Time synchronization method and device of GNSS (Global Navigation Satellite System) equipment

Abstract

The invention discloses a time synchronization method and device for GNSS (Global Navigation Satellite System) equipment, and relates to the field of time synchronization. According to the method, the pulse signal and the time synchronization standard signal are acquired, time synchronization is carried out through two different time synchronization types of signals, and the time parameters and the environment parameters of different time sources are fused, so that the time synchronization prediction parameters can more accurately represent the prediction condition of time synchronization of the GNSS equipment, and the accuracy of time synchronization prediction is improved. And then fusion calculation is carried out on the prediction condition of the time synchronization condition of the GNSS equipment and the environmental parameters through the cloud to obtain cloud time synchronization parameters, and when a single time source has an error, the time synchronization of the GNSS equipment can be corrected through the time synchronization prediction parameters and the cloud time synchronization parameters, so that the time synchronization accuracy of the GNSS equipment is improved. Therefore, the time synchronization result of the GNSS equipment considers the time parameters and the environment parameters of multiple time sources, the time synchronization deviation caused by a single time source is avoided, and the time synchronization accuracy of the GNSS equipment is improved.

Description

Technical Field

[0001] The present invention belongs to the field of time synchronization, and in particular relates to a time synchronization method and device for GNSS equipment. Background Art

[0002] With the continuous development of GNSS (Global Navigation Satellite System) technology, GNSS devices are also becoming more and more popular. GNSS devices, namely global navigation satellite system devices, are used to receive and process signals from GNSS satellites to determine information such as position, speed and time. GNSS devices can receive time information from GNSS satellite signals and calibrate local clocks through this time information to achieve time synchronization.

[0003] The time synchronization function of GNSS equipment provides a high-precision time reference and is widely used in aviation, navigation, geographic information systems, power systems and other fields. However, most of the current time synchronization technologies are implemented by using the pulse per second signal (PPS signal) provided by the GPS receiver. However, this time synchronization technology is too dependent on a single time source and lacks robustness and flexibility. When an error occurs in the time source, it will cause the time synchronization of the GNSS equipment to find deviations, reducing the accuracy of the time synchronization of the GNSS equipment. Therefore, there is an urgent need for a time synchronization method and device for GNSS equipment to solve the defects of the prior art. Summary of the invention

[0004] The present invention aims to provide a time synchronization method and device for GNSS equipment to solve the above technical problems, and to correct the time synchronization of GNSS equipment through adaptive learning algorithm and cloud time synchronization processing, thereby improving the time synchronization accuracy of GNSS equipment.

[0005] In order to solve the above technical problems, an embodiment of the present invention provides a time synchronization method for a GNSS device, comprising:

[0006] Acquire a pulse signal and a time synchronization standard signal, and perform time synchronization on the GNSS device according to the pulse signal and the time synchronization standard signal to obtain initial time synchronization parameters;

[0007] Acquire environmental parameters of the GNSS device and time parameters of several time sources, process the initial time synchronization parameters according to a preset adaptive learning algorithm, the environmental parameters and the time parameters, and obtain time synchronization prediction parameters;

[0008] Transmitting the initial time synchronization parameters, environmental parameters and time synchronization prediction parameters to the cloud, so that the cloud calculates the cloud time synchronization parameters according to the initial time synchronization parameters, environmental parameters and time synchronization prediction parameter set;

[0009] When the initial time synchronization parameters do not meet the preset requirements, the time synchronization of the GNSS device is corrected according to the time synchronization prediction parameters and the cloud time synchronization parameters to complete the time synchronization of the GNSS device.

[0010] It can be understood that, compared with the prior art, the present invention can improve the accuracy of GNSS device time synchronization by acquiring pulse signals and time synchronization standard signals, and by using two different types of time synchronization signals, and based on the initial time synchronization parameters obtained from the time synchronization, integrate the time parameters and environmental parameters of different time sources, so that the time synchronization prediction parameters can more accurately characterize the predicted situation of the time synchronization of the GNSS device, and then the predicted situation of the time synchronization of the GNSS device and the environmental parameters are integrated and calculated through the cloud to obtain the cloud time synchronization parameters. When an error occurs in a single time source, the time synchronization of the GNSS device can be corrected through the time synchronization prediction parameters and the cloud time synchronization parameters, so that the time synchronization result of the GNSS device takes into account the time parameters and environmental parameters of multiple time sources, avoids the time synchronization deviation caused by a single time source, and improves the time synchronization accuracy of the GNSS device.

[0011] As a preferred solution, the step of acquiring a pulse signal and a time synchronization standard signal, and performing time synchronization on a GNSS device according to the pulse signal and the time synchronization standard signal to obtain an initial time synchronization parameter specifically includes:

[0012] Obtain pulse signals and time synchronization standard signals;

[0013] Calculating the time delay between the pulse signal and the time synchronization standard signal;

[0014] Based on the pulse signal, the time synchronization standard signal and the time delay, the GNSS device is time synchronized, and the time offset and time delay in the GNSS device time synchronization process are calculated;

[0015] An initial time synchronization parameter is obtained according to the time offset and time delay in the time synchronization process of the GNSS device.

[0016] This preferred solution improves the accuracy of initial time synchronization of GNSS devices by calculating the time delay between the pulse signal and the time synchronization standard signal, using signals of different time synchronization types, obtains more accurate initial time synchronization parameters, and improves the accuracy of subsequent time synchronization prediction parameters and cloud time synchronization parameters in correcting the time synchronization of the GNSS device, thereby improving the time synchronization accuracy of GNSS devices.

[0017] As a preferred solution, the environmental parameters of the GNSS device and the time parameters of several time sources are obtained, and the initial time synchronization parameters are processed according to a preset adaptive learning algorithm, the environmental parameters and the time parameters to obtain the time synchronization prediction parameters, specifically including:

[0018] Perform multipath effect processing on the pulse signal and the time synchronization standard signal according to a preset signal processing method to obtain a first processing result;

[0019] Obtain environmental parameters of GNSS equipment and time parameters of several time sources;

[0020] According to a preset adaptive learning algorithm, the initial time synchronization parameters are processed in combination with the environmental parameters, the time parameters and the first processing result to obtain time synchronization prediction parameters.

