A receiver clock difference prediction method, device, equipment and storage medium

Abstract

The application discloses a receiver clock error prediction method, device and equipment and a storage medium. The method comprises the following steps: acquiring satellite navigation information and observation information of a current time node received by a receiver; performing standard single-point positioning calculation on the satellite navigation information and the observation information, and calculating a receiver clock error, an altitude and a time precision attenuation factor corresponding to the current time node; inputting the receiver clock error, the altitude and the time precision attenuation factor into a preset receiver clock error prediction model, predicting a receiver clock error of a future time node, and generating a receiver clock error corresponding to the future time node. The application can improve the prediction accuracy of the receiver clock error.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of satellite navigation positioning technology, and in particular to a receiver clock difference prediction method, device, equipment and storage medium. BACKGROUND

[0002] The receiver clock difference is an important GNSS parameter, and the theoretical research and application research on the receiver clock difference are of great significance. At present, the prediction methods of the receiver clock difference mainly include the receiver clock difference per epoch estimation model and the measured data mathematical modeling method.

[0003] Among them, the receiver clock difference per epoch estimation model solves the receiver clock difference together with the three-dimensional coordinates of the receiver to form four unknowns. At least four satellites need to be observed at the same time in each epoch to construct the pseudorange measurement equation set or the carrier phase measurement equation set to realize the receiver positioning and clock difference estimation. The measured data mathematical modeling method models the time series of the receiver clock difference by mathematical methods. In dynamic positioning, especially in complex environments such as tunnels and urban canyons, the short-term prediction accuracy is high. Specific methods include quadratic polynomial model, ARMA model, gray theory model, extended Kalman filter model, etc.

[0004] However, the prediction accuracy of these models is not high: (1) The per-epoch estimation model can only observe the satellites in the upper hemisphere of the horizon. This asymmetry leads to a clear correlation between the estimation of the receiver clock difference, the troposphere parameter and the height, so that the receiver clock difference estimation contains part of the pseudo-error related to the height. (2) The measured data mathematical modeling method: The traditional method is generally based on some physical characteristics of the receiver clock (such as noise type, frequency characteristics), excavates the internal rules of the clock difference sequence, makes statistical rule assumptions to model (such as quadratic polynomial model, extended Kalman filter model), or verifies that the prediction model conforms to an information processing system researched by experience (such as a gray system) to model the sequence. These methods often only model the receiver clock difference with a single factor, without considering the influence of height, troposphere delay and other factors, making it difficult to accurately model and predict the original measured sequence, resulting in low prediction accuracy. SUMMARY

[0005] The present application provides a receiver clock difference prediction method, device, equipment and storage medium to solve the technical problem of low prediction accuracy of existing receiver clock difference prediction models.

[0006] To solve the above technical problems, the present application embodiment provides a receiver clock difference prediction method, comprising:

[0007] Obtaining satellite navigation information and observation information received by the receiver at a current time node;

[0008] performing standard single point positioning calculation on the satellite navigation information and the observation information to obtain the receiver clock error, the height and the time precision attenuation factor corresponding to the current time node;

[0009] inputting the receiver clock error, the height and the time precision attenuation factor into a preset receiver clock error prediction model to predict the receiver clock error of a future time node, and generating the receiver clock error corresponding to the future time node.

[0010] As a preferred solution, the satellite navigation information comprises: orbit position and clock error of a satellite.

[0011] performing standard single point positioning calculation on the satellite navigation information and the observation information to obtain the receiver clock error, the height and the time precision attenuation factor corresponding to the current time node, comprising:

[0012] calculating the satellite observation pseudo-range according to the observation information;

[0013] calculating the receiver clock error and the approximate position coordinates of the receiver corresponding to the current time node according to the satellite observation pseudo-range, the orbit position and the clock error of at least four satellites, a preset troposphere error model and a preset ionosphere error model, and further calculating the receiver height and the time precision attenuation factor corresponding to the current time node according to the approximate position coordinates of the receiver.

[0014] As a preferred solution, before inputting the receiver clock error, the height and the time precision attenuation factor into the preset receiver clock error prediction model, the method further comprises:

[0015] preprocessing the receiver clock error, the height and the time precision attenuation factor; wherein the preprocessing comprises: data discontinuity processing, gross error detection and correction, and sampling time conversion.

[0016] As a preferred solution, the receiver comprises: a static base station receiver and a dynamic motion carrier receiver.

[0017] The receiver clock error prediction model comprises: a static receiver clock error prediction model and a dynamic receiver clock error prediction model.

