A Beidou positioning data intelligent fusion processing system

Abstract

The invention belongs to the technical field of high-precision positioning and navigation. The invention discloses a Beidou positioning data intelligent fusion processing system, which aims to enhance the accuracy of a positioning system in a complex environment and the adaptability to fast-moving targets and a changeable environment. The system comprises: obtaining a relative distance estimate by calculating a preliminary relative distance and correcting the preliminary relative distance with a correction value; analyzing a spatial geometric relationship according to the relative distance estimate to obtain a spatial feature vector; obtaining a positioning error model according to spatial feature vector training, predicting an expected positioning error under the current environment, judging whether a positioning error of a positioning terminal is abnormal according to the expected positioning error, and judging whether a multipath effect is an interference factor; making corrections according to the judgment result and adjusting the relative distance estimate; obtaining a high-precision positioning coordinate according to the adjusted relative distance estimate, and realizing high-precision and real-time positioning in a dynamic environment.

Description

Technical Field

[0001] The present invention relates to the field of high-precision positioning and navigation technology, and more specifically, to a Beidou positioning data intelligent fusion processing system. Background Art

[0002] The Beidou positioning and navigation satellite system is widely used in various fields, such as transportation, logistics, agriculture, etc. Traditional Beidou positioning methods mainly rely on single-point positioning, which calculates the terminal's position by receiving signals from multiple satellites. However, although these methods perform well in certain scenarios, there are still some technical challenges that have not been fully overcome in complex and dynamic environments.

[0003] Although existing technologies have made significant progress in improving positioning accuracy, stability and response speed, such as using Beidou positioning terminals combined with adaptive filters or spatial diversity reception technology to combat multipath effects, differential GPS to reduce errors, and dynamically adjusting sensor weights to improve positioning accuracy, when dealing with positioning in complex environments, such as urban canyons or changeable weather conditions, there is still a problem of difficulty in accurately distinguishing normal errors from interference factors such as multipath effects, resulting in unstable positioning results; for high-speed moving objects and in dynamic environments, existing methods fail to fully combine the motion characteristics of the positioning terminal (such as speed, acceleration, and angular velocity of rotation), and may not be able to provide timely and accurate positioning updates, thereby affecting positioning accuracy. These shortcomings limit the performance of high-precision positioning in complex and dynamic environments. Summary of the invention

[0004] In order to overcome the above-mentioned defects of the prior art and to achieve the above-mentioned purpose, the present invention provides the following technical solutions: a Beidou positioning data intelligent fusion processing system, comprising:

[0005] Relative distance calculation module: collects the initial position of the positioning terminal and the position of the communication satellite, and uses the Euclidean distance method to calculate the initial relative distance between the positioning terminal and each communication satellite; uses the correction value to correct the initial relative distance to obtain the relative distance estimation value between the positioning terminal and each communication satellite;

[0006] Feature vector acquisition module: used to obtain the estimated position of the positioning terminal according to the relative distance estimation value through the multi-point positioning algorithm, and calculate the spatial feature vector according to the estimated position;

[0007] Interference factor determination module: trains the positioning error model according to the spatial feature vector to obtain the expected positioning error in the current environment. Based on the expected positioning error, determines whether the positioning error of the positioning terminal is abnormal. If abnormal, determines whether the multipath effect is an interference factor.

[0008] Relative distance adjustment module: used to correct interference factors using smoothing technology according to the judgment results of interference factors, and adjust the relative distance estimation value according to the expected positioning error;

[0009] Positioning coordinate correction module: used to obtain the environment-corrected distance based on the adjusted relative distance estimate, and combine the environment-corrected distance with the position of the communication satellite to obtain an accurate position estimate as the high-precision positioning coordinates of the positioning terminal.

[0010] Furthermore, the relative distance estimate is calculated by:

[0011] Mark all satellites that are in communication connection with the positioning terminal as communication satellites, receive communication satellite signals through the positioning terminal, measure the time it takes for the signal of each communication satellite to propagate from the corresponding communication satellite to the positioning terminal, and multiply it by the speed of light to obtain a pseudo-range measurement value;

[0012] The pseudo-range measurement values ​​of all communication satellites at time k are horizontally spliced ​​to form a pseudo-range vector; the pseudo-range vector is subjected to the least square method to obtain the initial position of the positioning terminal;

[0013] The position of the communication satellite at time t is obtained using the satellite orbit model, and the initial relative distance between the positioning terminal and each communication satellite is calculated using the Euclidean distance formula in combination with the initial position of the positioning terminal and the position of the communication satellite.

[0014] According to the initial position and maximum communication range of the positioning terminal, the correction value of each differential base station within the maximum communication range of the positioning terminal at time t is obtained. According to the correction value, the initial relative distance is adjusted to obtain the relative distance estimate between the positioning terminal and each communication satellite. ;

[0015] in, represents the relative distance estimate, d represents the preliminary relative distance, n represents the number of differential base stations, represents the correction value of the pth differential base station, represents the weight of the pth differential base station correction value, , represents the variance of the pth differential base station correction value, and ϵ represents the regularization constant.

[0016] Furthermore, the correction value is obtained by:

[0017] Collect historical observation data sets, including preliminary relative distances, original correction values, and corresponding atmospheric condition data;

[0018] The original correction value is the sum of the satellite orbit error correction value and the clock error correction value downloaded from the differential base station through the network connection; the atmospheric condition data affecting the propagation of communication satellite signals, including the ionospheric activity index and tropospheric humidity, are obtained through ground meteorological stations;

[0019] Take the historical observation data set as input and the predicted correction value as output, and train a regression model to predict the correction value;

[0020] The initial relative distance at time t, the original correction value and the corresponding atmospheric condition data are input into the trained regression model to obtain the predicted correction value;

[0021] Based on the predicted correction value, an automatic adjustment factor based on atmospheric conditions is introduced to dynamically adjust the predicted correction value, and the state estimation is performed through the Kalman filter to obtain the correction value of the differential base station ;

[0022] in, Represents the correction value of the differential base station, represents the Kalman gain at time t, represents the predicted correction value obtained by the regression model, represents the observation vector, represents the system state vector, represents the automatic adjustment factor; , represents the regularization parameter of the atmospheric condition factor, exp represents the exponential function, It represents the average value of the atmospheric condition factor in dg moments before time t, where dg refers to the preset number of moments. represents the atmospheric condition factor at time t, represents the Euclidean norm, H represents the observation matrix;

[0023] The observation matrix H is obtained by calculating the difference between the observation vector and the observation noise and calculating the ratio of the difference to the system state vector. The observation noise is defined as zero-mean Gaussian white noise.

