Ionospheric error elimination methods and related equipment
By using meteorological data and total ionosphere electrons, combined with neural networks and clustering models, the total ionosphere electrons are predicted, and the problem of inaccurate estimation of traditional models is solved, improving the accuracy of ionosphere error elimination and phase synchronization speed.
Patent Information
- Application Number
- CN202210956212.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-08-10
- Publication Date
- 2024-12-17
- Estimated Expiration
- 2042-08-10
AI Technical Summary
The traditional ionosphere model is not accurate enough to estimate the ionosphere and lacks generalization ability, resulting in low accuracy in eliminating ionosphere errors.
By obtaining the meteorological data at tq+1 time and the total ionosphere electrons at tq time, combining long and short-term memory neural network and clustering model, the total ionosphere electrons at tq+1 time and tp time are predicted, and then the ionosphere is performed Error elimination.
It improves the accuracy of ionosphere error elimination, reduces the difficulty of acquiring astronomical observation data, and accelerates the phase synchronization process.
Smart Images
Figure CN115097492B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of navigation systems, and in particular to an ionospheric error elimination method and related equipment. Background Art
[0002] Ionospheric errors can cause large fluctuations in the sensitivity of the navigation system. Nowadays, the method of receiving dual-frequency signals is generally used to eliminate ionospheric delays. However, dual-frequency signals need to undergo signal capture and phase synchronization before they can be used for ionospheric error calculations, and the time it takes to enter phase synchronization is relatively long. The total amount of electrons in the ionosphere is not a constant value. The changes in the total amount of electrons in the ionosphere within a day often have a certain periodic law and are seriously affected by the weather. In the prior art, an ionospheric model is generally constructed based on the historical distribution of the total amount of electrons in the ionosphere, and then the total amount of electrons in the ionosphere at a certain moment is predicted and calculated through the ionospheric model. However, due to the influence of weather, the traditional ionospheric model is not accurate enough in estimating the ionosphere, and the generalization ability is insufficient, which makes the accuracy of ionospheric error elimination low. Summary of the invention
[0003] The embodiment of the present invention provides an ionospheric error elimination method, which aims to solve the problem that the traditional ionospheric model has inaccurate estimation of the ionosphere and insufficient generalization ability, thereby making the accuracy of ionospheric error elimination low. q+1 The meteorological data corresponding to the time is used to calculate the first astronomical observation data corresponding to the time. q The total amount of electrons in the ionosphere corresponding to time t q+1 The first astronomical observation data corresponding to time t is calculated q+1 The total amount of electrons in the ionosphere corresponding to the time t q+1 The total amount of electrons in the ionosphere corresponding to time t is calculated p The total amount of electrons in the ionosphere corresponding to the time t can be accurately obtained by taking into account the meteorological data. p The total amount of electrons in the ionosphere corresponding to the moment can be calculated, thereby improving the accuracy of eliminating ionospheric errors. In addition, since the first astronomical observation data is calculated based on meteorological data, the difficulty of obtaining astronomical observation data is reduced.
[0004] In a first aspect, an embodiment of the present invention provides an ionospheric error elimination method, comprising the following steps:
[0005] Get q The total amount of electrons in the ionosphere corresponding to the time t q+1 Meteorological data corresponding to the time;
[0006] According to the q+1 The meteorological data corresponding to the time is calculated to get t q+1The first astronomical observation data corresponding to the time;
[0007] Based on the q The total amount of electrons in the ionosphere corresponding to the time t q+1 The first astronomical observation data corresponding to the time is calculated to obtain the t q+1 The total amount of electrons in the ionosphere corresponding to the time;
[0008] According to the q+1 The total amount of electrons in the ionosphere corresponding to the time t is calculated p The total amount of electrons in the ionosphere corresponding to the time, where p is greater than q;
[0009] According to the p The total amount of ionospheric electrons corresponding to the moment is used to eliminate the ionospheric error.
[0010] Optionally, the meteorological data includes n types, the first astronomical observation data includes m types, and the q+1 The meteorological data corresponding to the time is calculated to get t q+1 The steps of obtaining the first astronomical observation data corresponding to the time include:
[0011] Through the preset m clustering models, the t q+1 Calculate the n types of meteorological data corresponding to the time and get t q+1 The first astronomical observation data of m types corresponding to the time.
[0012] Optionally, in the case of using a preset clustering model, q+1 The meteorological data corresponding to the time is calculated to obtain t q+1 Before the step of obtaining the first astronomical observation data corresponding to the time instant, the method further comprises:
[0013] Acquire n types of historical meteorological data, and acquire m types of historical astronomical observation data, wherein the historical meteorological data at each historical moment corresponds to the historical astronomical observation data at the same historical moment;
[0014] The historical astronomical observation data at the same historical moment are clustered according to the historical meteorological data to obtain m clustering models, wherein the input of the clustering model is n types of meteorological data.
[0015] Optionally, the step of clustering the historical astronomical observation data at the same historical moment according to the historical meteorological data to obtain m clustering models includes:
[0016] Sort each type of astronomical observation data by size and divide it into k i＝1，2，…m subclass, and the k i＝1，2，…m Subclasses serve as labels for meteorological data;
[0017] The k-NN clustering algorithm is used to perform cluster analysis on the historical astronomical observation data at the same time as the historical meteorological data to obtain m clustering models, each of which includes k i＝1，,2，…m Cluster centers.