[0018] inputting the receiver clock error, the height and the time precision attenuation factor into a preset receiver clock error prediction model, comprising:

[0019] when the receiver is a base station receiver, inputting the receiver clock error, the height and the time precision attenuation factor into the static receiver clock error prediction model; when the receiver is a motion carrier receiver, inputting the receiver clock error, the height and the time precision attenuation factor into the dynamic receiver clock error prediction model.

[0020] As a preferred solution, the receiver clock error prediction model comprises: an LSTM model;

[0021] The LSTM model comprises a plurality of LSTM units, each LSTM unit comprising:

[0022] f t =σ(W f [x t ,h t-1 ]+b f );

[0023] i t =σ(W i [x t ,h t-1 ]+b i );

[0024]

[0025] o t =σ(W o [x t ,h t-1 ]+b o );

[0026] h t =o t ⊙tanh(c t );

[0027] Wherein, f t represents the forgetting gate, indicating the proportion of past memory information that needs to be retained in the memory cell at the current time; i t represents the input gate, indicating the proportion of input information at the current time in the memory cell at the current time; c t represents the memory cell at the current time t, which selectively remembers past information and current input information; o t represents the output gate, indicating that the current memory information selectively flows into the current output information; h t represents the output at the current time; σ represents the sigmoid activation function, tanh represents the hyperbolic tangent activation function; W * represents the weight matrix; b * represents the bias vector; and represents matrix dot product.

[0028] As a preferred solution, the generation of the receiver clock error prediction model comprises:

[0029] Obtaining satellite navigation information and observation information received by the receiver at past time nodes;

[0030] The satellite navigation information and observation information are subjected to standard single-point positioning calculation to obtain receiver clock error, height and time precision attenuation factor corresponding to a past time node;

[0031] The receiver clock error, height and time precision attenuation factor are preprocessed, and the preprocessing includes data discontinuity processing, gross error detection and correction, and sampling time conversion;

[0032] Each sub-sequence in the preprocessed receiver clock error, height and time precision attenuation factor is subjected to normalization processing and is divided into a plurality of samples;

[0033] The samples are divided into a training set, a validation set and a test set, and then the LSTM model is trained according to the training set, the validation set, the test set, a preset input time step, a preset output time step and a preset hyperparameter in each sample, to obtain a receiver clock error prediction model.

[0034] As a preferred solution, after the receiver clock error, height and time precision attenuation factor are input into a preset receiver clock error prediction model to predict the receiver clock error of a future time node and generate the receiver clock error corresponding to the future time node, the method further includes:

[0035] The mean absolute error and the root mean square error of the receiver clock error are calculated;

[0036] The mean absolute error and the root mean square error of the receiver clock error are compared with the mean absolute error and the root mean square error of the receiver clock error based on a preset quadratic polynomial model and a gray theory model, and then the prediction performance of the receiver clock error prediction model is evaluated according to the comparison result.

[0037] On the basis of the above-mentioned embodiments, another embodiment of the receiver clock error prediction provides a device, which comprises a satellite navigation information and observation information acquisition module, a standard single-point positioning calculation module and a receiver clock error prediction module;

[0038] The satellite navigation information and observation information acquisition module is configured to acquire satellite navigation information and observation information received by a receiver at a current time node;

[0039] The standard single-point positioning calculation module is configured to calculate the receiver clock error, height and time precision attenuation factor corresponding to the current time node by performing standard single-point positioning calculation on the satellite navigation information and observation information;

[0040] The receiver clock difference prediction module is configured to input the receiver clock difference, the altitude, and the time precision attenuation factor into a preset receiver clock difference prediction model, predict the receiver clock difference at a future time node, and generate the receiver clock difference corresponding to the future time node.

[0041] On the basis of the above-mentioned embodiments, a further embodiment of the application provides an electronic device, which comprises a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, and the processor implements the receiver clock difference prediction method described in the above-mentioned embodiments of the application when executing the computer program.

[0042] On the basis of the above-mentioned embodiments, a further embodiment of the application provides a storage medium, which comprises a stored computer program, wherein the receiver clock difference prediction method described in the above-mentioned embodiments of the application is controlled to be executed by a device in which the storage medium is located when the computer program is executed.

[0043] Compared with the prior art, the embodiments of the application have the following beneficial effects:

[0044] The application provides a receiver clock difference prediction method, which comprises the following steps: acquiring satellite navigation information and observation information received by a receiver at a current time node; performing standard single-point positioning calculation based on the satellite navigation information and the observation information, and calculating a receiver clock difference, an altitude, and a time precision attenuation factor corresponding to the current time node; and inputting the receiver clock difference, the altitude, and the time precision attenuation factor into a preset receiver clock difference prediction model, predicting a receiver clock difference at a future time node, and generating a receiver clock difference corresponding to the future time node.

