Train operation state space-time synchronization system and method based on real-time Beidou information

Abstract

The invention relates to a train operation state space-time synchronization system and method based on real-time Beidou information, and the system comprises a space-time synchronization module, a time calibration module based on an MPC algorithm, a Beidou receiving module, and a Kalman filtering module. The time-space synchronization module is respectively connected with the time calibration module based on the MPC algorithm, the Beidou receiving module and the Kalman filtering module, and the Kalman filtering module is respectively connected with the time calibration module based on the MPC algorithm and the Beidou receiving module. The method comprises the steps that S1, a time calibration module based on the MPC algorithm sets a local time clock; s2, receiving and processing Beidou information; s3, adopting Kalman filtering to control the period so as to ensure that the time interval is integrally controlled in the period; and S4, performing time-space synchronization and data transmission. According to the method, the influence of the error between the time clock in the local system and the actual Beidou receiving time on the data sending period can be greatly reduced, the train running state information can be accurately obtained, and the time-space synchronization effect of the train running state is achieved.

Description

Technical Field

[0001] This invention relates to the field of rail transit operation safety technology, specifically to a train operation status spatiotemporal synchronization system and method based on real-time BeiDou information. Background Technology

[0002] To ensure operational safety and efficient dispatching, trains need to transmit critical status information such as position and speed to the ground control center at regular intervals during operation. However, the crystal oscillator used for timing on the train's local system can experience frequency drift due to temperature changes and aging, leading to accumulated errors in timekeeping. This can cause disordered information transmission intervals or misaligned timestamps, severely affecting the ground system's accurate judgment and tracking of the train's operational status. Therefore, the train relies on the high-precision, high-stability standard time signal transmitted by the BeiDou Navigation Satellite System to periodically calibrate its local clock in real time, correcting the inherent deviations of the crystal oscillator and providing a reliable time reference for train safety control and efficient management. However, because the BeiDou Navigation Satellite System's time calibration period for the train's local system is longer than the period during which the train transmits status information, the calibration causes a disturbance to the original train status information transmission period. This results in the calibrated train status information period deviating significantly from the prescribed period, leading to instability in the train status information transmission period.

[0003] Chinese utility model patent CN204669363U relates to an onboard equipment system based on BeiDou satellite timing signals. The system includes: a BeiDou NTP server for outputting a corrected clock signal based on the BeiDou satellite timing signal; a LAIS electrically connected to the BeiDou NTP server via a first serial bus; an LKJ electrically connected to the LAIS; an ATP device electrically connected to the LKJ via a second serial bus; a DMS electrically connected to the BeiDou NTP server via a third serial bus; and a CIR electrically connected to the DMS via a fourth serial bus. The LAIS, LKJ, ATP device, DMS, and CIR are each used to operate according to the corrected clock signal. In this onboard equipment system, the onboard equipment operates based on a more accurate and reliable clock signal corrected by the BeiDou satellite timing signal, thereby improving train operation safety. However, this utility model does not address application scenarios where the timing cycle of the BeiDou satellite navigation system is inconsistent with the clock cycle required by the on-board equipment. Furthermore, while it mentions switching to an atomic clock when the BeiDou signal is lost, atomic clocks themselves have a time-varying nature, and the longer the loss of signal, the greater the accumulated error. Therefore, this solution still fails to solve the technical problem of unstable train status information cycles caused by the existing BeiDou navigation system's time calibration cycle for the train's local system being longer than the train's status information transmission cycle, resulting in a significant deviation from the prescribed train status information cycle after BeiDou system time calibration. Summary of the Invention

[0004] To address the aforementioned problems in the prior art, this invention employs a Model Predictive Control (MPC) algorithm to obtain a precise reference period clock and uses Kalman filtering to reduce the adverse effects on the train's local system's status information transmission period caused by the BeiDou Navigation Satellite System's time calibration period being greater than the train's status information transmission period. Simultaneously, to reduce the adverse effects of the error between the local system's internal time and the BeiDou system's receiving time on the data transmission period, control over the data transmission period is achieved. This invention provides a train operation status spatiotemporal synchronization system based on real-time BeiDou information, including a spatiotemporal synchronization module, an MPC-based time calibration module, a BeiDou receiving module, and a Kalman filtering module. The spatiotemporal synchronization module is connected to the MPC-based time calibration module, the BeiDou receiving module, and the Kalman filtering module, respectively. The Kalman filtering module is connected to both the MPC-based time calibration module and the BeiDou receiving module. The spatiotemporal synchronization module receives processed BeiDou data and matches it with the calibrated time, then... The data is transmitted periodically; the time calibration module based on the MPC algorithm is used to obtain an accurate reference period through the MPC algorithm. Each cycle involves performing a time accumulation operation and recording the accumulated time. The cycle is completed when the condition is met. Data is transmitted at that time; the Beidou receiving module is used to receive data at a period of... The system collects, processes, and records the BeiDou information; the Kalman filter module is used to reduce period instability caused by time calibration, ensuring that the period is no longer affected by sudden disturbances at the calibration time.