[0021] This preferred solution can effectively eliminate the errors caused by the multipath effect by processing the multipath effect of the pulse signal and the time synchronization standard signal, thereby improving the accuracy of the initial time synchronization parameter processing, and through the environmental parameters of the GNSS device and the time parameters of several time sources, the time synchronization prediction parameters contain different time sources and environmental parameters, thereby improving the prediction accuracy of the time synchronization prediction parameters, avoiding the time synchronization deviation caused by a single time source, and improving the time synchronization accuracy of the GNSS device.

[0022] As a preferred solution, the initial time synchronization parameter is processed according to a preset adaptive learning algorithm in combination with the environmental parameter, the time parameter and the first processing result to obtain the time synchronization prediction parameter, specifically including:

[0023] After removing abnormal values ​​and noise points in the environmental parameters, time parameters, the first processing result and the initial time synchronization parameters according to a preset preprocessing method, extracting time synchronization features in the environmental parameters, time parameters, the first processing result and the initial time synchronization parameters to obtain a time synchronization feature set;

[0024] Inputting the time synchronization feature set into a preset adaptive learning model group to obtain a predicted time jitter value, a predicted time delay value and a predicted time offset value output by the adaptive learning model group;

[0025] A time synchronization prediction parameter is determined according to the predicted time jitter value, the predicted time delay value and the predicted time offset value.

[0026] This preferred scheme improves the prediction accuracy of the adaptive learning model group by extracting time synchronization features from environmental parameters, time parameters, the first processing results and initial time synchronization parameters, and predicts through an adaptive learning model group. This makes it possible to implicitly include different time sources and environmental parameters in the time synchronization prediction parameters, thereby improving the prediction accuracy of the time synchronization prediction parameters, avoiding time synchronization deviations caused by a single time source, and improving the time synchronization accuracy of GNSS equipment.

[0027] As a preferred solution, the step of enabling the cloud to calculate the cloud time synchronization parameters according to the initial time synchronization parameters, the environmental parameters and the time synchronization prediction parameter set specifically includes:

[0028] Building a global environment model according to the environmental parameters, and inputting the environmental parameters into the global environment model, predicting the change trend of the multipath effect of the GNSS device, and obtaining a multipath effect prediction value;

[0029] Extracting time features from the initial time synchronization parameters, environmental parameters, and time synchronization prediction parameter sets, and obtaining cloud time synchronization prediction parameters based on the time features;

[0030] Generate a cloud time synchronization parameter according to the multipath effect prediction value and the cloud time synchronization prediction parameter.

[0031] This preferred solution constructs a global environmental model to predict the changing trend of the multipath effect in the GNSS device, and then merges the time synchronization prediction parameters through the time characteristics of the environmental parameters and the time synchronization prediction parameter set, so that the generated cloud time synchronization parameters can better meet the actual time synchronization situation and future trends of the GNSS device, thereby improving the time synchronization accuracy of the GNSS device.

[0032] Accordingly, an embodiment of the present invention provides a time synchronization device for a GNSS device, comprising: a preliminary time synchronization parameter acquisition module, an adaptive learning module, a cloud time synchronization module, and a time synchronization correction module;

[0033] The preliminary time synchronization parameter acquisition module is used to acquire a pulse signal and a time synchronization standard signal, and perform time synchronization on the GNSS device according to the pulse signal and the time synchronization standard signal to obtain initial time synchronization parameters;

[0034] The adaptive learning module is used to obtain the environmental parameters of the GNSS device and the time parameters of several time sources, and process the initial time synchronization parameters according to the preset adaptive learning algorithm, the environmental parameters and the time parameters to obtain the time synchronization prediction parameters;

[0035] The cloud time synchronization module is used to transmit the initial time synchronization parameters, environmental parameters and time synchronization prediction parameters to the cloud, so that the cloud calculates the cloud time synchronization parameters according to the initial time synchronization parameters, environmental parameters and time synchronization prediction parameter set;

[0036] The time synchronization correction module is used to correct the time synchronization of the GNSS device according to the time synchronization prediction parameters and the cloud time synchronization parameters when the initial time synchronization parameters do not meet the preset requirements, so as to complete the time synchronization of the GNSS device.

[0037] As a preferred solution, the preliminary time synchronization parameter acquisition module includes: a preliminary time synchronization parameter acquisition unit;

[0038] The preliminary time synchronization parameter acquisition unit is used to acquire a pulse signal and a time synchronization standard signal;

[0039] Calculating the time delay between the pulse signal and the time synchronization standard signal;

[0040] Based on the pulse signal, the time synchronization standard signal and the time delay, the GNSS device is time synchronized, and the time offset and time delay in the GNSS device time synchronization process are calculated;

[0041] An initial time synchronization parameter is obtained according to the time offset and time delay in the time synchronization process of the GNSS device.

[0042] As a preferred solution, the adaptive learning module includes: an adaptive learning unit;

[0043] The adaptive learning unit is used to perform multipath effect processing on the pulse signal and the time synchronization standard signal according to a preset signal processing method to obtain a first processing result;

[0044] Obtain environmental parameters of GNSS equipment and time parameters of several time sources;

[0045] According to a preset adaptive learning algorithm, the initial time synchronization parameters are processed in combination with the environmental parameters, the time parameters and the first processing result to obtain time synchronization prediction parameters.

[0046] As a preferred solution, the adaptive learning unit includes: a time synchronization prediction parameter acquisition subunit;

[0047] The time synchronization prediction parameter acquisition subunit is used to extract the time synchronization features in the environment parameters, time parameters, first processing results and initial time synchronization parameters after removing abnormal values ​​and noise points in the environment parameters, time parameters, first processing results and initial time synchronization parameters according to a preset preprocessing method to obtain a time synchronization feature set;

[0048] Inputting the time synchronization feature set into a preset adaptive learning model group to obtain a predicted time jitter value, a predicted time delay value and a predicted time offset value output by the adaptive learning model group;

[0049] A time synchronization prediction parameter is determined according to the predicted time jitter value, the predicted time delay value and the predicted time offset value.

[0050] As a preferred solution, the time synchronization correction module includes: a time synchronization correction unit;

[0051] The time synchronization correction unit is used to construct a global environment model according to the environmental parameters, and input the environmental parameters into the global environment model to predict the change trend of the multipath effect of the GNSS device to obtain a multipath effect prediction value;

[0052] Extracting time features from the initial time synchronization parameters, environmental parameters, and time synchronization prediction parameter sets, and obtaining cloud time synchronization prediction parameters based on the time features;

[0053] Generate a cloud time synchronization parameter according to the multipath effect prediction value and the cloud time synchronization prediction parameter.