[0024] Furthermore, the method for obtaining the observation vector includes:

[0025] The carrier phase measurement values ​​of each communication satellite and the positioning terminal are obtained through real-time dynamic positioning technology;

[0026] Read the inertial measurement unit data in the positioning terminal, including the linear acceleration and rotational angular velocity of the positioning terminal;

[0027] The carrier phase measurement value, pseudorange measurement value, linear acceleration, rotational angular velocity, ionospheric activity index, tropospheric humidity, satellite orbit error correction value and clock error correction value at time t are horizontally spliced ​​to form the observation vector at time t ;

[0028] The system state vector can be obtained by:

[0029] The initial position of the positioning terminal at time t is directly expressed as ECEF coordinates;

[0030] Get the speed of the positioning terminal at time t, and decompose the speed at time t into the three axes of the ECEF coordinate system;

[0031] The position and velocity of the positioning terminal in the ECEF coordinate system are horizontally spliced ​​to form the system state vector of the positioning terminal at time t ;

[0032] The atmospheric condition factors can be obtained by:

[0033] The atmospheric condition factor is calculated by obtaining the tropospheric wet delay and tropospheric dry delay data from ground meteorological stations and combining them with the ionospheric activity index.

[0034] ;

[0035] in, represents the ionospheric activity index at time t, represents the average value of the ionospheric activity index within dg moments before moment t, represents the standard deviation of the ionospheric activity index within dg moments before moment t, represents the regularization parameter of ionospheric activity optimized by cross-validation method; and represent the tropospheric wet delay and tropospheric dry delay at time t, respectively; and They represent the average values ​​of the tropospheric wet delay and the tropospheric dry delay within dg moments before moment t, represents the regularization parameter of tropospheric activity optimized by cross-validation.

[0036] Furthermore, the spatial feature vector is obtained in the following manner:

[0037] The position of each communication satellite is obtained through the communication satellite signal, and the estimated position of the positioning terminal is obtained by solving the multi-point positioning algorithm using the relative distance estimation value of all communication satellites. : ,in, Represents the i-th relative estimated distance , represents the position of the i-th communication satellite , I represents the number of communication satellites;

[0038] Based on the estimated location of the positioning terminal , take the difference between the position of each communication satellite and the estimated position of the positioning terminal as the vector r from the positioning terminal to the communication satellite;

[0039] For each communication satellite, calculate the line-of-sight angle and azimuth angle of the communication satellite relative to the positioning terminal;

[0040] Viewing angle , azimuth ;

[0041] in, represents the line-of-sight angle of the i-th communication satellite relative to the positioning terminal, represents the azimuth of the i-th communication satellite relative to the positioning terminal, represents the vector r from the positioning terminal to the i-th communication satellite, Represents vector The module length, They represent the three components of the x-axis, y-axis and z-axis of the vector from the positioning terminal to the i-th communication satellite in the ECEF coordinate system, , T represents transpose;

[0042] Using digital elevation models and urban building databases, reflective surfaces that may cause multipath effects are identified, including reflective surfaces of tall buildings, bridges, and water surfaces, and the distance of the reflective surface relative to the positioning terminal, the material of the reflective surface, and the angle at which the signal reaches the reflective surface from the communication satellite.

[0043] The multipath score of each reflector is calculated based on the distance of the reflector relative to the positioning terminal, the material of the reflector, and the angle at which the signal reaches the reflector from the communication satellite. ,in, represents the multipath score of the fs-th reflector, Represents the reflection coefficient of the fs-th reflection surface material, represents the distance from the positioning terminal to the fsth reflection surface, represents the incident angle of the communication satellite signal reaching the fsth reflection surface, cs and Indicates a preset constant;

[0044] According to the observation matrix H, the geometric dilution factor is obtained ,in, represents the trace of a matrix;

[0045] The line-of-sight angle and azimuth of each communication satellite, the multipath score of each reflection surface, and the geometric dilution factor are standardized and then horizontally spliced ​​to form a spatial feature vector.

[0046] Furthermore, the training method of the positioning error model includes:

[0047] Step 61: Collect a sample set, including the spatial feature vector of the YB group and the corresponding true positioning error, and divide the sample set into a training set and a validation set in proportion;

[0048] Among them, the actual positioning error is the position of the positioning terminal The difference between the actual physical position and the preset known fixed reference point is taken as the actual physical position;

[0049] Step 62: construct a positioning error model, taking the spatial feature vector as input and the expected positioning error as output, wherein the positioning error model is a gradient boosting tree model;

[0050] Step 63: Initialize the hyperparameters of the gradient boosting tree model, use network search to tune the hyperparameters, use k-fold cross validation to evaluate the cross validation scores of the model under different hyperparameter combinations, and select the optimal parameter combination;

[0051] Step 64: Use the optimal parameter combination as the initial parameters of the model, and define the mean absolute error as the loss function for evaluating the prediction accuracy of the model;

[0052] Step 65: Use the training set and loss function to adjust the model parameters according to the gradient in each iteration of the model;

[0053] Step 66: For each iteration, update the model parameters using the training set using R 2 The score is used as the evaluation indicator and R is calculated on the validation set. 2 Score value;

[0054] According to the R 2 Score value, calculate the R after the current iteration 2 The score value is the same as the R of the previous iteration. 2 The difference in the score values ​​is recorded as the iteration difference;

[0055] Step 67: setting an iteration difference threshold, if the iteration difference is greater than the iteration difference threshold, it is determined that the performance of the model is improved;

[0056] If the iteration difference is less than or equal to the iteration difference threshold, it is determined that the performance of the model has not been improved;

[0057] If the performance of the model on the validation set does not improve in consecutive DC iterations, the training is stopped to obtain a trained positioning error model.

[0058] Furthermore, the method of determining whether the positioning error of the positioning terminal is abnormal, and if abnormal, further determining whether the multipath effect is an interference factor includes:

[0059] Set a positioning error threshold and compare the expected positioning error at the current moment with the positioning error threshold;

[0060] If the expected positioning error is less than or equal to the positioning error threshold, the error of the positioning terminal is determined to be normal, and the current relative distance estimation value continues to be used;

[0061] If the expected positioning error is greater than the positioning error threshold, the error of the positioning terminal is determined to be abnormal;

[0062] According to the error anomaly of the positioning terminal, a multipath effect threshold is set, the average of the multipath scores of all reflection surfaces is taken, and the average multipath score is compared with the multipath effect threshold;

[0063] If the multipath score mean is less than or equal to the multipath effect threshold, it is determined that the multipath effect is not an interference factor;

[0064] If the multipath score mean is greater than the multipath effect threshold, the multipath effect is determined to be an interference factor.

[0065] Furthermore, the relative distance estimation value is adjusted in the following manner:

[0066] If the multipath effect is an interference factor, the pseudorange measurement value is corrected by combining the carrier phase measurement value with the smoothing technology to obtain the smoothed pseudorange value.

[0067] ;

[0068] in, represents the smoothed pseudorange value of D at the current moment, represents the pseudo-range measurement value of D at the current moment, represents the pseudorange measurement value at time D-1, Indicates the absolute difference between the D carrier phase measurement value at time D-1 and the current time; represents the smoothed pseudorange value at time D-1, and is defined as is the pseudo-range measurement value at the first moment in the monitoring period, Indicates the preset carrier phase smoothing factor, Represents the preset exponential smoothing factor;

[0069] Use the smoothed pseudorange value as the new pseudorange measurement value to re-obtain the expected positioning error, and adjust the relative distance estimate based on the re-obtained expected positioning error ;

[0070] in, represents the adjusted relative distance estimate, It represents the expected mean positioning error of the expected positioning error within dg moments before the current moment. It represents the expected standard value of the expected positioning error within dg moments before the current moment, and zz represents the preset adjustment scale factor.