[0005] To address the aforementioned technical problems, this invention also provides a method for spatiotemporal synchronization of train operation status based on real-time BeiDou information. This method is implemented using the aforementioned preferred train operation status spatiotemporal synchronization system based on real-time BeiDou information, and includes the following steps: S1: The time calibration module based on the MPC algorithm sets up a local time clock with a rolling optimization mechanism including measurement-prediction-optimization-execution and forms a closed-loop control to continuously correct timing deviations. High-precision acquisition of the cycle; S2: Reception and Processing of BeiDou Information: The BeiDou receiving module receives and transmits information at a period of [time period missing]. The system stores the BeiDou information and converts it into train status information data including time, latitude and longitude, location, and speed. Then, it extracts the train operation status information, including time, location, and speed, and assigns the time value from the train operation status information to a global time variable. This allows for time calibration, eliminating the impact of internal clock errors on the overall time. S3: Kalman filter is used to control the period, which solves the problem of large fluctuations and deviations in the new period after calibration. Periodicity issues, or time intervals that are too large or too small due to calibration, must be addressed to ensure that the overall time interval is controlled within a certain range. Within the period; S4: Spatiotemporal Synchronization and Data Transmission: The spatiotemporal synchronization module receives the processed BeiDou information and matches the BeiDou data with the calibrated time. Then, upon receiving a trigger from the time calibration module based on the MPC algorithm, it transmits the data every [time period missing]. Data is sent once per cycle. When no trigger is received, the spatiotemporal synchronization module is in a blocked state. After receiving a trigger from the time calibration module based on the MPC algorithm, the spatiotemporal synchronization module matches the calibrated time with the train status information and sends data. The sent data includes the train speed and position, as well as the corresponding time, thereby achieving spatiotemporal synchronization of the train's operating status. After sending the data, the data buffer is cleared, and then the receiving and sending process is repeated.

[0006] Further, S1 includes the following steps: S11: Set a period in the time calibration module based on the MPC algorithm. The local time clock, through the MPC algorithm, in each Measure the current timing error at the start of the cycle; S12: First, construct a state vector and prediction model that includes the current error, historical error, and error change rate. Then, based on the state vector and prediction model, predict the error evolution trend for several future periods. This includes the following steps: S121: Define state variables: ; The state variable x ( k ) is used to describe the system state and as the initial condition for prediction, where: For the current actual cycle Compared with the benchmark period The difference represents the current cycle error; This represents the error from the previous cycle. This represents the rate of change of error between the current period and the previous period. The state transition relation is: ; From the above state transition relation (1), we obtain the state transition matrix and the control input matrix, where: It is a 3×3 state transition matrix. To control the input matrix, In the first k The control quantity actively applied in each cycle, i.e., the control quantity applied to the reference cycle. Active adjustment amount, For system perturbations that cannot be controlled in the k-th cycle, this is used to enhance the robustness of the algorithm; The system state equation is then: ; S122: Construct a prediction model, which includes a prediction state sequence, a prediction control sequence, and a prediction equation: The predicted state sequence is: ; In formula (3): For predicting a state sequence, it refers to the state sequence in the current period. k Predicting the state of multiple future cycles; For the first N The predicted state of the step represents the state within the cycle. k For the period k + N State prediction value; The predictive control sequence is: ; In equation (4): For predictive control sequences, it refers to the sequence in the current period. k Control and forecasting for multiple future cycles; For the first N Predictive control of steps, meaning in the cycle k For the period k + N Control predicted values; The prediction equation is: ; Equation (5) represents state evolution = natural response + forced response; In equation (5): To expand the state transition matrix; This indicates the free evolution of the system state without any control input being cleared; To control the influence matrix; This represents the cumulative effect of the control sequence on the system state; ;

