A method, apparatus and equipment for processing multi-source ship positioning information

Abstract

This invention provides a method, apparatus, and device for processing multi-source ship positioning information. The method includes: acquiring multi-source positioning information of a ship; performing conversion processing on the multi-source positioning information to obtain first intermediate processing information; performing track association on the first intermediate processing information to obtain second intermediate processing information; obtaining the dynamic variance of the multi-source positioning information based on the multi-source positioning information; obtaining a comprehensive weight based on the dynamic variance and a preset static variance; and performing fusion processing on the second intermediate processing information based on the comprehensive weight to obtain the target positioning information of the ship. The solution of this invention can achieve the fusion of multiple ship positioning data sources, and the fusion result is close to the actual ship navigation information, which helps to improve the reliability of ship navigation decisions and the anti-interference capability of shipborne navigation aids, ensuring the safety and efficiency of maritime traffic.

Description

Technical Field

[0001] This invention relates to the field of ship positioning information processing technology, and in particular to a multi-source ship positioning information processing method, apparatus and equipment. Background Technology

[0002] With the widespread application of electronic and communication technologies in the field of maritime transportation, the types and quantities of maritime communication and navigation equipment are increasing day by day. In response to the incomplete and inaccurate information provided by various heterogeneous maritime dynamic data sources, data fusion technology is used to obtain a more accurate and complete description of ship information, providing strong support for ship navigation decisions or judgments, thereby effectively ensuring the safety and efficiency of maritime transportation, and also contributing to the protection of the marine ecological environment.

[0003] In ship navigation scenarios, using data fusion technology to process multi-source maritime data has many advantages. For example, multi-source maritime data has the characteristics of comprehensive description of target ships and complementary data. Fusion operations can greatly improve the credibility of ship navigation decisions and the anti-interference capability of shipborne navigation aids. When ships are at sea, resources are limited. Fusion of multi-source data can reduce the redundancy of multi-source data, reduce the waste of storage resources, and reduce unnecessary resource consumption during data transmission.

[0004] However, existing multi-source fusion algorithms for ship navigation and positioning information are mostly limited to automatic identification system data and radar data in terms of data source selection, which cannot meet the processing needs of multi-source maritime data. Summary of the Invention

[0005] This invention provides a multi-source ship positioning information processing method, apparatus, and equipment, which can integrate multiple ship positioning data sources. The fusion result is close to the actual ship navigation information, which helps to improve the credibility of ship navigation decisions and the anti-interference capability of shipborne navigation aids, and ensures the safety and efficiency of maritime traffic.

[0006] To solve the above-mentioned technical problems, the technical solution of the present invention is as follows:

[0007] A method for processing multi-source ship positioning information includes:

[0008] Acquire multi-source positioning information of ships;

[0009] The multi-source positioning information is converted and processed to obtain first intermediate processing information;

[0010] The first intermediate processing information is correlated with the flight path to obtain the second intermediate processing information;

[0011] Based on the multi-source positioning information, the dynamic variance of the multi-source positioning information is obtained;

[0012] Based on the dynamic variance and the preset static variance, the comprehensive weights are obtained;

[0013] Based on the comprehensive weights, the second intermediate processing information is fused to obtain the target positioning information of the ship.

[0014] Optionally, the multi-source positioning information is transformed to obtain first intermediate processing information, including:

[0015] The multi-source positioning information is denoised to obtain denoised preprocessed information;

[0016] The preprocessed information is subjected to coordinate transformation and time alignment to obtain the first intermediate processing information.

[0017] Optionally, the first intermediate processing information is correlated with flight paths to obtain second intermediate processing information, including:

[0018] Perform preliminary track association on the first intermediate processing information to obtain preliminary association information;

[0019] The preliminary association information is further associated with the flight path to obtain the second intermediate processing information.

[0020] Optionally, based on the multi-source positioning information, the dynamic variance of the multi-source positioning information is obtained, including:

[0021] Based on the multi-source positioning information, the arithmetic mean of the multi-source positioning information is obtained as a reference value;

[0022] The dynamic variance is obtained based on the multi-source positioning information and the reference value.

[0023] Optionally, based on the dynamic variance and the preset static variance, a comprehensive weight value is obtained, including:

[0024] Determine the distribution coefficients k of dynamic and static errors, where 0 ≤ k ≤ 1;

[0025] Based on the k value, the dynamic variance and static variance are processed to obtain the comprehensive variance;

[0026] Based on the comprehensive variance, the comprehensive weights are obtained.

[0027] Optionally, the value of k is the output result obtained by inputting the second intermediate processing information into the distribution coefficient determination model;

[0028] The training process for determining the distribution coefficient model includes:

[0029] Obtain training multi-source localization information;

[0030] The training multi-source localization information is transformed to obtain the first training intermediate processing information;

[0031] The first training intermediate processing information is correlated with the flight path to obtain the second training intermediate processing information.

[0032] The second training intermediate processing information is input into the input layer of the preset neural network model for processing to obtain the first layer output;

[0033] The first layer output and the first target parameter are input into the hidden layer for processing to obtain the second layer output.

