GNSS time service receiver hardware delay temperature compensation method based on thermal response hysteresis
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- BEIHANG UNIV
- Filing Date
- 2026-06-09
- Publication Date
- 2026-07-10
AI Technical Summary
Existing GNSS timing receivers, under dynamic temperature environments, use hardware delay compensation methods that neglect the thermal response hysteresis characteristics of devices, leading to increased time errors and affecting timing accuracy.
By acquiring time deviation and temperature change sequences, the Pearson correlation coefficient is calculated to determine the optimal thermal hysteresis time. A mapping relationship model is constructed and real-time compensation is performed, including the selection of linear and quadratic models, reducing the computational burden and adapting to different temperature characteristics.
It significantly reduces the temperature-dependent time difference error of 1PPS output under dynamic temperature conditions, improves timing accuracy and reliability, and is suitable for complex application environments.
Smart Images
Figure CN122362433A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of satellite navigation technology, and in particular to a method for compensating for hardware delay temperature in GNSS timing receivers based on thermal response hysteresis. Background Technology
[0002] In practical engineering applications, high-precision GNSS timing receivers are often deployed in outdoor sites, equipment cabinets, or complex environments with limited temperature control, making it difficult to maintain a stable thermal equilibrium state over long periods. Changes in ambient temperature not only cause drift in the physical characteristics of electronic components such as crystal oscillators and RF front-ends, leading to changes in the signal processing link delay within the receiver, but also, due to the thermal inertia and thermal conductivity of the components, the hardware delay exhibits significant asynchronous response and hysteresis characteristics to temperature changes. This results in a dynamic time difference offset in the 1PPS time output that varies with temperature.
[0003] Existing receiver hardware delay temperature compensation methods are typically based on static temperature calibration or function fitting models under steady-state conditions. These methods assume the system has reached thermal equilibrium at a given temperature and only establish a correspondence between hardware delay and a single temperature value. However, under continuously varying temperatures or unsteady-state operating conditions, hardware delay is often closely related not only to the current temperature value but also to the rate of temperature change and historical temperature conditions. Directly applying the aforementioned static compensation methods to dynamic temperature environments can easily lead to decreased compensation accuracy and even introduce new time errors.
[0004] To address the aforementioned issues, existing technologies generally assume that the receiver hardware delay is only related to the currently measured temperature value, neglecting the thermal response hysteresis characteristics of core components such as crystal oscillators, RF front-ends, and digital processing units during temperature changes. This results in a residual systematic time difference error related to the temperature change process in the 1PPS output signal under scenarios of heating, cooling, or periodic temperature fluctuations, thereby limiting high-precision time synchronization performance. Summary of the Invention
[0005] The purpose of this invention is to provide a hardware delay temperature compensation method for GNSS timing receivers based on thermal response hysteresis, aiming to solve or improve at least one of the above-mentioned technical problems.
[0006] To achieve the above objectives, the present invention provides the following solution:
[0007] Hardware delay temperature compensation methods for GNSS timing receivers based on thermal response hysteresis include:
[0008] Acquire the time deviation sequence of the GNSS timing receiver relative to the reference reference, synchronously acquire the temperature change sequence of the receiver's operating environment, and perform unified time alignment;
[0009] Iterate through each time lag value within the preset lag time search range, calculate the Pearson correlation coefficient between the current time deviation sequence and the historical temperature sequence after response time shift, and determine the optimal thermal lag time;
[0010] Based on the optimal thermal hysteresis time, a mapping relationship model between the 1PPS time deviation sequence and the hysteresis temperature sequence is constructed, and a real-time compensation mapping relationship model is selected based on the data characteristics.
[0011] Based on the selected mapping model, the timing receiver is subjected to signal compensation and correction to obtain the final timing signal.
[0012] Furthermore, the Pearson correlation coefficient is expressed as follows:
[0013]
[0014] In the formula, For the lag time is The Pearson correlation coefficient at time [-1, 1]; For the first Each sampling time; For at any time Corresponding process Temperature value after time shift; For at any time The measured 1PPS time difference; Lag time The arithmetic mean of the subsequent temperature change sequence; The mean of the 1PPS time difference sequence; This represents the number of valid samples.
[0015] Furthermore, determining the optimal thermal lag time includes:
[0016] The time lag value corresponding to the maximum absolute value of the Pearson correlation coefficient is selected as the optimal thermal lag time.
[0017] When the maximum value of the Pearson correlation coefficient is lower than the preset correlation threshold within the search range, the optimal thermal hysteresis time is... .