[0054] It can be understood that, compared with the prior art, the present device can improve the accuracy of GNSS device time synchronization by acquiring pulse signals and time synchronization standard signals, and by using two different types of time synchronization signals, and based on the initial time synchronization parameters obtained from the time synchronization, integrates the time parameters and environmental parameters of different time sources, so that the time synchronization prediction parameters can more accurately characterize the predicted situation of the time synchronization of the GNSS device, and then integrates the predicted situation of the time synchronization of the GNSS device and the environmental parameters through the cloud to obtain the cloud time synchronization parameters. When an error occurs in a single time source, the time synchronization of the GNSS device can be corrected through the time synchronization prediction parameters and the cloud time synchronization parameters, so that the time synchronization result of the GNSS device takes into account the time parameters and environmental parameters of multiple time sources, avoids the time synchronization deviation caused by a single time source, and improves the time synchronization accuracy of the GNSS device. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1: A flowchart of a method for synchronizing a GNSS device according to an embodiment of the present invention;

[0056] Figure 2 : A structural diagram of a GNSS device provided in an embodiment of the present invention;

[0057] Figure 3 : A schematic diagram of the structure of a time synchronization device for a GNSS device provided in an embodiment of the present invention;

[0058] Among them, 201: preliminary time synchronization parameter acquisition module; 202: adaptive learning module; 203: cloud time synchronization module; 204: time synchronization correction module. DETAILED DESCRIPTION

[0059] 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.

[0060] Embodiment 1

[0061] Please refer to Figure 1 , is a step flow chart of a time synchronization method for a GNSS device provided in an embodiment of the present invention, including steps S101 to S104.

[0062] Step S101: Acquire a pulse signal and a time synchronization standard signal, and perform time synchronization on a GNSS device according to the pulse signal and the time synchronization standard signal to obtain initial time synchronization parameters.

[0063] In this embodiment, the step of acquiring a pulse signal and a time synchronization standard signal, and performing time synchronization on a GNSS device according to the pulse signal and the time synchronization standard signal to obtain an initial time synchronization parameter specifically includes:

[0064] Obtain pulse signals and time synchronization standard signals;

[0065] Calculating the time delay between the pulse signal and the time synchronization standard signal;

[0066] Based on the pulse signal, the time synchronization standard signal and the time delay, the GNSS device is time synchronized, and the time offset and time delay in the GNSS device time synchronization process are calculated;

[0067] An initial time synchronization parameter is obtained according to the time offset and time delay in the time synchronization process of the GNSS device.

[0068] This embodiment improves the accuracy of initial time synchronization of GNSS devices by calculating the time delay between the pulse signal and the time synchronization standard signal, using signals of different time synchronization types, obtains more accurate initial time synchronization parameters, and improves the accuracy of subsequent time synchronization prediction parameters and cloud time synchronization parameters in correcting the time synchronization of the GNSS device, thereby improving the time synchronization accuracy of GNSS devices.

[0069] In an alternative embodiment, please refer to Figure 2 , is a structural diagram of a GNSS device provided by an embodiment of the present invention, such as Figure 2 As shown, in a GNSS device, a GNSS receiving module, a network module, a SHM shared memory, a GPSD module and a Chronyd module are usually provided, wherein the GNSS receiving module is used to receive NMEA data (i.e., a time synchronization standard signal) and output a PPS signal pulse (i.e., a pulse signal), and receives and stores NMEA data from the GNSS receiving module in the form of a serial port through the SHM shared memory; the network module is used to receive NMEA data and transmit it to the SHM shared memory through a TCP / UDP interface; the GPSD module is used to obtain and parse NMEA data and PPS signal pulses, and the Chronyd module is used to realize time synchronization of the GNSS device;

[0070] Specifically, the time synchronization standard signals (i.e., NMEA data) of different satellite navigation systems (such as GPS, GLONASS, Galileo, BeiDou) are received through the GNSS receiving module, and the network module receives the time synchronization standard signals (i.e., NMEA data) in the network resources, and then inputs the time synchronization standard signals into the SHM shared memory for storage. At the same time, the GNSS receiving module also outputs PPS signal pulses (i.e., pulse signals);

[0071] Input the PPS signal pulse (i.e., pulse signal) and the time synchronization standard signal (i.e., NMEA data) in the SHM shared memory into the GPSD module, control the interruption of the GPSD module receiving the PPS signal pulse (i.e., pulse signal), and solve the time delay between the pulse signal and the time synchronization standard signal, and transmit the time delay to the Chronyd module, and then configure the Chronyd module, and then synchronize the GNSS device according to the Chronyd module in combination with the pulse signal, the time synchronization standard signal and the time delay, and calculate the time offset and time delay in the time synchronization process of the GNSS device, and obtain the initial time synchronization parameters according to the time offset and time delay in the time synchronization process of the GNSS device;

[0072] Among them, edit the gpsd file in the GPSD module, add two command lines "DEVI CES=" / dev / ttyUSB0, / dev / pps0" and "GPSD_OPTIONS="-nF / var / run / gpsd.sock" in the gpsd file, through these two command lines, the GPSD module can solve the time delay between the pulse signal and the time synchronization standard signal;

[0073] Edit the Chrony.conf file in the Chronyd module and add the command line "refclock SHM 0offset 0.5delay 0.2refid GPS precision 1e-6" to enable the Chronyd module to use NMEA data (i.e., the time synchronization standard signal). "SHM 0" is used to specify the SHM shared memory segment number to be used, "offset 0.5" is used to indicate the estimated time offset (in seconds), "delay 0.2" is used to indicate the estimated delay (in seconds), "refid GPS" is used to specify the identifier for the time source, and "precision 1e-6" is used to indicate the precision of the time source. After the above command line configuration, the Chronyd module can synchronize time based on the time synchronization standard signal. Then add "refclock PPS / dev / pps0 lock NMEA refidPPS precision 1e-6" to the Chrony.conf file. 1e-9" is used to enable the Chronyd module to use the PPS signal (i.e., pulse signal), where " / dev / pps0" is used to specify the PPS device path, "lock NMEA" is used to instruct the Chronyd module to lock the PPS signal with the NMEA data to ensure that the relative time relationship between the two is correct, "refid PPS" is used to specify an identifier for the time source, and "precision 1e-9" is used to indicate the accuracy of the time source (nanosecond level); in the Chronyd module, through the configuration of the above two command lines "refclock SHM 0offset 0.5delay 0.2refid GPS precision 1e-6" and "refclock PPS / dev / pps0 lock NMEA refid PPS precision 1e-9", the Chronyd module can use the time delay between the pulse signal and the time synchronization standard signal.