[0018] Optionally, the t q The total amount of electrons in the ionosphere corresponding to the time t q+1 The first astronomical observation data corresponding to the time is calculated to obtain the t q+1 The steps of calculating the total ionospheric electron quantity corresponding to the time include:
[0019] The t q The total amount of electrons in the ionosphere corresponding to the time t q+1 The first astronomical observation data corresponding to the time is predicted to obtain the t q+1 The total amount of electrons in the ionosphere at the time.
[0020] Optionally, in the case of using a preset ionosphere model to q The total amount of electrons in the ionosphere corresponding to the time t q+1 The first astronomical observation data corresponding to the time is predicted to obtain the t q+1 Before the total ionospheric electron quantity corresponding to the time instant, the method further comprises:
[0021] Acquire historical meteorological data and second astronomical observation data, wherein the second astronomical observation data includes the total amount of electrons in the ionosphere corresponding to each historical moment and the first astronomical observation data corresponding to each historical moment;
[0022] Calculating the historical meteorological data through the clustering model to obtain the first astronomical observation data corresponding to each historical moment;
[0023] The first astronomical observation data corresponding to each of the historical moments and the total amount of ionospheric electrons corresponding to the same historical moment are used as sample data, and the total amount of ionospheric electrons corresponding to the next moment of the historical moment is used as a label to construct a sample data set;
[0024] A long short-term memory neural network is constructed, and the long short-term memory neural network is trained using the sample data set, and the ionosphere model is obtained after the training is completed.
[0025] Optionally, the p The steps of eliminating ionospheric errors based on the total ionospheric electron quantity corresponding to the time include:
[0026] According to the p The total amount of electrons in the ionosphere corresponding to the time delay is estimated to obtain tp Delay estimation at the moment;
[0027] According to the p The delay estimate at time t p The phase relationship of the moments;
[0028] According to the above p The phase relationship at each moment is used as a benchmark to capture the phase within a preset time range;
[0029] The ionospheric error is eliminated through the result of the phase capture.
[0030] In a second aspect, an embodiment of the present invention provides an ionospheric error elimination device, the device comprising:
[0031] The first acquisition module is used to obtain t q The total amount of electrons in the ionosphere corresponding to the time t q+1 Meteorological data corresponding to the time;
[0032] The first calculation module is used to calculate the q+1 The meteorological data corresponding to the time is calculated to get t q+1 The first astronomical observation data corresponding to the time;
[0033] The second calculation module is used to calculate the q The total amount of electrons in the ionosphere corresponding to the time t q+1 The first astronomical observation data corresponding to the time is calculated to obtain the t q+1 The total amount of electrons in the ionosphere corresponding to the time;
[0034] The third calculation module is used to calculate the q+1 The total amount of electrons in the ionosphere corresponding to the time t is calculated p The total amount of electrons in the ionosphere corresponding to the time, where p is greater than q;
[0035] The error elimination module is used to eliminate the error according to the t p The total amount of ionospheric electrons corresponding to the moment is used to eliminate the ionospheric error.
[0036] In a third aspect, an embodiment of the present invention provides an electronic device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the steps of the ionospheric error elimination method provided in the embodiment of the present invention are implemented.
[0037] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the ionospheric error elimination method provided in the embodiment of the invention are implemented.
[0038] In the embodiment of the present invention, t is obtained q The total amount of electrons in the ionosphere corresponding to the time t q+1 The meteorological data corresponding to the time; according to the t q+1 The meteorological data corresponding to the time is calculated to get t q+1 The first astronomical observation data corresponding to the time; based on the t q The total amount of electrons in the ionosphere corresponding to the time t q+1 The first astronomical observation data corresponding to the time is calculated to obtain the t q+1 The total amount of electrons in the ionosphere corresponding to the time; according to the t q+1 The total amount of electrons in the ionosphere corresponding to the time t is calculated p The total amount of electrons in the ionosphere corresponding to the time, where p is greater than q; according to the t p The total amount of ionospheric electrons corresponding to the time is used to eliminate the ionospheric error. q+1 The meteorological data corresponding to the time is used to calculate the first astronomical observation data corresponding to the time. q The total amount of electrons in the ionosphere corresponding to time t q+1 The first astronomical observation data corresponding to time t is calculated q+1 The total amount of electrons in the ionosphere corresponding to the time t q+1 The total amount of electrons in the ionosphere corresponding to time t is calculated p The total amount of electrons in the ionosphere corresponding to the time t can be accurately obtained by taking into account the meteorological data. p The total amount of electrons in the ionosphere corresponding to the moment can be calculated, thereby improving the accuracy of eliminating ionospheric errors. In addition, since the first astronomical observation data is calculated based on meteorological data, the difficulty of obtaining astronomical observation data is reduced. BRIEF DESCRIPTION OF THE DRAWINGS
[0039] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0040] Figure 1 is a flow chart of an ionospheric error elimination method provided by an embodiment of the present invention;
[0041] Figure 2 It is a structural schematic diagram of an ionospheric error elimination device provided by an embodiment of the present invention;
[0042] Figure 3 It is a structural schematic diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION
[0043] 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.
[0044] For details, see Figure 1 , Figure 1 is a flow chart of a method for eliminating ionospheric errors provided by an embodiment of the present invention, such as Figure 1 As shown, the ionospheric error elimination method includes:
[0045] 101. Get t q The total amount of electrons in the ionosphere corresponding to the time t q+1 Meteorological data corresponding to the time.
