Power terminal positioning method based on multi-frequency combination of Beidou

Abstract

The application relates to the technical field of satellite navigation and positioning, and discloses a power terminal positioning method based on a Beidou multi-frequency combination, which comprises the following steps: obtaining an original Beidou multi-frequency combination signal stream, inputting a pre-trained positioning parameter matching network to obtain a matching result set, and determining the positioning parameter matching result corresponding to each signal. The positioning parameter matching network is trained through historical positioning data, the historical data contains signal quality parameters, positioning abnormality labels and weight coefficients extracted from historical Beidou multi-frequency signals, and the signal quality parameters are determined in combination with a single signal and adjacent signals. The method further comprises the following steps: extracting first and second groups of signal quality parameters and fusing the same, determining the positioning abnormality labels and the weight coefficients, training the network by using historical joint positioning records, and adjusting the network parameters according to the training error. Through multi-frequency signal fusion and adaptive learning, the method improves the positioning precision and anti-interference ability of the power terminal in a complex environment, and is suitable for the precise positioning demand of a smart power grid.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of satellite navigation and positioning, in particular to a power terminal positioning method based on a Beidou multi-frequency combination. BACKGROUND

[0002] In the power system, accurate positioning of power terminal devices is crucial for power grid operation monitoring, fault repair, asset management, and other aspects. Traditional power terminal positioning methods mainly rely on single-band satellite signals or other positioning technologies. However, these methods face many challenges in complex power environments.

[0003] First, the environment where power facilities are located often has multiple interference sources, such as electromagnetic interference from high-voltage transmission lines and multipath effects in urban building groups. Single-band signals are easily disturbed, leading to decreased positioning accuracy or even positioning failure. In existing technologies, positioning algorithms based on single-band signals cannot effectively deal with signal distortion problems in complex environments, making it difficult to accurately analyze the effective information in the signal and affecting the reliability of the positioning result.

[0004] Second, with the development of smart grids, higher requirements are placed on the real-time, high-precision, and stability of power terminal positioning. Traditional positioning methods lack comprehensive analysis of signal timing correlation characteristics and frequency band distribution characteristics when dealing with multi-source heterogeneous data, making it difficult to fully utilize the complementary advantages of multi-frequency signals. For example, different frequency bands of Beidou signals have different characteristics during transmission. High-frequency signals may carry more detailed information but have weaker penetration, while low-frequency signals have stronger anti-interference ability but lower resolution. How to effectively fuse the characteristics of multi-frequency signals becomes a key issue in improving positioning performance.

[0005] In addition, existing positioning parameter matching algorithms are usually based on fixed model parameters and lack adaptive learning ability. When facing changing signal environments, they cannot dynamically adjust the matching strategy according to real-time signal characteristics, resulting in insufficient generalization ability of the positioning model. In the power field, the electromagnetic environment differs greatly in different regions, and traditional methods are difficult to meet the positioning needs in diversified scenarios. For example, in different geographical environments such as mountainous areas and urban centers, the propagation characteristics of Beidou signals differ significantly. Fixed-parameter positioning models cannot flexibly adapt to these changes, leading to unstable positioning accuracy. SUMMARY

[0006] The purpose of the present application is to provide a power terminal positioning method based on a Beidou multi-frequency combination to solve the problems raised in the background.

[0007] To achieve the above purpose, the present application provides the following technical solution: a power terminal positioning method based on a Beidou multi-frequency combination, which is applied to a power terminal positioning system, and the method comprises:

[0008] obtaining an original Beidou multi-frequency combined signal stream to be positioned;

[0009] inputting the original Beidou multi-frequency combined signal stream into a pre-trained positioning parameter matching network to obtain a positioning parameter matching result set;

[0010] determining, based on the positioning parameter matching result set, a positioning parameter matching result corresponding to each Beidou multi-frequency signal in the original Beidou multi-frequency combined signal stream;

[0011] The original Beidou multi-frequency combined signal stream includes target frequency band signal data to be parsed.

[0012] The positioning parameter matching network is a parameter matching network obtained by inputting historical positioning data into an initial parameter matching network to be trained, the historical positioning data including a batch of historical signal quality parameters extracted from a batch of historical Beidou multi-frequency signals, and a positioning anomaly label and a weight coefficient corresponding to each historical signal quality parameter, the positioning anomaly label reflecting an interference state identifier of a historical signal quality parameter, and the weight coefficient reflecting a correction factor corresponding to the interference state identifier, the batch of historical Beidou multi-frequency signals being a batch of continuously collected Beidou multi-frequency signals, and each historical signal quality parameter including a signal quality parameter determined by a historical Beidou multi-frequency signal and an adjacent historical Beidou multi-frequency signal of the historical Beidou multi-frequency signal.

[0013] The positioning parameter matching result is used to represent a parsing parameter identifier of the target frequency band signal data.

[0014] Preferably, the method further comprises:

[0015] performing first feature parameter extraction on the batch of historical Beidou multi-frequency signals to obtain a first group of signal quality parameters, wherein one signal quality parameter in the first group of signal quality parameters corresponds to one historical Beidou multi-frequency signal in the batch of historical Beidou multi-frequency signals.

[0016] performing second feature parameter extraction on the batch of historical Beidou multi-frequency signals to obtain a second group of signal quality parameters, wherein a first signal quality parameter in the second group of signal quality parameters corresponds to a first historical Beidou multi-frequency signal in the historical Beidou multi-frequency signals, an adjacent historical Beidou multi-frequency signal of the first historical Beidou multi-frequency signal being a second historical Beidou multi-frequency signal, and the first signal quality parameter representing a time sequence correlation feature between the first historical Beidou multi-frequency signal and the second historical Beidou multi-frequency signal.

[0017] fusing the first group of signal quality parameters and the second group of signal quality parameters to obtain the batch of historical signal quality parameters.

[0018] Preferably, the second characteristic parameter extraction on the batch of historical Beidou multi-frequency signals is performed to obtain a second group of signal quality parameters, including:

[0019] The second characteristic parameter extraction on the batch of historical Beidou multi-frequency signals is performed by the following steps, and a second group of signal quality parameters is obtained, wherein each time the historical Beidou multi-frequency signal subjected to the second characteristic parameter extraction is regarded as a current historical Beidou multi-frequency signal, and the obtained signal quality parameter is regarded as a current signal quality parameter, and the second group of signal quality parameters includes the current signal quality parameter;

[0020] A frequency band analysis operation is performed on the current historical Beidou multi-frequency signal to determine a first group of frequency band characteristic quantities, wherein the first group of frequency band characteristic quantities are used to describe the frequency band distribution of the target frequency band signal data in the current historical Beidou multi-frequency signal;

[0021] A frequency band analysis operation is performed on the adjacent historical Beidou multi-frequency signal of the current historical Beidou multi-frequency signal to determine a second group of frequency band characteristic quantities, wherein the second group of frequency band characteristic quantities are used to describe the frequency band distribution of the target frequency band signal data in the adjacent historical Beidou multi-frequency signal;

[0022] The current signal quality parameter is determined based on the first group of frequency band characteristic quantities and the second group of frequency band characteristic quantities, wherein the current signal quality parameter is used to represent the deviation between the corresponding frequency band characteristic quantities in the first group of frequency band characteristic quantities and the second group of frequency band characteristic quantities.