[0074] In an optional embodiment, before configuring the Chronyd module for the PPS signal, it is also necessary to ensure that the Linux kernel in the GNSS device has been compiled with support for the PPS signal, and then install the pps-too ls package to obtain the ppstest tool to test whether the PPS signal is working properly.

[0075] In an optional embodiment, the time source and status currently used by the Chronyd module can also be queried through the command line "chronyc sources -v", and the initial time synchronization parameters, including but not limited to time offset, time delay, time jitter, etc., can be viewed through the command line "chronyc tracking". Errors or warning information in the time synchronization process can also be queried by viewing the log file in the Chronyd module.

[0076] The above embodiment can make full use of NMEA data, PPS signal and time delay analyzed by GPSD module to perform high-precision time synchronization of GNSS equipment by performing relevant configuration in Chronyd module, which not only improves the accuracy of time synchronization, but also enhances the operational stability and reliability of GNSS equipment.

[0077] It should be noted that the PPS signal (Pulse Per Second) is an important time synchronization signal, which is characterized by emitting a pulse once per second and can provide accurate time marking; NMEA data refers to data transmitted in accordance with the NMEA (National Marine Electronics Association) protocol standard, which contains a variety of different types of time message data and can be used for time synchronization; the Chronyd module is a commonly used time synchronization component in GNSS devices, which is used to adjust the system clock (local clock) inside the GNSS device to achieve time synchronization of the GNSS device; the GPSD (GPSDaemon) module is a commonly used time data processing process in GNSS devices, which is used to parse different types of time data.

[0078] It should be noted that although high-precision time synchronization of GNSS devices can be achieved through pulse signals and time synchronization standard signals, when an error occurs in the time source that sends the pulse signal and the time synchronization standard signal, it will cause the pulse signal and the time synchronization standard signal to deviate, thereby causing inaccurate time synchronization of the GNSS device. Therefore, this embodiment corrects the time synchronization of the GNSS device again through steps S102 to S104 below, thereby improving the time synchronization accuracy of the GNSS device.

[0079] Step S102: Acquire environmental parameters of the GNSS device and time parameters of several time sources, process the initial time synchronization parameters according to a preset adaptive learning algorithm, the environmental parameters and the time parameters, and obtain time synchronization prediction parameters.

[0080] In this embodiment, the environmental parameters of the GNSS device and the time parameters of several time sources are obtained, and the initial time synchronization parameters are processed according to a preset adaptive learning algorithm, the environmental parameters and the time parameters to obtain the time synchronization prediction parameters, specifically including:

[0081] Perform multipath effect processing on the pulse signal and the time synchronization standard signal according to a preset signal processing method to obtain a first processing result;

[0082] Obtain environmental parameters of GNSS equipment and time parameters of several time sources;

[0083] According to a preset adaptive learning algorithm, the initial time synchronization parameters are processed in combination with the environmental parameters, the time parameters and the first processing result to obtain time synchronization prediction parameters.

[0084] This embodiment can effectively eliminate the errors caused by the multipath effect by processing the multipath effect of the pulse signal and the time synchronization standard signal, thereby improving the accuracy of the initial time synchronization parameter processing, and through the environmental parameters of the GNSS device and the time parameters of several time sources, the time synchronization prediction parameters contain different time sources and environmental parameters, thereby improving the prediction accuracy of the time synchronization prediction parameters, avoiding the time synchronization deviation caused by a single time source, and improving the time synchronization accuracy of the GNSS device.

[0085] In this embodiment, the initial time synchronization parameter is processed according to the preset adaptive learning algorithm in combination with the environmental parameter, the time parameter and the first processing result to obtain the time synchronization prediction parameter, specifically including:

[0086] After removing abnormal values ​​and noise points in the environmental parameters, time parameters, the first processing result and the initial time synchronization parameters according to a preset preprocessing method, extracting time synchronization features in the environmental parameters, time parameters, the first processing result and the initial time synchronization parameters to obtain a time synchronization feature set;

[0087] Inputting the time synchronization feature set into a preset adaptive learning model group to obtain a predicted time jitter value, a predicted time delay value and a predicted time offset value output by the adaptive learning model group;

[0088] A time synchronization prediction parameter is determined according to the predicted time jitter value, the predicted time delay value and the predicted time offset value.

[0089] This embodiment improves the prediction accuracy of the adaptive learning model group by extracting time synchronization features from environmental parameters, time parameters, the first processing results and initial time synchronization parameters, and predicts through an adaptive learning model group. As a result, different time sources and environmental parameters are implied in the time synchronization prediction parameters, the prediction accuracy of the time synchronization prediction parameters is improved, the time synchronization deviation caused by a single time source is avoided, and the time synchronization accuracy of the GNSS device is improved.

[0090] In an optional embodiment, an integrated sensor, such as a temperature sensor, a humidity sensor, etc., may be installed on the GNSS device, and environmental parameters of the GNSS device, such as temperature, humidity, electromagnetic interference, etc., may be collected through the integrated sensor; time parameters of several time sources, such as GPS, GLONASS, Galileo, etc., may be obtained through the GNSS receiving module; and a Kalman filter may be used to process multipath effects of the pulse signal and the time synchronization standard signal to obtain a first processing result;

[0091] Then, the environmental parameters, time parameters, the first processing result and the initial time synchronization parameters are preprocessed to remove abnormal values ​​and noise points. Then, the environmental parameters, time parameters, the first processing result and the initial time synchronization parameters are parsed and received using the same field to facilitate subsequent time synchronization feature extraction;

[0092] According to the analysis results, key features of the environment parameters, time parameters, the first processing results and the initial time synchronization parameters are extracted, including time offset, delay, jitter and period, etc., to obtain a time synchronization feature set;

[0093] Then, the time synchronization feature set is input into the adaptive learning model group, and the adaptive learning model group is used to identify the regular patterns in the time synchronization feature set, obtain the predicted time jitter value, predicted time delay value and predicted time offset value, and determine the time synchronization prediction parameters.

[0094] In an optional embodiment, the multipath effect in the pulse signal and the time synchronization standard signal is detected, and the errors therein are corrected using a Kalman filter to obtain a correction timestamp, an error estimation report, and signal quality, etc., to determine a first processing result.