[0007] S13: First, solve a constrained optimization problem, and then calculate the optimal control sequence that minimizes the prediction error while satisfying the physical constraints of the periodic adjustment. This includes the following steps: S131: The objective function is constructed as follows: ; In equation (8): The state weight matrix is ​​in the form of: ,Mode middle: Indicates the current error The larger the penalty weight, the more emphasis is placed on clearing the current error; Indicates historical error The penalty weights, which are typically small, are used to connect time series; Represents the rate of change of error The penalty weights are used to suppress oscillations and ensure smoothness; R = r To control the weighting coefficients, R The larger the value, the smaller the control action and the smoother the response; R The smaller the value, the larger the control action and the faster the response; S132: The constraints that need to be met for construction are: ; ; Equation (9) is the control constraint, defining the output range of the controller, i.e., the periodic adjustment amount. Need to Minimum allowable adjustment amount and Between the maximum allowable adjustment amounts; Equation (10) is a control rate constraint, used to limit the rate of change of the control quantity and ensure the smoothness of the control. Indicates the amount of periodic adjustment change; Indicates the maximum permissible rate of change; S133: Perform optimization and solve the problem, define... Based on the current state And the prediction model, calculate the future N Step state prediction value Substitute the above values ​​into the objective function (8) to solve the constrained quadratic programming problem to obtain the optimal solution, and then take the first element of the optimal solution as the control implementation to realize the control of the current cycle; In the next reference cycle, the steps S11, S12 and S13 are repeated in sequence to achieve closed-loop control: through the steps S11, S12, S13 and formulas (1) to (10), the MPC algorithm can predict future trends based on the system model in each reference cycle and calculate the optimal cycle adjustment strategy through constraint optimization, thereby achieving high-precision timing control. S14: Perform cyclic accumulation timing: After obtaining the precise reference period, accumulate the time and then store the accumulated time locally. Periodic time variable and global time variables Global time variables As time information in the transmitted data, it is the final output time; local Periodic time variable This serves as a flag to trigger data transmission locally. Periodic time variable Each time it is satisfied Data transmission is triggered once per cycle, then locally Periodic time variable Re-accumulate after resetting to zero.

[0008] Furthermore, the state weight matrix in S13 Set as .

[0009] Furthermore, the control weight coefficients in S13 R = r Set as R = r=0.001, to ensure that the period is targeted. It can respond quickly while ensuring smoothness.

[0010] Further, S3 includes the following steps: S31: Problem modeling, which includes the following steps: S311: Establish a mathematical model for the time interval, considering that the time interval is related to the change of the current actual interval value and the actual interval value, and use equation (11) to define the state variables and state transition model: First, define the state variable as The time interval between two data transmissions. The change in time interval within a data transmission cycle is defined as the rate of change being considered very slow, and therefore the new change is considered to remain essentially constant. Secondly, since the new time interval equals the old time interval plus the change in time interval within a period, the state transition matrix is: The state transition model is ; S312: Due to the random disturbances that the system may be subject to, process noise is set using equation (12) to describe this uncertainty, and a measurement model and measurement noise are set to measure and evaluate the measurement effect: Define the noise matrix To represent the random disturbances that the system may be subjected to, a reference noise matrix is ​​defined. ,but ; Since this system only measures intervals, then the definition is... Let the measurement matrix be defined, and define the observation noise covariance matrix. , used to control Kalman gain; S32: Kalman filtering procedure, including the following steps: S321: Initialize the state estimation and error covariance matrix using equation (13): The Kalman filter includes a prediction phase and an update phase. In the prediction phase, the filter uses the estimate of the previous state to make an estimate of the current state. In the update phase, the filter uses the observation of the current state to optimize the prediction value obtained in the prediction phase to obtain a more accurate new estimate. The prediction and update process is shown in equation (13): ; S322: Entering the prediction-update loop step: The prediction step is represented by Equation (14), which predicts the current state based on the previous state. The predicted value obtained at this time is based on the state transition model of the system and has a large uncertainty. During the prediction step, within the interval between two BeiDou calibrations, the filter continuously performs predictions at a high frequency. Simultaneously, based on the established state equation, it calculates the current clock error value and the change in time interval, and estimates the uncertainty of the predicted value. Since the predicted value does not consider the current observation value, it should be considered an estimate. The error covariance matrix of the predicted value is calculated using the error covariance matrix from the previous moment and the system noise covariance matrix. The error covariance matrix is ​​then initialized. ; ; The update step is represented by Equation (15), which aims to update the uncertainty estimate by optimally fusing unreliable predictions and noisy measurements after calculating the Kalman filter gain. In the update step, when a new BeiDou calibration signal arrives, the filter does not simply overwrite its own predicted value with this new value. Instead, it compares the new calibration signal with the filter's own predicted value and calculates the current state estimate based on the current observation and predicted values. This current state estimate is a more accurate estimate because it takes into account the current observation values. The error covariance matrix of the current state estimate is obtained through the error covariance matrix calculated in the prediction step. Observation noise covariance matrix and Kalman gain Calculated; ; The final output value of the filter is the optimal linear unbiased estimate after fusing the local crystal oscillator dynamic model and all historical and current observation information. This allows the train's local clock system to achieve both the long-term absolute accuracy of BeiDou time and the short-term smoothness of a highly stable crystal oscillator, thus ensuring the stability of the train status information transmission cycle.