[0045] The receiver clock difference prediction model of the prior art only models the receiver clock difference as a single factor, does not consider the influence of multiple factors such as altitude and tropospheric delay, and is difficult to accurately model and predict the original measured sequence. Compared with the prior art, the receiver clock difference prediction model of the application is formed by taking the receiver clock difference, the altitude, and the time precision attenuation factor corresponding to a past time node as input, taking the receiver clock difference corresponding to a future time node as output, and performing model training on a preset neural network model. Therefore, the receiver clock difference prediction model of the application can automatically learn the correlation between the receiver clock difference and the altitude, and obtain the uncertainty of the receiver clock difference from the time precision attenuation factor, and can model complex nonlinear relationships. When the receiver clock difference prediction model of the application is used to predict the receiver clock difference, the satellite navigation information and the observation information received by the receiver are acquired to perform standard single-point positioning calculation, the receiver clock difference, the altitude, and the time precision attenuation factor are input into the receiver clock difference prediction model, and a receiver clock difference with high prediction accuracy can be obtained, thereby improving the prediction accuracy of the receiver clock difference. BRIEF DESCRIPTION OF DRAWINGS

[0046] Figure 1 is a flowchart of a receiver clock bias prediction method provided by an embodiment of the present application;

[0047] Figure 2 is a unit structure diagram of LSTM;

[0048] Figure 3 is a receiver clock bias prediction model diagram;

[0049] Figure 4 is a flowchart of LSTM for predicting receiver clock bias and auxiliary GNSS positioning;

[0050] Figure 5 is a structural diagram of a receiver clock bias prediction device provided by an embodiment of the present application. DETAILED DESCRIPTION

[0051] In order to make the objects, technical solutions and advantages of the present application clearer, the technical solutions in the present application will be described clearly and completely below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the scope of protection of the present application.

[0052] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the present application belongs; the terms used herein are only for the purpose of describing specific embodiments and are not intended to limit the present application; the terms "include" and "have" and any variations thereof in the specification and claims of the present application and the above description of drawings are intended to cover non-exclusive inclusion.

[0053] In the description of the embodiments of the present application, the technical terms "first", "second", etc. are only used to distinguish different objects, and cannot be understood as indicating or implying relative importance or implicitly indicating the number, specific order or primary and secondary relationship of the indicated technical features. In the description of the embodiments of the present application, the meaning of "a plurality of" is two or more, unless otherwise explicitly and specifically limited.

[0054] Reference herein to "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment can be included in at least one embodiment of the present application. The appearance of the phrase in various places in the specification does not necessarily all refer to the same embodiment, nor is it necessarily independent or alternative embodiments to other embodiments. Those skilled in the art explicitly and implicitly understand that the embodiments described herein can be combined with other embodiments.

[0055] In the description of the embodiments of the present application, the term "and / or" is merely an association relationship of the associated objects, which means that there can be three relationships, for example, A and / or B, which can represent the three cases of A alone, A and B together, and B alone. In addition, the character " / " herein generally represents an "or" relationship between the associated objects before and after it.

[0056] In the description of the embodiments of the present application, the term "a plurality of" refers to two or more (including two), and similarly, "a plurality of groups" refers to two or more groups (including two groups), and "a plurality of pieces" refers to two or more pieces (including two pieces).

[0057] In the description of the embodiments of the present application, unless otherwise explicitly specified and limited, the technical terms "mounting", "connection", "connection", "fixing" and the like should be understood in a broad sense, for example, it can be fixedly connected, or it can be detachably connected, or it can be integrated; it can be mechanically connected, or it can be electrically connected; it can be directly connected, or it can be indirectly connected through an intermediate medium; it can be the internal communication of two elements or the interaction relationship between two elements. For those skilled in the art, the specific meanings of the above terms in the embodiments of the present application can be understood according to the specific circumstances.

[0058] Embodiment one

[0059] Please refer to Figure 1 A flowchart of a receiver clock difference prediction method provided by an embodiment of the present application, comprising the following specific steps:

[0060] S1, acquiring satellite navigation information and observation information received by the receiver at the current time node;

[0061] In view of the problems of the prior art: (1) only single-factor modeling of receiver clock difference is performed, without considering the influence of factors such as elevation and troposphere delay, so it is difficult to accurately model and predict the original measured sequence; (2) only based on fixed functions or systems, it is difficult to represent the complex nonlinear input-output relationship of the receiver clock difference, and it is impossible to significantly improve the prediction accuracy. The present application constructs a receiver clock difference prediction model with strong generalization ability, which can realize long-time high-precision prediction and the prediction result can assist GNSS to realize high-precision positioning when the satellite signal quality is poor, and the same model can be directly used to express the static and dynamic cases.