[0034] The second layer output and the second target parameter are input to the output layer for processing to obtain the distribution coefficient determination model.

[0035] Optionally, based on the comprehensive weights, the second intermediate processing information is fused to obtain the target positioning information of the ship, including:

[0036] Through the formula:

[0037] The second intermediate processing information is fused to obtain the target positioning information of the ship;

[0038] in, The result is the data fusion, where N is the number of sensor types fused, and y is the fusion result. j For each type of sensor, w represents the second intermediate processing information. j This represents the overall weighting value for each type of sensor.

[0039] The present invention also provides a multi-source ship positioning information processing device, comprising:

[0040] The acquisition module is used to acquire multi-source positioning information of the vessel;

[0041] The processing module is used to convert and process the multi-source positioning information to obtain first intermediate processing information; perform track association on the first intermediate processing information to obtain second intermediate processing information; obtain the dynamic variance of the multi-source positioning information based on the multi-source positioning information; obtain a comprehensive weight based on the dynamic variance and a preset static variance; and perform fusion processing on the second intermediate processing information based on the comprehensive weight to obtain the target positioning information of the ship.

[0042] The present invention also provides a computing device, comprising: a processor and a memory storing a computer program, wherein the computer program, when executed by the processor, performs the method described above.

[0043] The present invention also provides a computer-readable storage medium storing instructions that, when executed on a computer, cause the computer to perform the method described above.

[0044] The above-described solution of the present invention has at least the following beneficial effects:

[0045] The above-described solution of the present invention acquires multi-source positioning information of a ship; performs transformation processing on the multi-source positioning information to obtain first intermediate processing information; performs track association on the first intermediate processing information to obtain second intermediate processing information; obtains the dynamic variance of the multi-source positioning information based on the multi-source positioning information; obtains a comprehensive weight based on the dynamic variance and a preset static variance; and performs fusion processing on the second intermediate processing information based on the comprehensive weight to obtain the target positioning information of the ship. This solution enables the fusion of multiple or more ship positioning data sources, and the fusion result closely approximates the actual ship navigation information, which helps improve the reliability of ship navigation decisions and the anti-interference capability of shipborne navigation aids, thus ensuring the safety and efficiency of maritime traffic. Attached Figure Description

[0046] Figure 1 This is a flowchart of a multi-source ship positioning information processing method provided in an embodiment of the present invention;

[0047] Figure 2 This is a neural network structure diagram of the distribution coefficient determination model provided in an embodiment of the present invention;

[0048] Figure 3 This is a flowchart of the training process of the distribution coefficient determination model provided in an embodiment of the present invention;

[0049] Figure 4 This is a flowchart of the information processing process of the multi-source ship positioning information processing method provided in the embodiments of the present invention;

[0050] Figure 5 This is a flowchart of the selection process of the adaptive fusion algorithm provided in an embodiment of the present invention;

[0051] Figure 6 A block diagram of a multi-source ship positioning information processing device provided in an embodiment of the present invention. Detailed Implementation

[0052] Exemplary embodiments of the invention will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the invention are shown in the drawings, it should be understood that the invention may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this invention will be thorough and complete, and will fully convey the scope of the invention to those skilled in the art.

[0053] like Figure 1 As shown, an embodiment of the present invention proposes a multi-source ship positioning information processing method, including:

[0054] Step 11: Obtain the ship's multi-source positioning information;

[0055] Step 12: Convert the multi-source positioning information to obtain first intermediate processing information;

[0056] Step 13: Perform track association on the first intermediate processing information to obtain the second intermediate processing information;

[0057] Step 14: Obtain the dynamic variance of the multi-source positioning information based on the multi-source positioning information;

[0058] Step 15: Obtain the comprehensive weights based on the dynamic variance and the preset static variance;

[0059] Step 16: Based on the comprehensive weight, the second intermediate processing information is fused to obtain the target positioning information of the ship.

[0060] In this embodiment, multi-source positioning information of the ship is acquired; the multi-source positioning information is transformed to obtain first intermediate processing information; the first intermediate processing information is correlated with a flight path to obtain second intermediate processing information; the dynamic variance of the multi-source positioning information is obtained based on the multi-source positioning information; a comprehensive weight is obtained based on the dynamic variance and a preset static variance; and the second intermediate processing information is fused based on the comprehensive weight to obtain the target positioning information of the ship. This method can achieve the fusion of multiple or more ship positioning data sources, and the fusion result is close to the actual ship navigation information, which helps to improve the credibility of ship navigation decisions and the anti-interference capability of shipborne navigation aids, thus ensuring the safety and efficiency of maritime traffic.

[0061] In an optional embodiment of the present invention, step 12 includes:

[0062] Step 121: Denoise the multi-source positioning information to obtain denoised preprocessed information;

[0063] Step 122: Perform coordinate transformation and time alignment processing on the preprocessed information to obtain the first intermediate processing information.

