A triple redundant data synchronization method and system based on dynamic priority weighting

Abstract

The present invention proposes a triple redundant data synchronization method and system based on dynamic priority weighting. The method includes: collecting multi-source data; calculating the health index of each processor module based on the multi-source data; calculating the dynamic weight of each processor module based on the health index of each processor module, and normalizing the dynamic weight; calculating the weighted support of the input value based on the processor module input value and the normalized dynamic weight; voting to select the input value based on the weighted support as the voting output of the three-series processor module; if the normalized dynamic weight of a processor module is lower than a predefined threshold for a preset number of consecutive cycles, the processor module is automatically isolated. The solution proposed by the present invention can dynamically adjust the voting weight of the module by real-time evaluation of its health, and improve the system fault tolerance and resource utilization by combining weighted voting with fault pre-isolation mechanism.

Description

Technical Field

[0001] The present invention belongs to the field of automation, and in particular relates to a triple redundant data synchronization method and system based on dynamic priority weighting. Background Art

[0002] Triple Modular Redundancy (TMR) is a technology that improves system reliability and fault tolerance through redundant design. It is widely used in safety-critical fields such as aerospace, nuclear power, and rail transportation. A triple modular redundancy system consists of three identical functional modules running in parallel, each independently processing inputs and generating outputs. The outputs of the three modules are compared in real time using a 2-out-of-3 majority voting principle. The core concept is to use a majority voting mechanism to shield single faults and ensure continuous system operation. In a triple modular redundancy system (TMR), the synchronous voting mechanism is the core component that ensures that the outputs of the three redundant modules are correctly compared and the final result is generated. Its design directly affects the system's fault tolerance, real-time performance, and reliability.

[0003] The existing TMR synchronous voting mechanism is as follows: First, the three input modules receive the same input data, which is strictly synchronized using hardware clocks or software protocols. Then, three independent processor modules independently process the input and generate output results. Finally, the output results are sent to a voter, which compares the three results. If the three are completely consistent, the result is directly output. If two of them are consistent but one is different, the majority result is selected and the faulty module is marked. If all three are different, error handling is triggered, such as a system reset or switching to a backup module. The marked faulty module can be isolated, restarted, or repaired online, while the system continues to operate based on the remaining modules.

[0004] Triple Multi-Mode Redundancy (TMR) uses three independent modules to perform the same task and vote on the output, theoretically tolerating a single module failure. Traditional synchronous voting methods have the following drawbacks:

[0005] Existing safety control systems, such as those for gas turbine speed control, reactor temperature monitoring, and rail transit signaling systems, are subject to electromagnetic interference, vibration, and high temperatures at industrial sites, leading to transient module failures or data drift. Traditional TMRs employ simple majority voting, such as a two-out-of-three vote. If two modules simultaneously fail due to transient interference, the system will adopt the incorrect result. Furthermore, the static redundancy mode of traditional TMRs fails to account for the real-time health status of modules (such as temperature and voltage fluctuations). Faulty modules still have full voting power, potentially impacting system reliability over the long term. Summary of the Invention

[0006] In order to solve the above technical problems, the present invention proposes a technical solution of a triple redundant data synchronization method based on dynamic priority weighting to solve the above technical problems.

[0007] The first aspect of the present invention discloses a triple redundant data synchronization method based on dynamic priority weighting, including three series processor modules A, B and C, and the method includes:

[0008] Step S1: Collect multi-source data from the three processor modules A, B, and C in each cycle; the multi-source data includes: processor module input values, environmental sensor data, historical error records, and mean time between failures;

[0009] Step S2: Calculate the health index of each processor module based on the environmental sensor data, historical error records, and mean time between failures; the health index includes: logic error rate, environmental immunity, and historical reliability;

[0010] Step S3: Calculate the dynamic weight of each processor module according to the health index of each processor module, and normalize the dynamic weight;

[0011] Step S4: Calculate the weighted support of the input value based on the processor module input value and the normalized dynamic weight; and select the input value by voting based on the weighted support as the voting output of the three-series processor module;

[0012] Step S5: If the normalized dynamic weight of a processor module is lower than a predefined threshold for a preset number of consecutive cycles, an alarm will be triggered, the processor module will be automatically isolated and hot backup will be started.

[0013] According to the method of the first aspect of the present invention, in step S1,

[0014] The processor module input value is an industrial parameter;

[0015] The environmental sensor data includes: temperature, voltage and vibration amplitude;

[0016] The historical error record includes: the number of CRC check failures in a predefined sub-synchronization period of the processor module;

[0017] The mean time between failures includes: the mean time between failures extracted from the log.