[0071] Furthermore, the environment correction distance is obtained in the following manner:

[0072] The linear acceleration and rotational angular velocity of the positioning terminal are used as the motion characteristics of the positioning terminal;

[0073] By integrating the angular velocity, the rotation matrix is ​​obtained. According to the adjusted relative distance estimate, the environment-corrected distance is calculated by combining the rotation matrix with the linear acceleration. ,in, represents the environmental correction distance, Represents the distance correction calculated by combining the rotation matrix and linear acceleration;

[0074] ,in, represents the estimated position of the positioning terminal at the current time D, represents the linear acceleration of the positioning terminal at the current time D, represents the speed of the positioning terminal at the current time D, represents the rotation matrix, Represents the time interval from time D-1 to the current time D.

[0075] Furthermore, the method of obtaining the high-precision positioning coordinates of the positioning terminal includes:

[0076] The extended Kalman filter is used to calculate the precise position estimate by correcting the distance by the environment. ,in, represents the environmentally corrected distance between the ith communication satellite and the positioning terminal, represents the extended Kalman filter, Represents the precise position estimate of the positioning terminal;

[0077] The precise position estimate is used as the high-precision positioning coordinates of the positioning terminal.

[0078] The technical effects and advantages of the Beidou positioning data intelligent fusion processing system of the present invention are as follows:

[0079] The present invention effectively improves the high-precision positioning performance in complex and dynamic environments by intelligently identifying and processing interference factors, combining terminal motion characteristics, fine-tuning spatial environment modeling, and real-time pseudo-range adjustment, and provides more stable and accurate positioning results.

[0080] First, the correction value provided by the differential base station is used to further correct the initial relative distance, reduce the influence of atmospheric effect and satellite orbit error, and provide a more accurate initial relative distance estimate;

[0081] Next, by analyzing the spatial geometric relationship between the positioning terminal and each communication satellite, the spatial feature vector is obtained to provide accurate data support for the subsequent interference factor determination and adjustment, thus enhancing the reliability of the error model.

[0082] Subsequently, based on the positioning error model obtained through training, it intelligently identifies and processes interference factors such as multipath effects to ensure the stability and reliability of positioning results, especially in complex environments.

[0083] Then, the Kalman filter is used for dynamic adjustment to compensate the pseudo-range measurement value in real time by the expected positioning error, thus avoiding error accumulation and improving the anti-interference ability and stability of the system.

[0084] Finally, the environment-corrected distance is calculated based on the terminal's motion characteristics (such as speed, acceleration, and angular velocity of rotation), and combined with the satellite position to obtain the final high-precision positioning coordinates, achieving high-precision, real-time positioning in a dynamic environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0085] Figure 1 A schematic diagram of a Beidou positioning data intelligent fusion processing system of the present invention;

[0086] Figure 2 A schematic diagram of a Beidou positioning data intelligent fusion processing method of the present invention;

[0087] Figure 3 A schematic diagram of obtaining correction values ​​for a Beidou positioning data intelligent fusion processing system of the present invention. DETAILED DESCRIPTION

[0088] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0089] Embodiment 1

[0090] See also Figure 1 and Figure 3 As shown, the Beidou positioning data intelligent fusion processing system described in this embodiment includes:

[0091] Relative distance calculation module: collects the initial position of the positioning terminal and the position of the communication satellite, and uses the Euclidean distance method to calculate the initial relative distance between the positioning terminal and each communication satellite; uses the correction value to correct the initial relative distance to obtain the relative distance estimation value between the positioning terminal and each communication satellite;

[0092] Among them, based on the tropospheric wet delay data, the tropospheric dry delay data and the ionospheric activity index, the atmospheric condition factor is obtained by using the double exponential atmospheric effect correction method, the observation matrix is ​​constructed based on the observation vector and the system state vector, and the atmospheric condition factor and the observation matrix are combined to calculate the automatic adjustment factor using the real-time atmospheric condition adaptive correction method; based on the preliminary relative distance, the original correction value and the corresponding atmospheric condition data, the predicted correction value is predicted using the regression model, and based on the predicted correction value, the correction value is calculated using the Kalman filter through the automatic adjustment factor;

[0093] Feature vector acquisition module: used to obtain the estimated position of the positioning terminal through the multi-point positioning algorithm according to the relative distance estimation value, and calculate the spatial feature vector of the spatial environment around the positioning terminal according to the estimated position;

[0094] Interference factor determination module: obtains the positioning error model based on spatial feature vector training, predicts the expected positioning error in the current environment based on the positioning error model, and then determines whether the positioning error of the positioning terminal is abnormal. If abnormal, further determines whether the multipath effect is an interference factor;

[0095] Relative distance adjustment module: used to correct the pseudo-range measurement value according to the judgment result by using smoothing technology combined with the carrier phase measurement value to obtain a smoothed pseudo-range value, and adjust the relative distance estimation value according to the smoothed pseudo-range value through the expected positioning error;

[0096] Positioning coordinate correction module: Based on the adjusted relative distance estimate and the motion characteristics of the positioning terminal, the environment correction distance is calculated. The accurate position estimate is calculated by combining the environment correction distance and the position of the communication satellite as the high-precision positioning coordinates of the positioning terminal.

[0097] Relative distance estimates are calculated using:

[0098] Mark all satellites that are in communication connection with the positioning terminal as communication satellites, receive communication satellite signals through the positioning terminal, measure the time it takes for the signal of each communication satellite to propagate from the corresponding communication satellite to the positioning terminal, and multiply it by the speed of light to obtain a pseudo-range measurement value;

[0099] The pseudo-range measurement values ​​of all communication satellites at time k are horizontally spliced ​​to form a pseudo-range vector; the pseudo-range vector is subjected to the least square method to obtain the initial position of the positioning terminal;

[0100] It should be noted that in satellite positioning, the least squares method is used to perform least squares fitting on pseudo-range measurements of multiple satellites, gradually optimizing the position estimate of the positioning terminal, and finally obtaining the most likely position coordinates as the initial position of the positioning terminal;

[0101] Specifically, in satellite positioning, the positioning terminal determines its own position by calculating pseudo-range measurement values ​​from multiple communication satellites. Through the pseudo-range measurement values ​​of multiple communication satellites, a guessed position of the positioning terminal can be guessed. According to the current guessed position, the theoretical distance between the communication satellite and the guessed position can be obtained. Then, the difference between each pseudo-range measurement value and the theoretical distance is calculated, and the least squares method is used to fit the multiple differences to find out how to adjust the deviations to minimize these differences, adjust the deviations of the guessed position and update the guessed position. After multiple iterative adjustments, a final position estimate is obtained as the initial position of the positioning terminal.