[0064] In this embodiment, the multi-source positioning information includes: radar system positioning information, automatic identification system positioning information, and BeiDou satellite navigation system positioning information, wherein the radar system positioning information is observation data acquired by the shore-based radar system;

[0065] The multi-source positioning information is denoised, and erroneous data that does not conform to the format specifications or numerical range is directly discarded to prevent outliers, noise or erroneous data from affecting the track association results.

[0066] To facilitate subsequent track association and fusion processing, the denoised preprocessed information is transformed into a coordinate system, which is then unified into a common coordinate system, specifically a Cartesian coordinate system.

[0067] Specific coordinate transformation:

[0068] Coordinate transformation between automatic identification system positioning information and BeiDou satellite navigation system positioning information:

[0069] Through the formula:

[0070]

[0071] Perform coordinate transformation between the positioning information of the automatic identification system and the positioning information of the Beidou satellite navigation system;

[0072] Where X is the x-coordinate of the target vessel in the Cartesian coordinate system after transformation; Y is the y-coordinate of the target vessel in the Cartesian coordinate system after transformation; L is the longitude of the target vessel in the coordinate system of the identification system; B is the latitude of the target vessel in the coordinate system of the identification system; S is the meridian arc length from the equator to the latitude in the coordinate system of the identification system; N is the radius of curvature of the primordial circle at the latitude; g is the gravitational acceleration; and η is the second eccentricity of the Earth.

[0073] Specifically, when performing coordinate transformation of the positioning information of the automatic identification system, L is the longitude of the target ship in the WGS-84 coordinate system, B is the latitude of the target ship in the WGS-84 coordinate system, and S is the meridian arc length from the equator to the latitude in the WGS-84 coordinate system. The WGS-84 coordinate system is the 1984 World Geodetic Coordinate System.

[0074] When performing coordinate transformation of positioning information from the BeiDou Navigation Satellite System, L is the longitude of the target ship in the CGCS2000 coordinate system, B is the latitude of the target ship in the CGCS2000 coordinate system, and S is the meridian arc length from the equator to the latitude in the CGCS2000 coordinate system. The CGCS2000 coordinate system is the 2000 National Geodetic Coordinate System.

[0075] Coordinate transformation of radar system positioning information:

[0076] Through the formula:

[0077]

[0078] Perform coordinate transformation of radar system positioning information;

[0079] Where X is the x-coordinate of the target vessel detected by the radar in the Cartesian coordinate system after conversion; Y is the y-coordinate of the target vessel detected by the radar in the Cartesian coordinate system after conversion; R is the distance in the polar coordinate system; and θ is the azimuth angle value in the polar coordinate system.

[0080] Because the sampling periods of each system are different, time alignment processing is necessary. Radar has a fixed scanning period, which is generally 15 r / min to 30 r / min, and the time interval is generally 2s to 4s. However, the data sampling periods of AIS (Automatic Identification System) and BDS (BeiDou Navigation Satellite System) vary greatly and are related to the ship's status or the settings of shipboard equipment. Therefore, the sampling time of the radar is used as the starting time reference for time alignment, and time alignment processing is performed on the positioning information of AIS and BDS.

[0081] Specifically:

[0082] The two times before and after time T are T1 and T2, respectively, and the collected positioning information is (L1, B1) and (L2, B2) respectively. Then, the positioning information of the target ship at time T is (L1, B1) and (L2, B2) respectively. T B T )for:

[0083]

[0084] In an optional embodiment of the present invention, step 13 includes:

[0085] Step 131: Perform preliminary track association on the first intermediate processing information to obtain preliminary association information;

[0086] Step 132: Perform a second track association on the preliminary association information to obtain the second intermediate processing information.

[0087] In this embodiment, preliminary association information is obtained by performing preliminary trajectory association on the first intermediate processing information; the second intermediate processing information is obtained by performing secondary trajectory association on the preliminary association information; the secondary trajectory association after the preliminary trajectory association can reduce the overall computational load, improve the trajectory association efficiency, and help improve the overall performance of trajectory fusion.

[0088] The positioning information from the Automatic Identification System (AIS) and the BeiDou Navigation Satellite System (BDS) contains basic information about the target vessel, such as its name, MMSI (Mobile Maritime Service Identifier), call sign, and equipment identification number. However, the positioning information from the radar system does not contain this basic information. Therefore, it is necessary to perform track association on the first intermediate processing information to determine the target vessel that matches the first intermediate processing information.

[0089] The positioning information of the Automatic Identification System (AIS) and the BeiDou Navigation Satellite System (BDS) can be matched using the MMSI number. Once either the AIS or BDS positioning information is matched with the radar system positioning information, the trajectory association of the AIS, BDS, and radar systems can be completed. Therefore, in this embodiment, the trajectory association of the AIS and radar systems is used as an example.

[0090] Correlation of flight paths between automatic identification system positioning information and radar system positioning information:

[0091] Preliminary track association:

[0092] Using the time and location information in the first intermediate processing information, several AIS tracks that may be associated with the radar tracks are roughly correlated with the shore-based radar as the center.