[0046] In the embodiment of the present invention, the above t q The time can be the J time before the target time, where J is an integer greater than or equal to 1. When J is equal to 1, then t q+1 The target time is the target time. The target time can also be understood as the current time when the ionospheric error needs to be eliminated.
[0047] The above-mentioned ionospheric total electron content (TEC) can also be called the total ionospheric electron concentration content or the ionospheric electron concentration column content, integral content, etc. TEC is an important parameter to describe the ionospheric morphology and structure. During the signal propagation process, its time delay depends on the ionospheric total electron content.
[0048] t q The total amount of electrons in the ionosphere corresponding to time t can be q The total amount of ionospheric electrons calculated when the ionospheric error is eliminated at time t q+1 The meteorological data corresponding to the time is t q The meteorological data corresponding to the next moment of time. For example, t q The total amount of electrons in the ionosphere corresponding to the time is the total amount of electrons in the ionosphere corresponding to 12:21, then t q+1 The meteorological data corresponding to the time is the meteorological data corresponding to 12:22.
[0049] The above q+1 The meteorological data corresponding to the time can be obtained through the meteorological information released by the local weather department.
[0050] 102. According to t q+1 The meteorological data corresponding to the time is calculated to get tq+1 The first astronomical observation data corresponding to the time.
[0051] In the embodiment of the present invention, the network model can be used to predict t q+1 The meteorological data corresponding to the time is predicted and processed to obtain t q+1 The prediction network model can be constructed based on a recurrent neural network or a long short-term memory neural network. The prediction network model is trained with historical meteorological data and the first historical astronomical observation data, so that the prediction network model can learn the implicit distribution between the historical meteorological data and the first historical astronomical observation data, thereby training the prediction network model to output the first corresponding astronomical observation data according to the meteorological data. q+1 The meteorological data corresponding to the time is input into the trained prediction network model, and the output is t q+1 The first astronomical observation data corresponding to the time.
[0052] We can also use clustering models to classify t q+1 The meteorological data corresponding to the time is predicted and processed to obtain t q+1 The clustering model includes the cluster centers corresponding to the first astronomical observation data. The distance between the meteorological data and each cluster center is calculated. When the distance between the meteorological data and a cluster center is the shortest, the first astronomical observation data corresponding to the cluster center is taken as t q+1 The first astronomical observation data corresponding to the time.
[0053] 103. Based on t q The total amount of electrons in the ionosphere corresponding to time t q+1 The first astronomical observation data corresponding to time t is calculated q+1 The total amount of electrons in the ionosphere at the time.
[0054] In the embodiment of the present invention, since t q The time is t q+1 Before time, so t q+1 The total amount of electrons in the ionosphere corresponding to the time can be calculated by iteration. q+1 During the signal propagation process at time t, its delay depends on q+1 The total amount of electrons in the ionosphere at the time.
[0055] Unlike the general ionosphere model, which is constructed based on the historical distribution of the total electron amount in the ionosphere, and then the total electron amount in the ionosphere at a certain moment is predicted and calculated by the ionosphere model, the embodiment of the present invention uses the first astronomical observation data and the historical distribution of the total electron amount in the ionosphere to predict the total electron amount in the ionosphere at a certain moment. q+1Predicting the total ionospheric electron content corresponding to a moment has higher accuracy. Moreover, the first astronomical observation data is obtained from meteorological data, and the difficulty of data acquisition is low.
[0056] 104. According to t q+1 the total ionospheric electron content corresponding to the moment, calculate the total ionospheric electron content corresponding to t p moment, where p is greater than q.
[0057] In the embodiment of the present invention, the above-mentioned t p moment can be understood as the target moment. p being greater than q indicates that the t p moment is after the t q moment. Of course, when p = q + 1, it means that the t q+1 moment and the t p moment are the same moment.
[0058] Specifically, when p = q + 1, the total ionospheric electron content corresponding to the t p moment is the total ionospheric electron content corresponding to the t q+1 moment. When p < q + 1, then according to the above steps 101 to 103, iteratively calculate the total ionospheric electron content corresponding to the t q+2 moment, the total ionospheric electron content corresponding to the t q+3 moment until the total ionospheric electron content corresponding to the t p moment is obtained.
[0059] 105. Eliminate the ionospheric error according to the total ionospheric electron content corresponding to the t p moment.
[0060] In the embodiment of the present invention, after obtaining the total ionospheric electron content corresponding to the t p moment, the ionospheric delay can be eliminated by using the method of receiving dual-frequency signals.
[0061] In the embodiment of the present invention, obtain the total ionospheric electron content corresponding to the t q moment, and obtain the meteorological data corresponding to the t q+1 moment; according to the meteorological data corresponding to the t q+1 moment, calculate the first astronomical observation data corresponding to the t q+1 moment; based on the total ionospheric electron content corresponding to the t q moment and the first astronomical observation data corresponding to the t q+1 moment, calculate the total ionospheric electron content corresponding to the t q+1 moment; according to the total ionospheric electron content corresponding to the t q+1 moment, calculate the total ionospheric electron content corresponding to the t p moment, where p is greater than q; according to the total ionospheric electron content corresponding to the tp The total amount of ionospheric electrons corresponding to the time is used to eliminate the ionospheric error. q+1 The meteorological data corresponding to the time is used to calculate the first astronomical observation data corresponding to the time. q The total amount of electrons in the ionosphere corresponding to time t q+1 The first astronomical observation data corresponding to time t is calculated q+1 The total amount of electrons in the ionosphere corresponding to the time t q+1 The total amount of electrons in the ionosphere corresponding to time t is calculated p The total amount of electrons in the ionosphere corresponding to the time t can be accurately obtained by taking into account the meteorological data. p The total amount of electrons in the ionosphere corresponding to the moment can be calculated, thereby improving the accuracy of eliminating ionospheric errors. In addition, since the first astronomical observation data is calculated based on meteorological data, the difficulty of obtaining astronomical observation data is reduced.