[0023] Preferably, the method further includes:

[0024] The positioning anomaly label and the weight coefficient corresponding to each historical signal quality parameter in the batch of historical signal quality parameters are determined respectively by the following steps, wherein each time the historical signal quality parameter subjected to the determination of the positioning anomaly label and the weight coefficient is regarded as a current historical signal quality parameter, and the positioning anomaly label and the weight coefficient corresponding to the current historical signal quality parameter are regarded as a current positioning anomaly label and a current weight coefficient:

[0025] The current positioning anomaly label is obtained based on the current historical signal quality parameter, wherein the initial parameter matching network is a parameter matching network obtained by pre-executing initialization processing;

[0026] determining the current weight coefficient based on the current positioning anomaly label and a reference verification label corresponding to the current historical signal quality parameter, wherein the reference verification label carries a calibration label in a historical Beidou multi-frequency signal corresponding to the current historical signal quality parameter, the calibration label includes a target analysis parameter label of the target frequency band signal data, and the weight coefficient is used to represent whether the current positioning anomaly label is consistent with the target analysis parameter label.

[0027] Preferably, the method further comprises:

[0028] inputting each analysis parameter label in a pre-stored analysis parameter label set into a first matching algorithm branch of the initial parameter matching network in sequence together with the current historical signal quality parameter, to obtain a first set of matching degree indexes, wherein the first set of matching degree indexes includes a matching degree index corresponding to each analysis parameter label;

[0029] determining a candidate analysis parameter label corresponding to a first matching degree index with the highest value in the first set of matching degree indexes as a current analysis parameter label corresponding to the current positioning anomaly label.

[0030] Preferably, before the step of inputting each analysis parameter label in a pre-stored analysis parameter label set into a first matching algorithm branch of the initial parameter matching network in sequence together with the current historical signal quality parameter, to obtain a first set of matching degree indexes, the method further comprises:

[0031] obtaining the signal intensity distribution quantity by statistical analysis of signal intensity data of the target frequency band in the current historical signal quality parameter;

[0032] determining whether the signal intensity distribution quantity conforms to a set intensity distribution rule;

[0033] on the basis that the signal intensity distribution quantity conforms to the set intensity distribution rule, dynamically selecting a candidate analysis parameter label in the pre-stored analysis parameter label set as the current positioning anomaly label;

[0034] on the basis that the signal intensity distribution quantity does not conform to the set intensity distribution rule, inputting the current historical signal quality parameter into the initial parameter matching network to obtain the current positioning anomaly label.

[0035] Preferably, the step of determining a candidate analysis parameter label corresponding to a first matching degree index with the highest value in the first set of matching degree indexes as a current analysis parameter label corresponding to the current positioning anomaly label comprises:

[0036] On the basis that the candidate analysis parameter identifier corresponding to the first matching degree indicator with the highest value in the first set of matching degree indicators is the same as the target analysis parameter identifier, the current weight coefficient is determined as a first weight coefficient, where the first weight coefficient indicates that the initial parameter matching network successfully generates a matching identifier;

[0037] On the basis that the candidate analysis parameter identifier corresponding to the first matching degree indicator with the highest value in the first set of matching degree indicators is different from the target analysis parameter identifier, the current weight coefficient is determined as a second weight coefficient, where the second weight coefficient indicates that the initial parameter matching network does not successfully generate a matching identifier.

[0038] Preferably, the method further comprises:

[0039] filtering a plurality of sets of historical joint positioning records from the historical positioning data, where the jth set of historical joint positioning records in the plurality of sets of historical joint positioning records includes a jth historical signal quality parameter, a jth positioning anomaly label corresponding to the jth historical signal quality parameter, a jth weight coefficient, and a (j+1)th historical signal quality parameter, j being a positive integer;

[0040] training an initial parameter matching network to be trained based on the plurality of sets of historical joint positioning records to obtain the positioning parameter matching network, where on the basis that the number of iterations of training the initial parameter matching network reaches a set training threshold, the initial parameter matching network is determined as the positioning parameter matching network, and on the basis that the number of iterations of training the initial parameter matching network does not reach the set training threshold, network parameters of the initial parameter matching network are updated according to a preset training error function, and the input of each training process is a set of historical joint positioning records in the plurality of sets of historical joint positioning records.

[0041] Preferably, the method further comprises:

[0042] training an initial parameter matching network to be trained based on a plurality of sets of historical joint positioning records to obtain the positioning parameter matching network, where the historical joint positioning record input into the initial parameter matching network is the jth set of joint positioning records:

[0043] determining whether the historical Beidou multi-frequency signal corresponding to the (j+1)th historical signal quality parameter is a last historical Beidou multi-frequency signal in a historical Beidou multi-frequency signal stream;

[0044] on the basis that the historical Beidou multi-frequency signal corresponding to the (j+1)th historical signal quality parameter is the last historical Beidou multi-frequency signal, determining a training parameter generated by the initial parameter matching network based on the jth weight coefficient;

[0045] determining the training parameter based on the jth weight coefficient and an output of an associated parameter matching network, wherein the associated parameter matching network is a parameter matching network obtained by performing an initialization process in advance, and the network parameters of the associated parameter matching network are different from the network parameters of the initial parameter matching network;

[0046] determining an error amount of the training error function based on the output of the initial parameter matching network and the training parameter, and adjusting the network parameters of the initial parameter matching network based on the error amount of the training error function;

[0047] determining the initial parameter matching network as the positioning parameter matching network when the number of iterations of the above steps reaches the set training threshold;

[0048] The determining of the training parameter based on the jth weight coefficient and the output of the associated parameter matching network comprises:

[0049] inputting each analysis parameter identifier in a pre-stored analysis parameter identifier set into the second matching algorithm branch of the associated parameter matching network in sequence together with the j+1th historical signal quality parameter, to obtain a second group of matching degree indexes, wherein the second group of matching degree indexes comprises a matching degree index corresponding to each analysis parameter identifier;

[0050] determining the second matching degree index with the highest value in the second group of matching degree indexes and the weighted result of the jth weight coefficient as the training parameter.

[0051] Preferably, the method further comprises:

[0052] performing a joint training operation on each group of records in the plurality of groups of historical joint positioning records to generate an optimized parameter of the positioning parameter matching network, wherein the joint training operation of each group of records comprises adjusting an update amplitude of the network parameters based on the associated features between adjacent historical signal quality parameters;

[0053] after the joint training operation of each group of records is completed, verifying the analysis parameter identifier matching accuracy of the positioning parameter matching network for target frequency band signal data, and re-executing the training process when the accuracy does not reach a set standard.

[0054] Compared with the prior art, the method has the following beneficial effects:

[0055] The original Beidou multi-frequency combined signal stream is obtained and input into a pre-trained positioning parameter matching network to realize comprehensive processing of the multi-frequency signal. The positioning parameter matching network is trained using historical positioning data, which includes signal quality parameters determined in combination with adjacent signals, and can effectively capture the timing correlation characteristics and frequency band distribution characteristics of the signal. For example, when extracting the signal quality parameters, both the frequency band characteristic quantity of a single signal and the frequency band deviation quantity between adjacent signals are considered. This dual feature extraction method enables the model to more comprehensively understand the signal characteristics, thereby improving the accuracy of analyzing the target frequency band signal data.