[0018] Furthermore, the mapping relationship model includes:
[0019] The linear model is expressed as follows:
[0020]
[0021] The quadratic model is expressed as follows:
[0022]
[0023] In the formula, , The predicted 1PPS time deviation values are shown for the two models, respectively. for The hysteresis temperature value corresponding to time. ; , These are the regression coefficients to be estimated in the linear model; , , These are the regression coefficients to be estimated in the quadratic model.
[0024] Furthermore, based on data characteristics, a mapping relationship model for real-time compensation is selected, including:
[0025] Based on the optimal thermal hysteresis time, the least squares method is used to fit the parameters of the mapping relationship model, minimize the objective function of the residual sum of squares, and generate the optimal parameters.
[0026] Using the optimal parameters, calculate the root mean square error of the mapping relationship model;
[0027] Based on the root mean square error of the mapping relationship model, the mapping relationship model for real-time compensation is determined.
[0028] Furthermore, based on the optimal thermal hysteresis time, the least squares method is used to fit the parameters of the mapping relationship model, minimizing the objective function of the residual sum of squares to generate the optimal parameters, including:
[0029] Solving linear models Generate the optimal solution and ;
[0030] Solving the quadratic model Generate the optimal solution , , .
[0031] Furthermore, using the optimal parameters, the root mean square error of the mapping relationship model is calculated, including:
[0032] The root mean square error of a linear model is expressed as:
[0033]
[0034] The root mean square error of the quadratic model is expressed as:
[0035]
[0036] In the formula, and These are the root mean square errors for the linear model and the quadratic model, respectively. and The smaller the value, the higher the model's fitting accuracy to the receiver's true temperature drift characteristics; The total number of samples; , These are the predicted values for the linear model and the quadratic model using the optimal parameters, respectively.
[0037] Furthermore, based on the root mean square error of the mapping relationship model, the mapping relationship model for real-time compensation is determined, including:
[0038] Set improvement threshold ,like A linear model is chosen as the model for real-time compensation; if A quadratic model was chosen as the model for real-time compensation.
[0039] Furthermore, based on the selected mapping model, signal compensation and correction are performed on the timing receiver to obtain the final timing signal, including:
[0040] Open a space of length in the receiver firmware. A circular buffer that stores the most recent data in real time. Temperature data from each sampling point;
[0041] At time t, index and extract from the circular buffer. The historical temperature at any given moment is expressed as:
[0042]
[0043] In the formula, The effective temperature at the current moment that affects the timing receiver and causes time delay drift;
[0044] Based on the selected mapping model, the estimated time delay drift relative to the reference temperature is calculated, expressed as:
[0045]
[0046] In the formula, This is an estimate of the time delay drift. The selected mapping relationship model; Reference temperature;
[0047] Read the raw 1PPS measurement data at the current moment, compensate and correct it according to the estimated time delay drift value, and obtain the final time signal, expressed as:
[0048]
[0049] In the formula, This is the final timing signal; This is the original 1PPS measurement data.
[0050] According to specific embodiments provided by the present invention, the present invention discloses the following technical effects:
[0051] This invention discloses a method for compensating for hardware delay temperature in GNSS timing receivers based on thermal response hysteresis. The method introduces an equivalent thermal response hysteresis time parameter and quantitatively determines the hysteresis amount based on the temperature-delay correlation, enabling the compensation model to accurately reflect the dynamic impact of temperature change on hardware delay. This significantly reduces the temperature-related time difference error in 1PPS output under conditions of continuous temperature variation or rapid temperature fluctuation.
[0052] Instead of complex modeling of the receiver's internal structure, material parameters, or thermal coupling relationships among multiple components, this method establishes an equivalent hysteresis relationship between the actual observed temperature sequence and the timing deviation sequence to achieve an engineering description of the thermal inertia effect. This method requires no additional hardware modifications or computationally complex thermal simulation models, making it easy to implement directly on existing GNSS timing receiver platforms and demonstrating good versatility and deployability.
[0053] Simultaneously, linear and quadratic models are constructed, and model selection is performed based on root mean square error and an improved threshold. This ensures that a low-complexity model is prioritized when the temperature characteristics are approximately linear, while a high-precision model is automatically selected when significant nonlinear characteristics exist. Compared to temperature compensation methods using a single fixed model, this approach can adaptively select a suitable compensation model based on the actual receiver temperature characteristics, reducing the real-time computational burden while maintaining timing accuracy.