[0062] Optionally, the meteorological data includes n types, the first astronomical observation data includes m types, according to t q+1 The meteorological data corresponding to the time is calculated to get t q+1 In the step of the first astronomical observation data corresponding to time t, the preset m clustering models can be used to classify t q+1 Calculate the n types of meteorological data corresponding to the time and get t q+1 The first astronomical observation data of m types corresponding to the time.
[0063] In an embodiment of the present invention, the above-mentioned meteorological data may include n types such as temperature, wind direction, humidity, weather conditions, etc., and the above-mentioned first astronomical observation data includes m types such as F1Q.7, AP, UP and the southerly component of the interplanetary magnetic field.
[0064] Specifically, n types of meteorological data can be used as input data of the clustering model, so that the input data dimension of the clustering model is n-dimensional, and the output of the clustering model is m types of values such as F1Q.7, AP, UP and the southward component of the interplanetary magnetic field, that is, the output data dimension of the clustering model is m-dimensional.
[0065] Furthermore, t q+1 The meteorological data corresponding to the time is input into the clustering model, and after calculation by the clustering model, the output is t q+1 The first astronomical observation data corresponding to the time can be obtained from the meteorological data through the clustering model, which reduces the difficulty of obtaining the first astronomical observation data.
[0066] Optionally, by using a preset clustering model, q+1 The meteorological data corresponding to the time is calculated to obtain t q+1Before the step of obtaining the first astronomical observation data corresponding to the moment, n types of historical meteorological data and m types of historical astronomical observation data can also be obtained, and the historical meteorological data of each historical moment corresponds to the historical astronomical observation data of the same historical moment; the historical astronomical observation data of the same historical moment are clustered according to the historical meteorological data to obtain m clustering models, and the input of the clustering model is the n types of meteorological data.
[0067] In an embodiment of the present invention, n types of meteorological data for several years can be obtained as historical meteorological data, including temperature, wind direction, humidity, weather conditions and other types, and m types of first astronomical observation data for several years can be obtained as historical astronomical observation data, including F1Q.7, AP, UP and the southward component of the interplanetary magnetic field and other types.
[0068] The historical astronomical observation data corresponding to the historical meteorological data of the selected historical moments are clustered for analysis. For example, the historical astronomical observation data corresponding to the historical meteorological data of a certain hour are clustered for analysis to obtain m clustering models, each of which corresponds to the first astronomical observation data of a type.
[0069] A clustering model can be constructed through historical meteorological data and historical astronomical observation data, thereby establishing an implicit relationship between meteorological data and the first astronomical forecast data. The clustering model obtains the corresponding first astronomical observation data from the meteorological data, which reduces the difficulty of obtaining the first astronomical observation data.
[0070] Optionally, in the step of clustering the historical astronomical observation data at the same historical moment according to the historical meteorological data to obtain m clustering models, each type of astronomical observation data can be sorted by size and divided into k clustering models. i＝1，2，…m subclass, and k i＝1，2，…m The subclass is used as the label of meteorological data; the k-NN clustering algorithm is used to cluster the historical astronomical observation data at the same time as the historical meteorological data, and m clustering models are obtained, each of which includes k i＝1，2，…m Cluster centers.
[0071] In the embodiment of the present invention, the first astronomical prediction data of each type can be sorted by size and divided into k i＝1，2，…m Subclasses, such as dividing F1Q.7 type astronomical prediction data into k i＝1，2，…m subclasses, dividing AP type astronomical prediction data into k i＝1，2，…m subclasses, dividing the UP type astronomical prediction data into k i＝1，2，…m subclasses, and the astronomical prediction data of the interplanetary magnetic field southward component type are divided into k i＝1，2，…mEach subclass is used as a label for the corresponding meteorological data, and each group of meteorological data (including n types of meteorological data) corresponds to a label. For a new set of meteorological data, the k groups of meteorological data closest to it can be calculated. Among the labels corresponding to the k groups of meteorological data, the label with the largest number is selected as the label of the new meteorological data.
[0072] It can be understood that corresponding cluster centers can be formed according to the labels, and each label corresponds to a cluster center. For a type of astronomical prediction data, there are k i＝1，2，…m labels, corresponding to k i＝1，2，…m There are cluster centers, and each type of astronomical prediction data corresponds to a cluster model, so there are m cluster models.
[0073] A clustering model can be constructed through historical meteorological data and historical astronomical observation data, thereby establishing an implicit relationship between meteorological data and the first astronomical prediction data. By quantifying the astronomical observation data, the corresponding labels of the meteorological data are obtained, and the k-NN clustering algorithm is used to obtain m clustering models, which can be used to obtain the corresponding first astronomical observation data from the meteorological data, reducing the difficulty of obtaining the first astronomical observation data.