[0018] According to the method of the first aspect of the present invention, in step S2, calculating the health index of each processor module based on the environmental sensor data, historical error records, and mean time between failures includes:

[0019] The logic error rate is calculated according to the number of CRC check failures, specifically:

[0020]

[0021] in, Indicates the i The logic error rate of each processor module; Indicates the i Number of CRC check failures for each processor module; Indicates the predefined number of synchronization cycles; ;

[0022] Calculate the environmental immunity based on the temperature, voltage and vibration amplitude. The specific formula is:

[0023]

[0024] in, Indicates the i Environmental immunity of each processor module; Indicates the i The temperature of each processor module; Indicates the i The voltage of each processor module; Indicates the i The vibration amplitude of each processor module; represents the temperature weight; represents the voltage weight; represents the vibration amplitude weight; and Indicates the maximum operating temperature and set operating temperature; and Indicates the maximum operating voltage and rated operating voltage; Indicates the maximum working vibration amplitude;

[0025] The historical reliability is calculated based on the mean time between failures. The specific formula is:

[0026]

[0027] in, Indicates the i Historical reliability of each processor module; Indicates the i MTBF per processor module; Indicates the mean time between failures of processor module A; Indicates the mean time between failures of the B processor module; Indicates the mean time between failures of the C processor module.

[0028] According to the method of the first aspect of the present invention, in step S3, calculating the dynamic weight of each processor module according to the health index of each processor module includes:

[0029]

[0030] in, Indicates the i Dynamic weights of processor modules; Indicates the i The logic error rate of each processor module; Indicates the i Environmental immunity of each processor module; Indicates the i Historical reliability of each processor module; 、 and Represents the weights of logic error rate, environmental immunity, and historical reliability, α+β+γ=1, where α, β, and γ are adjustable coefficients; .

[0031] According to the method of the first aspect of the present invention, in step S3, normalizing the dynamic weight includes:

[0032]

[0033] in, Indicates the i The normalized dynamic weight of each processor module; Indicates the dynamic weight of the A processor module; Indicates the dynamic weight of the B processor module; Indicates the dynamic weight of the C processor module.

[0034] According to the method of the first aspect of the present invention, in step S4, calculating the weighted support of the input value based on the processor module input value and the normalized dynamic weight includes:

[0035]

[0036] in, Represents the weighted support of the input value; Indicates the i The normalized dynamic weight of each processor module; Indicates the i The input value of each processor module; Represents the input value of a processor module.

[0037] According to the method of the first aspect of the present invention, in step S4, the step of voting and selecting an input value according to the weighted support as the voting output of the three-series processor module includes:

[0038] The input value with the largest weighted support is selected as the voting output.

[0039] A second aspect of the present invention discloses a triple redundant data synchronization system based on dynamic priority weighting, the system comprising:

[0040] The first processing module is configured to collect multi-source data from the three processor modules A, B, and C in each cycle; the multi-source data includes: processor module input values, environmental sensor data, historical error records, and mean time between failures;

[0041] A second processing module is configured to calculate a health indicator of each processor module based on the environmental sensor data, historical error records, and mean time between failures; the health indicator includes: logic error rate, environmental noise immunity, and historical reliability;

[0042] A third processing module is configured to calculate a dynamic weight of each processor module according to the health index of each processor module and normalize the dynamic weight;

[0043] a fourth processing module configured to calculate a weighted support of the input value based on the processor module input value and the normalized dynamic weight; and select the input value by voting based on the weighted support as the voting output of the three-series processor module;

[0044] The fifth processing module is configured to trigger an alarm, automatically isolate the processor module and start hot backup if the normalized dynamic weight of a processor module is lower than a predefined threshold for a preset number of consecutive cycles.

[0045] A third aspect of the present invention discloses an electronic device. The electronic device includes a memory and a processor. The memory stores a computer program. When the processor executes the computer program, it implements the steps of any one of the triple redundant data synchronization methods based on dynamic priority weighting described in the first aspect of the present disclosure.

[0046] A fourth aspect of the present invention discloses a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of any one of the methods for triple redundant data synchronization based on dynamic priority weighting in the first aspect of the present disclosure.