[0056] The positioning exception label and weight coefficient are introduced into the historical positioning data, and dynamic calibration of the model output is realized by comparing the reference verification label with the target analysis parameter label. When the matching degree index corresponds to the candidate analysis parameter label consistent with the target analysis parameter label, a higher weight coefficient is assigned to strengthen the correct matching. When they are inconsistent, the weight is reduced to guide the model to optimize the matching strategy. This mechanism enhances the model's ability to identify interference signals, effectively suppresses the influence of environmental factors such as electromagnetic interference and multipath effect on the positioning result, and improves the anti-interference ability of the positioning.

[0057] During network training, multiple sets of historical joint positioning records are used for iterative training, and the output of the associated parameter matching network is used to dynamically adjust the training parameters. By analyzing the correlation between adjacent historical signal quality parameters, the update amplitude of the network parameters is optimized, enabling the positioning parameter matching network to adaptively learn the variation rules of signal characteristics. For example, based on the judgment of whether the adjacent signals are end signals, different training parameter calculation methods are used to ensure that the model can achieve accurate matching in different signal sequence scenarios, improving the model's generalization ability and adaptability to diverse power environments.

[0058] This method realizes joint analysis of signal spatial and temporal characteristics by fusing the first and second sets of signal quality parameters. The first set of parameters reflects the independent characteristics of a single signal, and the second set of parameters reflects the timing correlation of adjacent signals. The fusion of the two enables the model to analyze signals from multiple dimensions, effectively addressing the limitations of single feature analysis in traditional methods. This multi-dimensional feature fusion strategy not only improves the richness of signal quality parameters, but also provides a more solid foundation for accurate matching of positioning parameters, thereby improving the overall performance of power terminal positioning.

[0059] By setting a training threshold and an error function, the positioning parameter matching network is continuously optimized during training until the set accuracy standard is reached. This closed-loop training mechanism ensures the stability and reliability of the model, enabling the positioning method to continuously provide high-precision positioning services in actual power scenarios, meeting the demand for real-time and accurate positioning of power terminal devices in smart grids. BRIEF DESCRIPTION OF DRAWINGS

[0060] Fig. 1 The working principle diagram of the power terminal positioning method based on the Beidou multi-frequency combination of the application;

[0061] Fig. 2 The flowchart for extracting the historical signal quality parameter;

[0062] Fig. 3 The design diagram for extracting the second feature parameter. DETAILED DESCRIPTION

[0063] The technical solutions in the embodiments of the application will be clearly and completely described below with reference to the drawings in the embodiments of the application. Obviously, the described embodiments are only part of the embodiments of the application, rather than all the embodiments of the application. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative work fall within the protection scope of the application.

[0064] Please refer to Figs. 1-3 The power terminal positioning method based on the Beidou multi-frequency combination of the application involves the following specific implementation steps:

[0065] An original Beidou multi-frequency combination signal stream to be positioned is acquired, and the signal stream contains target frequency band signal data to be parsed.

[0066] The original Beidou multi-frequency combination signal stream is input into a pre-trained positioning parameter matching network to obtain a positioning parameter matching result set. The positioning parameter matching network is obtained by training in the following manner: historical positioning data is input into an initial parameter matching network to be trained, and the historical positioning data is generated by a batch of historical signal quality parameters extracted from a batch of historical Beidou multi-frequency signals and a positioning anomaly label and a weight coefficient corresponding to each historical signal quality parameter. The historical Beidou multi-frequency signals are continuously collected signals, and each historical signal quality parameter is determined by a historical Beidou multi-frequency signal and its adjacent historical Beidou multi-frequency signals. The positioning anomaly label reflects the interference state identifier of the historical signal quality parameter, and the weight coefficient reflects the correction factor corresponding to the interference state identifier.

[0067] Based on the positioning parameter matching result set, the positioning parameter matching result corresponding to each Beidou multi-frequency signal in the original Beidou multi-frequency combination signal stream is determined, and the result is used to represent the parsed parameter identifier of the target frequency band signal data.

[0068] The application will be further described below in combination with Examples 1 to 5:

[0069] Example 1:

[0070] The extraction process of historical signal quality parameters is refined in this embodiment. The specific steps are as follows:

[0071] A batch of continuously collected historical Beidou multi-frequency signals are subjected to first feature parameter extraction to obtain a first group of signal quality parameters, wherein each signal quality parameter corresponds to a corresponding single historical Beidou multi-frequency signal. The first feature parameter extraction process focuses on the independent features of a single signal, and the unique feature quantity of each signal is obtained through time domain and frequency domain analysis. For example, for each historical Beidou multi-frequency signal, the basic attribute parameters such as amplitude, frequency and phase of the signal can be extracted through Fourier transform and other signal processing methods to form the first group of signal quality parameters. In this process, the feature extraction of each signal is carried out independently, aiming to capture the state information of a single signal at a specific time and provide basic data support for subsequent analysis.

[0072] The same batch of historical Beidou multi-frequency signals are subjected to second feature parameter extraction to obtain a second group of signal quality parameters. The first signal quality parameter in the second group of signal quality parameters corresponds to the first historical Beidou multi-frequency signal, and the adjacent signal of the first historical Beidou multi-frequency signal is the second historical Beidou multi-frequency signal. The first signal quality parameter represents the time sequence correlation feature between the first signal and the second signal. Specifically, the second feature parameter extraction process needs to consider the correlation between adjacent signals to dynamically capture the change rule of the signal in the time sequence.

[0073] In the second feature parameter extraction process, the historical Beidou multi-frequency signal for each feature extraction is taken as the current historical Beidou multi-frequency signal, and the signal quality parameter obtained is taken as the current signal quality parameter. The second group of signal quality parameters includes all the current signal quality parameters. The specific operation is as follows: performing a frequency band analysis operation on the current historical Beidou multi-frequency signal to determine a first group of frequency band features, which are used to describe the frequency band distribution of the target frequency band signal data in the current signal. The frequency band analysis can be realized by a spectrum analysis algorithm, for example, calculating the energy proportion, frequency offset, bandwidth change and other parameters of the target frequency band in the current signal to characterize the distribution characteristics of the target frequency band in the current signal. Perform the same frequency band analysis operation on the adjacent historical Beidou multi-frequency signal of the current historical Beidou multi-frequency signal to determine a second group of frequency band features, which are used to describe the frequency band distribution of the target frequency band signal data in the adjacent signal. The selection of the adjacent signal follows the continuity of the time sequence, that is, the signal collected continuously before or after the current signal, which can be determined according to the time sequence direction of signal processing. Based on the first group of frequency band features and the second group of frequency band features, the current signal quality parameter is determined. The specific method is to calculate the deviation between the corresponding frequency band features of the two, which is taken as the current signal quality parameter, and is used to represent the difference degree of the adjacent signal in the frequency band distribution. For example, if the energy proportion of the target frequency band in the first group of frequency band features is A, and the corresponding value in the second group is B, then the deviation can be represented as |A-B|; if the frequency offset is and then the deviation can be represented as In this way, the frequency band distribution difference of the adjacent signal is quantified into a specific numerical parameter, so as to reflect the stability or change trend of the signal in the time sequence.

