A data processing method and system based on redundant processors

Abstract

The present application provides a data processing method and system based on redundant processors, which relate to the field of data processing technology. The method synchronizes data to a prediction model through a main processor according to a set period, and the prediction model predicts control instruction data within a future prediction time window based on the synchronized data. Therefore, when the main processor fails, the prediction data of the prediction model is preferentially called by the slave processor until the control right is successfully switched, thereby achieving uninterrupted control of the display terminal by the processor to cope with the scenario where the display terminal has no control signal due to an emergency, and ensuring the reliability of the data displayed on the display terminal.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to a data processing method and system based on redundant processors. Background Art

[0002] In the field of rail transit, on-board display devices usually rely on a single processing unit for data processing and display control. When the processor or software fails, it will cause the display device to fail, which in turn affects the normal operation and driving safety of the vehicle. With the increasing complexity of rail transit systems, the processing power requirements of processors are also getting higher and higher. Modern processor redundant equipment has powerful computing capabilities and can also support multi-tasking and real-time data processing to meet the needs of train operation monitoring, dispatching and command, etc. The data synchronization between the master and slave processors must meet the requirements of high real-time performance, high reliability and dynamic adaptability. However, in the existing technology, during the switching process between the master and slave processors, the control of the display module is often interrupted due to various reasons, which leads to system instability and other safety issues. Summary of the Invention

[0003] The purpose of this application is to provide a data processing method and system based on redundant processors to solve the problems described in the background technology section of this application.

[0004] To achieve the above objectives, this application provides the following technical solutions:

[0005] A first aspect of the present application provides a data processing method based on a redundant processor, comprising:

[0006] The main processor sends first synchronization data to the prediction model in a first synchronization period, wherein the first synchronization data includes control instruction data, system status data, environmental data, and configuration data of the vehicle, wherein the control instruction data is used to control the display terminal of the vehicle;

[0007] Acquiring historical synchronization data, wherein the historical synchronization data includes historical control instruction data;

[0008] The first synchronization data and the historical synchronization data are respectively input into the prediction model as input data and a first prediction control instruction is output. The prediction model is used to predict the control instruction data sent by the main processor within a prediction time window. The first prediction control instruction is used to control the display content of the display terminal within the prediction time window. The first prediction control instruction includes a vehicle speed prediction value, a braking state prediction value, a navigation instruction prediction, and a passenger information prediction. The prediction time window is greater than the sum of the main processor fault detection time, the processor control right switching time, and the control instruction transmission delay time.

[0009] The prediction model sends the generated first prediction control instruction to the slave processor, which is provided with a buffer area for storing the first prediction control instruction. When a failure of the main processor is detected, the slave processor calls the control instruction data corresponding to the current timestamp in the first prediction control instruction from the buffer area to control the display terminal.

[0010] Furthermore, the outputting, according to the first synchronization data, a first prediction control instruction of the main processor to the display terminal within the prediction time window further includes:

[0011] The prediction model receives second synchronization data sent by the main processor in a second synchronization period, and overwrites the control instruction data with the same timestamp in the first prediction control instruction according to the control instruction data in the second synchronization data as overwriting data;

[0012] The prediction time window takes the timestamp of receiving the second synchronization data as a starting point and slides forward one synchronization cycle;

[0013] The prediction model generates new prediction data for the remaining timestamps in the sliding prediction time window except for the timestamp of the coverage data based on the second synchronization data, combines the coverage data and the new prediction data to generate a second prediction control instruction, and sends the second prediction control instruction to the buffer area of ​​the slave processor.

[0014] Furthermore, the method further comprises:

[0015] The main processor outputs the current scene state of the vehicle based on the current control instruction data, the system state data, the environmental data and the configuration data, wherein the scene state includes an emergency braking state, a network high load state and a normal cruising state, and different data synchronization strategies are set for different scene states;

[0016] The master processor and the slave processor perform data synchronization according to the data synchronization strategy corresponding to the current scene state, wherein different data synchronization strategies have different synchronization frequencies and bandwidth ratios for the control instruction data, the system status data, the environment data, and the configuration data.