[0102] Use the satellite orbit model (such as SP3 file) to get the position of the communication satellite at time t, combine the initial position of the positioning terminal and the position of the communication satellite, and calculate the preliminary relative distance between the positioning terminal and each communication satellite using the Euclidean distance formula;

[0103] According to the initial position and maximum communication range of the positioning terminal, the correction value of each differential base station within the maximum communication range of the positioning terminal at time t is obtained. According to the correction value, the initial relative distance is adjusted to obtain the relative distance estimate between the positioning terminal and each communication satellite. ,in, represents the relative distance estimate, d represents the preliminary relative distance, n represents the number of differential base stations, represents the correction value of the pth differential base station, represents the weight of the pth differential base station correction value, , represents the variance of the pth differential base station correction value, ϵ represents the regularization constant to prevent the denominator from being zero;

[0104] It should be noted that the formula for calculating the relative distance estimate is based on the original relative distance calculation formula The improvement is that by introducing a weight term for each correction term according to the uncertainty and confidence of each correction term, the importance of each correction term can be more accurately reflected and the influence of uncertain correction terms can be reduced. This method not only improves the accuracy of the relative distance calculation after correction, but also maintains the relative simplicity and operability of the formula, which is suitable for most practical applications.

[0105] The correction value of the differential base station at time t is obtained by:

[0106] Collect historical observation data sets, including preliminary relative distances, original correction values, and corresponding atmospheric condition data;

[0107] The original correction value is the sum of the satellite orbit error correction value and the clock error correction value downloaded from the differential base station through the network connection; the atmospheric condition data affecting the propagation of communication satellite signals are obtained through ground meteorological stations, including the ionospheric activity index (ionospheric electron density that affects the propagation of communication satellite signals) and tropospheric humidity;

[0108] Take the historical observation data set as input and the predicted correction value as output, and train a regression model to predict the correction value;

[0109] The initial relative distance at time t, the original correction value and the corresponding atmospheric condition data are input into the trained regression model to obtain the predicted correction value;

[0110] Based on the predicted correction value, an automatic adjustment factor based on atmospheric conditions is introduced to dynamically adjust the predicted correction value, and the state estimation is performed through the Kalman filter to obtain the correction value of the differential base station ;

[0111] in, Represents the correction value of the differential base station, represents the Kalman gain at time t. The Kalman filter is used to improve positioning accuracy and robustness. It represents the predicted correction value obtained by the regression model, which reduces the dependence on manually set parameters and enables the formula to be adaptively adjusted according to actual data. represents the observation vector, represents the system state vector, Represents the automatic adjustment factor. By introducing the automatic adjustment factor based on real-time atmospheric conditions, it ensures that the correction value can dynamically adapt to environmental changes. , represents the regularization parameter of the atmospheric condition factor, which is obtained through cross-validation optimization. exp represents the exponential function. It represents the average value of the atmospheric condition factor in dg moments before time t. dg refers to the preset number of moments, which is used to represent the value of each moment in a time period composed of multiple moments before time t. represents the atmospheric condition factor at time t, represents the Euclidean norm, H represents the observation matrix; the difference between the observation vector and the observation noise is calculated, and the difference is ratioed with the system state vector to obtain the observation matrix H, and the observation noise is defined as zero-mean Gaussian white noise;

[0112] This method not only improves the automation level of correction number calculation, but also enhances the adaptability and reliability of the system, and is suitable for high-precision positioning needs in various complex environments.

[0113] The methods for obtaining the observation vector include:

[0114] Obtain carrier phase measurements of each communication satellite and positioning terminal through real-time dynamic positioning technology (such as RTK) (referring to carrier phase data, a more accurate distance measurement method than pseudorange, usually used in differential GNSS (DGPS) or real-time dynamic positioning (RTK) to achieve centimeter-level or even millimeter-level positioning accuracy, which is essential for applications that require high-precision positioning, such as self-driving cars and precision agriculture);

[0115] Read the inertial measurement unit (IMU, which is usually installed on the positioning terminal to provide acceleration and angular velocity measurements, helping to estimate position changes in the absence of positioning signals or weak signals, enhancing the robustness of the system, and is particularly useful in tunnels, underground parking lots, etc., while assisting positioning updates in fast motion states) data in the positioning terminal, including the linear acceleration and rotational angular velocity of the positioning terminal;

[0116] The carrier phase measurement value, pseudorange measurement value, linear acceleration, rotational angular velocity, ionospheric activity index, tropospheric humidity, satellite orbit error correction value and clock error correction value at time t are horizontally spliced ​​to form the observation vector at time t ;

[0117] The system state vector can be obtained by:

[0118] The initial position of the positioning terminal at time t is directly expressed as ECEF coordinates (i.e., the earth-fixed coordinate system, with the origin located at the center of mass of the earth, the x-axis pointing to the intersection of the prime meridian and the equator, the y-axis pointing to the intersection of 90 degrees east longitude and the equator, and the z-axis pointing to the North Pole);

[0119] Get the speed of the positioning terminal at time t, and decompose the speed at time t into the three axes of the ECEF coordinate system;

[0120] The position and velocity of the positioning terminal in the ECEF coordinate system are horizontally spliced ​​to form the system state vector of the positioning terminal at time t ;

[0121] The atmospheric condition factors can be obtained by:

[0122] The atmospheric condition factor is calculated by obtaining the tropospheric wet delay and tropospheric dry delay data from ground meteorological stations and combining them with the ionospheric activity index.

[0123] ;

[0124] in, represents the ionospheric activity index at time t, represents the average value of the ionospheric activity index within dg moments before moment t, represents the standard deviation of the ionospheric activity index within dg moments before moment t, The regularization parameter representing the ionospheric activity is obtained by cross-validation optimization. and denote the tropospheric wet delay and tropospheric dry delay at time t, respectively. and They represent the average values ​​of the tropospheric wet delay and the tropospheric dry delay within dg moments before moment t, The regularization parameter representing the tropospheric activity is obtained by cross-validation optimization (the regularization parameter can be adjusted by and to adapt to different application scenarios and environmental conditions);

[0125] It should be noted that represents the part that calculates the effects of ionospheric activity, The part that represents the calculation of the influence of tropospheric activity. The two are combined and simplified to obtain a complete formula for calculating the atmospheric condition factor. This formula can be automatically adjusted according to the real-time observed ionospheric activity and tropospheric humidity, without the need to manually set complex parameters. By introducing the exponential function exp, it is ensured that the influence of atmospheric conditions on the correction number is within a reasonable range and gradually decreases as the degree of deviation from the average value increases;

[0126] Methods for analyzing the spatial geometric relationship between the positioning terminal and each communication satellite and extracting the spatial feature vector of the spatial environment around the positioning terminal include:

[0127] The position of each communication satellite is obtained through the communication satellite signal, and the estimated position of the positioning terminal is obtained by solving the multi-point positioning algorithm using the relative distance estimation value of all communication satellites. : ,in, Represents the i-th relative estimated distance , represents the position of the i-th communication satellite, I represents the number of communication satellites, express and The Euclidean distance between

[0128] Based on the estimated location of the positioning terminal , take the difference between the position of each communication satellite and the estimated position of the positioning terminal as the vector r from the positioning terminal to the communication satellite;