[0011] The working principle of this invention is as follows: To achieve spatiotemporal synchronization of train operation status, this invention first uses the MPC algorithm to decompose complex online optimization calculations into continuous and extremely short time windows of local optimization. Within each cycle, it strictly completes a closed-loop process of "measurement-prediction-solution-execution." At the start of each timing cycle, the system collects current error data and, combined with a built-in dynamic error model, proactively simulates the system behavior for several future cycles. Secondly, it solves a multi-objective optimization problem, generating an optimal compensation command sequence while considering control smoothness and response speed constraints. Finally, it corrects timing parameters in real time by rolling the execution of the first element of the sequence, forming an intelligent control system with predictive compensation capabilities. This effectively suppresses various interference factors and ensures the high accuracy and stability of the timing reference cycle. Simultaneously, this invention employs Kalman filtering and utilizes its unique "prediction-update" mechanism to integrate the short-term stability of the local crystal oscillator with the long-term absolute accuracy of BeiDou time. When a calibration value experiences a significant jump, the filter judges its unreliability based on the model's historical trend, thereby significantly reducing its impact, smoothing out the step disturbance caused by calibration, and outputting a continuous, stable, and optimal estimate.

[0012] Furthermore, for application scenarios where the BeiDou time synchronization period and the required train time are inconsistent, this invention can meet the required train time by adopting a BeiDou time synchronization + local time clock scheme. When BeiDou time synchronization is not received, the train operation status synchronization system accumulates the local time through a time calibration module based on the MPC algorithm. The MPC algorithm is used to eliminate the influence of the local time clock's own error on the time. When BeiDou time synchronization is received, BeiDou time is used as the standard, thereby ensuring time accuracy.

[0013] This invention is applicable to trains operating in a periodic manner. After receiving BeiDou information, according to Applications that send data periodically can ensure that the overall data transmission interval is [missing information]. It can significantly reduce the impact of the error between the local system's internal clock and the actual BeiDou reception time on the data transmission cycle, and can accurately obtain train operation status information such as time, location and speed, achieving a spatiotemporal synchronization effect of train operation status. Attached Figure Description

[0014] Figure 1 This is a schematic diagram of the framework structure of the train operation status spatiotemporal synchronization system based on real-time BeiDou information of the present invention.

[0015] Figure 2 This is a schematic diagram illustrating the workflow of the train operation status spatiotemporal synchronization method based on real-time BeiDou information according to the present invention. Detailed Implementation

[0016] To more clearly illustrate the present invention, the following description, in conjunction with the accompanying drawings, provides further insight. Those skilled in the art should understand that the following detailed description is illustrative rather than restrictive and should not be construed as limiting the scope of protection of the present invention.

[0017] To address the aforementioned technical problems, this invention provides a train operation status spatiotemporal synchronization system based on real-time BeiDou information, such as... Figure 1 and Figure 2 As shown, the system includes a spatiotemporal synchronization module, an MPC-based time calibration module, a BeiDou receiving module, and a Kalman filter module. The spatiotemporal synchronization module is connected to the MPC-based time calibration module, the BeiDou receiving module, and the Kalman filter module, respectively. The Kalman filter module is connected to both the MPC-based time calibration module and the BeiDou receiving module. The spatiotemporal synchronization module receives processed BeiDou data and matches it with the calibrated time, then... The data is transmitted periodically; the time calibration module based on the MPC algorithm is used to obtain an accurate reference period through the MPC algorithm. Each cycle involves performing a time accumulation operation and recording the accumulated time. The cycle is completed when the condition is met. Data is transmitted at that time; the Beidou receiving module is used to receive data at a period of... The system collects, processes, and records the BeiDou information; the Kalman filter module is used to reduce period instability caused by time calibration, ensuring that the period is no longer affected by sudden disturbances at the calibration time.

[0018] To address the aforementioned problems in the prior art, this invention also provides a method for synchronizing train operation status based on real-time BeiDou information, such as... Figure 1 and Figure 2 As shown, the method is implemented through the preferred train operation status spatiotemporal synchronization system based on real-time BeiDou information, and includes the following steps: S1: The time calibration module based on the MPC algorithm sets up a local time clock with a rolling optimization mechanism including measurement-prediction-optimization-execution and forms a closed-loop control to continuously correct timing deviations. High-precision acquisition of the cycle; S2: Reception and Processing of BeiDou Information: The BeiDou receiving module receives and transmits information at a period of [time period missing]. The system stores the BeiDou information and converts it into train status information data including time, latitude and longitude, location, and speed. Then, it extracts the train operation status information, including time, location, and speed, and assigns the time value from the train operation status information to a global time variable. This allows for time calibration, eliminating the impact of internal clock errors on the overall time. S3: Kalman filter is used to control the period, which solves the problem of large fluctuations and deviations in the new period after calibration. Periodicity issues, or time intervals that are too large or too small due to calibration, must be addressed to ensure that the overall time interval is controlled within a certain range. Within the period; S4: Spatiotemporal Synchronization and Data Transmission: The spatiotemporal synchronization module receives the processed BeiDou information and matches the BeiDou data with the calibrated time. Then, upon receiving a trigger from the time calibration module based on the MPC algorithm, it transmits the data every [time period missing]. Data is sent once per cycle. When no trigger is received, the spatiotemporal synchronization module is in a blocked state. After receiving a trigger from the time calibration module based on the MPC algorithm, the spatiotemporal synchronization module matches the calibrated time with the train status information and sends data. The sent data includes the train speed and position, as well as the corresponding time, thereby achieving spatiotemporal synchronization of the train's operating status. After sending the data, the data buffer is cleared, and then the receiving and sending process is repeated.

