average processing on each segment of data separately, set a sliding window, and set the sliding direction of the window as follows: the front data adopts forward sliding, that is, taking data forward; the middle data adopts forward and backward sliding, that is, taking half of the front and back respectively; the rear data adopts backward sliding, that is, taking data backward; 3) The observation values ​​with the fixed constant deviation removed are stored to obtain the pseudorange MP combination residual and phase CGMC combination observation values ​​that only retain the multipath signal and random noise error.

7. The flood detection method based on BeiDou MP and CGMC dual combined observation values ​​according to claim 5 is characterized in that: The specific method of step S2-2 is as follows: 1) Perform singular value decomposition on the pseudorange MP combined observation value and the phase CGMC combined observation value; 2) Using variable threshold strategy to denoise its singular values; 3) Use the denoised singular values ​​to perform inverse transformation on the signal to complete signal reconstruction and obtain the denoised signal.

8. The flood detection method based on BeiDou MP and CGMC dual combined observation values ​​according to claim 7 is characterized in that: The singular value decomposition method is as follows: According to the matrix decomposition theory, for a matrix A∈R with m rows and n columns, m*n Assuming the rank of the matrix is ​​r, there exists an orthogonal matrix U∈R m*m and the orthogonal matrix V∈R n*n , so that it satisfies: In the formula, the singular value λ i Satisfies λ1≥λ2≥...≥λ r >0, the column vectors of U and V are AA T and A T The characteristic matrix of A, T is the transpose of the matrix, min(˙) means taking the minimum value, and from linear algebra theory we can get: Where u i is the column vector of U, v i is the column vector of V. Through the above decomposition, the signal can be decomposed into mutually unrelated sub-matrices and the corresponding singular values.

9. The flood detection method based on BeiDou MP and CGMC dual combined observation values ​​according to claim 8 is characterized in that: The variable threshold strategy is expressed as: Where β = n / m is the ratio limit of the dimension of matrix A, are the singular values ​​after denoising.

10. The flood detection method based on BeiDou MP and CGMC dual combined observation values ​​according to claim 9 is characterized in that: In step 2-2, the reconstruction of the signal is expressed as: This is the reconstructed signal after denoising.

11. The flood detection method based on BeiDou MP and CGMC dual combined observation values ​​according to claim 1 is characterized in that: The method for establishing the flood detection model in step S3 is as follows: Continuously collect the original pseudorange observations and carrier phase observations of the satellite on rainless days, process the original data according to steps S1 and S2, and finally obtain accurate pseudorange MP and phase CGMC combined observations, and model the above two detection quantities according to the altitude angle and azimuth angle information according to the orbit type and satellite, and finally obtain an accurate flood detection model of pseudorange MP and phase CGMC combined observations.

12. The flood detection method based on BeiDou MP and CGMC dual combined observation values ​​according to claim 11 is characterized in that: The step S4 includes S4-1, flood detection based on pseudo-range MP combined observations; S4-2. Flood detection is performed using flood detection data based on phase CGMC combined observations.

13. The flood detection method based on BeiDou MP and CGMC dual combined observation values ​​according to claim 12 is characterized in that: The flood detection method of step S4-1 is as follows: 1) Calculate the residual of the pseudo-range MP combined observation value of the current detection day, and the calculation process is as in step S1 and step S2, to obtain the satellite pseudo-range MP combined observation value detection quantity and the corresponding altitude angle azimuth information at the current epoch time; 2) Compare the pseudo-range MP combined observation value at this moment with the pseudo-range MP combined observation value flood detection model established in step S3. The comparison reference is mainly the altitude angle and azimuth angle of the satellite. If the difference is greater than the set threshold, a flood mark is performed. If not, no mark is performed. 3) Output the detection results, The flood detection method in step S4-2 is the same as that in step S4-1.

14. The flood detection method based on BeiDou MP and CGMC dual combined observation values ​​according to claim 13 is characterized in that: The setting of the threshold is that, for the pseudorange MP combination observation value, the average pseudorange MP combination observation value is taken as the center value. If the downward or upward deviation reaches 0.2m, it indicates the existence of strong flood correlation. For the phase CGMC combination observation value, the average pseudorange CGMC combination observation value is taken as the center value. If the downward or upward deviation reaches 0.2cm, it indicates the existence of strong flood correlation.

Images