[0017] Furthermore, based on the current scene state of the vehicle, the first synchronization period of the prediction model is dynamically determined, and the determination method includes:

[0018]

[0019] Among them, T model is the first synchronization period, T model_min is the minimum value of the first synchronization period, Kstate is the adjustment coefficient corresponding to different scene states.

[0020] Furthermore, when the scene state is a normal cruising state, 1.2≤K state ≤1.5; when the scene state is emergency braking state, 0.6≤K state ≤0.8; when the scenario is a high network load state, 1.5≤K state ≤2.0.

[0021] Preferably, the minimum value of the first synchronization period is equal to the value of the prediction time window.

[0022] Furthermore, the slave processor calls the control instruction data corresponding to the current timestamp in the first prediction control instruction from the buffer area to control the display terminal, further comprising:

[0023] When a failure of the master processor is detected, control of the display terminal is switched to the slave processor, and it is determined whether the control is successfully switched: if the switch is successful, control of the display terminal according to the first predictive control instruction is terminated;

[0024] The slave processor controls the display content of the display terminal according to the vehicle data received in real time.

[0025] A second aspect of the present application provides a data processing system based on redundant processors, comprising:

[0026] a data synchronization module, configured for the main processor to send first synchronization data to the prediction model in a first synchronization period, wherein the first synchronization data includes control instruction data, system status data, environmental data, and configuration data of the vehicle, wherein the control instruction data is used to control a display terminal of the vehicle;

[0027] a control instruction prediction module, configured to obtain historical synchronization data, the historical synchronization data including historical control instruction data; input the first synchronization data and the historical synchronization data as input data into the prediction model and output a first prediction control instruction, the prediction model being configured to predict the control instruction data sent by the main processor within a prediction time window; the first prediction control instruction being configured to control the display content of the display terminal within the prediction time window; the first prediction control instruction including a vehicle speed prediction value, a braking state prediction value, a navigation instruction prediction, and a passenger information prediction; wherein the prediction time window is greater than the sum of the main processor fault detection time, the processor control right switching time, and the control instruction transmission delay time;

[0028] A prediction data storage module is used for the prediction model to send the first prediction control instruction generated to the slave processor. The slave processor is provided with a buffer area, and the buffer area is used to store the first prediction control instruction. When a failure of the main processor is detected, the slave processor calls the control instruction data corresponding to the current timestamp in the first prediction control instruction from the buffer area to control the display terminal.

[0029] Furthermore, the control instruction prediction module is also used to:

[0030] The prediction model receives second synchronization data sent by the main processor in a second synchronization period, and overwrites the control instruction data with the same timestamp in the first prediction control instruction according to the control instruction data in the second synchronization data as overwriting data;

[0031] The prediction time window takes the timestamp of receiving the second synchronization data as a starting point and slides forward one synchronization cycle;

[0032] The prediction model generates new prediction data for the remaining timestamps in the sliding prediction time window except for the timestamp of the coverage data based on the second synchronization data, combines the coverage data and the new prediction data to generate a second prediction control instruction, and sends the second prediction control instruction to the buffer area of ​​the slave processor.

[0033] The present application provides the above-mentioned data processing method based on redundant processors, which can at least achieve the following technical effects:

[0034] This application uses a prediction model to predict the control instructions of a set time window and stores them in the buffer area of ​​the slave processor. Therefore, when the main processor fails, the slave processor controls the display of the display terminal by calling the predicted control instructions stored in the buffer until the control is successfully switched. This ensures that the system's control over the display terminal is uninterrupted, and is able to cope with scenarios where the display terminal fails to receive control signals due to emergencies, thereby ensuring the reliability of data display on the display terminal. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the specific implementation methods of the present application or the technical solutions in the prior art, the following is a brief introduction to the drawings required for use in the specific implementation methods or the description of the prior art. Obviously, the drawings described below are some implementation methods of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0036] Figure 1 A flowchart of a data processing method based on redundant processors provided in an embodiment of the present application;

[0037] Figure 2 A schematic diagram of the structure of a data processing system based on redundant processors provided in an embodiment of the present application;

[0038] Figure 3 A schematic diagram of a computer device provided in an embodiment of the present application;