[0019] Optionally, such as Figure 1 and Figure 2 As shown, S1 includes the following steps: S11: Set a period in the time calibration module based on the MPC algorithm. The local time clock, through the MPC algorithm, in each Measure the current timing error at the start of the cycle; S12: First, construct a state vector and prediction model that includes the current error, historical error, and error change rate. Then, based on the state vector and prediction model, predict the error evolution trend for several future periods. This includes the following steps: S121: Define state variables: ; The state variable x ( k ) is used to describe the system state and as the initial condition for prediction, where: For the current actual cycle Compared with the benchmark period The difference represents the current cycle error; This represents the error from the previous cycle. This represents the rate of change of error between the current period and the previous period. The state transition relation is: ; From the above state transition relation (1), we obtain the state transition matrix and the control input matrix, where: It is a 3×3 state transition matrix. To control the input matrix, In the first k The control quantity actively applied in each cycle, i.e., the control quantity applied to the reference cycle. Active adjustment amount, For system perturbations that cannot be controlled in the k-th cycle, this is used to enhance the robustness of the algorithm; The system state equation is then: ; S122: Construct a prediction model, which includes a prediction state sequence, a prediction control sequence, and a prediction equation: The predicted state sequence is: ; In formula (3): For predicting a state sequence, it refers to the state sequence in the current period. k Predicting the state of multiple future cycles; For the first N The predicted state of the step represents the state within the cycle. k For the period k + N State prediction value; The predictive control sequence is: ; In equation (4): For predictive control sequences, it refers to the sequence in the current period. k Control and forecasting for multiple future cycles; For the first N Predictive control of steps, meaning in the cycle k For the period k + N Control predicted values; The prediction equation is: ; Equation (5) represents state evolution = natural response + forced response; In equation (5): To expand the state transition matrix; This indicates the free evolution of the system state without any control input being cleared; To control the influence matrix; This represents the cumulative effect of the control sequence on the system state; ; ; S13: First, solve a constrained optimization problem, and then calculate the optimal control sequence that minimizes the prediction error while satisfying the physical constraints of the periodic adjustment. This includes the following steps: S131: The objective function is constructed as follows: ; In equation (8): The state weight matrix is ​​in the form of: ,Mode middle: Indicates the current error The larger the penalty weight, the more emphasis is placed on clearing the current error; Indicates historical error The penalty weights, which are typically small, are used to connect time series; Represents the rate of change of error The penalty weights are used to suppress oscillations and ensure smoothness; R = r To control the weighting coefficients, R The larger the value, the smaller the control action and the smoother the response; R The smaller the value, the larger the control action and the faster the response; S132: The constraints that need to be met for construction are: ; ; Equation (9) is the control constraint, defining the output range of the controller, i.e., the periodic adjustment amount. Need to Minimum allowable adjustment amount and Between the maximum allowable adjustment amounts; Equation (10) is a control rate constraint, used to limit the rate of change of the control quantity and ensure the smoothness of the control. Indicates the amount of periodic adjustment change; Indicates the maximum permissible rate of change; S133: Perform optimization and solve the problem, define... Based on the current state And the prediction model, calculate the future N Step state prediction value Substitute the above values ​​into the objective function (8) to solve the constrained quadratic programming problem to obtain the optimal solution, and then take the first element of the optimal solution as the control implementation to realize the control of the current cycle; In the next reference cycle, the steps S11, S12 and S13 are repeated in sequence to achieve closed-loop control. Through the steps S11, S12, S13 and formulas (1) to (10), the MPC algorithm can predict future trends based on the system model in each reference cycle and calculate the optimal cycle adjustment strategy through constraint optimization, thereby achieving high-precision timing control. S14: Perform cyclic accumulation timing: After obtaining the precise reference period, accumulate the time and then store the accumulated time locally. Periodic time variable and global time variables Global time variables As time information in the transmitted data, it is the final output time; local Periodic time variable This serves as a flag to trigger data transmission locally. Periodic time variable Each time it is satisfied Data transmission is triggered once per cycle, then locally Periodic time variable Re-accumulate after resetting to zero.