[0047] In summary, the solution proposed in the present invention can dynamically adjust the voting weight of a module by real-time evaluation of its health, and combine weighted voting with a fault pre-isolation mechanism to improve the system's fault tolerance and resource utilization. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the specific embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0049] Figure 1 Flowchart of a triple redundant data synchronization method based on dynamic priority weighting according to an embodiment of the present invention;

[0050] Figure 2 A structural diagram of a triple redundant data synchronization system based on dynamic priority weighting according to an embodiment of the present invention;

[0051] Figure 3 FIG. 4 is a structural diagram of an electronic device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0052] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0053] The first aspect of the present invention discloses a triple redundant data synchronization method based on dynamic priority weighting. Figure 1 FIG. 1 is a flow chart of a triple redundant data synchronization method based on dynamic priority weighting according to an embodiment of the present invention, as shown in FIG. Figure 1 As shown, the method includes:

[0054] Step S1: Collect multi-source data from the three processor modules A, B, and C in each cycle; the multi-source data includes: processor module input values, environmental sensor data, historical error records, and mean time between failures;

[0055] Step S2: Calculate the health index of each processor module based on the environmental sensor data, historical error records, and mean time between failures; the health index includes: logic error rate, environmental immunity, and historical reliability;

[0056] Step S3: Calculate the dynamic weight of each processor module according to the health index of each processor module, and normalize the dynamic weight;

[0057] Step S4: Calculate the weighted support of the input value based on the processor module input value and the normalized dynamic weight; and select the input value by voting based on the weighted support as the voting output of the three-series processor module;

[0058] Step S5: If the normalized dynamic weight of a processor module is lower than a predefined threshold for a preset number of consecutive cycles, an alarm will be triggered, the processor module will be automatically isolated and hot backup will be started.

[0059] In step S1, multi-source data is collected for the three series processor modules A, B and C in each cycle; the multi-source data includes: processor module input values, environmental sensor data, historical error records and mean time between failures.

[0060] In some embodiments, in step S1, the processor module input value is an industrial parameter, (such as temperature, pressure and other industrial parameters);

[0061] The environmental sensor data includes: temperature, voltage and vibration amplitude (each module is monitored independently);

[0062] The historical error record includes: the number of CRC check failures in a predefined sub-synchronization period of the processor module;

[0063] The mean time between failures includes: the mean time between failures extracted from the log.

[0064] In step S2, the health index of each processor module is calculated based on the environmental sensor data, historical error records and mean time between failures; the health index includes: logic error rate, environmental immunity and historical reliability.

[0065] In some embodiments, in step S2, calculating the health index of each processor module based on the environmental sensor data, historical error records, and mean time between failures includes:

[0066] The logic error rate is calculated according to the number of CRC check failures, specifically:

[0067]

[0068] in, Indicates the i The logic error rate of each processor module; Indicates the iNumber of CRC check failures for each processor module; Indicates the predefined number of synchronization cycles; ;

[0069] Logical error rate It directly reflects the current calculation or communication anomaly of the module and can detect transient errors in the first place, such as bit flips caused by electromagnetic pulses. The number of errors in the last N synchronization cycles is counted by sliding windows to avoid excessive interference of single occasional errors on the weight. Logical errors in industrial control are often caused by signal interference, software deadlock, etc. This indicator is the first line of defense for fault-tolerant systems. For example, if module A fails the CRC check for three consecutive cycles or outputs an out-of-range pressure value, its error rate is When it rises, the weight decreases immediately to prevent the error from spreading.

[0070] Calculate the environmental immunity based on the temperature, voltage and vibration amplitude. The specific formula is:

[0071]

[0072] in, Indicates the i Environmental immunity of each processor module; Indicates the i The temperature of each processor module; Indicates the i The voltage of each processor module; Indicates the i The vibration amplitude of each processor module; represents the temperature weight; represents the voltage weight; represents the vibration amplitude weight; and Indicates the maximum operating temperature and set operating temperature; and Indicates the maximum operating voltage and rated operating voltage; Indicates the maximum working vibration amplitude; , , The default value is 1 / 3;

[0073] Through temperature, voltage, vibration and other sensor data, the module's gradual physical degradation can be identified, such as performance degradation caused by poor heat dissipation, loose connectors, etc. In industrial environments (such as power plants and chemical plants), physical interference such as high temperature and vibration may affect multiple modules at the same time. This indicator can provide early warning of common cause risks. Continuous high vibration or over-temperature signals indicate hardware aging and the module needs to be replaced in advance. For example, module B's temperature rises to 70°C (exceeding the rated 50°C) due to a cooling fan failure. P BEven if the logic does not report an error, the weight will be reduced to avoid potential failures caused by high temperature.

[0074] The historical reliability is calculated based on the mean time between failures. The specific formula is:

[0075]

[0076] in, Indicates the i Historical reliability of each processor module; Indicates the i MTBF per processor module; Indicates the mean time between failures of processor module A; Indicates the mean time between failures of the B processor module; Indicates the mean time between failures of the C processor module.

[0077] Based on the MTBF (mean time between failures) quantification of the inherent reliability of the module, "high failure-prone" modules are identified. If a module has been stable for a long time (high MTBF), a single transient error will not significantly reduce its weight, avoiding overreaction. The system is guided to prioritize trust in modules with excellent historical performance, reducing overall maintenance costs. For example, module C has been trouble-free for the past year (MTBF = 10,000 hours), but a temporary error occurred due to interference. Higher, the weight only decreases slightly, and the system still tends to trust its data.