[0074] Finally, the first group of signal quality parameters and the second group of signal quality parameters are fused respectively to obtain a batch of complete historical signal quality parameters. The fusion process aims to combine the independent features of a single signal with the associated features of adjacent signals to form a more comprehensive signal quality description. The fusion method can adopt data splicing, feature weighting and other methods, for example, combining each signal quality parameter in the first group with the corresponding adjacent deviation parameter in the second group to form a multi-dimensional feature vector. Among them, data splicing is to arrange the feature parameters of different groups in order to form a feature space containing more dimensions; feature weighting is to assign corresponding weights according to the importance of different features, and then perform linear combination to highlight the contribution of key features to signal quality description. Through the fusion processing, the historical signal quality parameter can more comprehensively reflect the characteristics of the Beidou multi-frequency signal, which contains not only the static features of a single signal, but also the dynamic associated features of adjacent signals, providing rich feature input for the subsequent positioning parameter matching network training, so that it can learn more complex mapping relationships between signal patterns and positioning parameters.

[0075] During the entire implementation process, it is necessary to ensure the continuity of the historical Beidou multi-frequency signal collection to ensure that the timing correlation characteristics between adjacent signals have actual physical meaning. At the same time, the specific algorithm of feature extraction and fusion can be selected and adjusted according to the actual signal processing needs, but it needs to revolve around the core goal of obtaining parameters that can accurately reflect the signal quality, and ensure that the subsequent positioning parameter matching network can be effectively trained based on reliable historical data, so as to improve the analysis ability of the target frequency band signal data in the original Beidou multi-frequency combined signal stream and the accuracy of the positioning parameter matching.

[0076] Embodiment 2:

[0077] During the construction of historical positioning data, for each current historical signal quality parameter in a batch of historical signal quality parameters, a determination operation of positioning abnormal label and weight coefficient needs to be performed, and the specific implementation is as follows:

[0078] The current historical signal quality parameter is input into the initial parameter matching network initialized in advance to obtain the current positioning abnormal label. The initial parameter matching network is a neural network model constructed based on a deep learning architecture, and its network structure includes an input layer, a hidden layer and an output layer. Through learning a large amount of historical data, it can capture the potential mapping relationship between signal quality parameters and positioning abnormal labels.

[0079] When obtaining the current positioning abnormal label, a pre-stored analysis parameter identifier set is obtained, which contains a plurality of possible analysis parameter identifiers for representing different analysis states of the target frequency band signal data. For example, the analysis parameter identifier can include signal normal state, multipath effect interference state, ionospheric scintillation interference state and other types of positioning abnormal conditions. Each analysis parameter identifier in the set is input into the first matching algorithm branch of the initial parameter matching network in sequence together with the current historical signal quality parameter. The first matching algorithm branch is a sub-module in the initial parameter matching network, which contains multiple layers of neural network structure inside, for feature extraction and matching degree calculation of the input signal quality parameter and analysis parameter identifier. Through the calculation of this branch, a first set of matching degree indexes can be obtained, each set of matching degree indexes corresponding to an analysis parameter identifier, reflecting the matching degree of the identifier and the current signal quality parameter. The calculation of the matching degree index is based on the output result of the neural network, and is usually represented in the form of probability value or similarity score, and the higher the value, the better the matching degree.

[0080] The first matching degree index with the highest value is selected from the first set of matching degree indexes, and the corresponding candidate analysis parameter identifier is determined as the current analysis parameter identifier corresponding to the current positioning anomaly label. For example, if the analysis parameter identifier set contains three identifiers A, B, and C, and the calculation of the first matching algorithm branch obtains a matching degree of 0.8 for A, 0.6 for B, and 0.7 for C, the highest matching degree A is selected as the current analysis parameter identifier, i.e., the current positioning anomaly label.

[0081] Before performing the above matching operation, the signal strength distribution quantity needs to be determined through a preset mode. The signal strength distribution quantity is a parameter that describes the distribution of the signal strength of the target frequency band in the time or frequency dimension, and can be obtained by statistical analysis of the signal strength data of the target frequency band in the current historical signal quality parameter. For example, the mean, variance, maximum, minimum, and other statistical quantities of the signal strength are calculated, or a distribution graph of the signal strength with respect to time or frequency is drawn. Then, it is determined whether the signal strength distribution quantity conforms to the set strength distribution rule. The strength distribution rule is a pre-set judgment standard for evaluating the rationality of the signal strength distribution. For example, it is determined whether the signal strength is within a pre-set threshold range, whether there is an abnormal mutation or fluctuation, whether it conforms to a specific distribution pattern (such as a normal distribution), etc.

[0082] If the signal strength distribution quantity conforms to the set strength distribution rule, it means that the signal state is stable, and the candidate analysis parameter identifier in the pre-stored analysis parameter identifier set can be directly selected as the current positioning anomaly label without the need for complex calculation through the initial parameter matching network. In this case, the corresponding analysis parameter identifier can be quickly matched according to the characteristics of the signal strength distribution, improving the processing efficiency. If it does not conform to the rule, the current historical signal quality parameter must be input into the initial parameter matching network, and the current positioning anomaly label is obtained through algorithm calculation to ensure the accuracy of the label. This is because when the signal strength distribution is abnormal, directly selecting the analysis parameter identifier may lead to incorrect positioning results, and through the calculation of the initial parameter matching network, multiple factors can be considered to improve the reliability of the label.

[0083] The current weight coefficient is determined based on the reference verification identifier corresponding to the current positioning anomaly label and the current historical signal quality parameter. The reference verification identifier is a reference identifier pre-set for each historical signal quality parameter, which carries the calibration label in the historical Beidou multi-frequency signal corresponding to the current historical signal quality parameter. The calibration label contains the target analysis parameter identifier of the target frequency band signal data. The target analysis parameter identifier is an analysis parameter identifier that is known and accurately reflects the true state of the signal, which can be obtained through manual annotation or other reliable methods.

[0084] The specific judgment logic is: if the candidate analysis parameter identifier corresponding to the first matching degree index (i.e. the current analysis parameter identifier) is the same as the target analysis parameter identifier, it means that the initial parameter matching network successfully generates a matching identifier, and the current weight coefficient is determined as the first weight coefficient. The first weight coefficient is a pre-set value, which is used to represent the state of successful matching, and is usually set to a large value to emphasize the importance of this group of data to network training. If they are different, it means that the matching fails, and the current weight coefficient is determined as the second weight coefficient. The second weight coefficient is also a pre-set value, which is used to represent the state of matching failure, and is usually set to a small value to reduce the impact of this group of data on network training.

[0085] When determining the weight coefficient, the adjustment effect of the signal strength distribution quantity on the weight also needs to be considered. If the signal strength distribution quantity meets the set strength distribution rule, it means that the signal quality is good, at which time the weight coefficient can be appropriately increased to enhance the influence of this group of data in network training; if it does not meet the rule, it means that the signal quality is poor, and there is a large interference, at which time the weight coefficient can be appropriately reduced to reduce the interference of noise data on network training.