[0054] It is suitable not only for environments with slow temperature drift, but also for scenarios with rapid or periodic temperature changes caused by day-night cycles and cabinet ventilation variations. By introducing hysteresis temperature in real time and dynamically correcting hardware latency, it can effectively improve the timing accuracy and reliability of GNSS timing receivers in complex real-world application environments, demonstrating significant engineering application value. Attached Figure Description
[0055] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0056] Figure 1 This is a flowchart illustrating the method of the present invention. Detailed Implementation
[0057] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0058] The purpose of this invention is to provide a hardware delay temperature compensation method for GNSS timing receivers based on thermal response hysteresis, aiming to solve or improve at least one of the above-mentioned technical problems.
[0059] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.
[0060] like Figure 1 As shown, this invention provides a hardware delay temperature compensation method for GNSS timing receivers based on thermal response hysteresis, including:
[0061] Step 1: Obtain the time deviation sequence D(t) of the GNSS timing receiver relative to the reference standard, synchronously obtain the temperature change sequence T(t) of the receiver's operating environment, and perform unified time alignment, including:
[0062] Connect the GNSS time receiver to a high-precision time interval counter. Using a standard reference source as a reference, continuously acquire the time deviation sequence of the receiver's 1PPS output signal relative to the reference reference, denoted as D(t), where D(t) represents the 1PPS output time deviation observed at time t. The data changes in the time deviation sequence mainly reflect the relative drift of the receiver hardware delay caused by changes in environmental conditions.
[0063] Using temperature sensors located near key components of the receiver or inside the chassis, the temperature change sequence of the receiver's operating environment is collected synchronously and denoted as T(t);
[0064] Align D(t) and T(t) with a unified time to form the basis of subsequent analysis and modeling.
[0065] Step 2: Iterate through each time lag value within the preset lag time search range, and calculate the time deviation sequence D(t) at the current moment and the historical temperature sequence T(t-) after response time shift. Pearson correlation coefficient between ) Determining the optimal thermal lag time includes:
[0066] Set the search range for the subsequent time to 0-1 hour, and iterate through different time lag values within the search range;
[0067] Calculate the 1PPS time deviation sequence D(t) at the current moment and the historical temperature sequence T(t-) after the same time interval. Pearson correlation coefficient between ) The expression is:
[0068]
[0069] In the formula, For the lag time is The Pearson correlation coefficient at time [-1, 1]; For the first Each sampling time; For at any time Corresponding process Temperature value after time shift; For at any time The measured 1PPS time difference; Lag time The arithmetic mean of the subsequent temperature change sequence; The mean of the 1PPS time difference sequence; The number of valid samples;
[0070] When thermal response hysteresis exists, temperature change and hardware delay change show the strongest correlation in time at the hysteresis position.
[0071] The optimal thermal lag time is selected as the time lag value corresponding to the maximum absolute value of the Pearson correlation coefficient. ;
[0072] When the maximum value of the Pearson correlation coefficient is lower than the preset correlation threshold within the search range, the hysteresis effect of temperature change on hardware latency is considered negligible, and the optimal thermal hysteresis time is determined. .
[0073] Step 3: Based on the optimal thermal hysteresis time, construct the 1PPS time deviation sequence D(t) and the hysteresis temperature sequence T(t-). A mapping relationship model between data characteristics is established, and a real-time compensation mapping relationship model is selected based on the data features, including:
[0074] set up Let t be the effective operating temperature after hysteresis correction. Define a mapping model, including:
[0075] The linear model, applicable to weakly nonlinear devices, is expressed as follows:
[0076]
[0077] The quadratic model is applicable to devices such as crystal oscillators that exhibit significant nonlinear temperature-frequency characteristics, and its expression is:
[0078]
[0079] In the formula, , The predicted 1PPS time deviation values are shown for the two models, respectively. for The hysteresis temperature value corresponding to time. ; , These are the regression coefficients to be estimated in the linear model; , , These are the regression coefficients to be estimated in the quadratic model.
[0080] Based on the optimal thermal hysteresis time, the least squares method is used to fit the parameters of the mapping relationship model, minimizing the objective function of the sum of squared residuals, and generating the optimal parameters, including:
[0081] Solving linear models Generate the optimal solution and ;
[0082] Solving the quadratic model Generate the optimal solution , , .