[0078] Specifically, the collaborative design logic of the three health indicators is shown in Table 1.

[0079] Table 1

[0080]

[0081] Industrial TMR systems are required to cover all failure modes (physical, logical, and long-term / short-term). A three-indicator combination can achieve: short-term fault tolerance (E), medium-term warning (P), and long-term optimization (H). Covered fault types include random hardware failures, systematic failures, and environmental failures.

[0082] In step S3, the dynamic weight of each processor module is calculated according to the health index of each processor module, and the dynamic weight is normalized.

[0083] In some embodiments, in step S3, calculating the dynamic weight of each processor module according to the health index of each processor module includes:

[0084]

[0085] in, Indicates the i Dynamic weights of processor modules; Indicates the i The logic error rate of each processor module; Indicates the i Environmental immunity of each processor module; Indicates the i Historical reliability of each processor module; 、 and Represents the weights of logical error rate, environmental noise immunity, and historical reliability, α+β+γ=1, α, β, γ are adjustable coefficients, and α=0.6, β=0.3, γ=0.1 are recommended; .

[0086] Normalizing the dynamic weight includes:

[0087]

[0088] in, Indicates the i The normalized dynamic weight of each processor module; Indicates the dynamic weight of the A processor module; Indicates the dynamic weight of the B processor module; Indicates the dynamic weight of the C processor module.

[0089] Specifically, parameter determination needs to be optimized based on the unique requirements of industrial control scenarios. In industrial control systems, transient logic errors (such as communication packet loss and data verification failures) are the most common direct cause of control failure. Therefore, in environments with strong electromagnetic interference (such as near inverters), the weight of α (logic error rate) should be appropriately increased. Abnormal physical parameters such as temperature and vibration are often precursors to hardware failure, but their impact lags behind that of logic errors. Therefore, in harsh environments, such as high-temperature / high-vibration metallurgical workshops, the weight of β (environmental immunity) can be increased to 0.4; in constant-temperature, clean environments, such as laboratories, the weight of β (environmental immunity) can be reduced to 0.1. MTBF reflects the long-term stability of the module, but its contribution to real-time fault response is limited. Therefore, when module batches vary significantly, such as when mixing new and old hardware, γ (historical reliability) should be increased to 0.3 to differentiate reliability. When modules are highly homogeneous, such as from the same batch, γ can be reduced to 0.05.

[0090] The α, β, and γ coefficients can also be adaptively adjusted according to different operating stages. For example, during the startup phase, when environmental parameters are not stable, such as during the equipment warm-up period, β (environmental immunity) can be increased. During stable operation, the weighting of the α (logical error rate) standard can be restored. During system maintenance, γ (historical reliability) can be increased to identify modules nearing the end of their lifespan.

[0091] In step S4, the weighted support of the input value is calculated based on the processor module input value and the normalized dynamic weight; the input value is voted and selected based on the weighted support as the voting output of the three-series processor module.

[0092] In some embodiments, in step S4, calculating the weighted support of the input value according to the processor module input value and the normalized dynamic weight includes:

[0093]

[0094] in, Represents the weighted support of the input value; Indicates the i The normalized dynamic weight of each processor module; Indicates the i The input value of each processor module; Represents the input value of a processor module.

[0095] The step of voting and selecting an input value according to the weighted support as the voting output of the three-series processor module includes:

[0096] The input value with the largest weighted support is selected as the voting output.

[0097] Specifically, for example, the module input value , and the weight of module A =0.6 (healthy), module B weight =0.3 (recent occasional errors), module C weight =0.1 (high temperature alarm).

[0098] The data of A and B are consistent, but the data of C are different. W(100)=(0.6+0.3)>W(105)=0.1. The final result adopts the data of A / B. .

[0099] For another example, the module input value , and the weight of module A =0.6 (healthy), module B weight =0.3 (recent occasional errors), module C weight =0.1 (high temperature alarm).

[0100] Processors A, B, and C have different collection values. W(100)=0.6>W(102)=0.3>W(105)=0.1. The final result adopts data from A. .

[0101] In step S5, if the normalized dynamic weight of a processor module is lower than a predefined threshold for a preset number of consecutive cycles, an alarm will be triggered, the processor module will be automatically isolated and hot backup will be started.

[0102] Specifically, , which lasts for K cycles. Usually set to 0.2, K=5.