[0039] Figure numerals: 200, a data processing system based on redundant processors; 201, data synchronization module; 202, control instruction prediction module; 203, prediction data storage module; 301, memory; 302, processor. DETAILED DESCRIPTION

[0040] The following will clearly and completely describe the technical solutions of this application in conjunction with the embodiments. Obviously, the embodiments described are only part of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0041] The embodiment of the present application provides a data processing method based on redundant processors, which ensures that when the main processor fails, the slave processor's control of the display terminal is not interrupted, so as to cope with the scenario where the display terminal fails to receive control signals due to an emergency, thereby ensuring the reliability of the data displayed by the display terminal. Figure 1 As shown, the data processing method based on redundant processors provided in this embodiment specifically includes the following steps:

[0042] Step S100: The main processor sends first synchronization data to the prediction model in a first synchronization period, wherein the first synchronization data includes vehicle control instruction data, system status data, environmental data, and configuration data, wherein the control instruction data is used to control the vehicle display terminal;

[0043] Specifically, this embodiment employs redundant processors, with the master and slave processors independently executing the same tasks. This ensures that the slave processor can continue operating in the event of a master processor failure. Both the master and slave processors include a CPU, memory, a hard disk, and peripheral circuitry. They are the core components of the display terminal system, responsible for executing instructions, managing resources, coordinating tasks, and controlling data flow. Regular data synchronization is required between the master and slave processors to ensure that the operating status of the display terminal system remains unchanged when switching between processors.

[0044] Furthermore, the main processor collects various data of the vehicle in real time, including control instruction data, system status data, environmental data and configuration data, wherein the control instruction data is the instruction data sent to the vehicle display terminal. The main processor synchronizes data to the prediction model and the slave processor according to different cycles. The main processor sets up dual channels for data synchronization, model channel: high-frequency synchronization to ensure real-time prediction; slave processor channel: low-frequency synchronization to ensure that emergency data covers the complete switching window. In step S100, the main processor sends the above four types of synchronization data, i.e., the first synchronization data, to the prediction model according to the set first synchronization cycle.

[0045] Furthermore, in step S100, the master processor also sends synchronization data to the slave processor, including:

[0046] Step S101: The main processor outputs the current scene state of the vehicle based on the current control instruction data, the system state data, the environmental data, and the configuration data. The scene state includes an emergency braking state, a high network load state, and a normal cruising state. Different data synchronization strategies are set for different scene states.

[0047] Step S102: The master processor and the slave processor perform data synchronization according to the data synchronization strategy corresponding to the current scene state, wherein different data synchronization strategies have different synchronization frequencies and bandwidth ratios for the control instruction data, the system status data, the environment data, and the configuration data.

[0048] Specifically, the main processor monitors the changing patterns of various vehicle operating data based on real-time collected control command data, system status data, environmental data, and configuration data. It then divides the vehicle's current state into different scenarios, including emergency, high network load, and normal cruising. Based on the current scenario, it dynamically adjusts the synchronization priority (including synchronization frequency and bandwidth usage) of various data types when synchronizing data to the slave processors. Different judgment conditions are set for different scenarios. Once the vehicle is determined to be in a certain state, data synchronization is performed according to the preset data synchronization strategy for that state.

[0049] Preferably, the scene state determination conditions set in this embodiment are shown in Table 1 below:

[0050] Table 1

[0051]

[0052] Preferably, various data synchronization strategies under different scenarios are:

[0053] 1. Emergency braking state: Control command data is synchronized at a frequency of 200 Hz, accounting for 70% of the bandwidth; system status data is synchronized at a frequency of 50 Hz, accounting for 15% of the bandwidth; environmental data is synchronized at a frequency of 20 Hz, accounting for 10% of the bandwidth; configuration data synchronization is suspended. In an emergency braking state, priority is given to ensuring the real-time and accuracy of critical control commands.

[0054] 2. Under high network load conditions: The synchronization frequency for control command data is 50Hz to 30Hz, accounting for 50% to 30% of the bandwidth; the synchronization frequency for system status data is 20Hz to 10Hz, accounting for 20% to 10% of the bandwidth; the synchronization frequency for environmental data is 20Hz to 10Hz, accounting for 20% to 10% of the bandwidth; and the synchronization frequency for configuration data is 1Hz to 0.5Hz, accounting for 5% to 2% of the bandwidth. Under high network load conditions, data transmission volume is reduced to optimize bandwidth usage while maintaining the availability of critical functions.