[0129] For each communication satellite, calculate the line-of-sight angle and azimuth angle of the communication satellite relative to the positioning terminal;

[0130] Viewing angle , azimuth ;

[0131] in, represents the line-of-sight angle of the i-th communication satellite relative to the positioning terminal, represents the azimuth of the i-th communication satellite relative to the positioning terminal, represents the vector from the positioning terminal to the i-th communication satellite, Represents vector The module length, They represent the three components of the vector from the positioning terminal to the i-th communication satellite in the ECEF coordinate system (Earth Center Earth Fixed Coordinate System), , T represents transposition, x, y and z represent the x-axis, y-axis and z-axis in the ECEF coordinate system respectively;

[0132] It should be noted that These components are used to calculate the elevation angle and azimuth angle to describe the spatial geometric relationship between the satellite and the positioning terminal. This can accurately describe the relative position between the positioning terminal and the communication satellite, which helps to further analyze their spatial geometric relationship. The use of the ECEF coordinate system ensures global consistency and comparability and is suitable for various positioning applications. This method not only improves the accuracy of spatial feature vector extraction, but also enhances the adaptability and reliability of the system. It is suitable for high-precision positioning needs in various complex environments. In addition, it provides solid technical support for future intelligent transportation, autonomous driving and other fields.

[0133] Using digital elevation models and urban building databases, reflective surfaces that may cause multipath effects are identified, including reflective surfaces of tall buildings, bridges, and water surfaces, and the distance of the reflective surface relative to the positioning terminal, the material of the reflective surface, and the angle at which the signal reaches the reflective surface from the communication satellite.

[0134] It should be noted that in a high-precision positioning system, the multipath effect is one of the important factors affecting the positioning accuracy. The multipath effect refers to the satellite signal reaching the positioning terminal after being reflected by the reflecting surface, resulting in additional path delay. In order to mitigate the impact of this effect, the digital elevation model (DEM) and the Urban Building Database (UBD) can be used to identify the reflecting surface and obtain relevant parameters, including the distance of the reflecting surface relative to the positioning terminal, the material of the reflecting surface, and the angle from the communication satellite to the reflecting surface;

[0135] Digital Elevation Model (DEM): provides terrain height information and is suitable for identifying reflective surfaces in natural environments; Urban Building Database (UBD): contains information such as the location, height, and shape of buildings and is suitable for identifying reflective surfaces in urban environments;

[0136] Extract the terrain height information around the positioning terminal from the DEM, generate a three-dimensional terrain model, calculate the line of sight (LOS) between the positioning terminal and each satellite, determine which terrain features may reflect the signal, and estimate the position of the reflection point for each possible reflection surface. This is usually based on the principle of geometric optics, assuming that the signal path propagates in a straight line and follows the law of reflection;

[0137] Extract information about buildings around the positioning terminal from the UBD, including location, height, and shape, and convert the building information into a three-dimensional geometric model. According to the location and direction of the building, screen out surfaces that may reflect the signal (for example, the wall or roof edge facing the positioning terminal). For each possible reflection surface, estimate the position of the reflection point, also based on the principle of geometric optics.

[0138] Once the position of the reflection point is determined, the distance formula in three-dimensional space can be used to calculate the distance of the reflection surface relative to the positioning terminal, and then the line of sight angle from the communication satellite to the reflection point, including the azimuth and elevation angle, is calculated. According to the law of reflection, the angle of the reflected signal from the reflection point to the positioning terminal is calculated. The reflection angle is equal to the incident angle, but in the opposite direction.

[0139] The multipath score of each reflector is calculated based on the distance of the reflector relative to the positioning terminal, the material of the reflector, and the angle at which the signal reaches the reflector from the communication satellite. ,in, represents the multipath score of the fs-th reflector, Represents the reflection coefficient of the fs-th reflection surface material, represents the distance from the positioning terminal to the fsth reflection surface, represents the incident angle of the communication satellite signal reaching the fsth reflection surface, cs and represents a constant used for adjustment and to ensure numerical stability and to prevent the denominator from being zero;

[0140] It should be noted that by considering the characteristics of the reflecting surfaces around the receiver, the impact of multipath effects on positioning accuracy can be more accurately evaluated. Combined with real-time monitoring data, it can dynamically adapt to environmental changes, improve positioning accuracy and robustness, and automatically identify and evaluate multipath effects in complex environments, reducing the need for manual intervention.

[0141] According to the observation matrix H, the influence of the geometric distribution of communication satellites on positioning accuracy is quantified to obtain the geometric dilution factor ,in, represents the trace of the matrix (i.e., the sum of the main diagonal elements);

[0142] The line-of-sight angle and azimuth angle of each communication satellite, the multipath score of each reflection surface, and the geometric dilution factor are standardized and then horizontally spliced ​​to form a spatial feature vector;

[0143] The training methods of the positioning error model include:

[0144] Step 1: Collect a sample set, including the spatial feature vector of the YB group and the corresponding true positioning error, and divide the sample set into a training set and a validation set in proportion;

[0145] Among them, the actual positioning error is the position of the positioning terminal The difference between the actual physical position and the preset known fixed reference point is taken as the actual physical position;

[0146] Step 2: construct a positioning error model, taking the spatial feature vector as input and the expected positioning error as output, wherein the positioning error model is a gradient boosting tree model;

[0147] Step 3: Initialize the hyperparameters of the gradient boosting tree model, use network search to tune the hyperparameters, use k-fold cross validation to evaluate the cross-validation scores of the model under different hyperparameter combinations, and select the optimal parameter combination;

[0148] Step 4: Use the optimal parameter combination as the initial parameters of the model and define the mean absolute error as the loss function to evaluate the prediction accuracy of the model ,in, represents the expected positioning error of the vth training sample, Represents the true positioning error of the vth training sample;

[0149] Step 5: Use the training set and loss function to adjust the model parameters according to the gradient in each iteration of the model;

[0150] Step 6: For each iteration, update the model parameters using the training set using R 2 The score is used as the evaluation indicator and R is calculated on the validation set. 2 Score value;

[0151] According to the R 2 Score value, calculate the R after the current iteration 2 The score value is the same as the R of the previous iteration. 2 The difference in the score values ​​is recorded as the iteration difference;

[0152] Step 7: Set an iteration difference threshold. If the iteration difference is greater than the iteration difference threshold, it is determined that the performance of the model has improved.