[0086] In this way, each historical signal quality parameter is assigned a weight coefficient corresponding to its matching state and signal quality, so that in the subsequent training process of the positioning parameter matching network, more attention can be paid to data with successful matching and good signal quality, improving the learning efficiency and generalization ability of the network. At the same time, the introduction of the weight coefficient can also effectively handle noise data and abnormal data, enhance the robustness of the network, and enable it to accurately analyze Beidou multi-frequency signals in complex electromagnetic environments, providing reliable positioning services for power terminals.

[0087] During the entire implementation process, the accuracy and reliability of the reference verification identifier need to be ensured, and its quality directly affects the reasonableness of the weight coefficient and the effect of network training. At the same time, the initialization process of the initial parameter matching network is also crucial, and appropriate initialization methods (such as random initialization, pre-training initialization, etc.) should be used, and reasonable network parameters (such as learning rate, iteration times, etc.) should be set to ensure that the network can effectively learn the mapping relationship between the signal quality parameter and the positioning abnormal label. In addition, the construction of the analysis parameter identifier set should comprehensively cover possible positioning abnormal situations to ensure accurate identification of signal states in various scenarios.

[0088] Embodiment 3:

[0089] In the training link of the positioning parameter matching network, a plurality of groups of historical joint positioning records need to be constructed based on historical positioning data, and the performance of the initial parameter matching network is optimized through iterative training, and the specific implementation manner is as follows:

[0090] A plurality of sets of historical joint positioning records are screened from historical positioning data. Each set of historical joint positioning records (the jth set, j being a positive integer) includes a jth historical signal quality parameter, a jth positioning anomaly label corresponding to the parameter, a jth weight coefficient, and a (j+1)th historical signal quality parameter. The records are constructed in accordance with the continuity of the time sequence, i.e., the historical Beidou multi-frequency signal corresponding to the (j+1)th historical signal quality parameter is the next continuously collected signal of the jth signal, ensuring that the time sequence correlation characteristics between adjacent signals can be effectively captured. By screening consecutive signal pairs, joint records containing signal characteristics, matching labels, and weights are formed, providing sample data with time sequence dependence for network training.

[0091] The training process takes a plurality of sets of historical joint positioning records as input and iteratively trains the initial parameter matching network. The specific steps are as follows:

[0092] After inputting the jth set of historical joint positioning records, it is first determined whether the historical Beidou multi-frequency signal corresponding to the (j+1)th historical signal quality parameter is the last signal in the historical Beidou multi-frequency signal stream. The last signal is determined based on the time sequence of signal collection, i.e., there is no subsequent continuously collected signal after this signal. If it is the last signal, the training parameter generated by the initial parameter matching network is determined based on the jth weight coefficient. At this time, the determination of the training parameter only depends on the weight coefficient of the current set, which has reflected the matching state (such as successful or failed matching) of the jth signal quality parameter and the positioning anomaly label. For example, if the jth weight coefficient is the first weight coefficient (successful matching), the training parameter can be set to a value positively related to the weight coefficient; if it is the second weight coefficient (failed matching), the training parameter can be set to a value negatively related to the weight coefficient, to guide the network to update the parameter in the direction of reducing the matching error.

[0093] If the historical Beidou multi-frequency signal corresponding to the j+1th historical signal quality parameter is not the last signal, the training parameter is determined based on the jth weight coefficient and the output of the associated parameter matching network. The associated parameter matching network is another parameter matching network with the same structure as the initial parameter matching network but different network parameters, which are set through an independent initialization process and are designed to provide different feature matching perspectives. The specific operation is as follows: each analytical parameter identifier in the pre-stored analytical parameter identifier set is input into the second matching algorithm branch of the associated parameter matching network in sequence together with the j+1th historical signal quality parameter. The structure of the second matching algorithm branch is similar to that of the first matching algorithm branch of the initial parameter matching network, and a second set of matching degree indicators is calculated through a multi-layer neural network. Each set of indicators corresponds to an analytical parameter identifier and reflects the matching degree of the identifier with the j+1th signal quality parameter. The second matching degree indicator with the highest value is selected from the second set of matching degree indicators, and the result of the weighted calculation of the second matching degree indicator and the jth weight coefficient is taken as the training parameter. The weighted calculation method can be linear combination, for example, training parameter = a x second matching degree indicator + b x jth weight coefficient, where a and b are pre-set weighting coefficients used to balance the influence of the matching state of adjacent signals and the current signal weight on the training parameter.

[0094] Based on the training parameter and the output of the initial parameter matching network, the error amount of the training error function is determined. The training error function usually adopts a loss function (such as mean square error loss function, cross-entropy loss function, etc.) to measure the difference between the network's predicted positioning anomaly label and the actual label (i.e. the positioning anomaly label). The calculation method of the error amount is determined according to the selected loss function, for example, the mean square error loss function quantifies the error by calculating the average of the square of the difference between the predicted value and the actual value. According to the error amount, the network parameters (such as weight matrix, bias vector, etc.) of the initial parameter matching network are adjusted using the backpropagation algorithm to minimize the error and optimize the network's feature matching capability. The backpropagation algorithm transmits the error layer by layer through the chain rule, calculates the gradient of each parameter, and updates the parameter according to the gradient direction, so that the network can more accurately predict the positioning anomaly label in subsequent training.

[0095] The above steps of inputting records, judging signal positions, determining training parameters, calculating errors and updating network parameters are repeated until the number of iterations reaches the set training threshold. The set training threshold is the maximum number of iterations determined in advance to control the termination condition of the training process. When the number of iterations reaches the threshold, if the network parameters have converged (i.e. the error amount does not decrease significantly in continuous iterations), the initial parameter matching network is determined as the positioning parameter matching network, and the training is completed; if the threshold is not reached, the iterative training is continued until the condition is met.

[0096] During the training process, attention should be paid to the temporal continuity between adjacent historical joint positioning records to ensure that the network can learn the dynamic change rules of signals in time series. For example, the j+1th signal quality parameter in the previous record is used as the jth signal quality parameter in the next record to form a continuous signal pair sequence, enabling the network to capture the evolution trend of signal characteristics. In addition, the introduction of the association parameter matching network increases the diversity of the training process, and by processing adjacent signals from different network parameter perspectives, it can avoid the network from falling into a local optimal solution and improve the generalization ability of the model.

[0097] During the training process, the weighting coefficients a and β should also be reasonably set to balance the influence of adjacent signal matching degree and current signal weight. If more attention is paid to the temporal correlation characteristics of adjacent signals, the value of a can be appropriately increased; if more attention is paid to the matching state reliability of the current signal, the value of β can be increased. At the same time, the learning rate and other hyperparameters should be dynamically adjusted according to the training progress of the network to optimize the training efficiency and convergence speed.

[0098] The entire training process revolves around the use of temporal correlation characteristics and matching state information in historical joint positioning records to gradually learn the mapping relationship between Beidou multi-frequency signal characteristics and positioning parameters through iterative optimization of the initial parameter matching network, ultimately forming a positioning parameter matching network that can accurately analyze the original Beidou multi-frequency combined signal stream. This process does not rely on hypothetical experimental data, but through strict signal processing logic and network training mechanism, it ensures the effectiveness and reliability of the model.