[0095] The above-mentioned embodiments can reduce the time synchronization error caused by the multipath effect, so that the GNSS still has accurate time synchronization in multi-reflection environments such as cities with tall buildings or mountainous areas.

[0096] In an optional embodiment, the time synchronization prediction parameters include Stratum, Poll Interval, Offset, Delay, Jitter and Dispersion. Stratum is used to indicate the level of the time source, for example, 0 indicates the reference clock, 1 indicates the server directly connected to the reference clock, and so on. Poll Interval indicates the time interval for communication between the GNSS device and the time server (time source), which is usually expressed in powers of 2 (such as 4=16 seconds, 6=64 seconds). Offset is used to indicate the difference between the local clock and the reference clock. Delay is used to indicate the time delay caused by network transmission. Jitter indicates the standard deviation of the time offset measured several times in a row, reflecting the stability of time synchronization. Dispersion indicates the additional delay caused by different paths.

[0097] In an optional embodiment, the adaptive learning model group includes a recursive neural network (RNN), a long short-term memory network (LSTM) and a Kalman filter. The recursive neural network (RNN), the long short-term memory network (LSTM) and the Kalman filter have all been pre-trained. Since the training process of the recursive neural network (RNN), the long short-term memory network (LSTM) and the Kalman filter currently has relatively mature technology, it will not be expanded upon here.

[0098] In an optional embodiment, the prediction process of the adaptive learning model group is regulated by adaptive rate, feedback weight and threshold, wherein the adaptive rate is used to adjust the speed of parameter update in the adaptive learning model group to balance the requirements of fast response and stability; the feedback weight is used to determine the influence of information from system feedback on parameter update; the threshold is used to set the conditions for triggering specific behaviors, such as starting a self-correction mechanism when the error exceeds a certain threshold.

[0099] It should be noted that in the adaptive learning model group, the essence of the Kalman filter is a recursive algorithm. In the time synchronization process of the GNSS device, the Kalman filter can predict the most likely real time of the GNSS device, and can also predict the future time synchronization state of the GNSS device and use it to correct the current time synchronization estimation result, thereby reducing the error accumulation in the time synchronization. The following are the relevant parameters of the Kalman filter, including the initial state vector (Initial State Vector), process noise covariance matrix (Process Noise Covariance Matrix, Q), measurement noise covariance matrix (Measurement Noise Covariance Matrix, R), state transition matrix (State Transition Matrix, A), control input matrix (Control Input Matrix, B), observation matrix (Observation Matrix, H) and prediction error covariance matrix (Error Covariance Matrix, P); the initial state vector is used to initialize the state estimate of the Kalman filter; the process noise covariance matrix is ​​used to describe the degree of uncertainty within the system; the observation noise covariance matrix is ​​used to describe the uncertainty of the observation data; the state transition matrix is ​​used to describe the way the system transfers from one state to the next; the control input matrix is ​​used to represent the influence of external control input on the system; the observation matrix is ​​used to map the state space to the observation space matrix; the prediction error covariance matrix is ​​used to update the error distribution of the state estimate.

[0100] It should be noted that in the adaptive learning model group, both the recurrent neural network (RNN) and the long short-term memory network (LSTM) are used to identify which factors in the time synchronization feature set will affect the accuracy of time synchronization and output the optimal time synchronization strategy; the following are the relevant parameters contained in both the recurrent neural network (RNN) and the long short-term memory network (LSTM), including: the number of hidden units, the number of layers, the learning rate, the batch size, the activation function and the regularization parameter. Parameters); the number of hidden layer units is the number of neurons in each layer, which affects the complexity and fitting ability of the model; the number of layers is used to indicate that multiple layers of RNN or LSTM units are stacked in the recursive neural network (RNN) and the long short-term memory network (LSTM), thereby increasing the depth of the model; the learning rate is used to determine the size of the gradient descent step, which affects the training speed and convergence; the batch size is used to indicate the number of samples used in each iteration, which affects the training efficiency and generalization performance; the activation function is used to determine the output characteristics of each neuron, such as ReLU, Sigmoid, etc.; the regularization parameter is used to prevent the model from overfitting, such as the L2 regularization coefficient.

[0101] Step S103: transmitting the initial time synchronization parameters, environmental parameters and time synchronization prediction parameters to the cloud, so that the cloud calculates the cloud time synchronization parameters according to the initial time synchronization parameters, environmental parameters and time synchronization prediction parameter set.

[0102] In this embodiment, the step of enabling the cloud to calculate the cloud time synchronization parameter according to the initial time synchronization parameter, the environmental parameter and the time synchronization prediction parameter set specifically includes:

[0103] Building a global environment model according to the environmental parameters, and inputting the environmental parameters into the global environment model, predicting the change trend of the multipath effect of the GNSS device, and obtaining a multipath effect prediction value;

[0104] Extracting time features from the initial time synchronization parameters, environmental parameters, and time synchronization prediction parameter sets, and obtaining cloud time synchronization prediction parameters based on the time features;

[0105] Generate a cloud time synchronization parameter according to the multipath effect prediction value and the cloud time synchronization prediction parameter.

[0106] This embodiment constructs a global environmental model to predict the changing trend of the multipath effect in the GNSS device, and then merges the time synchronization prediction parameters through the time characteristics of the environmental parameters and the time synchronization prediction parameter set, so that the generated cloud time synchronization parameters can better meet the actual time synchronization situation and future trends of the GNSS device, thereby improving the time synchronization accuracy of the GNSS device.

[0107] In an optional embodiment, it is also possible to extract time features (such as time offset, time delay, time jitter, etc.) from the initial time synchronization parameters, environmental parameters, and time synchronization prediction parameter sets, compress these time features, and then transmit them to the cloud through a low-latency communication protocol (such as MQTT, CoAP).

[0108] The above embodiment can reduce the amount of data transmitted from the GNSS device to the cloud device by compressing the time features after extraction in the GNSS device, thereby reducing the loss and required resources during the data transmission process.

[0109] In an optional embodiment, a new adaptive learning model group and multipath effect correction strategy may be generated in the cloud based on the time characteristics of the initial time synchronization parameters, environmental parameters and time synchronization prediction parameter set, and the new adaptive learning model group and multipath effect correction strategy may be returned to the GNSS device, thereby updating the adaptive learning model group and multipath effect processing process in the GNSS device.