[0055] 3. Normal cruise mode: Control command data is synchronized at a 50Hz frequency, accounting for 40% of the bandwidth; system status data is synchronized at a 20Hz frequency, accounting for 20% of the bandwidth; environmental data is synchronized at a 10Hz frequency, accounting for 25% of the bandwidth; and configuration data is synchronized at a 2Hz frequency, accounting for 10% of the bandwidth. In normal cruise mode, data integrity is maximized to provide a complete display experience.

[0056] Through the data synchronization strategy corresponding to the scenario status described above, the changing patterns of various types of train operation data are monitored in real time, different scenario states are divided, and the synchronization priority of various types of data between the master and slave processors is dynamically adjusted in combination with the current scenario, realizing multi-rate data integration and dynamic scenario adaptation, and improving the high reliability of the rail transit control system.

[0057] Furthermore, after the main processor determines the current scene state of the vehicle, it determines the data synchronization frequency sent to the prediction model based on the current scene state, i.e., the first cycle. The data synchronization cycle of the prediction model is dynamically adjusted by determining the current scene state, and the prediction duration and prediction frequency of the prediction model are flexibly adjusted according to the control requirements, real-time requirements, and network resource conditions in different scenarios, thereby achieving the best data prediction strategy. The adjustment method of the first synchronization cycle based on the current scene state is expressed as the following formula (1):

[0058] (1)

[0059] In formula (1), T model is the first synchronization period, T model_min is the minimum value of the first synchronization period, K state is the adjustment coefficient corresponding to different scene states.

[0060] Optionally, when the scene state is a normal cruising state, 1.2≤K state ≤1.5; when the scene state is emergency braking state, 0.6≤K state ≤0.8; when the scenario is a high network load state, 1.5≤K state ≤2.0.

[0061] Preferably, T model_min The value is based on the fault detection cycle by the main processor , the control switching time between the master processor and the slave processor and control command delay time Joint decision, and T model_min is greater than the sum of the above three, as shown in the following formula (2).

[0062] (2)

[0063] In formula (2), is the safety factor, To reserve margin.

[0064] Step S200: Acquire historical synchronization data, where the historical synchronization data includes historical control instruction data;

[0065] Step S300: Input the first synchronization data and the historical synchronization data as input data to the prediction model and output a first prediction control instruction. The prediction model is used to predict the control instruction data sent by the main processor within a prediction time window. The first prediction control instruction is used to control the display content of the display terminal within the prediction time window. The first prediction control instruction includes a vehicle speed prediction value, a braking state prediction value, a navigation instruction prediction, and a passenger information prediction. The prediction time window is greater than the sum of the main processor fault detection time, the processor control right switching time, and the control instruction transmission delay time.

[0066] Specifically, after receiving the first synchronization data during the first synchronization period, the prediction model inputs the first synchronization data into the prediction model for data training, and outputs a first prediction control instruction for the processor to send to the display terminal within a future period, i.e., a prediction time window. The prediction model is a neural network model, trained using historical synchronization data as a training dataset, and the received first synchronization data is stored in the historical synchronization data. The first prediction control instruction includes a predicted vehicle speed value, a predicted braking state value, a navigation instruction prediction, and a passenger information prediction.

[0067] Preferably, T model_minThe value of is equal to the value of the prediction time window. The prediction time window / first synchronization period is greater than the sum of the main processor fault detection time, the processor control switching time and the control instruction transmission delay time. Among them, the fault detection time depends on the main processor's heartbeat interval and confirmation logic, the control switching time is determined by the hardware switching speed and software initialization efficiency, and the instruction transmission delay is affected by the bus type and protocol overhead. By integrating the three and reserving a safety margin, the synchronization period and prediction time window of the prediction model are designed to ensure that the rail transit system can still switch seamlessly when the main processor fails, thereby ensuring driving safety and passenger experience.