[0099] Embodiment 4:

[0100] During the training and optimization phase of the positioning parameter matching network, joint training operations should be performed on multiple groups of historical joint positioning records, and the network performance should be ensured through an accuracy verification mechanism. The specific implementation is as follows:

[0101] The joint training operation is performed on each set of historical joint positioning records, aiming to improve the rationality of network parameter updating through correlation analysis of adjacent signal features. Each set of records contains two consecutive historical signal quality parameters (the jth and the j+1th), corresponding positioning anomaly labels, and weight coefficients. The core of the joint training operation is to analyze the correlation features between adjacent historical signal quality parameters, such as the frequency band distribution deviation amount of adjacent signals, the time sequence feature difference value, etc. These correlation features are obtained by calculating the difference or ratio of corresponding feature values in two signal quality parameters, such as the difference in target frequency band energy proportion, the frequency offset change rate, etc. Based on the above correlation features, the update amplitude of the initial parameter matching network when updating the network parameters is adjusted. Specifically, if the correlation feature difference of adjacent signals is large (such as the frequency band distribution deviation exceeds the preset threshold), it indicates that the signal state has changed significantly, and the update amplitude of the network parameters is increased to speed up the response of the network to new features; if the difference is small (such as the deviation is within the preset threshold), the update amplitude is reduced to avoid excessive adjustment of network parameters due to minor fluctuations, affecting the stability of the model. The adjustment of the parameter update amplitude can be realized by introducing a dynamic learning rate in the backpropagation algorithm, for example, multiplying the difference value of the correlation features by a preset scaling factor to amplify or reduce the learning rate, thereby controlling the step size of each parameter update.

[0102] After the joint training operation of each set of records is completed, the performance of the positioning parameter matching network needs to be verified, focusing on the accuracy of the matching of the target frequency band signal data. The verification process uses a verification data set independent of the training data set, which contains historical Beidou multi-frequency signal data with known target analysis parameter identifiers. The specific steps are as follows: input the original Beidou multi-frequency combined signal stream in the verification data set into the positioning parameter matching network of the current training stage, and obtain the analysis parameter identifier matching result set output by the network; compare each matching result in the result set with the corresponding target analysis parameter identifier, and calculate the matching accuracy (i.e. the proportion of the number of matching successful samples to the total number of samples). The calculation method of the matching accuracy is as follows: count the number of samples whose analysis parameter identifier output by the network is consistent with the target analysis parameter identifier in all verification samples, and divide the total number of verification samples to obtain the accuracy value in percentage form.

[0103] If the matching accuracy meets the set standard (such as a preset accuracy threshold, for example, 90%), it is considered that the current network parameters have met the actual application requirements, and the training process is terminated; if the standard is not met, the retraining mechanism is triggered. During retraining, the following measures can be taken: adjusting the training parameters, such as increasing the number of iterations, modifying the learning rate, or optimizing the weighting coefficients; optimizing the network structure, such as increasing the number of hidden layers, adjusting the number of neurons; reselecting or expanding the historical joint positioning records, and supplementing sample data containing complex signal characteristics. Through the above measures, the training process is re-executed until the matching accuracy of the network meets the requirements.

[0104] In the joint training operation, the analysis of the correlation characteristics of adjacent signals needs to be combined with the physical meaning of signal processing. For example, a large frequency band distribution deviation may indicate that the signal is subject to sudden interference, at which time the network needs to quickly adjust the parameters to adapt to the interference changes; while a small deviation may belong to normal signal fluctuations, and there is no need to make large adjustments to the parameters. This dynamic adjustment mechanism based on signal characteristics can enable the network to more intelligently learn signal patterns and improve its adaptability to complex electromagnetic environments.

[0105] During the verification process, the reasonableness of the set standard is crucial. The standard needs to be determined according to the actual application scenarios of power terminal positioning, for example, in scenarios where positioning accuracy is required, the set standard can be correspondingly increased; in complex interference environments, the standard can be appropriately relaxed to balance accuracy and robustness. In addition, the verification data set needs to be representative, covering signal samples of different interference levels and different frequency band distributions, to comprehensively evaluate the generalization ability of the network.

[0106] The entire optimization process is through the cycle mechanism of "joint training-accuracy verification-adjustment optimization", ensuring that the positioning parameter matching network learns the timing correlation characteristics of adjacent signals while continuously improving the matching accuracy of the parameter identification. This process does not rely on hypothetical experimental effect descriptions, but through specific feature analysis, parameter adjustment, and accuracy calculation, gradually optimizes the network performance, providing reliable model support for power terminal positioning systems.

[0107] Embodiment 5:

[0108] In the training and optimization phase of the positioning parameter matching network, joint training operations need to be performed on multiple groups of historical joint positioning records, and the verification mechanism needs to ensure the matching accuracy of the parameter identification of the target frequency band signal data by the network. The specific implementation is as follows:

[0109] When performing the joint training operation on each set of historical joint positioning records, the correlation characteristics between adjacent historical signal quality parameters need to be analyzed. These correlation characteristics include, but are not limited to, frequency band distribution deviation of adjacent signals, timing feature change, etc. The frequency band distribution deviation can be obtained by calculating the energy distribution difference of adjacent signals in the target frequency band, the frequency offset change, etc. For example, for the i th and i + 1 th historical signal quality parameters, the energy spectrum density functions of them in the target frequency band are calculated respectively, and then the frequency band distribution deviation is determined by comparing the difference between the two functions. The timing feature change can be obtained by analyzing the feature evolution of adjacent signals in the time dimension, such as the change trend of signal intensity, the continuity of phase, etc.

[0110] Based on the above correlation characteristics, the update amplitude of the network parameters is dynamically adjusted. When the correlation characteristics of adjacent signals are significantly different, it means that the signal state has changed significantly, and at this time the update amplitude of the network parameters is increased to speed up the response speed of the network to new features. Specifically, in the back propagation algorithm, for the case where the correlation characteristic difference exceeds the preset threshold, the learning rate is multiplied by an adjustment factor greater than 1, thereby increasing the step size of parameter update. On the contrary, when the correlation characteristic difference is small, it means that the signal state is relatively stable, and at this time the update amplitude is reduced to avoid excessive adjustment of the network parameters. For the case where the correlation characteristic difference is less than the preset threshold, the learning rate is multiplied by an adjustment factor less than 1, thereby reducing the step size of parameter update.

[0111] After the joint training operation of each set of records is completed, the matching accuracy of the positioning parameter matching network to the parsed parameter identifier of the target frequency band signal data needs to be verified. The verification process uses an independent verification data set, which contains multiple historical Beidou multi-frequency signals with known parsed parameter identifiers. These signals are input into the trained positioning parameter matching network to obtain the parsed parameter identifier output by the network, and compared with the actual label.

[0112] To more accurately evaluate the network performance, the verification process uses a multi-level matching accuracy calculation method. First, the basic matching accuracy is calculated, that is, the proportion of the number of samples whose parsed parameter identifier output by the network is completely consistent with the actual label to the total number of samples. Second, considering that some parsed parameter identifiers may have multiple subcategories, the subcategory matching accuracy is introduced, that is, for each main category of parsed parameter identifier, the proportion of correctly matched subcategories is calculated. Finally, considering the importance difference of the parsed parameter identifiers, different weights are assigned to different types of parsed parameter identifiers, and the weighted matching accuracy is calculated.