[0020] Optionally, the state weight matrix in S13 Set as .

[0021] Optionally, the control weighting coefficients in S13 are... R = r Set as R = r =0.001, to ensure that the period is targeted. It can respond quickly while ensuring smoothness.

[0022] Optionally, such as Figure 1 and Figure 2 As shown, step S3 includes the following steps: S31: Problem modeling, which includes the following steps: S311: Establish a mathematical model for the time interval, considering that the time interval is related to the change of the current actual interval value and the actual interval value, and use equation (11) to define the state variables and state transition model: First, define the state variable as The time interval between two data transmissions. The change in time interval within a data transmission cycle is defined as the rate of change being considered very slow, and therefore the new change is considered to remain essentially constant. Secondly, since the new time interval equals the old time interval plus the change in time interval within a period, the state transition matrix is: The state transition model is as follows: ; S312: Due to the random disturbances that the system may be subject to, process noise is set using equation (12) to describe this uncertainty, and a measurement model and measurement noise are set to measure and evaluate the measurement effect: Define the noise matrix To represent the random disturbances that the system may be subjected to, a reference noise matrix is ​​defined. ,but ; Since this system only measures intervals, then the definition is... Let the measurement matrix be defined, and define the observation noise covariance matrix. , used to control Kalman gain; S32: Kalman filtering procedure, including the following steps: S321: Initialize the state estimation and error covariance matrix using equation (13): The Kalman filter includes a prediction phase and an update phase. In the prediction phase, the filter uses the estimate of the previous state to make an estimate of the current state. In the update phase, the filter uses the observation of the current state to optimize the prediction value obtained in the prediction phase to obtain a more accurate new estimate. The prediction and update process is shown in equation (13):

[0023] S322: Entering the prediction-update loop step: The prediction step is represented by Equation (14), which predicts the current state based on the previous state. The predicted value obtained at this time is based on the state transition model of the system and has a large uncertainty. During the prediction step, within the interval between two BeiDou calibrations, the filter continuously performs predictions at a high frequency. Simultaneously, based on the established state equation, it calculates the current clock error value and the change in time interval, and estimates the uncertainty of the predicted value. Since the predicted value does not consider the current observation value, it should be considered an estimate. The error covariance matrix of the predicted value is calculated using the error covariance matrix from the previous moment and the system noise covariance matrix. The error covariance matrix is ​​then initialized. ; ; The update step is represented by Equation (15), which aims to update the uncertainty estimate by optimally fusing unreliable predictions and noisy measurements after calculating the Kalman filter gain. In the update step, when a new BeiDou calibration signal arrives, the filter does not simply overwrite its own predicted value with this new value. Instead, it compares the new calibration signal with the filter's own predicted value and calculates the current state estimate based on the current observation and predicted values. This current state estimate is a more accurate estimate because it takes into account the current observation values. The error covariance matrix of the current state estimate is obtained through the error covariance matrix calculated in the prediction step. Observation noise covariance matrix and Kalman gain Calculated; ; The final output value of the filter is the optimal linear unbiased estimate after fusing the local crystal oscillator dynamic model and all historical and current observation information. This allows the train's local clock system to achieve both the long-term absolute accuracy of BeiDou time and the short-term smoothness of a highly stable crystal oscillator, thus ensuring the stability of the train status information transmission cycle.

[0024] This invention is applicable to trains operating in a periodic manner. After receiving BeiDou information, according to Applications that send data periodically can ensure that the overall data transmission interval is [missing information]. It can significantly reduce the impact of the error between the local system's internal clock and the actual BeiDou reception time on the data transmission cycle, and can accurately obtain train operation status information such as time, location and speed, achieving a spatiotemporal synchronization effect of train operation status.

[0025] The above preferred embodiments of the present invention are merely examples to clearly illustrate the present invention and are not intended to limit the implementation of the present invention. For those skilled in the art, other variations or modifications can be made based on the above description. It is impossible to exhaustively list all the implementation methods here. All obvious variations or modifications derived from the technical solutions of the present invention should still be within the protection scope of the present invention.

Claims

1. A train operation state space-time synchronization system based on real-time Beidou information, characterized in that: The system includes a spatiotemporal synchronization module, an MPC-based time calibration module, a BeiDou receiving module, and a Kalman filter module. The spatiotemporal synchronization module is connected to the MPC-based time calibration module, the BeiDou receiving module, and the Kalman filter module, respectively. The Kalman filter module is connected to both the MPC-based time calibration module and the BeiDou receiving module. The spatiotemporal synchronization module receives processed BeiDou data and matches it with the calibrated time, then... The data is transmitted periodically; the time calibration module based on the MPC algorithm is used to obtain an accurate reference period through the MPC algorithm. Each cycle involves performing a time accumulation operation and recording the accumulated time. The cycle is completed when the condition is met. Data is transmitted at that time; the Beidou receiving module is used to receive data at a period of... The system collects, processes, and records the BeiDou information; the Kalman filter module is used to reduce period instability caused by time calibration, ensuring that the period is no longer affected by sudden disturbances at the calibration time.