[0093] Specifically:

[0094] When multiple radars operate together, their coverage area is large, but only a portion of it falls within the shared surveillance zone. If all tracks within the entire area are traversed when calculating track association, the computational load and time consumption will be substantial. Furthermore, tracks may terminate, and if the time interval between the latest states of two tracks is too long, changes in speed and heading information can easily lead to false associations. In these cases, association is unnecessary. Therefore, preliminary association thresholds are set for distance and time to exclude targets outside these thresholds.

[0095] Suppose that at time n, the j-th target track information of radar i is received, and the track state is represented as:

[0096]

[0097] If the track has not been associated with any system tracks before, then a coarse correlation determination based on time and position features is performed pairwise with all system tracks, and the time interval threshold is updated to σ. t The distance association threshold is σ z ;

[0098] The system track at this time is:

[0099]

[0100] in,

[0101] System track The current latest system track status is:

[0102]

[0103] If the following conditions are met:

[0104]

[0105] It is assumed that the target track received at time n from radar i is the same as the system track. Related;

[0106] Through the above process, preliminary track association is performed on the first intermediate processing information to obtain preliminary association information; preliminary track association can reduce the amount of computation for subsequent track association and improve the efficiency of subsequent track association.

[0107] Perform further track association on the aforementioned preliminary association information:

[0108] The preliminary association information is re-associated using a modified K-Nearest Neighbor (MK-NN) trajectory association algorithm to obtain the second intermediate processing information;

[0109] Specifically:

[0110] The association status of a track is divided into an association period, an inspection period, and a maintenance period, and track association is handled through track association quality and disengagement quality.

[0111] During the association period, track association quality and disengagement quality are introduced. Within a continuous time period, all candidate tracks near the target track are evaluated one by one. If a candidate track meets the association threshold K times in the past N evaluations, it is considered to be associated. If multiple candidate tracks meet the condition, ambiguity processing is performed. When the time is greater than N, the inspection period begins. The target track and the selected tracks are subjected to three consecutive chi-square distribution tests. If all three tests pass, the two tracks become a fixed association pair, and the association is completed, entering the maintenance period. If one test fails, two more tests are performed. If three out of five tests pass, the association is considered complete; otherwise, the process returns to the association period.

[0112] In an optional embodiment of the present invention, step 14 includes:

[0113] Step 141: Based on the multi-source positioning information, obtain the arithmetic mean of the multi-source positioning information as a reference value;

[0114] Step 142: Obtain the dynamic variance based on the multi-source positioning information and the reference value.

[0115] In this embodiment, the multi-source positioning information is taken as radar system positioning information, automatic identification system positioning information, and BeiDou satellite navigation system positioning information;

[0116] Initially, the arithmetic mean of multi-source positioning information is used as the baseline value. Subsequently, the prediction result of the trajectory at the next time step is used by Kalman filtering as the baseline value for calculating the variance at the next time step. This can reduce the deviation of the baseline value and help improve the accuracy of the fusion result.

[0117] Specifically:

[0118]

[0119] Where y0(t) is the state at time t predicted by Kalman filtering at time t-1; y a (t) represents the positioning information observed by the automatic identification system; y r (t) represents the positioning information observed by the radar system; y b (t) represents the positioning information observed by the BeiDou Navigation Satellite System; To automatically identify the dynamic variance of the system; The dynamic variance of the radar system; This represents the dynamic variance of the BeiDou Navigation Satellite System.

[0120] In an optional embodiment of the present invention, step 15 includes:

[0121] Step 151: Determine the distribution coefficients k of the dynamic error and the static error, where 0 ≤ k ≤ 1;

[0122] Step 152: Based on the k value, process the dynamic variance and static variance to obtain the comprehensive variance;

[0123] Step 153: Obtain the comprehensive weights based on the comprehensive variance.

[0124] In this embodiment, the static variances of the preset automatic identification system, radar system, and BeiDou satellite navigation system, calculated based on historical data or nominal accuracy, are respectively... and

[0125] Among them, static error is the slow drift of the measurement result (or output value) when the measured value (or input value) does not change over time. It belongs to systematic error and is given directly.

[0126] Based on the k-value, the dynamic and static variances are processed to obtain the comprehensive variance; specifically:

[0127]

[0128] Based on the k value, the proportions of dynamic variance and static variance are adjusted to obtain the comprehensive variance, which can reduce the bias of the comprehensive variance and thus improve the accuracy of subsequent fusion results.

[0129] After obtaining the comprehensive variance, the weights are calculated according to the principle that the larger the error, the smaller the weight, and the sum of the weights is 1:

[0130]

[0131] Among them, w a The comprehensive weights for the automatic identification system; w r The overall weighting of the radar system; w b This represents the comprehensive weighting of the BeiDou Navigation Satellite System.

[0132] In an optional embodiment of the present invention, the value of k is the output result obtained by inputting the second intermediate processing information into the distribution coefficient determination model;

[0133] The training process for determining the distribution coefficient model includes:

[0134] Obtain training multi-source localization information;

[0135] The training multi-source localization information is transformed to obtain the first training intermediate processing information;

[0136] The first training intermediate processing information is correlated with the flight path to obtain the second training intermediate processing information.