[0110] The above-mentioned embodiment can improve the accuracy of the adaptive learning model group and multipath effect processing in the GNSS device by updating the adaptive learning model group and multipath effect processing process in the cloud, thereby improving the time synchronization accuracy of the GNSS device.

[0111] In an optional embodiment, a global environmental model can also be constructed based on the environmental parameters, and the environmental parameters can be input into the global environmental model to predict the changing trend of the multipath effect of the GNSS device to obtain a multipath effect prediction value. The global environmental model can be a recursive neural network (RNN), a long short-term memory network (LSTM) and a Kalman filter.

[0112] The above embodiment can predict the changing trend of the multipath effect of the GNSS device by constructing a global environment model, thereby avoiding potential problems in the time synchronization process of the GNSS device in advance, and improving the accuracy of the time synchronization of the GNSS device. In particular, there are relatively mature technologies for the construction and training of the global environment model, which will not be elaborated in detail here.

[0113] In an optional embodiment, a neural network model (such as an LSTM and RNN model) may also be set up on the cloud, and the time features of the initial time synchronization parameters, environmental parameters, and time synchronization prediction parameter sets may be extracted through the neural network model, and cloud time synchronization prediction parameters may be obtained based on the time features; wherein the cloud time synchronization prediction parameters include: cloud time offset prediction value, cloud time delay prediction value, and cloud time jitter prediction value.

[0114] The above-mentioned embodiment can obtain more accurate time synchronization parameters by fusing the initial time synchronization parameters, environmental parameters and time characteristics of the time synchronization prediction parameter set in the cloud, thereby improving the time synchronization accuracy of the GNSS device.

[0115] Step S104: When the initial time synchronization parameters do not meet the preset requirements, the time synchronization of the GNSS device is corrected according to the time synchronization prediction parameters and the cloud time synchronization parameters to complete the time synchronization of the GNSS device.

[0116] In an optional embodiment, when the initial time synchronization parameters do not meet the preset requirements, for example, when the time offset and time delay in the GNSS device time synchronization process exceed the preset threshold, the predicted time jitter value, predicted time delay value and predicted time offset value in the time synchronization prediction parameters can be obtained, and the predicted time jitter value, predicted time delay value and predicted time offset value can be input into the Chronyd module, so as to synchronize the GNSS device again; the cloud time synchronization parameters (including cloud time offset prediction value, cloud time delay prediction value and cloud time jitter prediction value) can also be input into the Chronyd module, so as to synchronize the GNSS device again.

[0117] It should be noted that the Chronyd module in the GNSS device performs time synchronization based on the predicted values ​​of time offset, time delay and time jitter, which is a routine operation for those skilled in the art and will not be elaborated on here.

[0118] In an optional embodiment, the cloud described in this embodiment is provided with a private NTP server, which can be synchronized according to the NTP network time and provide a network time synchronization source for the Chronyd module. Therefore, this embodiment also provides a fault-tolerant mechanism; when the main time source fails, that is, when the pulse signal and the time synchronization standard signal obtained by the GNSS device are wrong, the NTP server in the cloud can be used as a backup time source to maintain the time synchronization of the GNSS device; specifically, the Chrony.conf file in the Chronyd module is edited, and "server0.pool.ntp.org iburst maxdistance1.0" is added to the Chrony.conf file, so that the Chronyd module can obtain relevant time synchronization data from the cloud NTP server.

[0119] In an optional embodiment, when multiple GNSS devices are time synchronized, this embodiment improves the time synchronization accuracy and efficiency of the GNSS devices through a dynamic load balancing method;

[0120] Specifically, when GNSS devices perform time synchronization, they need to obtain pulse signals and time synchronization standard signals from the time source (i.e., the time synchronization server), and the pulse signals and time synchronization standard signals need to be transmitted through the time synchronization path. Congestion in the time synchronization path will lead to inaccurate time synchronization of GNSS devices. When a time synchronization server fails to send, it will also lead to inaccurate time synchronization of GNSS devices. In addition, since an adaptive learning model group is set up in the GNSS device, corresponding computing resources are required, and the GNSS device also needs to occupy resources when sending data to the cloud. Therefore, load balancing of GNSS devices is very important for the time synchronization accuracy of GNSS devices.

[0121] In an optional embodiment, a lightweight monitoring task is arranged on each GNSS device to continuously collect key performance indicators such as CPU usage, memory usage, and network bandwidth of the GNSS device, and resource calls of the GNSS device are allocated based on these key performance indicators;

[0122] Specifically, the key performance indicators can be input into the adaptive learning model group, and the adaptive learning model group can be used to predict the resource usage of the GNSS equipment in the future. According to the predicted resource usage, the computing resources of the adaptive learning model group of the GNSS equipment, the resources required to send data to the cloud, the resources required to obtain pulse signals and time synchronization standard signals from the time source, and the time synchronization path selection of the pulse signals and time synchronization standard signals can be allocated.

[0123] It should be noted that dynamic load balancing is currently commonly used in resource scheduling of various devices. This embodiment uses dynamic load balancing in GNSS devices, which is a routine operation of those skilled in the art and will not be elaborated herein.

[0124] This embodiment obtains a pulse signal and a time synchronization standard signal. Through two different types of time synchronization signals, the accuracy of GNSS device time synchronization can be improved. Based on the initial time synchronization parameters obtained from the time synchronization, the time parameters and environmental parameters of different time sources are integrated, so that the time synchronization prediction parameters can more accurately characterize the predicted situation of the time synchronization of the GNSS device. After that, the predicted situation of the time synchronization of the GNSS device and the environmental parameters are integrated and calculated through the cloud to obtain the cloud time synchronization parameters. When an error occurs in a single time source, the time synchronization of the GNSS device can be corrected through the time synchronization prediction parameters and the cloud time synchronization parameters, so that the time synchronization result of the GNSS device takes into account the time parameters and environmental parameters of multiple time sources, avoids the time synchronization deviation caused by a single time source, and improves the time synchronization accuracy of the GNSS device.

[0125] Embodiment 2

[0126] Please refer to Figure 3 , which is a structural diagram of a time synchronization device for a GNSS device provided in an embodiment of the present invention, comprising: a preliminary time synchronization parameter acquisition module 201, an adaptive learning module 202, a cloud time synchronization module 203 and a time synchronization correction module 204.

[0127] The preliminary time synchronization parameter acquisition module 201 is used to acquire a pulse signal and a time synchronization standard signal, and perform time synchronization on the GNSS device according to the pulse signal and the time synchronization standard signal to obtain initial time synchronization parameters.