[0062] When the receiver clock difference prediction model constructed by the present application is used to predict the receiver clock difference, the prediction steps are as follows:

[0063] 1. First, acquire the static satellite navigation information and observation information received by the base station receiver (static), and the dynamic satellite navigation information and observation information received by the moving carrier receiver (dynamic).

[0064] S2, performing a standard single point positioning calculation on the satellite navigation information and the observation information to obtain a receiver clock error, an altitude, and a time dilution of precision corresponding to a current time node;

[0065] Preferably, the satellite navigation information comprises orbit positions and clock errors of satellites; the standard single point positioning calculation on the satellite navigation information and the observation information to obtain a receiver clock error, an altitude, and a time dilution of precision corresponding to a current time node comprises: calculating satellite observation pseudoranges according to the observation information; calculating a receiver clock error and approximate position coordinates of a receiver corresponding to the current time node according to the satellite observation pseudoranges, orbit positions and clock errors of at least four satellites, a preset troposphere error model, and a preset ionosphere error model, and further calculating a receiver altitude corresponding to the approximate position coordinates of the receiver and a time dilution of precision corresponding to the current time node.

[0066] 2, after obtaining static satellite navigation information and observation information received by a base station receiver and dynamic satellite navigation information and observation information received by a moving carrier receiver, performing a standard single point positioning (SPP) calculation on each of the satellite navigation information and the observation information to obtain a receiver clock error, an altitude, and a time dilution of precision (TDOP) corresponding to a current time node, and the steps of calculating the receiver clock error, the altitude, and the TDOP by the SPP calculation are as follows:

[0067] (1) Obtain a signal transmission duration according to a transmission time and a receiving time of a satellite signal, multiply the duration by a speed of light to obtain a signal transmission distance measurement, which is called a pseudorange. Due to the asynchronism of satellite clocks and receiver clocks, orbit errors, troposphere and ionosphere delays, multipath effects, hardware noise, etc., the pseudorange is not equal to the actual distance. Orbit errors and satellite clock errors are corrected by using navigation information, and the effects of troposphere and ionosphere delays are eliminated by modeling, so the pseudorange is expressed as:

[0068] ρ i = d i + ct R + v i ;

[0069] wherein ρ i is a pseudorange between an i-th satellite and a receiver, d i is an actual distance, c is a speed of light, t R is a receiver clock error, and v i is an error caused by multipath effects, hardware noise, etc.

[0070] (2) Let the position coordinates of the i-th satellite and the receiver be denoted as and (x u ,y u ,z u ), respectively, then

[0071]

[0072] (3) According to the broadcast ephemeris, the satellite coordinates can be obtained, and only four unknown parameters (x u ,y u ,z u ,t R ) can be solved by observing at least four satellites.

[0073] According to the approximate position coordinates of the receiver and the Taylor series expansion, the following can be obtained:

[0074]

[0075] wherein, If n satellites are observed, Δρ can be expressed as:

[0076]

[0077] wherein, (dx, dy, dz) is the difference between the actual position coordinates and the approximate position coordinates of the receiver. Therefore, when the number of visible satellites is equal to or greater than four satellites, x = (dx, dy, dz, t R ) can be solved. The covariance matrix of the solution value and the actual value of x is

[0078]

[0079] TDOP is:

[0080]

[0081] Preferably, before inputting the receiver clock error, height and time precision attenuation factor into the preset receiver clock error prediction model, the method further comprises: preprocessing the receiver clock error, height and time precision attenuation factor; wherein the preprocessing comprises: data discontinuity processing, gross error detection and correction, and sampling time conversion.

[0082] 3. After the receiver clock error, height and TDOP are solved by SPP, the data solved by SPP is preprocessed, and the preprocessing steps are as follows:

[0083] (1) To address the discontinuity of the data, the data is first divided into a series of continuous subsequences. When the time interval between two subsequences is less than or equal to 10s, a quadratic polynomial is used to fit and supplement the missing data points, connecting them into a subsequence. When the time interval is greater than 10s, we assume that the two subsequences may have been obtained under different environmental conditions, and we cannot supplement their missing data points.