[0137] The second training intermediate processing information is input into the input layer of the preset neural network model for processing to obtain the first layer output;

[0138] The first layer output and the first target parameter are input into the hidden layer for processing to obtain the second layer output.

[0139] The second layer output and the second target parameter are input to the output layer for processing to obtain the distribution coefficient determination model.

[0140] In this embodiment, the distribution coefficient k represents the proportion of the random component and the fixed component of the error during the error calculation process. According to the error calculation principle, the smoothness of the system's observed data over a period of time is used to judge the error situation. The principle is as follows: if the observed data is smooth over a period of time, it means that the data is less affected by noise. When calculating the error, the proportion of the random component should be reduced and the proportion of the fixed component should be increased. That is, when the value of k is in the range of 0 to 1, k should be appropriately reduced to be closer to 0. Similarly, if the observed data changes significantly over a period of time, the proportion of the random component should be increased, and k will be closer to 1.

[0141] The smoothness of the observed data can be reflected by the standard deviation of the slope, calculated using the following formula:

[0142] sr i =x i+1 -x i

[0143]

[0144] Among them, sr i The slope is calculated by subtracting elements from a shifted array. denoted as the mean of the slope, sd as the standard deviation of the slope, and N as the number of observations.

[0145] To ensure the stability of the fusion process and the accuracy of the fusion results, this application inputs the second intermediate processing information into the distribution coefficient determination model, and uses the output of the distribution coefficient determination model as the distribution coefficient k value of the dynamic error and the static error.

[0146] like Figure 2 As shown, the distribution coefficient determination model has three layers: an input layer, a hidden layer, and an output layer.

[0147] The training multi-source localization information is transformed to obtain the first training intermediate processing information, which has been described in step 12 and will not be repeated here.

[0148] The first training intermediate processing information is correlated with the flight path to obtain the second training intermediate processing information, which has been described in step 13 and will not be repeated here.

[0149] like Figure 3 As shown, the specific training process of the distribution coefficient determination model is as follows:

[0150] S1. Initialize network parameters: including network structure, number of layers, number of nodes per layer; initialize input / output vectors (X, D), and connection weights ω from the input layer to the hidden layer. ih The connection weight ω from the hidden layer to the output layer hj The network learning rate η, the momentum coefficient for weight changes α, the network tolerance error accuracy eps, and the number of iterations K;

[0151] S2. Select a pattern sample and begin the forward propagation calculation:

[0152] First, calculate the hidden layer output:

[0153]

[0154] in,

[0155] Then calculate the output of the output layer:

[0156]

[0157] in, Or φ(x) = x;

[0158] S3, Reverse calculation of neuron error:

[0159] First, calculate the output layer error:

[0160]

[0161] Then calculate the hidden layer error:

[0162]

[0163] S4, Update weights:

[0164] New weights from hidden layer to output layer:

[0165]

[0166] New weights from the input layer to the hidden layer:

[0167]

[0168] S5. Calculate the network output error:

[0169]

[0170] If the result is greater than eps, proceed to S2; otherwise, the program ends and training is complete.

[0171] Since the value of k is limited to between 0 and 1, even if the output of the distribution coefficient determination model differs greatly from the actual situation, or even if k takes the value of 1 or 0 in extreme cases, the error is entirely composed of random or fixed components, and its impact on the final result is limited. The model can still output a result that ensures a certain level of accuracy. Therefore, by inputting the second intermediate processing information into the distribution coefficient determination model and using the output of the distribution coefficient determination model as the distribution coefficient k value for dynamic and static errors, the stability of the fusion result can be guaranteed. This reduces the amount of computation and makes full use of historical and real-time information, thereby further improving the accuracy of ship positioning.

[0172] In an optional embodiment of the present invention, step 16 includes:

[0173] Step 161, using the formula:

[0174] The second intermediate processing information is fused to obtain the target positioning information of the ship;

[0175] in, The result is the data fusion, where N is the number of sensor types fused, and y is the fusion result. j For each type of sensor, w represents the second intermediate processing information. j This represents the overall weighting value for each type of sensor.

[0176] In this embodiment, static and dynamic weights are combined. While historical data or nominal accuracy reflects the accuracy of each positioning system, the dynamic error is obtained by calculating the real-time variance. The proportion of static error for each system is determined by identifying the k-value. Based on the k-value, the dynamic and static variances are processed to obtain the comprehensive variance, ultimately yielding the comprehensive weight. This comprehensive weight is then used for fusion processing to obtain the ship's target positioning information. This approach simultaneously considers both historical and real-time system data, maximizing the extraction of effective information from the positioning data of each system and improving the accuracy of the positioning fusion results.

[0177] like Figure 4 As shown, the information processing process of the multi-source ship positioning information processing method is as follows:

[0178] The system acquires the ship's positioning information from the Automatic Identification System, radar system, and BeiDou Navigation Satellite System, which serves as the ship's multi-source positioning information.