[0113] If the weighted matching accuracy reaches the set standard, it is considered that the training of the positioning parameter matching network has been completed and can be used for actual power terminal positioning applications. If the standard is not reached, the training process is re-executed. When retraining, training parameters can be adjusted, such as optimizing the learning rate scheduling strategy, modifying the batch size, or adjusting the regularization parameters. The network structure can also be optimized, such as introducing an attention mechanism to enhance attention to key features or using residual connections to improve the network's deep expression ability.

[0114] In the joint training operation, the analysis of the correlation features between adjacent historical signal quality parameters needs to be combined with the physical characteristics of signal processing. For example, a large frequency band distribution deviation may correspond to a signal affected by multipath effects or ionospheric disturbances, in which case the network needs to quickly adapt to such changes; while a small change in the timing feature may indicate that the signal is in a stable propagation environment, in which case the relative stability of the network parameters should be maintained. Through this adjustment strategy based on physical characteristics, the network can better adapt to the complex electromagnetic environment in actual applications.

[0115] The construction of the verification data set needs to follow the principles of representativeness and diversity. It should include signal samples collected in different geographic locations, different time periods, and different weather conditions to cover various scenarios that may be encountered in actual applications. At the same time, to avoid one-sidedness of the verification results, the verification data set should be independent of the training data set to ensure that the generalization ability of the model on unknown data is truly evaluated.

[0116] In the retraining process, parameter adjustment and network structure optimization need to follow the principle of step-by-step iteration. Each adjustment should focus on solving specific performance bottlenecks, such as increasing the weight of related samples or optimizing the extraction method of corresponding features to address the low matching accuracy of certain types of analytical parameter identification. At the same time, to avoid falling into a local optimal solution, random search or Bayesian optimization methods can be used to explore a wider parameter space.

[0117] The entire process of the embodiment continuously optimizes the performance of the positioning parameter matching network through the joint training and verification mechanism. This mechanism does not rely on hypothetical experimental effect data, but through rigorous algorithm design and training process, it ensures the effectiveness and stability of the network in actual applications. By analyzing the correlation features of adjacent signals to dynamically adjust the learning strategy and using multi-level verification evaluation methods, the positioning parameter matching network can accurately analyze the target frequency band signal data in the original Beidou multi-frequency combined signal stream, providing reliable positioning services for power terminals.

[0118] In practical applications, this mechanism can be customized according to the specific needs of the power system. For example, for scenarios with high positioning accuracy requirements, the verification standard can be appropriately improved; for areas with complex interference environments, challenging samples can be added to enhance the robustness of the network. Through this flexible implementation, the positioning parameter matching network can adapt to different power terminal positioning application scenarios, providing strong support for the safe and stable operation of the power system.

[0119] To more clearly describe the process of dynamically adjusting the learning rate, the formula is introduced:

[0120]

[0121] wherein is the original learning rate, is the adjusted learning rate, is the adjustment coefficient (which can be set according to actual conditions), is the correlation feature difference value of the adjacent signal. This formula shows that when the correlation feature difference value is large, the learning rate will be increased accordingly, thereby speeding up the update speed of network parameters; on the contrary, when the difference value is small, the learning rate will be reduced to avoid excessive adjustment. This formula is only used to describe the logical relationship of the adjustment strategy and does not involve specific experimental data or effect quantification.

[0122] It should be noted that, in this document, relational terms such as first and second and the like can be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises", "comprising", or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can include other elements not expressly listed or inherent to such process, method, article, or apparatus.

[0123] Although embodiments of the present application have been shown and described, it is to be understood that various modifications, substitutions, replacements and changes can be made to these embodiments without departing from the principles and spirit of the present application, and the scope of the present application is defined by the appended claims and their equivalents.

Claims

1. A power terminal positioning method based on a multi-frequency combination of Beidou, characterized in that, The method is applied to a power terminal positioning system, and the method comprises: obtaining an original Beidou multi-frequency combined signal stream to be positioned; inputting the original Beidou multi-frequency combined signal stream into a pre-trained positioning parameter matching network to obtain a positioning parameter matching result set; determining a positioning parameter matching result corresponding to each Beidou multi-frequency signal in the original Beidou multi-frequency combined signal stream based on the positioning parameter matching result set; the original Beidou multi-frequency combined signal stream comprises target frequency band signal data to be parsed; the positioning parameter matching network is a parameter matching network obtained by inputting historical positioning data into an initial parameter matching network to be trained, the historical positioning data comprises a batch of historical signal quality parameters extracted from a batch of historical Beidou multi-frequency signals respectively, and a positioning anomaly label and a weight coefficient corresponding to each historical signal quality parameter, the positioning anomaly label reflects an interference state identifier of a historical signal quality parameter, the weight coefficient reflects a correction factor corresponding to the interference state identifier, the batch of historical Beidou multi-frequency signals are a batch of continuously collected Beidou multi-frequency signals, and each historical signal quality parameter comprises a signal quality parameter determined by a historical Beidou multi-frequency signal and an adjacent historical Beidou multi-frequency signal of the historical Beidou multi-frequency signal; the positioning parameter matching result is used to represent a parsing parameter identifier of the target frequency band signal data.

2. The BeiDou multi-frequency combination-based power terminal positioning method according to claim 1, characterized in that, The method further comprises: performing first feature parameter extraction on the batch of historical Beidou multi-frequency signals to obtain a first group of signal quality parameters, wherein one signal quality parameter in the first group of signal quality parameters corresponds to one historical Beidou multi-frequency signal in the batch of historical Beidou multi-frequency signals; performing second feature parameter extraction on the batch of historical Beidou multi-frequency signals to obtain a second group of signal quality parameters, wherein a first signal quality parameter in the second group of signal quality parameters corresponds to a first historical Beidou multi-frequency signal in the historical Beidou multi-frequency signals, an adjacent historical Beidou multi-frequency signal of the first historical Beidou multi-frequency signal is a second historical Beidou multi-frequency signal, and the first signal quality parameter represents a time sequence correlation feature between the first historical Beidou multi-frequency signal and the second historical Beidou multi-frequency signal; fusing the first group of signal quality parameters and the second group of signal quality parameters to obtain the batch of historical signal quality parameters.

3. The BeiDou multi-frequency combination-based power terminal positioning method according to claim 2, characterized in that, The second feature parameter extraction on the batch of historical Beidou multi-frequency signals to obtain the second group of signal quality parameters comprises: performing second feature parameter extraction on the batch of historical Beidou multi-frequency signals to obtain a second group of signal quality parameters by the following steps, wherein each historical Beidou multi-frequency signal subjected to the second feature parameter extraction each time is regarded as a current historical Beidou multi-frequency signal, a signal quality parameter obtained each time is regarded as a current signal quality parameter, and the second group of signal quality parameters comprises the current signal quality parameter; performing a frequency band analysis operation on the current historical Beidou multi-frequency signal to determine a first set of frequency band feature quantities, wherein the first set of frequency band feature quantities are used to describe a frequency band distribution of the target frequency band signal data in the current historical Beidou multi-frequency signal; performing a frequency band analysis operation on a neighboring historical Beidou multi-frequency signal of the current historical Beidou multi-frequency signal to determine a second set of frequency band feature quantities, wherein the second set of frequency band feature quantities are used to describe a frequency band distribution of the target frequency band signal data in the neighboring historical Beidou multi-frequency signal; determining the current signal quality parameter based on the first set of frequency band feature quantities and the second set of frequency band feature quantities, wherein the current signal quality parameter is used to represent a deviation between corresponding frequency band feature quantities in the first set of frequency band feature quantities and the second set of frequency band feature quantities.