[0128] In this embodiment, the preliminary time synchronization parameter acquisition module 201 includes: a preliminary time synchronization parameter acquisition unit;

[0129] The preliminary time synchronization parameter acquisition unit is used to acquire a pulse signal and a time synchronization standard signal;

[0130] Calculating the time delay between the pulse signal and the time synchronization standard signal;

[0131] Based on the pulse signal, the time synchronization standard signal and the time delay, the GNSS device is time synchronized, and the time offset and time delay in the GNSS device time synchronization process are calculated;

[0132] An initial time synchronization parameter is obtained according to the time offset and time delay in the time synchronization process of the GNSS device.

[0133] The adaptive learning module 202 is used to obtain the environmental parameters of the GNSS device and the time parameters of several time sources, and process the initial time synchronization parameters according to a preset adaptive learning algorithm, the environmental parameters and the time parameters to obtain the time synchronization prediction parameters.

[0134] In this embodiment, the adaptive learning module 202 includes: an adaptive learning unit;

[0135] The adaptive learning unit is used to perform multipath effect processing on the pulse signal and the time synchronization standard signal according to a preset signal processing method to obtain a first processing result;

[0136] Obtain environmental parameters of GNSS equipment and time parameters of several time sources;

[0137] According to a preset adaptive learning algorithm, the initial time synchronization parameters are processed in combination with the environmental parameters, the time parameters and the first processing result to obtain time synchronization prediction parameters.

[0138] In this embodiment, the adaptive learning unit includes: a time synchronization prediction parameter acquisition subunit;

[0139] The time synchronization prediction parameter acquisition subunit is used to extract the time synchronization features in the environment parameters, time parameters, first processing results and initial time synchronization parameters after removing abnormal values ​​and noise points in the environment parameters, time parameters, first processing results and initial time synchronization parameters according to a preset preprocessing method to obtain a time synchronization feature set;

[0140] Inputting the time synchronization feature set into a preset adaptive learning model group to obtain a predicted time jitter value, a predicted time delay value and a predicted time offset value output by the adaptive learning model group;

[0141] A time synchronization prediction parameter is determined according to the predicted time jitter value, the predicted time delay value and the predicted time offset value.

[0142] The cloud time synchronization module 203 is used to transmit the initial time synchronization parameters, environmental parameters and time synchronization prediction parameters to the cloud, so that the cloud calculates the cloud time synchronization parameters according to the initial time synchronization parameters, environmental parameters and time synchronization prediction parameter set.

[0143] In this embodiment, the time synchronization correction module includes: a time synchronization correction unit;

[0144] The time synchronization correction unit is used to construct a global environment model according to the environmental parameters, and input the environmental parameters into the global environment model to predict the change trend of the multipath effect of the GNSS device to obtain a multipath effect prediction value;

[0145] Extracting time features from the initial time synchronization parameters, environmental parameters, and time synchronization prediction parameter sets, and obtaining cloud time synchronization prediction parameters based on the time features;

[0146] Generate a cloud time synchronization parameter according to the multipath effect prediction value and the cloud time synchronization prediction parameter.

[0147] The time synchronization correction module 204 is used to correct the time synchronization of the GNSS device according to the time synchronization prediction parameters and the cloud time synchronization parameters when the initial time synchronization parameters do not meet the preset requirements, so as to complete the time synchronization of the GNSS device.

[0148] This embodiment obtains a pulse signal and a time synchronization standard signal. Through two different types of time synchronization signals, the accuracy of GNSS device time synchronization can be improved. Based on the initial time synchronization parameters obtained from the time synchronization, the time parameters and environmental parameters of different time sources are integrated, so that the time synchronization prediction parameters can more accurately characterize the predicted situation of the time synchronization of the GNSS device. After that, the predicted situation of the time synchronization of the GNSS device and the environmental parameters are integrated and calculated through the cloud to obtain the cloud time synchronization parameters. When an error occurs in a single time source, the time synchronization of the GNSS device can be corrected through the time synchronization prediction parameters and the cloud time synchronization parameters, so that the time synchronization result of the GNSS device takes into account the time parameters and environmental parameters of multiple time sources, avoids the time synchronization deviation caused by a single time source, and improves the time synchronization accuracy of the GNSS device.

[0149] To summarize, the embodiments of the present invention can improve the accuracy of time synchronization of GNSS devices by acquiring pulse signals and time synchronization standard signals through two different types of time synchronization signals, and based on the initial time synchronization parameters obtained from the time synchronization, fuse the time parameters and environmental parameters of different time sources, so that the time synchronization prediction parameters can more accurately characterize the predicted situation of time synchronization of GNSS devices, and then fuse the predicted situation of time synchronization of GNSS devices and environmental parameters through the cloud to obtain cloud time synchronization parameters. When an error occurs in a single time source, the time synchronization of the GNSS device can be corrected through the time synchronization prediction parameters and the cloud time synchronization parameters, so that the time synchronization result of the GNSS device takes into account the time parameters and environmental parameters of multiple time sources, avoids the time synchronization deviation caused by a single time source, and improves the time synchronization accuracy of the GNSS device.

[0150] The specific embodiments described above further illustrate the purpose, technical solutions and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. It is particularly pointed out that for those skilled in the art, any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the scope of protection of the present invention.

Claims

1. A time synchronization method for a GNSS device, characterized in that: include: Acquire a pulse signal and a time synchronization standard signal, and perform time synchronization on the GNSS device according to the pulse signal and the time synchronization standard signal to obtain initial time synchronization parameters; Acquire environmental parameters of the GNSS device and time parameters of several time sources, process the initial time synchronization parameters according to a preset adaptive learning algorithm, the environmental parameters and the time parameters, and obtain time synchronization prediction parameters; Transmitting the initial time synchronization parameters, environmental parameters and time synchronization prediction parameters to the cloud, so that the cloud calculates the cloud time synchronization parameters according to the initial time synchronization parameters, environmental parameters and time synchronization prediction parameter set; When the initial time synchronization parameters do not meet the preset requirements, the time synchronization of the GNSS device is corrected according to the time synchronization prediction parameters and the cloud time synchronization parameters to complete the time synchronization of the GNSS device.