[0179] The multi-source positioning information is denoised as preprocessing to obtain denoised preprocessed information; the preprocessed information is then subjected to coordinate transformation and time alignment to obtain first intermediate processing information.

[0180] Perform preliminary track association on the first intermediate processing information to obtain preliminary association information; perform further track association on the preliminary association information to obtain the second intermediate processing information;

[0181] Based on the multi-source positioning information, the arithmetic mean of the multi-source positioning information is obtained as a reference value; based on the multi-source positioning information and the reference value, the dynamic variance is obtained; the second intermediate processing information is input into the distribution coefficient determination model, and the output result is used as the distribution coefficient k of the dynamic error and the static error; based on the value of k, the dynamic variance and the preset static variance are processed to obtain the comprehensive variance;

[0182] Based on the comprehensive variance, the comprehensive weights are obtained;

[0183] Based on the comprehensive weights, the second intermediate processing information is fused to obtain a preliminary fusion result;

[0184] The preliminary fusion results are filtered and calculated to obtain the fusion result at the current time and the predicted value at the next time. The predicted value at the next time is used as the reference value for the next time to calculate the dynamic variance at the next time. The fusion result at the current time is used as the target positioning information of the ship at the current time.

[0185] Specific filtering calculations:

[0186] S1. Update the state transition matrix F, whose value is related to the previous update time l of the system track and the current radar track time k.时 Related to the interval:

[0187]

[0188] Where T = k 时 -1;

[0189] S2. Predicting the state and calculating the innovation (residual):

[0190]

[0191] in, The track status of the m-th system track at the previous update time l; For it in the current k 时 The predicted trajectory status at any given moment;

[0192] The predicted flight path position is:

[0193] The observation matrix is:

[0194]

[0195] The position value of the j-th track of radar i is:

[0196] z ij (k 时 )=Hx ij (k 时 );

[0197] Where, x ij (k 时 () represents the system track of radar i at the previous update time k. 时 The flight path status;

[0198] The new information can be obtained by subtracting the two position values:

[0199]

[0200] S3. Update and correct the observation noise covariance matrix R:

[0201] The measurement error covariance matrix of radar i is:

[0202]

[0203] Among them, the range and azimuth errors of the radar are both zero-mean Gaussian white noise and are independent of each other; and These represent their standard deviations. When radar track information is transformed from polar coordinates to rectangular coordinates, there is a nonlinear transformation relationship as follows. There is also a nonlinear transformation relationship between the actual radar measurement error and the measurement error in the rectangular coordinate system. Measurement error plays a crucial role in the Kalman filtering process, so it is also important to transform the measurement error and use a more accurate measurement error covariance matrix.

[0204] The position value of radar i in the rectangular coordinate system for the j-th track is:

[0205] z ij =[x ij ;y ij ];

[0206] The observation noise covariance matrix is:

[0207]

[0208] Expanding the above equation:

[0209]

[0210] Among them, (x' ij ,y' ij () represents the actual Cartesian coordinate system position;

[0211] According to trigonometric identities, the mean square error of the rectangular coordinate system is:

[0212]

[0213] The two are not independent of each other; the measurement error of each direct coordinate system is determined by the actual target distance, orientation, distance error, and orientation error.

[0214] When the polar coordinate measurement noise is zero-mean Gaussian white noise, we have:

[0215]

[0216] The average error is:

[0217]

[0218] After performing algebraic transformations, the elements of R are obtained as follows:

[0219]

[0220] S4. Prediction error autocorrelation matrix and autocorrelation matrix for calculating innovation (residuals):

[0221]

[0222] in, Let m be the autocorrelation matrix of the state error of the system trajectory m at time 1; For the system trajectory m in k 时 The autocorrelation matrix of the predicted state error at time t; Q is the covariance matrix of the process noise;

[0223] The corresponding information autocorrelation matrix is:

[0224]

[0225] S5. Calculate the filter gain:

[0226]

[0227] S6. Update the state and error autocorrelation matrix:

[0228]

[0229] Only one radar track can be updated at a time. Also, updating the state error autocorrelation matrix is ​​more suitable for updating tracks based on a single measurement:

[0230]

[0231] In practical applications, to reduce the computational load on navigating vessels and identify fault data, an adaptive fusion algorithm can be selected before performing fusion calculations; for example... Figure 5 As shown, the relationship between system data is judged based on real-time observation values, and it is determined whether the track distance is too close or too far: (1) When the track distance is too close, no fusion is performed, and the positioning information of the system with the highest accuracy is directly output; (2) When the track difference is too far, if the data source is found to be faulty, the faulty data source is removed and the calculation is performed; (3) If there is a certain difference between the tracks, but it does not reach the threshold, the fusion calculation process is entered and the fusion result is output.