[0103] Example 1: The following is a complete numerical simulation of temperature control for a chemical reactor, demonstrating how a dynamic weighting algorithm can cope with sensor drift and sudden disturbances. The system needs to maintain a setpoint temperature of 150°C, including detailed data changes over five consecutive cycles.

[0104] Initial parameter setting

[0105] Module health status:

[0106] Module A: MTBF = 9,000 hours, initial temperature = 148°C (slight negative sensor drift)

[0107] Module B: MTBF = 6,000 hours, initial temperature = 150°C (sensitive to voltage fluctuations)

[0108] Module C: MTBF = 12,000 hours, initial temperature = 151°C

[0109] Weight coefficients: α=0.6 (logical error rate), β=0.25 (environmental immunity), γ=0.15 (historical reliability)

[0110] Dynamic threshold: Temperature tolerance ±2°C, voltage fluctuation >10% triggers downgrade

[0111] Error rate window: N=5 (counts the number of errors in the last five cycles)

[0112] Alarm rules: (If it falls below this value for 2 consecutive cycles, the module will be isolated).

[0113] The gas turbine speed control embodiment cycles 1-5 are deduced as shown in Table 2.

[0114] Table 2

[0115]

[0116] Cycle 2: Voltage sag event

[0117] The supply voltage of module B drops by 10%, and the output temperature is abnormally 145°C (5°C lower than the set value).

[0118] Health calculation:

[0119] Logical error rate: =1 / 5=0.2

[0120] Environmental immunity: =1-[0.5×0.1(voltage)+0.5×(5 / 150)(temperature)]=0.933

[0121] Historical reliability: =6,000 / (9,000+6,000+12,000)=0.222

[0122] =0.6×(1-0.2)+0.25×0.933+0.15×0.222=0.747

[0123] Weight update: =0.35, =0.29, =0.36

[0124] Voting results: W(148)=0.35, W(145)=0.29, W(151)=0.36, output C value 151℃.

[0125] Cycle 3: Sensor drift and voltage fault superposition

[0126] Module A experiences a continuous -2°C drift, and the voltage of module B does not recover.

[0127] Weight changes: because =0.2 to 0.33, because =0.4 further dropped to 0.27.

[0128] System action: Module B weight is < 0.3 for two consecutive cycles, triggering the isolation process.

[0129] Cycle 5: Backup module activation

[0130] The new module D (high MTBF) is added, with the weight initialized to 0.40, and the system restores balanced voting.

[0131] In summary, the solution proposed in this paper proposes a dynamic priority-weighted synchronization algorithm that dynamically adjusts voting weights by evaluating module health in real time. A multi-dimensional health model is employed to quantify module health by integrating logical error rates, physical environment parameters, and historical reliability. Combining weighted voting with a fault isolation mechanism improves system fault tolerance and resource utilization.

[0132] The health status of the processor modules is quantified through a dynamic weight mechanism. Healthy modules receive higher voting weights, suppressing the impact of low-health modules. This significantly improves the reliability of the TMR system in complex industrial environments. It combines real-time and maintainability advantages and is suitable for safety-critical control areas.

[0133] Furthermore, the present invention can automatically isolate modules with persistently low performance and activate backup units. Faulty modules can be repaired or replaced in the background without affecting system operation. Potential faulty modules can be identified in advance, avoiding unexpected downtime.

[0134] The second aspect of the present invention discloses a triple redundant data synchronization system based on dynamic priority weighting. Figure 2 FIG. 1 is a structural diagram of a triple redundant data synchronization system based on dynamic priority weighting according to an embodiment of the present invention; FIG. Figure 2 As shown, the system 100 includes:

[0135] The first processing module 101 is configured to collect multi-source data from the three processor modules A, B, and C in each cycle; the multi-source data includes: processor module input values, environmental sensor data, historical error records, and mean time between failures;

[0136] The second processing module 102 is configured to calculate the health index of each processor module based on the environmental sensor data, historical error records and mean time between failures; the health index includes: logic error rate, environmental noise immunity and historical reliability;

[0137] The third processing module 103 is configured to calculate the dynamic weight of each processor module according to the health index of each processor module and normalize the dynamic weight;

[0138] The fourth processing module 104 is configured to calculate the weighted support of the input value based on the processor module input value and the normalized dynamic weight; and select the input value based on the weighted support as the voting output of the three-series processor module;

[0139] The fifth processing module 105 is configured to trigger an alarm, automatically isolate the processor module and start hot backup if the normalized dynamic weight of a processor module is lower than a predefined threshold for a preset number of consecutive cycles.