[0074] Optionally, in the q The total amount of electrons in the ionosphere corresponding to time t q+1 The first astronomical observation data corresponding to time t is calculated q+1 In the step of calculating the total amount of electrons in the ionosphere corresponding to the time instant, the preset ionosphere model can be used to calculate the total amount of electrons in the ionosphere corresponding to the time instant. q The total amount of electrons in the ionosphere corresponding to time t q+1 The first astronomical observation data corresponding to the time is predicted to get t q+1 The total amount of electrons in the ionosphere at the time.
[0075] In the embodiment of the present invention, the ionosphere model can be obtained based on a recurrent neural network or a long short-term memory neural network. q The total amount of electrons in the ionosphere corresponding to time t q+1 The first astronomical observation data corresponding to the time is predicted as input data and input into the preset ionosphere model. The preset ionosphere model outputs t q+1 The total amount of electrons in the ionosphere at the time.
[0076] Specifically, the preset ionosphere model learns t through the training process q The total amount of electrons in the ionosphere corresponding to time t q+1 The first astronomical observation data corresponding to time t q+1 The implicit relationship between the total amount of electrons in the ionosphere corresponding to the time t qThe total amount of electrons in the ionosphere corresponding to time t q+1 The first astronomical observation data corresponding to time is used to predict t q+1 The total amount of electrons in the ionosphere at the time.
[0077] The ionosphere model is preset to q+1 The total amount of electrons in the ionosphere corresponding to the time can be predicted and calculated to obtain t more accurately and quickly. q+1 The total amount of electrons in the ionosphere corresponding to the time instant, and combined with t q+1 The first astronomical observation data corresponding to the time further improved t q+1 The accuracy of the total amount of electrons in the ionosphere corresponding to the moment.
[0078] Optionally, the ionospheric model can be used to calculate the q The total amount of electrons in the ionosphere corresponding to time t q+1 The first astronomical observation data corresponding to the time is predicted to get t q+1 Before obtaining the total ionospheric electron quantity corresponding to the moment, historical meteorological data and second astronomical observation data can also be obtained, the second astronomical observation data including the total ionospheric electron quantity corresponding to each historical moment and the first astronomical observation data corresponding to each historical moment; the historical meteorological data are calculated by a clustering model to obtain the first astronomical observation data corresponding to each historical moment; the first astronomical observation data corresponding to each historical moment and the total ionospheric electron quantity corresponding to the same historical moment are used as sample data, and the total ionospheric electron quantity corresponding to the next moment of the historical moment is used as a label to construct a sample data set; a long short-term memory neural network is constructed, and the long short-term memory neural network is trained with the sample data set, and the ionospheric model is obtained after the training is completed.
[0079] In an embodiment of the present invention, the above-mentioned historical moment can be every historical hour or every historical minute. The embodiment of the present invention takes the historical moment as every historical hour as an example. The first astronomical observation data corresponding to every historical hour can be used as input, and the total amount of ionospheric electrons corresponding to the next hour can be used as a label to train the long short-term memory neural network. The network training iteration is performed once every 24 hours for the corresponding first astronomical observation data until the long short-term memory neural network converges or reaches a preset number of iterations, thereby obtaining a trained long short-term memory neural network as an ionosphere model.
[0080] By training a long short-term memory neural network as an ionosphere model, the ionosphere model can be enabled to capture implicit information of longer time series, thereby making the training results more accurate and improving the accuracy of the ionosphere model.
[0081] Optional, according to t p In the step of eliminating the ionospheric error based on the total ionospheric electron quantity corresponding to the time, the ionospheric error can be calculated based on tp The total amount of electrons in the ionosphere corresponding to the time delay is estimated to obtain t p The delay estimation at time t p The delay estimate at time t p The phase relationship at time t p The phase relationship at each moment is used as a benchmark, and phase capture is performed within a preset time range; the ionospheric error is eliminated based on the result of phase capture.
[0082] In the embodiment of the present invention, t p The total amount of electrons in the ionosphere corresponding to the time delay is estimated according to t p The delay estimate at time t p The phase relationship at time t p The phase relationship at each moment is used as a benchmark, and phase capture is performed within a preset time range, thereby narrowing the phase search range, accelerating the speed of phase synchronization, and further increasing the speed of eliminating ionospheric delay using the method of receiving dual-frequency signals.
[0083] The ionospheric error elimination method provided in the embodiment of the present invention constructs a relatively accurate ionospheric model, which can bring more accurate results for the estimation of ionospheric TEC; the data used are general meteorological data, which are more common and easy to obtain in actual operation; the phase synchronization method used can narrow the phase search range and accelerate the phase synchronization process.
[0084] It should be noted that the ionospheric error elimination method provided in the embodiment of the present invention can be applied to smart phones, navigation devices, computers, servers and other devices.
[0085] Optional, see Figure 2 , Figure 2 is a schematic diagram of the structure of an ionospheric error elimination device provided by an embodiment of the present invention, such as Figure 2 As shown, the device comprises:
[0086] The first acquisition module 201 is used to acquire t q The total amount of electrons in the ionosphere corresponding to the time t q+1 Meteorological data corresponding to the time;
[0087] The first calculation module 202 is used to calculate the q+1 The meteorological data corresponding to the time is calculated to get t q+1 The first astronomical observation data corresponding to the time;
[0088] The second calculation module 203 is used to calculate the q The total amount of electrons in the ionosphere corresponding to the time t q+1The first astronomical observation data corresponding to the time is calculated to obtain the t q+1 The total amount of electrons in the ionosphere corresponding to the time;
[0089] The third calculation module 204 is used to calculate the q+1 The total amount of electrons in the ionosphere corresponding to the time t is calculated p The total amount of electrons in the ionosphere corresponding to the time, where p is greater than q;
[0090] The error elimination module 205 is used to eliminate the error according to the t p The total amount of ionospheric electrons corresponding to the moment is used to eliminate the ionospheric error.