[0153] If the iteration difference is less than or equal to the iteration difference threshold, it is determined that the performance of the model has not been improved;

[0154] If the performance of the model on the validation set does not improve in consecutive DC iterations, the training is stopped to obtain the trained positioning error model;

[0155] Based on the expected positioning error, it is determined whether the positioning error of the positioning terminal is abnormal. If abnormal, the method of further determining whether the multipath effect is an interference factor includes:

[0156] Set a positioning error threshold and compare the expected positioning error at the current moment with the positioning error threshold;

[0157] If the expected positioning error is less than or equal to the positioning error threshold, the error of the positioning terminal is determined to be normal, and the current relative distance estimation value continues to be used;

[0158] If the expected positioning error is greater than the positioning error threshold (through actual tests by industry experts and historical data, the positioning error distribution under normal conditions is statistically analyzed, and the average and standard deviation of the positioning error under normal conditions are calculated. Usually, the average error plus several times the standard deviation can be selected as the threshold. For example, the average error plus 2 times the standard deviation (95% confidence interval) is selected as the positioning error threshold), the positioning terminal is judged to have an error abnormality;

[0159] According to the error anomaly of the positioning terminal, set the multipath effect threshold (determine a reasonable multipath effect threshold through actual tests of industry experts and historical data. For example, through experiments, it is found that when the multipath average value exceeds a certain value, the impact of the multipath effect on the positioning accuracy increases significantly, and this value can be used as the multipath effect threshold). Take the average of the multipath scores of all reflection surfaces and compare the average multipath score with the multipath effect threshold;

[0160] If the multipath score mean is less than or equal to the multipath effect threshold, it is determined that the multipath effect is not an interference factor;

[0161] If the multipath score mean is greater than the multipath effect threshold, the multipath effect is determined to be an interference factor;

[0162] According to the judgment result, the interference factor correction is performed by using smoothing technology, and the relative distance estimation value is adjusted by the expected positioning error, including:

[0163] If the multipath effect is an interference factor, the pseudorange measurement value is corrected by combining the carrier phase measurement value with the smoothing technology to obtain the smoothed pseudorange value.

[0164] ;

[0165] in, represents the smoothed pseudorange value of D at the current moment, represents the pseudo-range measurement value of D at the current moment, represents the pseudorange measurement value at time D-1, It represents the absolute difference between the D carrier phase measurement value at time D-1 and the current time. represents the smoothed pseudorange value at time D-1, and is defined as is the pseudo-range measurement value at the first moment in the monitoring period, represents the preset carrier phase smoothing factor (usually between 0.7 and 0.9), Represents the preset exponential smoothing factor (usually between 0 and 1);

[0166] It should be noted that the carrier phase smoothing factor Used for carrier phase smoothing. Usually set to a higher value (0.7 to 0.9) to take full advantage of the high accuracy of the carrier phase. Exponential smoothing factor Used for exponential smoothing, can be adjusted according to the application scenario, larger More emphasis is placed on the new carrier phase smoothing value, the smaller the It relies more on historical smoothing values;

[0167] Use the smoothed pseudorange value as the new pseudorange measurement value to re-obtain the expected positioning error, and adjust the relative distance estimate based on the re-obtained expected positioning error ,in, represents the adjusted relative distance estimate, It represents the expected mean positioning error of the expected positioning error within dg moments before the current moment. It represents the expected standard value of the expected positioning error within dg moments before the current moment, and zz represents the preset adjustment scale factor, which is set according to actual needs and is used to weigh the impact of the standard deviation on the adjustment amount;

[0168] Based on the adjusted relative distance estimate and the motion characteristics of the positioning terminal, the environment correction distance can be calculated in the following ways:

[0169] The linear acceleration and rotational angular velocity of the positioning terminal are used as the motion characteristics of the positioning terminal;

[0170] By integrating the angular velocity, the rotation matrix is ​​obtained. According to the adjusted relative distance estimate, the environment-corrected distance is calculated by combining the rotation matrix with the linear acceleration. ,in, represents the environmental correction distance, Represents the distance correction calculated by combining the rotation matrix and linear acceleration;

[0171] ,in, represents the estimated position of the positioning terminal at the current time D, represents the linear acceleration of the positioning terminal at the current time D, represents the speed of the positioning terminal at the current time D, represents the rotation matrix, Represents the time interval from time D-1 to the current time D;

[0172] Combining the environmental correction distance with the position of the communication satellite, an accurate position estimate is calculated as the high-precision positioning coordinates of the positioning terminal, including:

[0173] The extended Kalman filter is used to calculate the precise position estimate by correcting the distance by the environment. ,in, represents the environmentally corrected distance between the ith communication satellite and the positioning terminal, represents the extended Kalman filter, Represents the precise position estimate of the positioning terminal;

[0174] The precise position estimate is used as the high-precision positioning coordinates of the positioning terminal.

[0175] This embodiment effectively improves the high-precision positioning performance in complex and dynamic environments by intelligently identifying and processing interference factors, combining terminal motion characteristics, fine-tuning spatial environment modeling, and real-time pseudorange adjustment, and provides more stable and accurate positioning results.

[0176] First, the correction value provided by the differential base station is used to further correct the initial relative distance, reduce the influence of atmospheric effects and satellite orbit errors, and provide a more accurate initial relative distance estimate;

[0177] Next, by analyzing the spatial geometric relationship between the positioning terminal and each communication satellite, the spatial feature vector is obtained to provide accurate data support for the subsequent interference factor determination and adjustment, thus enhancing the reliability of the error model.

[0178] Subsequently, based on the positioning error model obtained through training, it intelligently identifies and processes interference factors such as multipath effects to ensure the stability and reliability of positioning results, especially in complex environments.

[0179] Then, the Kalman filter is used for dynamic adjustment to compensate the pseudo-range measurement value in real time by the expected positioning error, thus avoiding error accumulation and improving the anti-interference ability and stability of the system.

[0180] Finally, the environment-corrected distance is calculated based on the terminal's motion characteristics (such as speed, acceleration, and angular velocity of rotation), and combined with the satellite position to obtain the final high-precision positioning coordinates, achieving high-precision, real-time positioning in a dynamic environment.

[0181] Embodiment 2

[0182] See also Figure 2 As shown, the part not described in detail in this embodiment is described in Example 1, which provides a Beidou positioning data intelligent fusion processing method, including:

[0183] S1. Collect the initial position of the positioning terminal and the position of the communication satellite, and use the Euclidean distance method to calculate the initial relative distance between the positioning terminal and each communication satellite; use the correction value to correct the initial relative distance to obtain the relative distance estimation value between the positioning terminal and each communication satellite;

[0184] S2. According to the relative distance estimation value, the estimated position of the positioning terminal is obtained by solving the multi-point positioning algorithm, and the spatial feature vector is calculated according to the estimated position;

[0185] S3. Train the positioning error model according to the spatial feature vector to obtain the expected positioning error in the current environment. According to the expected positioning error, determine whether the positioning error of the positioning terminal is abnormal. If abnormal, determine whether the multipath effect is an interference factor.

[0186] S4. According to the judgment result of the interference factor, the interference factor is corrected by using the smoothing technology, and the relative distance estimation value is adjusted by the expected positioning error;

[0187] S5. Obtain the environment-corrected distance based on the adjusted relative distance estimate, and obtain an accurate position estimate by combining the environment-corrected distance with the position of the communication satellite as the high-precision positioning coordinates of the positioning terminal.

[0188] Embodiment 3

[0189] This embodiment discloses an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the operation mode of the Beidou positioning data intelligent fusion processing method provided above is implemented.