4. The power terminal positioning method based on multi-frequency combination of Beidou according to claim 1, characterized in that, The method further comprises: determining a positioning anomaly label and a weight coefficient corresponding to each historical signal quality parameter in the batch of historical signal quality parameters respectively, wherein each time the historical signal quality parameter for determining the positioning anomaly label and the weight coefficient is taken as a current historical signal quality parameter, and the positioning anomaly label and the weight coefficient corresponding to the current historical signal quality parameter are taken as a current positioning anomaly label and a current weight coefficient: obtaining the current positioning anomaly label based on the current historical signal quality parameter, wherein the initial parameter matching network is a parameter matching network obtained by pre-executing an initialization process; determining the current weight coefficient based on the current positioning anomaly label and a benchmark verification identifier corresponding to the current historical signal quality parameter, wherein the benchmark verification identifier carries a calibration label in a historical Beidou multi-frequency signal corresponding to the current historical signal quality parameter, the calibration label includes a target analysis parameter identifier of the target frequency band signal data, and the weight coefficient is used to represent whether the current positioning anomaly label is consistent with the target analysis parameter identifier.

5. The BeiDou multi-frequency combination-based power terminal positioning method according to claim 4, characterized in that, The method further comprises: inputting each analysis parameter identifier in a pre-stored analysis parameter identifier set into a first matching algorithm branch of the initial parameter matching network in sequence together with the current historical signal quality parameter to obtain a first set of matching degree indicators, wherein the first set of matching degree indicators includes a matching degree indicator corresponding to each analysis parameter identifier; determining a candidate analysis parameter identifier corresponding to a first matching degree indicator with the highest value in the first set of matching degree indicators as a current analysis parameter identifier corresponding to the current positioning anomaly label.

6. The BeiDou multi-frequency combination-based power terminal positioning method according to claim 5, characterized in that, Before the step of inputting each analysis parameter identifier in a pre-stored analysis parameter identifier set into a first matching algorithm branch in sequence together with the current historical signal quality parameter to obtain a first set of matching degree indicators, the method further comprises: obtaining a signal strength distribution quantity by statistically analyzing signal strength data of a target frequency band in the current historical signal quality parameter; determining whether the signal strength distribution quantity conforms to a set strength distribution rule; on the basis that the signal strength distribution quantity conforms to the set strength distribution rule, dynamically selecting a candidate analysis parameter identifier in the pre-stored analysis parameter identifier set as the current positioning anomaly label; On the basis that the signal strength distribution quantity does not conform to the set strength distribution rule, input the current historical signal quality parameter into the initial parameter matching network to obtain the current positioning anomaly label.

7. The BeiDou multi-frequency combination-based power terminal positioning method according to claim 5, characterized in that, The method further comprises: On the basis that the candidate resolution parameter identifier corresponding to the first matching degree indicator with the highest value in the first group of matching degree indicators is the same as the target resolution parameter identifier, the current weight coefficient is determined as a first weight coefficient, wherein the first weight coefficient indicates that the initial parameter matching network successfully generates a matching identifier; On the basis that the candidate resolution parameter identifier corresponding to the first matching degree indicator with the highest value in the first group of matching degree indicators is different from the target resolution parameter identifier, the current weight coefficient is determined as a second weight coefficient, wherein the second weight coefficient indicates that the initial parameter matching network does not successfully generate a matching identifier.

8. The BeiDou multi-frequency combination-based power terminal positioning method according to claim 1, characterized in that, The method further comprises: screening a plurality of groups of historical joint positioning records from the historical positioning data, wherein the jth group of historical joint positioning records in the plurality of groups of historical joint positioning records comprises a jth historical signal quality parameter, a jth positioning anomaly label corresponding to the jth historical signal quality parameter, a jth weight coefficient, and a (j+1)th historical signal quality parameter, j being a positive integer; training an initial parameter matching network to be trained based on the plurality of groups of historical joint positioning records to obtain the positioning parameter matching network, wherein, on the basis that the number of iterations of training the initial parameter matching network reaches a set training threshold, the initial parameter matching network is determined as the positioning parameter matching network, and on the basis that the number of iterations of training the initial parameter matching network does not reach the set training threshold, the network parameters of the initial parameter matching network are updated according to a preset training error function, and the input of each training process is a group of historical joint positioning records in the plurality of groups of historical joint positioning records.

9. The BeiDou multi-frequency combination-based power terminal positioning method according to claim 8, characterized in that, The method further comprises: training an initial parameter matching network to be trained based on a plurality of groups of historical joint positioning records to obtain the positioning parameter matching network, wherein the historical joint positioning record input into the initial parameter matching network is the jth group of joint positioning records: determining whether the historical Beidou multi-frequency signal corresponding to the (j+1)th historical signal quality parameter is a last historical Beidou multi-frequency signal in a historical Beidou multi-frequency signal stream; on the basis that the historical Beidou multi-frequency signal corresponding to the (j+1)th historical signal quality parameter is the last historical Beidou multi-frequency signal, determining a training parameter generated by the initial parameter matching network based on the jth weight coefficient; on the basis that the j+1th historical signal quality parameter corresponds to a historical Beidou multi-frequency signal that is not the last historical Beidou multi-frequency signal, determining the training parameter based on the jth weight coefficient and an output of an associated parameter matching network, wherein the associated parameter matching network is a parameter matching network obtained by performing initialization processing in advance, and the associated parameter matching network and the initial parameter matching network have different network parameters; determining an error amount of the training error function based on the training parameter and the output of the initial parameter matching network, and adjusting the network parameters of the initial parameter matching network based on the error amount of the training error function; on the basis that the number of iterations of the above steps reaches the set training threshold, determining the initial parameter matching network as the positioning parameter matching network; the determining of the training parameter based on the jth weight coefficient and the output of the associated parameter matching network comprises: inputting each analysis parameter identifier in a pre-stored analysis parameter identifier set into a second matching algorithm branch of the associated parameter matching network in sequence together with the j+1th historical signal quality parameter, to obtain a second group of matching degree indexes, wherein the second group of matching degree indexes comprises a matching degree index corresponding to each analysis parameter identifier; determining a second matching degree index with the highest value in the second group of matching degree indexes and a weighted result of the jth weight coefficient as the training parameter.

10. The BeiDou multi-frequency combination-based power terminal positioning method according to claim 9, characterized in that, The method further comprises: performing joint training operations on each group of records in the plurality of groups of historical joint positioning records to generate optimized parameters of the positioning parameter matching network, wherein the joint training operation of each group of records comprises adjusting an update amplitude of the network parameters based on associated features between adjacent historical signal quality parameters; after the joint training operation of each group of records is completed, verifying the matching accuracy of the positioning parameter matching network for analysis parameter identifier of target frequency band signal data, and re-executing the training process when the accuracy does not reach a set standard.

Images