[0083] Using the optimal parameters, calculate the root mean square error of the mapping relationship model, including:
[0084] The root mean square error of a linear model is expressed as:
[0085]
[0086] The root mean square error of the quadratic model is expressed as:
[0087]
[0088] In the formula, and These are the root mean square errors for the linear model and the quadratic model, respectively. and The smaller the value, the higher the model's fitting accuracy to the receiver's true temperature drift characteristics; The total number of samples; , These are the predicted values for the linear model and the quadratic model using the optimal parameters, respectively.
[0089] Based on the root mean square error of the mapping relationship model, the mapping relationship model for real-time compensation is determined, including:
[0090] Set improvement threshold In this embodiment, it is 5%;
[0091] like If the accuracy improvement of the quadratic model is not significant compared to the linear model, then the linear model is chosen as the model for real-time compensation in order to reduce the computational load.
[0092] like This indicates that the quadratic model significantly reduces the fitting error, suggesting that the system exhibits obvious nonlinear characteristics, thus the quadratic model is chosen as the model for real-time compensation.
[0093] Step 4: Based on the selected mapping model, perform signal compensation and correction on the timing receiver to obtain the final timing signal, including:
[0094] Open a space of length in the receiver firmware. A circular buffer that stores the most recent data in real time. Temperature data from each sampling point;
[0095] At time t, index and extract from the circular buffer. The historical temperature at any given moment is expressed as:
[0096]
[0097] In the formula, The effective temperature at the current moment that affects the timing receiver and causes time delay drift;
[0098] To eliminate fixed deviations and focus on temperature-induced fluctuations, a reference temperature is introduced. (The steady-state temperature during calibration is usually taken.)
[0099] Based on the selected mapping model, the estimated time delay drift relative to the reference temperature is calculated, expressed as:
[0100]
[0101] In the formula, This is an estimate of the time delay drift. The selected mapping relationship model; Reference temperature;
[0102] Read the raw 1PPS measurement data at the current moment, compensate and correct it according to the estimated time delay drift value, and obtain the final time signal, expressed as:
[0103]
[0104] In the formula, This is the final timing signal; This is the original 1PPS measurement data.
[0105] Real-time temperature compensation for hardware delay in GNSS timing receivers is achieved, making the timing output of the receiver in a dynamic temperature environment equivalent to the stable state under the reference temperature condition, thereby realizing hardware delay temperature compensation in high-precision timing.
[0106] This invention introduces an equivalent thermal response lag time parameter and quantitatively determines the lag amount based on the temperature-delay correlation, enabling the compensation model to accurately reflect the dynamic impact of temperature change on hardware delay, thereby significantly reducing the temperature-related time difference error in 1PPS output under continuous temperature variation or rapid temperature fluctuation conditions.
[0107] This invention does not involve complex modeling of the receiver's internal structure, material parameters, or the thermal coupling relationships of multiple components. Instead, it establishes an equivalent hysteresis relationship between the actual observed temperature sequence and the timing deviation sequence to achieve an engineering description of the thermal inertia effect. This method requires no additional hardware modifications or computationally complex thermal simulation models, making it easy to implement directly on existing GNSS timing receiver platforms and exhibiting good versatility and deployability.
[0108] This invention simultaneously constructs linear and quadratic models and selects the optimal model based on root mean square error and an improved threshold. This allows for the priority use of low-complexity models when temperature characteristics are approximately linear, while automatically selecting high-precision models when significant nonlinear characteristics exist. Compared to temperature compensation methods using a single fixed model, this invention can adaptively select a suitable compensation model based on the actual receiver temperature characteristics, reducing the real-time computational burden while ensuring timing accuracy.
[0109] The temperature compensation method described in this invention is applicable not only to environments with slow temperature drift, but also to scenarios with rapid or periodic temperature changes caused by day-night cycles, cabinet ventilation variations, etc. By introducing hysteresis temperature in real time and dynamically correcting hardware delays, the timing accuracy and reliability of GNSS timing receivers in complex real-world application environments can be effectively improved, demonstrating significant engineering application value.
[0110] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.
[0111] This document uses specific examples to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the core ideas of the present invention. Furthermore, those skilled in the art will recognize that, based on the ideas of the present invention, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of the present invention.
Claims
1. A hardware delay temperature compensation method for GNSS timing receivers based on thermal response hysteresis, characterized in that, include: Acquire the time deviation sequence of the GNSS timing receiver relative to the reference reference, synchronously acquire the temperature change sequence of the receiver's operating environment, and perform unified time alignment; Iterate through each time lag value within the preset lag time search range, calculate the Pearson correlation coefficient between the current time deviation sequence and the historical temperature sequence after response time shift, and determine the optimal thermal lag time; Based on the optimal thermal hysteresis time, a mapping relationship model between the 1PPS time deviation sequence and the hysteresis temperature sequence is constructed, and a real-time compensation mapping relationship model is selected based on the data characteristics. Based on the selected mapping model, the timing receiver is subjected to signal compensation and correction to obtain the final timing signal.