[0140] According to the system of the second aspect of the present invention, the first processing module 101 is specifically configured as follows: the input value of the processor module is an industrial parameter; (such as temperature, pressure and other industrial parameters);

[0141] The environmental sensor data includes: temperature, voltage and vibration amplitude (each module is monitored independently);

[0142] The historical error record includes: the number of CRC check failures in a predefined sub-synchronization period of the processor module;

[0143] The mean time between failures includes: the mean time between failures extracted from the log.

[0144] According to the system of the second aspect of the present invention, the second processing module 102 is specifically configured to calculate the health index of each processor module based on the environmental sensor data, historical error records and mean time between failures, including:

[0145] The logic error rate is calculated according to the number of CRC check failures, specifically:

[0146]

[0147] in, Indicates the i The logic error rate of each processor module; Indicates the i Number of CRC check failures for each processor module; Indicates the predefined number of synchronization cycles; ;

[0148] Logical error rate It directly reflects the current calculation or communication anomaly of the module and can detect transient errors in the first place, such as bit flips caused by electromagnetic pulses. The number of errors in the last N synchronization cycles is counted by sliding windows to avoid excessive interference of single occasional errors on the weight. Logical errors in industrial control are often caused by signal interference, software deadlock, etc. This indicator is the first line of defense for fault-tolerant systems. For example, if module A fails the CRC check for three consecutive cycles or outputs an out-of-range pressure value, its error rate is When it rises, the weight decreases immediately to prevent the error from spreading.

[0149] Calculate the environmental immunity based on the temperature, voltage and vibration amplitude. The specific formula is:

[0150]

[0151] in, Indicates the i Environmental immunity of each processor module; Indicates the i The temperature of each processor module; Indicates the i The voltage of each processor module; Indicates the i The vibration amplitude of each processor module; represents the temperature weight; represents the voltage weight; represents the vibration amplitude weight; and Indicates the maximum operating temperature and set operating temperature; and Indicates the maximum operating voltage and rated operating voltage; Indicates the maximum working vibration amplitude; , , The default value is 1 / 3;

[0152] Through temperature, voltage, vibration and other sensor data, the module's gradual physical degradation can be identified, such as performance degradation caused by poor heat dissipation, loose connectors, etc. In industrial environments (such as power plants and chemical plants), physical interference such as high temperature and vibration may affect multiple modules at the same time. This indicator can provide early warning of common cause risks. Continuous high vibration or over-temperature signals indicate hardware aging and the module needs to be replaced in advance. For example, module B's temperature rises to 70°C (exceeding the rated 50°C) due to a cooling fan failure. P B Even if the logic does not report an error, the weight will be reduced to avoid potential failures caused by high temperature.

[0153] The historical reliability is calculated based on the mean time between failures. The specific formula is:

[0154]

[0155] in, Indicates the i Historical reliability of each processor module; Indicates the i MTBF per processor module; Indicates the mean time between failures of processor module A; Indicates the mean time between failures of the B processor module; Indicates the mean time between failures of the C processor module.

[0156] Based on the MTBF (mean time between failures) quantification of the inherent reliability of the module, "high failure-prone" modules are identified. If a module has been stable for a long time (high MTBF), a single transient error will not significantly reduce its weight, avoiding overreaction. The system is guided to prioritize trust in modules with excellent historical performance, reducing overall maintenance costs. For example, module C has been trouble-free for the past year (MTBF = 10,000 hours), but a temporary error occurred due to interference. Higher, the weight only decreases slightly, and the system still tends to trust its data.

[0157] Specifically, industrial TMR systems are required to cover all failure modes (physical, logical, and long-term / short-term). A three-indicator combination can achieve: short-term fault tolerance (E), medium-term warning (P), and long-term optimization (H). Covered fault types include random hardware failures, systematic failures, and environmental failures.

[0158] According to the system of the second aspect of the present invention, the third processing module 103 is specifically configured to calculate the dynamic weight of each processor module according to the health index of each processor module, including:

[0159]

[0160] in, Indicates the i Dynamic weights of processor modules; Indicates the i The logic error rate of each processor module; Indicates the i Environmental immunity of each processor module; Indicates the i Historical reliability of each processor module; 、 and Represents the weight of logical error rate, environmental noise immunity and historical reliability, α+β+γ=1, α, β, γ are adjustable coefficients, and α=0.6, β=0.3, γ=0.1 are recommended; .

[0161] Normalizing the dynamic weight includes:

[0162]

[0163] in, Indicates the i The normalized dynamic weight of each processor module; Indicates the dynamic weight of the A processor module; Indicates the dynamic weight of the B processor module; Indicates the dynamic weight of the C processor module.