2. A time synchronization method for GNSS equipment as claimed in claim 1, characterized in that: The step of acquiring a pulse signal and a time synchronization standard signal, and performing time synchronization on a GNSS device according to the pulse signal and the time synchronization standard signal to obtain an initial time synchronization parameter specifically includes: Obtain pulse signals and time synchronization standard signals; Calculating the time delay between the pulse signal and the time synchronization standard signal; Based on the pulse signal, the time synchronization standard signal and the time delay, the GNSS device is time synchronized, and the time offset and time delay in the GNSS device time synchronization process are calculated; An initial time synchronization parameter is obtained according to the time offset and time delay in the time synchronization process of the GNSS device.

3. A time synchronization method for GNSS equipment as claimed in claim 1, characterized in that: The acquiring of the environmental parameters of the GNSS device and the time parameters of several time sources, and processing the initial time synchronization parameters according to a preset adaptive learning algorithm, the environmental parameters and the time parameters to obtain the time synchronization prediction parameters specifically includes: Perform multipath effect processing on the pulse signal and the time synchronization standard signal according to a preset signal processing method to obtain a first processing result; Obtain environmental parameters of GNSS equipment and time parameters of several time sources; According to a preset adaptive learning algorithm, the initial time synchronization parameters are processed in combination with the environmental parameters, the time parameters and the first processing result to obtain time synchronization prediction parameters.

4. A time synchronization method for GNSS equipment as claimed in claim 3, characterized in that: The processing of the initial time synchronization parameter according to the preset adaptive learning algorithm in combination with the environmental parameter, the time parameter and the first processing result to obtain the time synchronization prediction parameter specifically includes: After removing abnormal values ​​and noise points in the environmental parameters, time parameters, the first processing result and the initial time synchronization parameters according to a preset preprocessing method, extracting time synchronization features in the environmental parameters, time parameters, the first processing result and the initial time synchronization parameters to obtain a time synchronization feature set; Inputting the time synchronization feature set into a preset adaptive learning model group to obtain a predicted time jitter value, a predicted time delay value and a predicted time offset value output by the adaptive learning model group; A time synchronization prediction parameter is determined according to the predicted time jitter value, the predicted time delay value and the predicted time offset value.

5. A time synchronization method for GNSS equipment as claimed in claim 1, characterized in that: The method of enabling the cloud to calculate the cloud time synchronization parameters according to the initial time synchronization parameters, the environmental parameters and the time synchronization prediction parameter set specifically includes: Building a global environment model according to the environmental parameters, and inputting the environmental parameters into the global environment model, predicting the change trend of the multipath effect of the GNSS device, and obtaining a multipath effect prediction value; Extracting time features from the initial time synchronization parameters, environmental parameters, and time synchronization prediction parameter sets, and obtaining cloud time synchronization prediction parameters based on the time features; Generate a cloud time synchronization parameter according to the multipath effect prediction value and the cloud time synchronization prediction parameter.

6. A time synchronization device for a GNSS device, characterized in that: include: Preliminary time synchronization parameter acquisition module, adaptive learning module, cloud time synchronization module and time synchronization correction module; The preliminary time synchronization parameter acquisition module is used to acquire a pulse signal and a time synchronization standard signal, and perform time synchronization on the GNSS device according to the pulse signal and the time synchronization standard signal to obtain initial time synchronization parameters; The adaptive learning module is used to obtain the environmental parameters of the GNSS device and the time parameters of several time sources, and process the initial time synchronization parameters according to the preset adaptive learning algorithm, the environmental parameters and the time parameters to obtain the time synchronization prediction parameters; The cloud time synchronization module is used to transmit the initial time synchronization parameters, environmental parameters and time synchronization prediction parameters to the cloud, so that the cloud calculates the cloud time synchronization parameters according to the initial time synchronization parameters, environmental parameters and time synchronization prediction parameter set; The time synchronization correction module is used to correct the time synchronization of the GNSS device according to the time synchronization prediction parameters and the cloud time synchronization parameters when the initial time synchronization parameters do not meet the preset requirements, so as to complete the time synchronization of the GNSS device.

7. A time synchronization device for GNSS equipment as claimed in claim 6, characterized in that: The preliminary time synchronization parameter acquisition module includes: a preliminary time synchronization parameter acquisition unit; The preliminary time synchronization parameter acquisition unit is used to acquire a pulse signal and a time synchronization standard signal; Calculating the time delay between the pulse signal and the time synchronization standard signal; Based on the pulse signal, the time synchronization standard signal and the time delay, the GNSS device is time synchronized, and the time offset and time delay in the GNSS device time synchronization process are calculated; An initial time synchronization parameter is obtained according to the time offset and time delay in the time synchronization process of the GNSS device.

8. A time synchronization device for GNSS equipment as claimed in claim 6, characterized in that: The adaptive learning module comprises: an adaptive learning unit; The adaptive learning unit is used to perform multipath effect processing on the pulse signal and the time synchronization standard signal according to a preset signal processing method to obtain a first processing result; Obtain environmental parameters of GNSS equipment and time parameters of several time sources; According to a preset adaptive learning algorithm, the initial time synchronization parameters are processed in combination with the environmental parameters, the time parameters and the first processing result to obtain time synchronization prediction parameters.

9. A time synchronization device for GNSS equipment as claimed in claim 8, characterized in that: The adaptive learning unit includes: a time synchronization prediction parameter acquisition subunit; The time synchronization prediction parameter acquisition subunit is used to extract the time synchronization features in the environment parameters, time parameters, first processing results and initial time synchronization parameters after removing abnormal values ​​and noise points in the environment parameters, time parameters, first processing results and initial time synchronization parameters according to a preset preprocessing method to obtain a time synchronization feature set; Inputting the time synchronization feature set into a preset adaptive learning model group to obtain a predicted time jitter value, a predicted time delay value and a predicted time offset value output by the adaptive learning model group; A time synchronization prediction parameter is determined according to the predicted time jitter value, the predicted time delay value and the predicted time offset value.

10. A time synchronization device for GNSS equipment as claimed in claim 6, characterized in that: The time synchronization correction module comprises: a time synchronization correction unit; The time synchronization correction unit is used to construct a global environment model according to the environmental parameters, and input the environmental parameters into the global environment model to predict the change trend of the multipath effect of the GNSS device to obtain a multipath effect prediction value; Extracting time features from the initial time synchronization parameters, environmental parameters, and time synchronization prediction parameter sets, and obtaining cloud time synchronization prediction parameters based on the time features; Generate a cloud time synchronization parameter according to the multipath effect prediction value and the cloud time synchronization prediction parameter.

Images