[0068] Furthermore, step S300 further includes:

[0069] Step S301: The prediction model receives second synchronization data sent by the main processor in a second synchronization period, and overwrites the control instruction data with the same timestamp in the first predicted control instruction according to the control instruction data in the second synchronization data as overwritten data;

[0070] Step S302: The prediction time window starts from the timestamp of receiving the second synchronization data and slides forward by one synchronization period.

[0071] Step S303: The prediction model generates new prediction data for the remaining timestamps in the sliding prediction time window except for the timestamp of the coverage data based on the second synchronization data, combines the coverage data and the new prediction data to generate a second prediction control instruction, and sends the second prediction control instruction to the buffer area of ​​the slave processor.

[0072] Specifically, after the prediction time window slides through one cycle, during the second synchronization cycle, the prediction model receives the second synchronization data and directly replaces the predicted values ​​with the same timestamps in the first prediction control instructions with the actual control instruction values ​​(vehicle speed prediction value, braking status prediction value, navigation instruction prediction, and passenger information prediction) in the second synchronization data. That is, the predicted values ​​with corresponding timestamps in the old predictions are replaced by actual data, which no longer relies on model predictions but instead comes from real input from the main processor. Based on the second synchronization data, the prediction model only predicts the time periods in the sliding prediction time window that have not yet been covered by actual data, generates new prediction values, and retains the actual data and the new prediction values ​​as the second prediction control instructions. By overwriting the old predictions with new data, the actual data takes precedence over the predicted values, avoiding error accumulation. Furthermore, the predictions are only made for time periods that have not yet occurred, reducing the amount of computation. Even if the old predictions have deviations, they can be corrected immediately using actual data.

[0073] Step S400: The prediction model sends the generated first prediction control instruction to the slave processor. The slave processor is provided with a buffer area, and the buffer area is used to store the first prediction control instruction. When a failure of the main processor is detected, the slave processor calls the control instruction data corresponding to the current timestamp in the first prediction control instruction from the buffer area to control the display terminal.

[0074] Specifically, the prediction control instructions generated by the prediction model are sent and stored in the buffer area of ​​the slave processor. When the master processor is normal, the slave processor continues to receive prediction data but does not actively output it, only keeping the buffer area updated. When the control right switch is triggered, the slave processor extracts the prediction data corresponding to the current timestamp from the buffer for display control to cope with the scenario where the display terminal does not receive the control signal due to an emergency, ensuring the reliability of the data display. When the control right is successfully switched, the slave processor directly receives data from sensors and other data, and sends the synchronization data to the prediction model according to the original synchronization cycle.

[0075] Furthermore, in step S400, the slave processor calls the control instruction data corresponding to the current timestamp in the first prediction control instruction from the buffer area to control the display terminal, further comprising:

[0076] Step S401: When a failure of the master processor is detected, control of the display terminal is switched to the slave processor, and it is determined whether the control is successfully switched; if so, control of the display terminal according to the first predictive control instruction is terminated;

[0077] Step S402: The slave processor controls the display content of the display terminal according to the vehicle data received in real time.

[0078] Specifically, the method for detecting faults in the main processor includes: the main processor periodically sends heartbeat signals (e.g., every 5ms), and the slave processor monitors their continuity; sensors monitor the main processor's hardware status, such as temperature, voltage, and memory errors, in real time; and diagnostic programs (such as memory self-tests and task response delay statistics) are regularly run. Fault determination conditions are set. If the main processor determines a fault, control is switched while the slave processor extracts predicted instruction data from the buffer area to control the display terminal until the slave processor takes control. The control switching process includes hardware switching, software initialization, and data reading (reading the latest vehicle data).

[0079] Furthermore, the display terminal supports full-screen display of the control screen of the main processor or the slave processor, or independent display of the screens of the main processor and the slave processor respectively. By flexibly adjusting the display mode, the display terminal can cope with fault switching and provide redundant display, thereby enhancing the fault tolerance of the equipment.