[0164] Specifically, parameter determination needs to be optimized based on the unique requirements of industrial control scenarios. In industrial control systems, transient logic errors (such as communication packet loss and data verification failures) are the most common direct cause of control failure. Therefore, in environments with strong electromagnetic interference (such as near inverters), the weight of α (logic error rate) should be appropriately increased. Abnormal physical parameters such as temperature and vibration are often precursors to hardware failure, but their impact lags behind that of logic errors. Therefore, in harsh environments, such as high-temperature / high-vibration metallurgical workshops, the weight of β (environmental immunity) can be increased to 0.4; in constant-temperature, clean environments, such as laboratories, the weight of β (environmental immunity) can be reduced to 0.1. MTBF reflects the long-term stability of the module, but its contribution to real-time fault response is limited. Therefore, when module batches vary significantly, such as when mixing new and old hardware, γ (historical reliability) should be increased to 0.3 to differentiate reliability. When modules are highly homogeneous, such as from the same batch, γ can be reduced to 0.05.

[0165] The α, β, and γ coefficients can also be adaptively adjusted according to different operating stages. For example, during the startup phase, when environmental parameters are not stable, such as during the equipment warm-up period, β (environmental immunity) can be increased. During stable operation, the weighting of the α (logical error rate) standard can be restored. During system maintenance, γ (historical reliability) can be increased to identify modules nearing the end of their lifespan.

[0166] According to the system of the second aspect of the present invention, the fourth processing module 104 is specifically configured to calculate the weighted support of the input value based on the processor module input value and the normalized dynamic weight, including:

[0167]

[0168] in, Represents the weighted support of the input value; Indicates the i The normalized dynamic weight of each processor module; Indicates the i The input value of each processor module; Represents the input value of a processor module.

[0169] The step of voting and selecting an input value according to the weighted support as the voting output of the three-series processor module includes:

[0170] The input value with the largest weighted support is selected as the voting output.

[0171] Specifically, for example, the module input value , and the weight of module A =0.6 (healthy), module B weight =0.3 (recent occasional errors), module C weight =0.1 (high temperature alarm).

[0172] The data of A and B are consistent, but the data of C are different. W(100)=(0.6+0.3)>W(105)=0.1. The final result adopts the data of A / B. .

[0173] For another example, the module input value , and the weight of module A =0.6 (healthy), module B weight =0.3 (recent occasional errors), module C weight =0.1 (high temperature alarm).

[0174] Processors A, B, and C have different collection values. W(100)=0.6>W(102)=0.3>W(105)=0.1. The final result adopts data from A. .

[0175] According to the system of the second aspect of the present invention, the fifth processing module 105 is specifically configured to: , which lasts for K cycles. Usually set to 0.2, K=5.

[0176] A third aspect of the present invention discloses an electronic device. The electronic device includes a memory and a processor. The memory stores a computer program. When the processor executes the computer program, it implements the steps of any one of the triple redundant data synchronization methods based on dynamic priority weighting disclosed in the first aspect of the present invention.

[0177] Figure 3 FIG. 1 is a structural diagram of an electronic device according to an embodiment of the present invention. Figure 3 As shown, the electronic device includes a processor, memory, a communication interface, a display, and an input device connected via a system bus. The processor of the electronic device is used to provide computing and control capabilities. The memory of the electronic device includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The communication interface of the electronic device is used to communicate with an external terminal via wired or wireless communication. The wireless communication method can be achieved through Wi-Fi, a carrier network, near-field communication (NFC), or other technologies. The display of the electronic device can be a liquid crystal display or an electronic ink display. The input device of the electronic device can be a touch layer covering the display, or it can be buttons, a trackball, or a touchpad provided on the electronic device housing, or it can be an external keyboard, touchpad, or mouse.

[0178] Those skilled in the art will understand that Figure 3The structure shown in the figure is only a structural diagram of the part related to the technical solution of the present disclosure, and does not constitute a limitation on the electronic device to which the solution of the present application is applied. The specific electronic device may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0179] A fourth aspect of the present invention discloses a computer-readable storage medium having a computer program stored thereon. When executed by a processor, the computer program implements any one of the steps in the method for synchronizing triple redundant data based on dynamic priority weighting as disclosed in the first aspect of the present invention.

[0180] Please note that the technical features of the above embodiments can be combined arbitrarily. In order to make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification. The above embodiments only express several implementation methods of the present application. The description is relatively specific and detailed, but it cannot be understood as a limitation on the scope of the invention patent. It should be pointed out that for ordinary technicians in this field, without departing from the concept of this application, several variations and improvements can be made, which all fall within the scope of protection of this application. Therefore, the scope of protection of the patent in this application shall be based on the attached claims.