2. A train operation state space-time synchronization method based on real-time Beidou information, the method is implemented by the train operation state space-time synchronization system based on real-time Beidou information of claim 1, characterized in that: Comprising the following steps: S1: Time calibration module based on MPC algorithm sets local time clock, including rolling optimization mechanism of measurement-prediction-optimization-execution and forms closed loop control, for continuously correcting timing deviation, realizing high-precision acquisition of period ​ S2: Reception and Processing of BeiDou Information: The BeiDou receiving module receives and transmits information at a period of [time period missing]. The system stores the BeiDou information and converts it into train status information data including time, latitude and longitude, location, and speed. Then, it extracts the train operation status information, including time, location, and speed, and assigns the time value from the train operation status information to a global time variable. This allows for time calibration, eliminating the impact of internal clock errors on the overall time. S3: adopt Kalman filter control period, solve the problem of large fluctuation of new period after calibration and deviation from period, or the problem of too large or too small time interval caused by calibration, to ensure the overall control of time interval within period; S4: Spatiotemporal Synchronization and Data Transmission: The spatiotemporal synchronization module receives the processed BeiDou information and matches the BeiDou data with the calibrated time. Then, upon receiving a trigger from the time calibration module based on the MPC algorithm, it transmits the data every [time period missing]. Data is sent once per cycle. When no trigger is received, the spatiotemporal synchronization module is in a blocked state. After receiving a trigger from the time calibration module based on the MPC algorithm, the spatiotemporal synchronization module matches the calibrated time with the train status information and sends data. The sent data includes the train speed and position, as well as the corresponding time, thereby achieving spatiotemporal synchronization of the train's operating status. After sending the data, the data buffer is cleared, and then the receiving and sending process is repeated.

3. The train operation state space-time synchronization method based on real-time Beidou information according to claim 2, characterized in that: The S1 comprises the following steps: S11: Set a period in the time calibration module based on the MPC algorithm. The local time clock, through the MPC algorithm, in each Measure the current timing error at the start of the cycle; S12: first constructing a state vector containing current error, historical error and error change rate and a prediction model, and then predicting the error evolution trend of future periods based on the state vector and the prediction model, specifically comprising the following steps: S121: Define state variables: ; the state variable x ( k ) for describing the system state and as initial conditions for the prediction, wherein: is the current actual period the difference to the reference period , which represents the current period error; is the previous period error; is the error change rate of the current period to the previous period; The state transition relationship is: ； From the above state transition relation (1), we obtain the state transition matrix and the control input matrix, where: It is a 3×3 state transition matrix. To control the input matrix, In the first k The control quantity actively applied in each cycle, i.e., the control quantity applied to the reference cycle. Active adjustment amount, For system perturbations that cannot be controlled in the k-th cycle, this is used to enhance the robustness of the algorithm; The system state equation is: ； S122: constructing a prediction model, the prediction model comprising a prediction state sequence, a prediction control sequence and a prediction equation: The prediction state sequence is: ； In formula (3): is the predicted state sequence, which means predicting the state of the current period k for a plurality of future periods; is the predicted state of the first N step, which means predicting the state of the period k for the period k + N ; The prediction control sequence is: ； In formula (4): is the predicted control sequence, which means that in the current period k the control is predicted for a number of future periods; is the predicted control of the first N step, which means that in the period k the control is predicted for the period k + N ; The prediction equation is: ； Formula (5) represents state evolution = natural response + forced response; In formula (5): is the extended state transition matrix; denotes the free evolution of the system state in the absence of control input; is the control influence matrix; denotes the cumulative influence of the control sequence on the system state; ; ; S13: first solving a constrained optimization problem, and then calculating the optimal control sequence that can minimize the prediction error under the premise of meeting the physical limit of the periodic adjustment amount, specifically comprising the following steps: S131: constructing a target function: ; In formula (8), is the state weight matrix, in the form of , formula In formula (8), represents the penalty weight of the current error , the larger the more emphasis on eliminating the current error; represents the penalty weight of the historical error , usually small, used to connect the time series; represents the penalty weight of the error change rate , used to suppress oscillation and ensure smoothness; R = r , the greater the value, the smaller the control action, and the smoother the response; R R , the smaller the value, the greater the control action, and the faster the response;​ S132: constructing the constraint conditions required to be met: ; ; Equation (9) is a control quantity constraint, which defines the output range of the controller, i.e. the periodic adjustment quantity need to be within the minimum allowed adjustment quantity and the maximum allowed adjustment quantity; The formula (10) is a control rate of change constraint for limiting the change speed of the control quantity, ensuring the smoothness of the control; represents a cycle adjustment change amount; represents a maximum allowable rate of change; S133: Perform optimization and solve the problem, define... Based on the current state And the prediction model, calculate the future N Step state prediction value Substitute the above values ​​into the objective function (8) to solve the constrained quadratic programming problem to obtain the optimal solution, and then take the first element of the optimal solution as the control implementation to realize the control of the current cycle; In the next reference period, the steps S11, S12 and S13 are repeated in turn to realize closed-loop control: through the steps S11, S12, S13 and formulas (1) to (10), the MPC algorithm can predict future trends based on the system model in each reference period, and calculate the optimal periodic adjustment strategy through constrained optimization, thereby realizing high-precision timing control; S14: carry out cycle accumulation timing: after obtaining the accurate reference period, carry out accumulation of time, after accumulation, store the time into local cycle time variable and global time variable ; global time variable as time information in sending data, is the last output time; local cycle time variable is the flag as the trigger of data sending, local cycle time variable every time cycle is satisfied, data sending is triggered once, after that, local cycle time variable is re-accumulated after being cleared.