[0080] Furthermore, by adopting a distributed computing architecture, the edge nodes and the cloud collaborate to perform data consistency checks on the control instruction prediction data of the prediction model, the display data of the display terminal, and the control instruction data sent to the display terminal in real time by the main processor. This ensures that the "control instruction prediction results of the prediction model," the "real-time control instructions of the main processor," and the "final data displayed by the display terminal" are logically consistent or verifiably consistent, thereby improving the fault tolerance and robustness of the system. In this embodiment, the prediction model preferably runs on the edge node (such as the internal slave processor or dedicated inference unit of the vehicle), and the cloud holds complete historical data for verifying model accuracy, training, updating, and further analyzing the prediction results.

[0081] Furthermore, the consistency check results include: main processor control failure: using the prediction data of the prediction model + the control instructions that have passed the verification; display terminal abnormality: alarm, lock the current status, and record logs for cloud tracing.

[0082] In addition, the present application also provides a redundant processor-based data processing system 200, as described in the following embodiments. Since the principle of solving problems in a redundant processor-based data processing system is similar to that of a redundant processor-based data processing method, the implementation of a redundant processor-based data processing system can refer to the implementation of a redundant processor-based data processing method, and the repeated parts are not repeated here. As used below, the terms "unit" or "module" can be a combination of software and / or hardware that implements a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, implementation in hardware, or a combination of software and hardware, is also possible and conceivable.

[0083] Figure 2 This is a structural block diagram of a data processing system 200 based on redundant processors according to an embodiment of the present application. Figure 2 Shown, including:

[0084] A data synchronization module 201 is configured to cause the main processor to send first synchronization data to the prediction model in a first synchronization period, wherein the first synchronization data includes vehicle control instruction data, system status data, environmental data, and configuration data, wherein the control instruction data is used to control the vehicle display terminal;

[0085] A control instruction prediction module 202 is configured to obtain historical synchronization data, the historical synchronization data including historical control instruction data; input the first synchronization data and the historical synchronization data as input data into the prediction model, and output a first prediction control instruction. The prediction model is configured to predict the control instruction data sent by the main processor within a prediction time window. The first prediction control instruction is configured to control the display content of the display terminal within the prediction time window. The first prediction control instruction includes a vehicle speed prediction value, a braking state prediction value, a navigation instruction prediction, and a passenger information prediction. The prediction time window is greater than the sum of the main processor fault detection time, the processor control right switching time, and the control instruction transmission delay time.

[0086] The prediction data storage module 203 is used for the prediction model to send the first prediction control instruction generated to the slave processor. The slave processor is provided with a buffer area, and the buffer area is used to store the first prediction control instruction. When a failure of the main processor is detected, the slave processor calls the control instruction data corresponding to the current timestamp in the first prediction control instruction from the buffer area to control the display terminal.

[0087] Furthermore, the control instruction prediction module 202 is further configured to:

[0088] The prediction model receives second synchronization data sent by the main processor in a second synchronization period, and overwrites the control instruction data with the same timestamp in the first prediction control instruction according to the control instruction data in the second synchronization data as overwriting data;

[0089] The prediction time window takes the timestamp of receiving the second synchronization data as a starting point and slides forward one synchronization cycle;

[0090] The prediction model generates new prediction data for the remaining timestamps in the sliding prediction time window except for the timestamp of the coverage data based on the second synchronization data, combines the coverage data and the new prediction data to generate a second prediction control instruction, and sends the second prediction control instruction to the buffer area of ​​the slave processor.

[0091] Furthermore, the data synchronization module 201 is further configured to:

[0092] The main processor outputs the current scene state of the vehicle based on the current control instruction data, the system state data, the environmental data and the configuration data, wherein the scene state includes an emergency braking state, a network high load state and a normal cruising state, and different data synchronization strategies are set for different scene states;

[0093] The master processor and the slave processor perform data synchronization according to the data synchronization strategy corresponding to the current scene state, wherein different data synchronization strategies have different synchronization frequencies and bandwidth ratios for the control instruction data, the system status data, the environment data, and the configuration data.

[0094] Furthermore, the data synchronization module 201 is further configured to dynamically determine a first synchronization period of the prediction model based on the current scene state of the vehicle, wherein the determination method includes:

[0095]

[0096] Among them, T model is the first synchronization period, T model_min is the minimum value of the first synchronization period, K state is the adjustment coefficient corresponding to different scene states.