2. The hardware delay temperature compensation method for GNSS timing receivers based on thermal response hysteresis according to claim 1, characterized in that, The Pearson correlation coefficient is expressed as follows: In the formula, For the lag time is The Pearson correlation coefficient at time [-1, 1]; For the first Each sampling time; For at any time Corresponding process Temperature value after time shift; For at any time The measured 1PPS time difference; Lag time The arithmetic mean of the subsequent temperature change sequence; The mean of the 1PPS time difference sequence; This represents the number of valid samples.
3. The hardware delay temperature compensation method for GNSS timing receivers based on thermal response hysteresis according to claim 1, characterized in that, Determining the optimal thermal hysteresis time includes: The time lag value corresponding to the maximum absolute value of the Pearson correlation coefficient is selected as the optimal thermal lag time. When the maximum value of the Pearson correlation coefficient is lower than the preset correlation threshold within the search range, the optimal thermal hysteresis time is... .
4. The hardware delay temperature compensation method for GNSS timing receivers based on thermal response hysteresis according to claim 1, characterized in that, The mapping relationship model includes: The linear model is expressed as follows: The quadratic model is expressed as follows: In the formula, , The predicted 1PPS time deviation values are shown for the two models, respectively. for The hysteresis temperature value corresponding to time. ; , These are the regression coefficients to be estimated in the linear model; , , These are the regression coefficients to be estimated in the quadratic model.
5. The hardware delay temperature compensation method for GNSS timing receivers based on thermal response hysteresis according to claim 4, characterized in that, The mapping relationship model for selecting real-time compensation based on data characteristics includes: Based on the optimal thermal hysteresis time, the least squares method is used to fit the parameters of the mapping relationship model, minimize the objective function of the residual sum of squares, and generate the optimal parameters. Using the optimal parameters, calculate the root mean square error of the mapping relationship model; Based on the root mean square error of the mapping relationship model, the mapping relationship model for real-time compensation is determined.
6. The hardware delay temperature compensation method for GNSS timing receiver based on thermal response hysteresis according to claim 5, characterized in that, The process involves using the least squares method to fit parameters to the mapping relationship model based on the optimal thermal hysteresis time, minimizing the objective function of the sum of squared residuals, and generating optimal parameters, including: Solving linear models Generate the optimal solution and ; Solving the quadratic model Generate the optimal solution , , .
7. The hardware delay temperature compensation method for GNSS timing receivers based on thermal response hysteresis according to claim 5, characterized in that, The step of calculating the root mean square error of the mapping relationship model using optimal parameters includes: The root mean square error of a linear model is expressed as: The root mean square error of the quadratic model is expressed as: In the formula, and These are the root mean square errors for the linear model and the quadratic model, respectively. and The smaller the value, the higher the model's fitting accuracy to the receiver's true temperature drift characteristics; The total number of samples; , These are the predicted values for the linear model and the quadratic model using the optimal parameters, respectively.
8. The hardware delay temperature compensation method for GNSS timing receiver based on thermal response hysteresis according to claim 5, characterized in that, The step of determining the real-time compensation mapping relationship model based on the root mean square error of the mapping relationship model includes: Set improvement threshold ,like A linear model is chosen as the model for real-time compensation; if A quadratic model was chosen as the model for real-time compensation.
9. The hardware delay temperature compensation method for GNSS timing receiver based on thermal response hysteresis according to claim 1, characterized in that, The step of performing signal compensation and correction on the timing receiver according to the selected mapping relationship model to obtain the final timing signal includes: Open a space of length in the receiver firmware. A circular buffer that stores the most recent data in real time. Temperature data from each sampling point; At time t, index and extract from the circular buffer. The historical temperature at any given moment is expressed as: In the formula, The effective temperature at the current moment that affects the timing receiver and causes time delay drift; Based on the selected mapping model, the estimated time delay drift relative to the reference temperature is calculated, expressed as: In the formula, This is an estimate of the time delay drift. The selected mapping relationship model; Reference temperature; Read the raw 1PPS measurement data at the current moment, compensate and correct it according to the estimated time delay drift value, and obtain the final time signal, expressed as: In the formula, This is the final timing signal; This is the original 1PPS measurement data.