[0190] Since the electronic device introduced in this embodiment is an electronic device used to implement a Beidou positioning data intelligent fusion processing method in the embodiment of the present application, based on the Beidou positioning data intelligent fusion processing method introduced in the embodiment of the present application, the technical personnel of this field can understand the specific implementation of the electronic device of this embodiment and its various variations, so how the electronic device implements the method in the embodiment of the present application is not described in detail here. As long as the technical personnel of this field implement the electronic device used in the Beidou positioning data intelligent fusion processing method in the embodiment of the present application, it belongs to the scope of protection of this application.

[0191] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters and thresholds in the formula are set by technicians in this field according to actual conditions.

[0192] The above is only a preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above embodiments. All technical solutions under the concept of the present invention belong to the protection scope of the present invention. It should be pointed out that for ordinary technical users in this technical field, some improvements and modifications without departing from the principle of the present invention should also be regarded as the protection scope of the present invention.

Claims

1. A Beidou positioning data intelligent fusion processing system, characterized in that: include: Relative distance calculation module: collects the initial position of the positioning terminal and the position of the communication satellite, and uses the Euclidean distance method to calculate the initial relative distance between the positioning terminal and each communication satellite; uses the correction value to correct the initial relative distance to obtain the relative distance estimation value between the positioning terminal and each communication satellite; Among them, based on the tropospheric wet delay data, the tropospheric dry delay data and the ionospheric activity index, the atmospheric condition factor is obtained by using the double exponential atmospheric effect correction method, the observation matrix is ​​constructed based on the observation vector and the system state vector, and the atmospheric condition factor and the observation matrix are combined to calculate the automatic adjustment factor using the real-time atmospheric condition adaptive correction method; based on the preliminary relative distance, the original correction value and the corresponding atmospheric condition data, the predicted correction value is predicted using the regression model, and based on the predicted correction value, the correction value is calculated using the Kalman filter through the automatic adjustment factor; Feature vector acquisition module: used to obtain the estimated position of the positioning terminal according to the relative distance estimation value through the multi-point positioning algorithm, and calculate the spatial feature vector according to the estimated position; Interference factor determination module: trains the positioning error model according to the spatial feature vector to obtain the expected positioning error in the current environment. Based on the expected positioning error, determines whether the positioning error of the positioning terminal is abnormal. If abnormal, determines whether the multipath effect is an interference factor. Relative distance adjustment module: used to correct interference factors using smoothing technology according to the judgment results of interference factors, and adjust the relative distance estimation value according to the expected positioning error; Positioning coordinate correction module: used to obtain the environment-corrected distance based on the adjusted relative distance estimate, and combine the environment-corrected distance with the position of the communication satellite to obtain an accurate position estimate as the high-precision positioning coordinates of the positioning terminal.

2. The Beidou positioning data intelligent fusion processing system according to claim 1 is characterized in that: The relative distance estimation value is calculated by: Mark all satellites that are in communication connection with the positioning terminal as communication satellites, receive communication satellite signals through the positioning terminal, measure the time it takes for the signal of each communication satellite to propagate from the corresponding communication satellite to the positioning terminal, and multiply it by the speed of light to obtain a pseudo-range measurement value; The pseudo-range measurement values ​​of all communication satellites at time k are horizontally spliced ​​to form a pseudo-range vector; the pseudo-range vector is subjected to the least square method to obtain the initial position of the positioning terminal; The position of the communication satellite at time t is obtained using the satellite orbit model, and the initial relative distance between the positioning terminal and each communication satellite is calculated using the Euclidean distance formula in combination with the initial position of the positioning terminal and the position of the communication satellite. According to the initial position and maximum communication range of the positioning terminal, the correction value of each differential base station within the maximum communication range of the positioning terminal at time t is obtained. According to the correction value, the initial relative distance is adjusted to obtain the relative distance estimate between the positioning terminal and each communication satellite. ; in, represents the relative distance estimate, represents the initial relative distance, represents the number of differential base stations, represents the correction value of the pth differential base station, represents the weight of the pth differential base station correction value, , represents the variance of the pth differential base station correction value, and ϵ represents the regularization constant.

3. The Beidou positioning data intelligent fusion processing system according to claim 2 is characterized in that: The correction value is obtained by: Collect historical observation data sets, including preliminary relative distances, original correction values, and corresponding atmospheric condition data; The original correction value is the sum of the satellite orbit error correction value and the clock error correction value downloaded from the differential base station through the network connection; the atmospheric condition data affecting the propagation of communication satellite signals, including the ionospheric activity index and tropospheric humidity, are obtained through ground meteorological stations; Take the historical observation data set as input and the predicted correction value as output, and train a regression model to predict the correction value; The initial relative distance at time t, the original correction value and the corresponding atmospheric condition data are input into the trained regression model to obtain the predicted correction value; Based on the predicted correction value, an automatic adjustment factor based on atmospheric conditions is introduced to dynamically adjust the predicted correction value, and the state estimation is performed through the Kalman filter to obtain the correction value of the differential base station ; in, Represents the correction value of the differential base station, represents the Kalman gain at time t, represents the predicted correction value obtained by the regression model, represents the observation vector, represents the system state vector, represents the automatic adjustment factor; , represents the regularization parameter of the atmospheric condition factor, represents the exponential function, It represents the average value of the atmospheric condition factor in dg moments before time t, where dg refers to the preset number of moments. represents the atmospheric condition factor at time t, represents the Euclidean norm, represents the observation matrix; The observation matrix is ​​obtained by calculating the difference between the observation vector and the observation noise and calculating the ratio of the difference to the system state vector. , the observation noise is defined as zero-mean Gaussian white noise.

4. The Beidou positioning data intelligent fusion processing system according to claim 3 is characterized in that: The method for obtaining the observation vector includes: The carrier phase measurement values ​​of each communication satellite and the positioning terminal are obtained through real-time dynamic positioning technology; Read the inertial measurement unit data in the positioning terminal, including the linear acceleration and rotational angular velocity of the positioning terminal; The carrier phase measurement value, pseudorange measurement value, linear acceleration, rotational angular velocity, ionospheric activity index, tropospheric humidity, satellite orbit error correction value and clock error correction value at time t are horizontally spliced ​​to form the observation vector at time t ; The system state vector can be obtained by: The initial position of the positioning terminal at time t is directly expressed as ECEF coordinates; Get the speed of the positioning terminal at time t, and decompose the speed at time t into the three axes of the ECEF coordinate system; The position and velocity of the positioning terminal in the ECEF coordinate system are horizontally spliced ​​to form the system state vector of the positioning terminal at time t ; The atmospheric condition factors can be obtained by: The atmospheric condition factor is calculated by obtaining the tropospheric wet delay and tropospheric dry delay data from ground meteorological stations and combining them with the ionospheric activity index. ； in, represents the ionospheric activity index at time t, represents the average value of the ionospheric activity index within dg moments before moment t, represents the standard deviation of the ionospheric activity index within dg moments before moment t, represents the regularization parameter of ionospheric activity optimized by cross-validation method; and represent the tropospheric wet delay and tropospheric dry delay at time t, respectively; and They represent the average values ​​of the tropospheric wet delay and the tropospheric dry delay within dg moments before moment t, represents the regularization parameter of tropospheric activity optimized by cross-validation.