[0097] Furthermore, the data synchronization module 201 is further configured to: when the scene state is the normal cruise state, 1.2≤K state ≤1.5; when the scene state is emergency braking state, 0.6≤K state ≤0.8; when the scenario is a high network load state, 1.5≤K state ≤2.0.

[0098] Furthermore, in the data synchronization module 201, the minimum value of the first synchronization period is equal to the value of the prediction time window.

[0099] Furthermore, the prediction data storage module 203 is further configured to:

[0100] When a failure of the master processor is detected, control of the display terminal is switched to the slave processor, and it is determined whether the control is successfully switched: if the switch is successful, control of the display terminal according to the first predictive control instruction is terminated;

[0101] The slave processor controls the display content of the display terminal according to the vehicle data received in real time.

[0102] In this embodiment, a computer device is also provided, such as Figure 3 As shown, it includes a memory 301, a processor 302 and a computer program stored in the memory 301 and executable on the processor 302. When the processor 302 executes the computer program, any of the above-mentioned data processing methods based on redundant processors is implemented.

[0103] Specifically, the computer device may be a computer terminal, a server or a similar computing device.

[0104] In this embodiment, a computer-readable storage medium is further provided, wherein the computer-readable storage medium stores a computer program for executing any of the above redundant processor-based data processing methods.

[0105] Specifically, computer-readable storage media include permanent and non-permanent, removable and non-removable media that can be used to store information by any method or technology. The information can be computer-readable instructions, data structures, program modules or other data. Examples of computer-readable storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer-readable storage media does not include temporary computer-readable media (transitory media), such as modulated data signals and carrier waves.

[0106] The embodiments of the present invention achieve the following technical effects:

[0107] 1. This application uses a prediction model to predict control instructions within a set time window and stores them in a buffer area of ​​a slave processor. Therefore, when a master processor fails, the slave processor controls the display terminal by calling the predicted control instructions stored in the buffer until control is successfully switched. This ensures uninterrupted control of the display terminal by the system, copes with scenarios where the display terminal fails to receive control signals due to emergencies, and ensures the reliability of data display on the display terminal.

[0108] 2. This application ensures that the rail transit system can seamlessly switch when the main processor fails, thereby ensuring driving safety and passenger experience by rationally designing the synchronization period and prediction time window of the prediction model;

[0109] 3. This application solves the difficult problems of multi-rate data integration and dynamic scene adaptation by real-time monitoring of the changing patterns of various types of train operation data, dividing different scene states, and dynamically adjusting the synchronization priority of various types of data between the master and slave processors based on the current scene. It is suitable for high-reliability rail transit control systems.

[0110] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A data processing method based on redundant processors, characterized in that: include: The main processor sends first synchronization data to the prediction model in a first synchronization period, wherein the first synchronization data includes control instruction data, system status data, environmental data, and configuration data of the vehicle, wherein the control instruction data is used to control the display terminal of the vehicle; Acquiring historical synchronization data, wherein the historical synchronization data includes historical control instruction data; The first synchronization data and the historical synchronization data are respectively input into the prediction model as input data and a first prediction control instruction is output. The prediction model is used to predict the control instruction data sent by the main processor within a prediction time window. The first prediction control instruction is used to control the display content of the display terminal within the prediction time window. The first prediction control instruction includes a vehicle speed prediction value, a braking state prediction value, a navigation instruction prediction, and a passenger information prediction. The prediction time window is greater than the sum of the main processor fault detection time, the processor control right switching time, and the control instruction transmission delay time. The prediction model sends the generated first prediction control instruction to the slave processor, which is provided with a buffer area for storing the first prediction control instruction. When a failure of the main processor is detected, the slave processor calls the control instruction data corresponding to the current timestamp in the first prediction control instruction from the buffer area to control the display terminal.

2. The data processing method based on redundant processors according to claim 1, characterized in that: Also includes: The prediction model receives second synchronization data sent by the main processor in a second synchronization period, and overwrites the control instruction data with the same timestamp in the first prediction control instruction according to the control instruction data in the second synchronization data as overwriting data; The prediction time window takes the timestamp of receiving the second synchronization data as a starting point and slides forward one synchronization cycle; The prediction model generates new prediction data for the remaining timestamps in the sliding prediction time window except for the timestamp of the coverage data based on the second synchronization data, combines the coverage data and the new prediction data to generate a second prediction control instruction, and sends the second prediction control instruction to the buffer area of ​​the slave processor.