[0084] (2) For gross errors, the 3σ criterion is used to detect gross errors and a quadratic polynomial fitting is used to correct them.

[0085] (3) The sampling time for dynamic data used in this experiment is 1s, and the sampling time for static data is 5s. In order to maintain consistency, the dynamic data is extracted so that its sampling time becomes 5s.

[0086] S3. Input the receiver clock bias, elevation and time accuracy attenuation factor into the preset receiver clock bias prediction model to predict the receiver clock bias at future time nodes and generate the receiver clock bias corresponding to the future time nodes.

[0087] Preferably, the receiver includes: a static base station receiver and a dynamic moving vehicle receiver; the receiver clock bias prediction model includes: a static receiver clock bias prediction model and a dynamic receiver clock bias prediction model; inputting the receiver clock bias, elevation, and time accuracy attenuation factor into the preset receiver clock bias prediction model includes: when the receiver is a base station receiver, inputting the receiver clock bias, elevation, and time accuracy attenuation factor into the static receiver clock bias prediction model; when the receiver is a moving vehicle receiver, inputting the receiver clock bias, elevation, and time accuracy attenuation factor into the dynamic receiver clock bias prediction model.

[0088] 4. After preprocessing the receiver clock bias, elevation, and TDOP calculated by SPP, the receiver clock bias, elevation, and time accuracy attenuation factor of the base station receiver are input into the corresponding static receiver clock bias prediction model; the receiver clock bias, elevation, and time accuracy attenuation factor of the running carrier receiver are input into the corresponding dynamic and static receiver clock bias prediction model, so that the static receiver clock bias prediction model and the dynamic receiver clock bias prediction model can predict the receiver clock bias at future time nodes based on the input receiver clock bias, elevation, and time accuracy attenuation factor, respectively, and generate the receiver clock bias corresponding to the future time nodes.

[0089] Preferably, the receiver clock bias prediction model includes an LSTM model; wherein the LSTM model includes a plurality of LSTM units, and each LSTM unit includes:

[0090] f t =σ(W f [x t ,ht-1 ]+b f );

[0091] i t =σ(W i [x t ,h t-1 ]+b i );

[0092]

[0093] o t =σ(W o [x t ,h t-1 ]+b o );

[0094] h t =o t ⊙tanh(c t );

[0095] wherein f t represents a forget gate, indicating the proportion of past memory information that needs to be retained in the memory cell at the current moment; i t represents an input gate, indicating the proportion of input information at the current moment in the memory cell at the current moment; c t represents the memory cell at the current moment t, selectively memorizing past information and current input information; o t represents an output gate, indicating that the current memory information selectively flows into the current output information; h t represents the output at the current moment; σ represents a sigmoid activation function, and tanh represents a hyperbolic tangent activation function; W * represents a weight matrix; b * represents a bias vector; and represents matrix point multiplication.

[0096] Preferably, the generation of the receiver clock error prediction model comprises: obtaining satellite navigation information and observation information received by a receiver at past time nodes; performing standard single point positioning calculation on the satellite navigation information and observation information to obtain receiver clock errors, elevations and time dilution of precision factors corresponding to the past time nodes; preprocessing the receiver clock errors, elevations and time dilution of precision factors; wherein the preprocessing comprises: data discontinuity processing, gross error detection and correction, and sampling time conversion; performing normalization processing on each sub-sequence in the preprocessed receiver clock errors, elevations and time dilution of precision factors respectively, and dividing the sub-sequences into a plurality of samples; dividing the samples into a training set, a validation set and a test set, and then training the LSTM model according to the training set, the validation set, the test set, a preset input time step, an output time step in each sample and a preset hyperparameter to obtain a receiver clock error prediction model.

[0097] Preferably, after inputting the receiver clock errors, elevations and time dilution of precision factors into a preset receiver clock error prediction model to predict receiver clock errors at future time nodes and generating receiver clock errors corresponding to the future time nodes, the method further comprises: calculating mean absolute errors and root mean square errors of the receiver clock errors; comparing the mean absolute errors and root mean square errors of the receiver clock errors with mean absolute errors and root mean square errors of receiver clock errors based on a preset quadratic polynomial model and a gray theory model respectively, and then evaluating the prediction performance of the receiver clock error prediction model according to the comparison results.

[0098] Preferably, the training process of the receiver clock error prediction model is as follows:

[0099] 1. Obtain satellite navigation information and observation information received by a base station receiver (static) and a moving carrier receiver (dynamic) in a past period of time, and perform SPP calculation on the satellite navigation information and observation information to obtain corresponding receiver clock errors, elevations and TDOP.