5. The Beidou positioning data intelligent fusion processing system according to claim 4 is characterized in that: The spatial feature vector is obtained in the following manner: The position of each communication satellite is obtained through the communication satellite signal, and the estimated position of the positioning terminal is obtained by solving the multi-point positioning algorithm using the relative distance estimation value of all communication satellites. : ,in, Represents the i-th relative estimated distance , represents the position of the i-th communication satellite , Indicates the number of communication satellites; Based on the estimated location of the positioning terminal , take the difference between the position of each communication satellite and the estimated position of the positioning terminal as the vector r from the positioning terminal to the communication satellite; For each communication satellite, calculate the line-of-sight angle and azimuth angle of the communication satellite relative to the positioning terminal; Viewing angle , azimuth ; in, represents the line-of-sight angle of the i-th communication satellite relative to the positioning terminal, represents the azimuth of the i-th communication satellite relative to the positioning terminal, represents the vector r from the positioning terminal to the i-th communication satellite, Represents vector The module length, They represent the three components of the x-axis, y-axis and z-axis of the vector from the positioning terminal to the i-th communication satellite in the ECEF coordinate system, , T represents transpose; Using digital elevation models and urban building databases, reflective surfaces that may cause multipath effects are identified, including reflective surfaces of tall buildings, bridges, and water surfaces, and the distance of the reflective surface relative to the positioning terminal, the material of the reflective surface, and the angle at which the signal reaches the reflective surface from the communication satellite. The multipath score of each reflector is calculated based on the distance of the reflector relative to the positioning terminal, the material of the reflector, and the angle at which the signal reaches the reflector from the communication satellite. ,in, represents the multipath score of the fs-th reflector, Represents the reflection coefficient of the fs-th reflection surface material, represents the distance from the positioning terminal to the fsth reflection surface, represents the incident angle of the communication satellite signal reaching the fsth reflection surface, cs and Indicates a preset constant; According to the observation matrix H, the geometric dilution factor is obtained ,in, represents the trace of a matrix; The line-of-sight angle and azimuth of each communication satellite, the multipath score of each reflection surface, and the geometric dilution factor are standardized and then horizontally spliced ​​to form a spatial feature vector.

6. The Beidou positioning data intelligent fusion processing system according to claim 5, characterized in that: The training method of the positioning error model includes: Step 61: Collect a sample set, including the spatial feature vector of the YB group and the corresponding true positioning error, and divide the sample set into a training set and a validation set in proportion; Among them, the actual positioning error is the position of the positioning terminal The difference between the actual physical position and the preset known fixed reference point is taken as the actual physical position; Step 62: construct a positioning error model, taking the spatial feature vector as input and the expected positioning error as output, wherein the positioning error model is a gradient boosting tree model; Step 63: Initialize the hyperparameters of the gradient boosting tree model, use network search to tune the hyperparameters, use k-fold cross validation to evaluate the cross validation scores of the model under different hyperparameter combinations, and select the optimal parameter combination; Step 64: Use the optimal parameter combination as the initial parameters of the model, and define the mean absolute error as the loss function for evaluating the prediction accuracy of the model; Step 65: Use the training set and loss function to adjust the model parameters according to the gradient in each iteration of the model; Step 66: For each iteration, update the model parameters using the training set using R 2 The score is used as the evaluation indicator and R is calculated on the validation set. 2 Score value; According to the R 2 Score value, calculate the R after the current iteration 2 The score value is the same as the R of the previous iteration. 2 The difference in the score values ​​is recorded as the iteration difference; Step 67: setting an iteration difference threshold, if the iteration difference is greater than the iteration difference threshold, it is determined that the performance of the model is improved; If the iteration difference is less than or equal to the iteration difference threshold, it is determined that the performance of the model has not been improved; If the performance of the model on the validation set does not improve in consecutive DC iterations, the training is stopped to obtain a trained positioning error model.

7. The Beidou positioning data intelligent fusion processing system according to claim 6 is characterized in that: The method of determining whether the multipath effect is an interference factor includes: Set a positioning error threshold and compare the expected positioning error at the current moment with the positioning error threshold; If the expected positioning error is less than or equal to the positioning error threshold, the error of the positioning terminal is determined to be normal, and the current relative distance estimation value continues to be used; If the expected positioning error is greater than the positioning error threshold, the error of the positioning terminal is determined to be abnormal; According to the error anomaly of the positioning terminal, a multipath effect threshold is set, the average of the multipath scores of all reflection surfaces is taken, and the average multipath score is compared with the multipath effect threshold; If the multipath score mean is less than or equal to the multipath effect threshold, it is determined that the multipath effect is not an interference factor; If the multipath score mean is greater than the multipath effect threshold, the multipath effect is determined to be an interference factor.

8. The Beidou positioning data intelligent fusion processing system according to claim 7, characterized in that: The adjustment method of the relative distance estimation value includes: If the multipath effect is an interference factor, the pseudorange measurement value is corrected by combining the carrier phase measurement value with the smoothing technology to obtain the smoothed pseudorange value. ； in, represents the smoothed pseudorange value of D at the current moment, represents the pseudo-range measurement value of D at the current moment, represents the pseudorange measurement value at time D-1, Indicates the absolute difference between the D carrier phase measurement value at time D-1 and the current time; represents the smoothed pseudorange value at time D-1, and is defined as is the pseudo-range measurement value at the first moment in the monitoring period, Indicates the preset carrier phase smoothing factor, Represents the preset exponential smoothing factor; Use the smoothed pseudorange value as the new pseudorange measurement value to re-obtain the expected positioning error, and adjust the relative distance estimate based on the re-obtained expected positioning error ; in, represents the adjusted relative distance estimate, It represents the expected mean positioning error of the expected positioning error within dg moments before the current moment. It represents the expected standard value of the expected positioning error within dg moments before the current moment, and zz represents the preset adjustment scale factor.

9. The Beidou positioning data intelligent fusion processing system according to claim 8, characterized in that: The method of obtaining the environment correction distance includes: The linear acceleration and rotational angular velocity of the positioning terminal are used as the motion characteristics of the positioning terminal; By integrating the angular velocity, the rotation matrix is ​​obtained. According to the adjusted relative distance estimate, the environment-corrected distance is calculated by combining the rotation matrix with the linear acceleration. ,in, Indicates the environmental correction distance, Represents the distance correction calculated by combining the rotation matrix and linear acceleration; ,in, represents the estimated position of the positioning terminal at the current time D, represents the linear acceleration of the positioning terminal at the current time D, represents the speed of the positioning terminal at the current time D, represents the rotation matrix, Represents the time interval from time D-1 to the current time D.

10. A Beidou positioning data intelligent fusion processing system according to claim 9, characterized in that: The high-precision positioning coordinates are obtained by: The extended Kalman filter is used to calculate the precise position estimate by correcting the distance by the environment. ,in, represents the environmentally corrected distance between the ith communication satellite and the positioning terminal, represents the extended Kalman filter, Represents the precise position estimate of the positioning terminal; The precise position estimate is used as the high-precision positioning coordinates of the positioning terminal.

Images