3. The data processing method based on redundant processors according to claim 1, characterized in that: Also includes: The main processor outputs the current scene state of the vehicle based on the current control instruction data, the system state data, the environmental data and the configuration data, wherein the scene state includes an emergency braking state, a network high load state and a normal cruising state, and different data synchronization strategies are set for different scene states; The master processor and the slave processor perform data synchronization according to the data synchronization strategy corresponding to the current scene state, wherein different data synchronization strategies have different synchronization frequencies and bandwidth ratios for the control instruction data, the system status data, the environment data, and the configuration data.

4. The data processing method based on redundant processors according to claim 3, characterized in that: Also includes: Based on the current scene state of the vehicle, a first synchronization period of the prediction model is dynamically determined.

5. The data processing method based on redundant processors according to claim 4, characterized in that: The adjustment method of the first synchronization period based on the scene state is expressed as follows: Among them, T model is the first synchronization period, T model_min is the minimum value of the first synchronization period, K state is the adjustment coefficient corresponding to different scene states.

6. The data processing method based on redundant processors according to claim 5, characterized in that: When the scene state is normal cruising state, 1.2≤K state ≤1.5; when the scene state is emergency braking state, 0.6≤K state ≤0.8; when the scenario is a high network load state, 1.5≤K state ≤2.

0.

7. The data processing method based on redundant processors according to claim 1, characterized in that: The minimum value of the first synchronization period is equal to the value of the prediction time window.

8. The data processing method based on redundant processors according to claim 1, characterized in that: The slave processor calls the control instruction data corresponding to the current timestamp in the first prediction control instruction from the buffer area to control the display terminal, further comprising: When a failure of the main processor is detected, the control of the display terminal is switched to the slave processor, and it is determined whether the control is successfully switched: if the switch is successful, the control of the display terminal by the slave processor according to the first predictive control instruction is terminated, and the slave processor controls the display content of the display terminal according to the vehicle data received in real time.

9. A data processing system based on redundant processors, characterized in that: include: a data synchronization module, configured for the main processor to send first synchronization data to the prediction model in a first synchronization period, wherein the first synchronization data includes control instruction data, system status data, environmental data, and configuration data of the vehicle, wherein the control instruction data is used to control a display terminal of the vehicle; a control instruction prediction module, configured to obtain historical synchronization data, the historical synchronization data including historical control instruction data; input the first synchronization data and the historical synchronization data as input data into the prediction model and output a first prediction control instruction, the prediction model being configured to predict the control instruction data sent by the main processor within a prediction time window; the first prediction control instruction being configured to control the display content of the display terminal within the prediction time window; the first prediction control instruction including a vehicle speed prediction value, a braking state prediction value, a navigation instruction prediction, and a passenger information prediction; wherein the prediction time window is greater than the sum of the main processor fault detection time, the processor control right switching time, and the control instruction transmission delay time; A prediction data storage module is used for the prediction model to send the first prediction control instruction generated to the slave processor. The slave processor is provided with a buffer area, and the buffer area is used to store the first prediction control instruction. When a failure of the main processor is detected, the slave processor calls the control instruction data corresponding to the current timestamp in the first prediction control instruction from the buffer area to control the display terminal.

10. The data processing system based on redundant processors according to claim 9, characterized in that: The control instruction prediction module is further used for: The prediction model receives second synchronization data sent by the main processor in a second synchronization period, and overwrites the control instruction data with the same timestamp in the first prediction control instruction according to the control instruction data in the second synchronization data as overwriting data; The prediction time window takes the timestamp of receiving the second synchronization data as a starting point and slides forward one synchronization cycle; The prediction model generates new prediction data for the remaining timestamps in the sliding prediction time window except for the timestamp of the coverage data based on the second synchronization data, combines the coverage data and the new prediction data to generate a second prediction control instruction, and sends the second prediction control instruction to the buffer area of ​​the slave processor.

Images