[0100] 2. Preprocess the data calculated by SPP: data discontinuity processing, gross error detection and correction, and sampling time conversion.

[0101] 3. Construct a receiver clock error prediction model based on LSTM; wherein the principle of LSTM is as follows:

[0102] Please refer to Figure 2, the unit structure diagram of LSTM, LSTM has a gating mechanism, which can solve the problem of gradient disappearance or explosion of recurrent neural network (RNN), capture the long-term dependence of time series, and is often used in time series data processing tasks. It has a recurrent connection feature, so that information can be transmitted between multiple time steps, and the output of each unit will become the input of the next time step, which can effectively learn the relationship between elements of the sequence. LSTM is composed of multiple units, each unit includes a forgetting gate, an input gate and an output gate, which can effectively control the flow of information and the retention of memory, and the LSTM unit structure is as shown in Figure 2

[0103] f t = σ (W f [x t , h t-1 ] + b f ) ;

[0104] i t = σ (W i [x t , h t-1 ] + b i ) ;

[0105]

[0106] o t = σ (W o [x t , h t-1 ] + b o ) ;

[0107] h t = o t ⊙ tanh (c t ) ;

[0108] Where f t represents the forgetting gate, which represents the proportion of past memory information that needs to be retained in the memory cell at the current time; i t represents the input gate, which represents the proportion of input information at the current time in the memory cell at the current time; c t represents the memory cell at the current time t, which selectively remembers past information and current input information; o t represents the output gate, which represents the selective flow of current memory information into the current output information; h t represents the output at the current time; σ represents the sigmoid activation function, tanh represents the hyperbolic tangent activation function; W * represents the weight matrix; b * represents the bias vector; and ⊙ represents matrix point multiplication.​

[0109] 4, Integrated pre-processed data: The data set obtained after preprocessing is: receiver clock error, height, TDOP, each sub-sequence divided in the data set is respectively normalized and divided into multiple samples in the form of sliding window (sliding one epoch at a time), the normalized receiver clock error, height and TDOP in the first L in epoch in each sample are used as model input, the anti-normalized receiver clock error of the last L out epoch is used as model label, that is, the feature dimension of the model input is 3, the time step is L in , the output feature dimension is 1, and the time step is L out , and then 6:2:2 is used to divide multiple samples into training, validation and test subsets, and then these subsets are combined to form training set, validation set and test set. The receiver clock error prediction model is shown in Figure 3 ;

[0110] It should be noted that the static and dynamic cases are separately trained models, which should be divided into two data sets, and the models are divided into static models and dynamic models.

[0111] 5, The present application sets multiple input time steps and output time steps, which are L in = 36 (3 min), 48 (4 min), 60 (5 min), L out = 24 (2 min), 30 (2.5 min), 36 (3 min). For each case, keep the structure and hyperparameters of the static model and the dynamic model consistent, and train and test them respectively. The specific hyperparameter settings are shown in Table 1 below.

[0112]

[0113]

[0114] Table 1: LSTM hyperparameter setting table

[0115] 6, Calculate the prediction accuracy (mean absolute error MAE and root mean square error RMSE) of the constructed receiver clock error prediction model, and compare the results with the prediction accuracy of the quadratic polynomial model and the gray theory model, to evaluate the prediction performance of the receiver clock error prediction model. The calculation of the mean absolute error MAE and the root mean square error RMSE is as follows:

[0116]

[0117] 7. Use the receiver clock biases predicted by the LSTM model, quadratic polynomial model, and grey theory model for positioning experiments. Calculate the root mean square error of the position coordinates in the North, East, and Up directions, and evaluate the performance of the receiver clock bias predictions in assisting GNSS positioning. Please refer to... Figure 4 This is a flowchart for LSTM prediction of receiver clock bias and auxiliary GNSS positioning.

[0118] Therefore, this invention provides a receiver clock bias prediction method. Based on a pre-constructed receiver clock bias prediction model, the receiver clock bias, elevation, and TDOP for a current period are normalized and used as model inputs to predict the receiver clock bias sequence for a future period as output. This model can automatically learn the correlation between receiver clock bias and elevation, and obtain the uncertainty of receiver clock bias from TDOP, enabling it to model complex nonlinear relationships and improve prediction accuracy. Furthermore, the model considers both static and dynamic cases, training and testing subsequences for different time periods to ensure model usability under different dynamics and environmental conditions, improving the robustness and generalization ability of the prediction model. In addition, this model can reduce interference from noise and elevation errors on receiver clock bias, enabling the predicted receiver clock bias to assist GNSS positioning in complex urban environments, thus improving positioning accuracy.