[0091] Optionally, the meteorological data includes n types, the first astronomical observation data includes m types, and the first calculation module 202 is further used to calculate the t q+1 Calculate the n types of meteorological data corresponding to the time and get t q+1 The first astronomical observation data of m types corresponding to the time.
[0092] Optionally, the device further comprises:
[0093] A second acquisition module is used to acquire n types of historical meteorological data and m types of historical astronomical observation data, wherein the historical meteorological data at each historical moment corresponds to the historical astronomical observation data at the same historical moment;
[0094] The clustering module is used to cluster the historical astronomical observation data at the same historical moment according to the historical meteorological data to obtain m clustering models, wherein the input of the clustering model is n types of meteorological data.
[0095] Optionally, the clustering module is further used to sort each type of astronomical observation data by size and divide it into k i＝1，2，…m subclass, and the k i＝1，2，…m The subclass is used as the label of the meteorological data; the historical astronomical observation data at the same time as the historical meteorological data are clustered using the k-NN clustering algorithm to obtain m clustering models, each of which includes k i＝1，,2，…m Cluster centers.
[0096] Optionally, the second calculation module 203 is further configured to calculate the t q The total amount of electrons in the ionosphere corresponding to the time t q+1 The first astronomical observation data corresponding to the time is predicted to obtain the t q+1 The total amount of electrons in the ionosphere at the time.
[0097] Optionally, the device further comprises:
[0098] A third acquisition module is used to acquire historical meteorological data and second astronomical observation data, wherein the second astronomical observation data includes the total ionospheric electron quantity corresponding to each historical moment and the first astronomical observation data corresponding to each historical moment;
[0099] A fourth calculation module, used to calculate the historical meteorological data by using the clustering model to obtain the first astronomical observation data corresponding to each historical moment;
[0100] A first construction module is used to construct a sample data set using the first astronomical observation data corresponding to each of the historical moments and the total amount of ionospheric electrons corresponding to the same historical moment as sample data, and using the total amount of ionospheric electrons corresponding to the next moment of the historical moment as a label;
[0101] The second construction module is used to construct a long short-term memory neural network, and train the long short-term memory neural network through the sample data set, and obtain the ionosphere model after the training is completed.
[0102] Optionally, the error elimination module 205 is further configured to: p The total amount of electrons in the ionosphere corresponding to the time delay is estimated to obtain t p The delay estimation at time t p The delay estimate at time t p The phase relationship at the moment; p The phase relationship at each moment is used as a reference, and phase capture is performed within a preset time range; and the ionospheric error is eliminated based on the result of the phase capture.
[0103] It should be noted that the ionospheric error elimination device provided in the embodiment of the present invention can be applied to smart phones, navigation devices, computers, servers and other devices that can perform navigation and positioning.
[0104] The ionospheric error elimination device provided in the embodiment of the present invention can implement each process implemented by the ionospheric error elimination method in the above method embodiment, and can achieve the same beneficial effect. To avoid repetition, it will not be described here.
[0105] See also Figure 3 , Figure 3 is a schematic diagram of the structure of an electronic device provided by an embodiment of the present invention, such as Figure 3 As shown, it includes: a memory 302, a processor 301, and a computer program of the ionospheric error elimination method stored in the memory 302 and executable on the processor 301, wherein:
[0106] The processor 301 is used to call the computer program stored in the memory 302 and execute the following steps:
[0107] Get q The total amount of electrons in the ionosphere corresponding to the time t q+1 Meteorological data corresponding to the time;
[0108] According to the q+1 The meteorological data corresponding to the time is calculated to get t q+1 The first astronomical observation data corresponding to the time;
[0109] Based on the q The total amount of electrons in the ionosphere corresponding to the time t q+1 The first astronomical observation data corresponding to the time is calculated to obtain the t q+1 The total amount of electrons in the ionosphere corresponding to the time;
[0110] According to the q+1 The total amount of electrons in the ionosphere corresponding to the time t is calculated p The total amount of electrons in the ionosphere corresponding to the time, where p is greater than q;
[0111] According to the p The total amount of ionospheric electrons corresponding to the moment is used to eliminate the ionospheric error.
[0112] Optionally, the meteorological data includes n types, the first astronomical observation data includes m types, and the processor 301 executes the method according to the t q+1 The meteorological data corresponding to the time is calculated to get t q+1 The steps of obtaining the first astronomical observation data corresponding to the time include:
[0113] Through the preset m clustering models, the t q+1 Calculate the n types of meteorological data corresponding to the time and get t q+1 The first astronomical observation data of m types corresponding to the time.