[0232] like Figure 6 As shown, embodiments of the present invention also provide a multi-source ship positioning information processing device 60, comprising:

[0233] Module 61 is used to acquire multi-source positioning information of the ship;

[0234] The processing module 62 is used to convert and process the multi-source positioning information to obtain first intermediate processing information; perform track association on the first intermediate processing information to obtain second intermediate processing information; obtain the dynamic variance of the multi-source positioning information based on the multi-source positioning information; obtain a comprehensive weight based on the dynamic variance and a preset static variance; and perform fusion processing on the second intermediate processing information based on the comprehensive weight to obtain the target positioning information of the ship.

[0235] Optionally, the multi-source positioning information is transformed to obtain first intermediate processing information, including:

[0236] The multi-source positioning information is denoised to obtain denoised preprocessed information;

[0237] The preprocessed information is subjected to coordinate transformation and time alignment to obtain the first intermediate processing information.

[0238] Optionally, the first intermediate processing information is correlated with flight paths to obtain second intermediate processing information, including:

[0239] Perform preliminary track association on the first intermediate processing information to obtain preliminary association information;

[0240] The preliminary association information is further associated with the flight path to obtain the second intermediate processing information.

[0241] Optionally, based on the multi-source positioning information, the dynamic variance of the multi-source positioning information is obtained, including:

[0242] Based on the multi-source positioning information, the arithmetic mean of the multi-source positioning information is obtained as a reference value;

[0243] The dynamic variance is obtained based on the multi-source positioning information and the reference value.

[0244] Optionally, based on the dynamic variance and the preset static variance, a comprehensive weight value is obtained, including:

[0245] Determine the distribution coefficients k of dynamic and static errors, where 0 ≤ k ≤ 1;

[0246] Based on the k value, the dynamic variance and static variance are processed to obtain the comprehensive variance;

[0247] Based on the comprehensive variance, the comprehensive weights are obtained.

[0248] Optionally, the value of k is the output result obtained by inputting the second intermediate processing information into the distribution coefficient determination model;

[0249] The training process for determining the distribution coefficient model includes:

[0250] Obtain training multi-source localization information;

[0251] The training multi-source localization information is transformed to obtain the first training intermediate processing information;

[0252] The first training intermediate processing information is correlated with the flight path to obtain the second training intermediate processing information.

[0253] The second training intermediate processing information is input into the input layer of the preset neural network model for processing to obtain the first layer output;

[0254] The first layer output and the first target parameter are input into the hidden layer for processing to obtain the second layer output.

[0255] The second layer output and the second target parameter are input to the output layer for processing to obtain the distribution coefficient determination model.

[0256] Optionally, based on the comprehensive weights, the second intermediate processing information is fused to obtain the target positioning information of the ship, including:

[0257] Through the formula:

[0258] The second intermediate processing information is fused to obtain the target positioning information of the ship;

[0259] in, The result is the data fusion, where N is the number of sensor types fused, and y is the fusion result. j For each type of sensor, w represents the second intermediate processing information. j This represents the overall weighting value for each type of sensor.

[0260] It should be noted that this device is the same as the method described above. All implementations in the above method embodiments are applicable to the embodiments of this device and can achieve the same technical effect.

[0261] An embodiment of the present invention also provides a computing device, including: a processor and a memory storing a computer program, wherein the computer program, when executed by the processor, performs the method described in the above embodiments. All implementations in the above method embodiments are applicable to this embodiment and can achieve the same technical effect.

[0262] In this embodiment of the invention, a computer-readable storage medium is also provided, storing instructions that, when executed on a computer, cause the computer to perform the method described in the above embodiments. All implementations of the methods described in the above embodiments are applicable to this embodiment and can achieve the same technical effect.

[0263] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0264] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0265] In the embodiments provided by this invention, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative. For instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0266] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0267] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0268] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, essentially, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, ROM, RAM, magnetic disks, or optical disks.

[0269] Furthermore, it should be noted that in the apparatus and method of the present invention, it is obvious that the components or steps can be decomposed and / or recombined. These decompositions and / or recombinations should be considered equivalent solutions of the present invention. Moreover, the steps performing the above-described series of processes can naturally be executed in the order described, but are not necessarily required to be executed in chronological order; some steps can be executed in parallel or independently of each other. Those skilled in the art will understand that all or any step or component of the method and apparatus of the present invention can be implemented in any computing device (including processors, storage media, etc.) or network of computing devices, in hardware, firmware, software, or a combination thereof. This is something that those skilled in the art can achieve by using their basic programming skills after reading the description of the present invention.

[0270] Therefore, the object of the present invention can also be achieved by running a program or a set of programs on any computing device. The computing device can be a known general-purpose device. Therefore, the object of the present invention can also be achieved simply by providing a program product containing program code implementing the method or apparatus. That is, such a program product also constitutes the present invention, and the storage medium storing such a program product also constitutes the present invention. Obviously, the storage medium can be any known storage medium or any storage medium developed in the future. It should also be noted that in the apparatus and method of the present invention, it is obvious that the components or steps can be decomposed and / or recombined. These decompositions and / or recombinations should be considered equivalent to the present invention. Furthermore, the steps performing the above series of processes can naturally be performed in the order described, but are not necessarily required to be performed in chronological order. Some steps can be performed in parallel or independently of each other.