[0119] Example 2

[0120] Please refer to Figure 5 This is a schematic diagram of a receiver clock bias prediction device according to an embodiment of the present invention. The device includes: a satellite navigation information and observation information acquisition module, a standard single-point positioning calculation module, and a receiver clock bias prediction module.

[0121] The satellite navigation information and observation information acquisition module is used to acquire the satellite navigation information and observation information received by the receiver at the current time node;

[0122] The standard point positioning solution module is used to perform standard point positioning solution on the satellite navigation information and observation information, and calculate the receiver clock error, elevation and time accuracy attenuation factor corresponding to the current time node.

[0123] The receiver clock bias prediction module is used to input the receiver clock bias, elevation and time accuracy attenuation factor into a preset receiver clock bias prediction model, predict the receiver clock bias at future time nodes, and generate the receiver clock bias corresponding to the future time nodes.

[0124] It should be noted that the apparatus embodiments described above are only illustrative, and the units described as separate components can or can not be physically separated, and the components displayed as units can or can not be physical units, i.e., can be located in one place or distributed on multiple network units. Part or all of the modules can be selected to achieve the purpose of the embodiment according to actual needs. In addition, the connection relationship between the modules in the apparatus embodiment provided by the present application indicates that there is a communication connection between them, which can be implemented as one or more communication buses or signal lines. Those skilled in the art can understand and implement it without creative labor.

[0125] Those skilled in the art can clearly understand that, for the convenience and brevity, the specific working process of the apparatus described above can refer to the corresponding process in the foregoing method embodiments, which will not be repeated here.

[0126] Embodiment three

[0127] Correspondingly, the embodiment of the present application provides an electronic device, which comprises a processor, a memory and a computer program stored in the memory and configured to be executed by the processor, and the processor implements the receiver clock difference prediction method described in the above-mentioned embodiment of the application when executing the computer program.

[0128] The electronic device can be a desktop computer, a notebook computer, a palm computer, a cloud server and the like. The device can include but is not limited to a processor and a memory.

[0129] The processor can be a central processing unit (CPU), and can also be other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor, etc. The processor is the control center of the device, and connects various parts of the device through various interfaces and lines.

[0130] Embodiment four

[0131] Correspondingly, the embodiment of the present application provides a storage medium, the storage medium comprising a stored computer program, wherein the computer program controls a device where the storage medium is located to perform the receiver clock difference prediction method when the computer program is running.

[0132] The memory can be configured to store the computer program, and the processor can realize various functions of the device by running or executing the computer program stored in the memory and calling data stored in the memory. The memory can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, at least one application required by a function, and the like; and the data storage area can store data created according to use of the mobile phone and the like. In addition, the memory can include a high-speed random access memory, and can also include a nonvolatile memory, for example, a hard disk, a memory, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, at least one disk storage device, a flash memory device, or other volatile solid-state memory devices.

[0133] The storage medium is a computer readable storage medium, and the computer program is stored in the computer readable storage medium. The computer program can realize the steps of each method embodiment when executed by the processor. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or some intermediate forms, etc. The computer readable medium can include any entity or device capable of carrying the computer program code, recording medium, U disk, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal, and software distribution medium, etc. It should be noted that the computer readable medium can include or exclude contents according to requirements of legislation and patent practice in a jurisdiction, for example, in some jurisdictions, according to legislation and patent practice, the computer readable medium does not include electrical carrier signals and telecommunication signals.

[0134] The above describes the preferred embodiments of the present application. It should be noted that, for those skilled in the art, without departing from the principles of the present application, a number of improvements and refinements can be made, and these improvements and refinements are also considered to be within the protection scope of the present application.