[0114] Optionally, in the case of using a preset clustering model, q+1 The meteorological data corresponding to the time is calculated to obtain t q+1 Before the step of obtaining the first astronomical observation data corresponding to the time instant, the method executed by the processor 301 further includes:
[0115] Acquire n types of historical meteorological data, and acquire m types of historical astronomical observation data, wherein the historical meteorological data at each historical moment corresponds to the historical astronomical observation data at the same historical moment;
[0116] The historical astronomical observation data at the same historical moment are clustered according to the historical meteorological data to obtain m clustering models, wherein the input of the clustering model is n types of meteorological data.
[0117] Optionally, the step of clustering the historical astronomical observation data at the same historical moment according to the historical meteorological data to obtain m clustering models performed by the processor 301 includes:
[0118] Sort each type of astronomical observation data by size and divide it into k i＝1，2，…m subclass, and the k i＝1，2，…m Subclasses serve as labels for meteorological data;
[0119] The k-NN clustering algorithm is used to perform cluster analysis on the historical astronomical observation data at the same time as the historical meteorological data to obtain m clustering models, each of which includes k i＝1，,2，…m Cluster centers.
[0120] Optionally, the processor 301 executes the q The total amount of electrons in the ionosphere corresponding to the time t q+1 The first astronomical observation data corresponding to the time is calculated to obtain the t q+1 The steps of calculating the total ionospheric electron quantity corresponding to the time include:
[0121] The t q The total amount of electrons in the ionosphere corresponding to the time t q+1 The first astronomical observation data corresponding to the time is predicted to obtain the t q+1 The total amount of electrons in the ionosphere at the time.
[0122] Optionally, in the case of using a preset ionospheric model to q The total amount of electrons in the ionosphere corresponding to the time t q+1 The first astronomical observation data corresponding to the time is predicted to obtain the t q+1 Before the total ionospheric electron quantity corresponding to the time instant, the method executed by the processor 301 further includes:
[0123] Acquire historical meteorological data and second astronomical observation data, wherein the second astronomical observation data includes the total amount of electrons in the ionosphere corresponding to each historical moment and the first astronomical observation data corresponding to each historical moment;
[0124] Calculating the historical meteorological data through the clustering model to obtain the first astronomical observation data corresponding to each historical moment;
[0125] The first astronomical observation data corresponding to each of the historical moments and the total amount of ionospheric electrons corresponding to the same historical moment are used as sample data, and the total amount of ionospheric electrons corresponding to the next moment of the historical moment is used as a label to construct a sample data set;
[0126] A long short-term memory neural network is constructed, and the long short-term memory neural network is trained using the sample data set, and the ionosphere model is obtained after the training is completed.
[0127] Optionally, the processor 301 executes the p The steps of eliminating ionospheric errors based on the total ionospheric electron quantity corresponding to the time include:
[0128] According to the p The total amount of electrons in the ionosphere corresponding to the time delay is estimated to obtain t p Delay estimation at the moment;
[0129] According to the p The delay estimate at time t p The phase relationship of the moments;
[0130] According to the above p The phase relationship at each moment is used as a benchmark to capture the phase within a preset time range;
[0131] The ionospheric error is eliminated through the result of the phase capture.
[0132] It should be noted that the electronic device provided in the embodiment of the present invention can be applied to smart phones, navigation devices, computers, servers and other devices that can eliminate ionospheric errors.
[0133] The electronic device provided in the embodiment of the present invention can implement each process implemented by the ionospheric error elimination method in the above method embodiment, and can achieve the same beneficial effect. To avoid repetition, it will not be described here.
[0134] An embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the computer program implements the various processes of the ionospheric error elimination method or the application-end ionospheric error elimination method provided in the embodiment of the present invention, and can achieve the same technical effect. To avoid repetition, it will not be repeated here.
[0135] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program, and the program can be stored in a computer-readable storage medium, and when the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, the storage medium can be a disk, an optical disk, a read-only memory (ROM) or a random access memory (RAM), etc.
[0136] The above disclosure is only the preferred embodiment of the present invention, which certainly cannot be used to limit the scope of the present invention. Therefore, equivalent changes made according to the claims of the present invention are still within the scope of the present invention.