[0271] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for processing multi-source ship positioning information, characterized in that, include: Acquire multi-source positioning information of ships; The multi-source positioning information is converted and processed to obtain first intermediate processing information; The first intermediate processing information is correlated with the flight path to obtain the second intermediate processing information; Based on the multi-source positioning information, the dynamic variance of the multi-source positioning information is obtained; Based on the dynamic variance and the preset static variance, the comprehensive weights are obtained; Based on the comprehensive weights, the second intermediate processing information is fused to obtain the target positioning information of the ship; The comprehensive weights are obtained based on the dynamic variance and the preset static variance, including: Determine the distribution coefficients k of dynamic and static errors, where 0 ≤ k ≤ 1; Based on the k value, the dynamic variance and static variance are processed to obtain the comprehensive variance; Based on the comprehensive variance, the comprehensive weights are obtained; Wherein, k is the output result obtained by inputting the second intermediate processing information into the distribution coefficient determination model; The training process for determining the distribution coefficient model includes: Obtain training multi-source localization information; The training multi-source localization information is transformed to obtain the first training intermediate processing information; The first training intermediate processing information is correlated with the flight path to obtain the second training intermediate processing information. The second intermediate training information is input into the input layer of the preset neural network model for processing to obtain the first layer output; The first layer output and the first target parameter are input into the hidden layer for processing to obtain the second layer output. The second layer output and the second objective parameter are input into the output layer for processing to obtain the distribution coefficient determination model. Specifically, based on the comprehensive weights, the second intermediate processing information is fused to obtain the target positioning information of the ship, including: Through the formula: The second intermediate processing information is fused to obtain the target positioning information of the ship; in, The result is the data fusion, where N is the number of sensor types fused. This is the second intermediate processing information corresponding to each sensor. This represents the overall weighting value for each type of sensor.

2. The multi-source ship positioning information processing method according to claim 1, characterized in that, The multi-source positioning information is transformed to obtain first intermediate processing information, including: The multi-source positioning information is denoised to obtain denoised preprocessed information; The preprocessed information is subjected to coordinate transformation and time alignment to obtain the first intermediate processing information.

3. The multi-source ship positioning information processing method according to claim 1, characterized in that, The first intermediate processing information is correlated with flight paths to obtain the second intermediate processing information, including: Perform preliminary track association on the first intermediate processing information to obtain preliminary association information; The preliminary association information is further associated with the flight path to obtain the second intermediate processing information.

4. The multi-source ship positioning information processing method according to claim 1, characterized in that, Based on the multi-source positioning information, the dynamic variance of the multi-source positioning information is obtained, including: Based on the multi-source positioning information, the arithmetic mean of the multi-source positioning information is obtained as a reference value; The dynamic variance is obtained based on the multi-source positioning information and the reference value.

5. A multi-source ship positioning information processing device, characterized in that, include: The acquisition module is used to acquire multi-source positioning information of the vessel; The processing module is used to convert and process the multi-source positioning information to obtain first intermediate processing information; The first intermediate processing information is correlated with a flight path to obtain the second intermediate processing information; the dynamic variance of the multi-source positioning information is obtained based on the multi-source positioning information; a comprehensive weight is obtained based on the dynamic variance and a preset static variance; and the second intermediate processing information is fused based on the comprehensive weight to obtain the target positioning information of the ship. The comprehensive weights are obtained based on the dynamic variance and the preset static variance, including: Determine the distribution coefficients k of dynamic and static errors, where 0 ≤ k ≤ 1; Based on the k value, the dynamic variance and static variance are processed to obtain the comprehensive variance; Based on the comprehensive variance, the comprehensive weights are obtained; Wherein, k is the output result obtained by inputting the second intermediate processing information into the distribution coefficient determination model; The training process for determining the distribution coefficient model includes: Obtain training multi-source localization information; The training multi-source localization information is transformed to obtain the first training intermediate processing information; The first training intermediate processing information is correlated with the flight path to obtain the second training intermediate processing information. The second intermediate training information is input into the input layer of the preset neural network model for processing to obtain the first layer output; The first layer output and the first target parameter are input into the hidden layer for processing to obtain the second layer output. The second layer output and the second objective parameter are input into the output layer for processing to obtain the distribution coefficient determination model. Specifically, based on the comprehensive weights, the second intermediate processing information is fused to obtain the target positioning information of the ship, including: Through the formula: The second intermediate processing information is fused to obtain the target positioning information of the ship; in, The result is the data fusion, where N is the number of sensor types fused. This is the second intermediate processing information corresponding to each sensor. This represents the overall weighting value for each type of sensor.

6. A computing device, characterized in that, include: A processor, a memory storing a computer program, wherein the computer program, when executed by the processor, performs the method as described in any one of claims 1 to 4.

7. A computer-readable storage medium, characterized in that, The system stores instructions that, when executed on a computer, cause the computer to perform the method as described in any one of claims 1 to 4.

Images