Claims

1. A method of receiver clock bias prediction, characterized by, The method comprises the following steps: obtaining satellite navigation information and observation information received by a receiver at a current time node; performing standard single-point positioning calculation on the satellite navigation information and observation information to obtain receiver clock bias, height and time precision attenuation factor corresponding to the current time node; inputting the receiver clock bias, height and time precision attenuation factor into a preset receiver clock bias prediction model to predict the receiver clock bias at a future time node and generate the receiver clock bias corresponding to the future time node; wherein the receiver clock bias prediction model comprises an LSTM model; the LSTM model comprises a plurality of LSTM units, and each LSTM unit comprises: ； ； ； ； ； ； wherein, denotes a forget gate, and denotes a proportion of past memory information that needs to be retained in the memory cell at the current time point; denotes an input gate, and denotes a proportion of input information at the current time point in the memory cell at the current time point; denotes a memory cell at the current time point, selectively memorizes past information and current input information; denotes an output gate, and denotes that the current memory information selectively flows into the current output information; denotes an output at the current time point; denotes a sigmoid activation function, denotes a hyperbolic tangent activation function; denotes a weight matrix; denotes a bias vector; denotes a matrix dot product; the generation of the receiver clock bias prediction model comprises: obtaining satellite navigation information and observation information received by a receiver at a past time node; performing standard single-point positioning calculation on the satellite navigation information and observation information to obtain receiver clock bias, height and time precision attenuation factor corresponding to the past time node; preprocessing the receiver clock bias, height and time precision attenuation factor; wherein the preprocessing comprises data discontinuity processing, gross error detection and correction, and sampling time conversion; normalizing each sub-sequence in the preprocessed receiver clock bias, height and time precision attenuation factor and dividing them into a plurality of samples; dividing the samples into a training set, a validation set and a test set, and then training the LSTM model according to the training set, the validation set, the test set, a preset input time step, a preset output time step and a preset hyperparameter to obtain a receiver clock bias prediction model.

2. The receiver clock error prediction method of claim 1, wherein, The satellite navigation information comprises the orbital position and clock bias of the satellite. The standard single-point positioning calculation on the satellite navigation information and observation information to obtain receiver clock bias, height and time precision attenuation factor corresponding to the current time node comprises: calculating satellite observation pseudoranges according to the observation information; calculating receiver clock bias and approximate position coordinates of the receiver corresponding to the current time node according to the satellite observation pseudoranges, the orbital position and clock bias of at least four satellites, a preset troposphere error model and a preset ionosphere error model, and then calculating the corresponding receiver height and time precision attenuation factor corresponding to the current time node according to the approximate position coordinates of the receiver.

3. The receiver clock error prediction method of claim 1, wherein, Before inputting the receiver clock bias, height and time precision attenuation factor into the preset receiver clock bias prediction model, the method further comprises: preprocessing the receiver clock bias, height and time precision attenuation factor; wherein the preprocessing comprises data discontinuity processing, gross error detection and correction, and sampling time conversion.

4. The receiver clock bias prediction method of claim 1, wherein, The receiver comprises a static base station receiver and a dynamic motion carrier receiver. The receiver clock bias prediction model comprises a static receiver clock bias prediction model and a dynamic receiver clock bias prediction model. Inputting the receiver clock bias, height and time precision attenuation factor into the preset receiver clock bias prediction model comprises: When the receiver is a base station receiver, the receiver clock error, altitude and time precision attenuation factor are input into the static receiver clock error prediction model; when the receiver is a moving carrier receiver, the receiver clock error, altitude and time precision attenuation factor are input into the dynamic receiver clock error prediction model.

5. The receiver clock error prediction method of claim 4, wherein, After inputting the receiver clock error, altitude and time precision attenuation factor into the preset receiver clock error prediction model, the receiver clock error at a future time node is predicted to generate the receiver clock error corresponding to the future time node, and the method further comprises: calculating the mean absolute error and root mean square error of the receiver clock error; comparing the mean absolute error and root mean square error of the receiver clock error with the mean absolute error and root mean square error of the receiver clock error based on the preset quadratic polynomial model and the grey theory model respectively, and then evaluating the prediction performance of the receiver clock error prediction model according to the comparison result.

6. A receiver clock error prediction apparatus, characterized by, The receiver clock error prediction method is suitable for the receiver clock error prediction method in any one of claims 1-5, comprising: a satellite navigation information and observation information acquisition module, a standard single point positioning calculation module and a receiver clock error prediction module; The satellite navigation information and observation information acquisition module is configured to acquire satellite navigation information and observation information received by a receiver at a current time node; The standard single point positioning calculation module is configured to perform standard single point positioning calculation on the satellite navigation information and observation information to calculate receiver clock error, altitude and time precision attenuation factor corresponding to the current time node; The receiver clock error prediction module is configured to input the receiver clock error, altitude and time precision attenuation factor into a preset receiver clock error prediction model to predict the receiver clock error at a future time node and generate the receiver clock error corresponding to the future time node.

7. An electronic device, comprising: The storage medium comprises a stored computer program, wherein the computer program controls the device where the storage medium is located to perform the receiver clock error prediction method in any one of claims 1-5 when the computer program is running.

8. A storage medium, characterized by The storage medium comprises a stored computer program, wherein the computer program controls the device where the storage medium is located to perform the receiver clock error prediction method in any one of claims 1-5 when the computer program is running.

Images