Claims
1. A method for eliminating ionospheric errors, characterized in that: The following steps are involved: Get q The total amount of electrons in the ionosphere corresponding to the time t q+1 Meteorological data corresponding to the time; According to the q+1 The meteorological data corresponding to the time is calculated to get t q+1 The first astronomical observation data corresponding to the time; Based on the q The total amount of electrons in the ionosphere corresponding to the time t q+1 The first astronomical observation data corresponding to the time is calculated to obtain the t q+1 The total amount of ionospheric electrons corresponding to the time; the q The total amount of electrons in the ionosphere corresponding to the time t q+1 The first astronomical observation data corresponding to the time is calculated to obtain the t q+1 The step of calculating the total amount of electrons in the ionosphere corresponding to the time instant comprises: using a preset ionosphere model to calculate the total amount of electrons in the ionosphere corresponding to the time instant t q The total amount of electrons in the ionosphere corresponding to the time t q+1 The first astronomical observation data corresponding to the time is predicted to obtain the t q+1 The total amount of electrons in the ionosphere corresponding to the time; According to the q+1 The total amount of electrons in the ionosphere corresponding to the time t is calculated p The total amount of electrons in the ionosphere corresponding to the time, where p is greater than q; According to the p The total amount of ionospheric electrons corresponding to the moment is used to eliminate the ionospheric error; The meteorological data includes n types, the first astronomical observation data includes m types, and the q+1 The meteorological data corresponding to the time is calculated to get t q+1 The steps of obtaining the first astronomical observation data corresponding to the time include: Through the preset m clustering models, the t q+1 Calculate the n types of meteorological data corresponding to the time and get t q+1 The first astronomical observation data of m types corresponding to the time; In the based on the t q The total amount of electrons in the ionosphere corresponding to the time t q+1 The first astronomical observation data corresponding to the time is calculated to obtain the t q+1 Before the total ionospheric electron quantity corresponding to the time instant, the method further comprises: Acquire historical meteorological data and second astronomical observation data, wherein the second astronomical observation data includes the total amount of electrons in the ionosphere corresponding to each historical moment and the first astronomical observation data corresponding to each historical moment; Calculating the historical meteorological data through the clustering model to obtain the first astronomical observation data corresponding to each historical moment; The first astronomical observation data corresponding to each of the historical moments and the total amount of ionospheric electrons corresponding to the same historical moment are used as sample data, and the total amount of ionospheric electrons corresponding to the next moment of the historical moment is used as a label to construct a sample data set; Constructing a long short-term memory neural network, and training the long short-term memory neural network using the sample data set, and obtaining the ionosphere model after the training is completed; According to the t p The steps of eliminating ionospheric errors based on the total ionospheric electron quantity corresponding to the time include: According to the p The total amount of electrons in the ionosphere corresponding to the time delay is estimated to obtain t p Delay estimation at the moment; According to the p The delay estimate at time t p The phase relationship of the moments; According to the t p The phase relationship at each moment is used as a benchmark to capture the phase within a preset time range; The ionospheric error is eliminated through the result of the phase capture.
2. The method according to claim 1, characterized in that In the preset clustering model, the t q+1 The meteorological data corresponding to the time is calculated to obtain t q+1 Before the step of obtaining the first astronomical observation data corresponding to the time instant, the method further comprises: Acquire n types of historical meteorological data, and acquire m types of historical astronomical observation data, wherein the historical meteorological data at each historical moment corresponds to the historical astronomical observation data at the same historical moment; The historical astronomical observation data at the same historical moment are clustered according to the historical meteorological data to obtain m clustering models, wherein the input of the clustering model is n types of meteorological data.
3. The method according to claim 2, characterized in that The step of clustering the historical astronomical observation data at the same historical moment according to the historical meteorological data to obtain m clustering models comprises: Sort each type of astronomical observation data by size and divide it into k i=1，2，…m subclass, and the k i=1,2，…m Subclasses serve as labels for meteorological data; The k-NN clustering algorithm is used to perform cluster analysis on the historical astronomical observation data at the same time as the historical meteorological data to obtain m clustering models, each of which includes k i=1，2，…m Cluster centers.
4. An ionospheric error elimination device, characterized in that: include: The first acquisition module is used to obtain t q The total amount of electrons in the ionosphere corresponding to the time t q+1 Meteorological data corresponding to the time; The first calculation module is used to calculate the q+1 The meteorological data corresponding to the time is calculated to get t q+1 The first astronomical observation data corresponding to the time; The second calculation module is used to calculate the q The total amount of electrons in the ionosphere corresponding to the time t q+1 The first astronomical observation data corresponding to the time is calculated to obtain the t q+1 The second calculation module is also used to calculate the total amount of ionospheric electrons corresponding to the time t through a preset ionospheric model q The total amount of electrons in the ionosphere corresponding to the time t q+1 The first astronomical observation data corresponding to the time is predicted to obtain the t q+1 The total amount of electrons in the ionosphere corresponding to the time; The third calculation module is used to calculate the q+1 The total amount of electrons in the ionosphere corresponding to the time t is calculated p The total amount of electrons in the ionosphere corresponding to the time, where p is greater than q; The error elimination module is used to eliminate the error according to the t p The total amount of ionospheric electrons corresponding to the moment is used to eliminate the ionospheric error; The meteorological data includes n types, the first astronomical observation data includes m types, and the first calculation module is further used to calculate the t q+1 Calculate the n types of meteorological data corresponding to the time and get t q+1 The first astronomical observation data of m types corresponding to the time; The device also includes: A third acquisition module is used to acquire historical meteorological data and second astronomical observation data, wherein the second astronomical observation data includes the total ionospheric electron quantity corresponding to each historical moment and the first astronomical observation data corresponding to each historical moment; A fourth calculation module, used to calculate the historical meteorological data by using the clustering model to obtain the first astronomical observation data corresponding to each historical moment; A first construction module is used to construct a sample data set using the first astronomical observation data corresponding to each of the historical moments and the total amount of ionospheric electrons corresponding to the same historical moment as sample data, and using the total amount of ionospheric electrons corresponding to the next moment of the historical moment as a label; A second construction module is used to construct a long short-term memory neural network, and train the long short-term memory neural network through the sample data set, and obtain the ionosphere model after the training is completed; The error elimination module is also used to p The total amount of electrons in the ionosphere corresponding to the time delay is estimated to obtain t p The delay estimation at time t p The delay estimate at time t p The phase relationship at time t p The phase relationship at each moment is used as a reference, phase capture is performed within a preset time range, and the ionospheric error is eliminated based on the result of the phase capture.
5. An electronic device, characterized in that: include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the steps in the ionospheric error elimination method according to any one of claims 1 to 3 are implemented.
6. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps in the ionospheric error elimination method according to any one of claims 1 to 3 are implemented.