4. The train operation state space-time synchronization method based on real-time Beidou information according to claim 3, characterized in that: The state weight matrix in S13 is set as .​ 5. The train operation state space-time synchronization method based on real-time Beidou information according to claim 3, characterized in that: The control weight coefficient in S13 R = r Set as R = r =0.

001.

6. The train operation state space-time synchronization method based on real-time Beidou information according to claim 2, characterized in that: The S3 comprises the following steps: S31: problem modeling, specifically comprising the following steps: S311: establishing a mathematical model for the time interval, considering that the time interval is related to the current actual interval value and the change amount of the actual interval value, and using formula (11) to define the state variable and the state transition model: First, define the state variable as is the time interval between two data transmissions, is the change in time interval within one data transmission cycle, where the change rate is considered to be very slow, so the new change is considered to be substantially constant; Secondly, since the new time interval is equal to the old time interval plus the change of the time interval in a period, the state transition matrix is and the state transition model is ; S312: due to random disturbances that the system may be subjected to, formula (12) is used to set the process noise to describe this uncertainty, and a measurement model and measurement noise are set to measure and evaluate the measurement effect: Define the noise matrix to represent the random disturbances that the system can be subjected to, define a reference noise matrix then ; Since the system only measures the intervals, define the measurement matrix, and define the observation noise covariance matrix for controlling the Kalman gain; S32: Kalman filtering process, comprising the following steps: S321: using formula (13) to initialize state estimation and error covariance matrix: The Kalman filter comprises a prediction stage and a correction stage, in the prediction stage, the filter uses the estimation of the last state to make an estimation of the current state; in the correction stage, the filter optimizes the prediction value obtained in the prediction stage by using the observation value of the current state to obtain a more accurate new estimation value, the prediction and update process is shown in formula (13): ; S322: entering the prediction-update loop step: The prediction step uses formula (14) to represent the prediction of the current state based on the last state, at this time, the prediction value obtained is based on the state transition model of the system, and the uncertainty is large; In the prediction step, the filter constantly makes prediction at high frequency within the interval between two Beidou calibrations, and calculates the clock error value at the current time and the change value of the time interval according to the established state equation, and estimates the uncertainty of the prediction value. The prediction value should be an estimated value because the observation value at the current time is not considered. The error covariance matrix of the prediction value is calculated by the error covariance matrix at the last time and the system noise covariance matrix. The error covariance matrix is initialized as follows ; ; The update step uses formula (15) to represent, the purpose is to update the uncertainty estimation by calculating the Kalman filter gain to optimally fuse the unreliable prediction and the measurement with noise; In the updating step, when a new Beidou calibration signal arrives, the filter does not simply use this new value to cover the self-predicted value, but compares the new calibration signal with the self-predicted value, and calculates a state estimation value at the current time according to the observation value and the predicted value at the current time, the state estimation value at the current time being a more accurate estimation value, because the state estimation value at the current time has considered the observation value at the current time, and the error covariance matrix of the state estimation value at the current time is calculated through the error covariance matrix calculated in the predicting step , an observation noise covariance matrix and a Kalman gain . ; The filter final output value is the optimal linear unbiased estimation value after fusing the local crystal oscillator dynamic model and all historical and current observation information Thus, the train local clock system can obtain the long-term absolute accuracy of Beidou time and maintain the short-term smoothness of high-stability crystal oscillator, so as to ensure the stability of the train state information transmission period.

Images