Claims

1. A triple redundant data synchronization method based on dynamic priority weighting, comprising: The three series processor modules A, B and C are characterized in that the method includes: Step S1: Collect multi-source data from the three processor modules A, B, and C in each cycle; the multi-source data includes: processor module input values, environmental sensor data, historical error records, and mean time between failures; Step S2: Calculate the health index of each processor module based on the environmental sensor data, historical error records, and mean time between failures; the health index includes: logic error rate, environmental immunity, and historical reliability; Step S3: Calculate the dynamic weight of each processor module according to the health index of each processor module, and normalize the dynamic weight; Step S4: Calculate the weighted support of the input value based on the processor module input value and the normalized dynamic weight; and select the input value by voting based on the weighted support as the voting output of the three-series processor module; Step S5: If the normalized dynamic weight of a processor module is lower than a predefined threshold for a preset number of consecutive cycles, an alarm is triggered, the processor module is automatically isolated, and hot backup is started; In the step S1, The processor module input value is an industrial parameter; The environmental sensor data includes: temperature, voltage and vibration amplitude; The historical error record includes: the number of CRC check failures in a predefined sub-synchronization period of the processor module; The mean time between failures includes: the mean time between failures extracted from the log; In step S2, calculating the health index of each processor module based on the environmental sensor data, historical error records, and mean time between failures includes: The logic error rate is calculated according to the number of CRC check failures, specifically: in, Indicates the i The logic error rate of each processor module; Indicates the i Number of CRC check failures for each processor module; Indicates the predefined number of synchronization cycles; ; Calculate the environmental immunity based on the temperature, voltage and vibration amplitude. The specific formula is: in, Indicates the i Environmental immunity of each processor module; Indicates the i The temperature of each processor module; Indicates the i The voltage of each processor module; Indicates the i The vibration amplitude of each processor module; represents the temperature weight; represents the voltage weight; represents the vibration amplitude weight; and Indicates the maximum operating temperature and set operating temperature; and Indicates the maximum operating voltage and rated operating voltage; Indicates the maximum working vibration amplitude; The historical reliability is calculated based on the mean time between failures. The specific formula is: in, Indicates the i Historical reliability of each processor module; Indicates the i MTBF per processor module; Indicates the mean time between failures of processor module A; Indicates the mean time between failures of the B processor module; Indicates the mean time between failures of the C processor module; In step S3, calculating the dynamic weight of each processor module according to the health index of each processor module includes: in, Indicates the i Dynamic weights of processor modules; Indicates the i The logic error rate of each processor module; Indicates the i Environmental immunity of each processor module; Indicates the i Historical reliability of each processor module; 、 and Represents the weights of logic error rate, environmental immunity, and historical reliability, α+β+γ=1, where α, β, and γ are adjustable coefficients; ; In step S3, normalizing the dynamic weight includes: in, Indicates the i The normalized dynamic weight of each processor module; Indicates the dynamic weight of the A processor module; Indicates the dynamic weight of the B processor module; Indicates the dynamic weight of the C processor module; In step S4, calculating the weighted support of the input value according to the processor module input value and the normalized dynamic weight includes: in, Represents the weighted support of the input value; Indicates the i The normalized dynamic weight of each processor module; Indicates the i The input value of each processor module; Represents the input value of a processor module.

2. The triple redundant data synchronization method based on dynamic priority weighting according to claim 1, characterized in that: In step S4, the voting to select an input value according to the weighted support as the voting output of the three-series processor module includes: The input value with the largest weighted support is selected as the voting output.

3. A triple redundant data synchronization system based on dynamic priority weighting, characterized in that: The system adopts the method according to any one of claims 1 to 2, and the system includes: The first processing module is configured to collect multi-source data from the three processor modules A, B, and C in each cycle; the multi-source data includes: processor module input values, environmental sensor data, historical error records, and mean time between failures; A second processing module is configured to calculate a health indicator of each processor module based on the environmental sensor data, historical error records, and mean time between failures; the health indicator includes: logic error rate, environmental noise immunity, and historical reliability; A third processing module is configured to calculate a dynamic weight of each processor module according to the health index of each processor module and normalize the dynamic weight; a fourth processing module configured to calculate a weighted support of the input value based on the processor module input value and the normalized dynamic weight; and select the input value by voting based on the weighted support as the voting output of the three-series processor module; The fifth processing module is configured to trigger an alarm, automatically isolate the processor module and start hot backup if the normalized dynamic weight of a processor module is lower than a predefined threshold for a preset number of consecutive cycles.

4. An electronic device, characterized in that: The electronic device includes a memory and a processor, the memory stores a computer program, and when the processor executes the computer program, the steps of the triple redundant data synchronization method based on dynamic priority weighting according to any one of claims 1 to 2 are implemented.

5. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the computer program implements the steps of the triple redundant data synchronization method based on dynamic priority weighting according to any one of claims 1 to 2.

Images