A heterogeneous dual-redundant cooperative control system and method
By using a heterogeneous dual-redundant collaborative control system, data synchronization and transient hot switching are achieved by connecting heterogeneous CPUs via an SRIO bus. This solves the problems of data synchronization difficulties and control jitter in heterogeneous redundant systems, and improves the reliability and accuracy of control.
Patent Information
- Application Number
- CN202611338119.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-08-31
- Publication Date
- 2026-11-17
AI Technical Summary
Existing heterogeneous redundant systems suffer from difficulties in data synchronization and control jitter, especially during transient switching, where it is difficult to guarantee the continuity and reliability of control.
A heterogeneous dual-redundant collaborative control system is adopted, which connects the host and standby units with the same architecture through the SRIO bus. Different models of CPUs are used to collect sensor data in parallel, and the software output state machine is synchronized through the SRIO bus to realize transient hot switching and multi-level dynamic fault-tolerant filtering, ensuring data synchronization and control continuity.
It solves the problems of data synchronization difficulties and control jitter in heterogeneous redundant systems, realizes seamless switching in the event of CPU failure, improves the reliability and accuracy of control, and ensures the normal operation of the product under abnormal conditions.
Smart Images

Figure CN123113285A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of dual-redundancy design technology, specifically relating to a heterogeneous dual-redundancy cooperative control system and method. Background Technology
[0002] In fields such as aerospace, military equipment, and rail transportation, extremely high product reliability is required; errors at critical points can lead to serious consequences. To improve product reliability and safety, most products incorporate varying degrees of redundancy design. Current mainstream redundancy design approaches include: Homogeneous dual-machine hot standby: This is currently the most common redundancy method. The primary and standby machines use the same processor, software, and architecture, and the switchover is determined by heartbeat status. The disadvantage is that the risk of common-mode failure is high, and the heartbeat status-based switchover inevitably leads to lower real-time switching. Although this can be compensated for by the standby machine repeating the work done before the primary machine switched over after the switchover is completed, it is still not possible to achieve "switching invisibility" for the user.
[0003] Heterogeneous redundancy: The primary and backup machines use different models of processors, which solves the common mode problem of homogeneous dual-machine hot standby. However, this also brings new technical bottlenecks: the clock and execution timing of different models of processors are different, making data synchronization difficult; during transient switching, the output source is switched from the primary machine to the backup machine, which brings control jitter.
[0004] The Chinese invention patent, with publication number CN117348373A, entitled "A Redundant Flight Control and Navigation System," primarily comprises multiple sensors connected in sequence, a redundancy data fusion and filtering system, a sensor information adjudication system, a redundancy flight control and navigation computer, and a control information adjudication system. Multiple sensors are used to collect various types of information; the redundancy data fusion and filtering system is used to fuse and filter the various information collected by the multiple sensors to obtain sensor information. It belongs to the traditional homogeneous hot standby redundancy method and does not involve heterogeneous redundancy or transient hot switching. Summary of the Invention
[0005] The purpose of this invention is to provide a heterogeneous dual-redundant cooperative control system and method to overcome the technical problems of data synchronization difficulties and control jitter in existing heterogeneous system control.
[0006] To solve the above problems, the present invention adopts the following technical solution: A heterogeneous dual-redundant cooperative control system includes a primary and a standby machine with consistent architecture: The host includes a first SRIO switch. The input of the first SRIO switch is connected to a first heterogeneous CPU and multiple interface boards via an SRIO bus. Each interface board is connected to a sensor and is interconnected with the first heterogeneous CPU via an SRIO bus. The backup unit includes a second SRIO switch. The input of the second SRIO switch is connected to a second heterogeneous CPU and multiple interface boards via an SRIO bus. Each interface board is connected to a sensor and is interconnected with the second heterogeneous CPU via an SRIO bus. The first SRIO switch and the second SRIO switch are interconnected via an SRIO bus. The first heterogeneous CPU is used for output, and the second heterogeneous CPU is used for calculation and backup output.
[0007] Furthermore, the interface boards in the host machine and the interface boards in the standby machine are in one-to-one correspondence and are redundant to each other.
[0008] Furthermore, an FPGA is provided on the interface board, and the sensor is connected to the interface board through the FPGA.
[0009] Furthermore, the first heterogeneous CPU and the second heterogeneous CPU synchronize the software output state machine via the SRIO bus; after each output is completed, the first heterogeneous CPU notifies the second heterogeneous CPU of the output success via SRIO; the second heterogeneous CPU monitors the output result of the first heterogeneous CPU via SRIO in each control cycle.
[0010] Furthermore, when the collected value of any sensor is determined to be invalid by both the first heterogeneous CPU and the second heterogeneous CPU for three consecutive control cycles, the sensor is set to a failed state and will no longer be collected from that sensor.
[0011] Secondly, the present invention also provides a heterogeneous dual-redundancy cooperative control method, comprising: Dual-redundant collaborative control steps: The first heterogeneous CPU and the second heterogeneous CPU are respectively connected to multiple interface boards via SRIO bus to form a network, so that the first heterogeneous CPU and the second heterogeneous CPU can simultaneously collect information from all interface boards. Transient hot switching steps: The status is synchronously output between the first heterogeneous CPU and the second heterogeneous CPU via the SRIO bus. When the first heterogeneous CPU fails to complete the output within a preset control cycle, the second heterogeneous CPU takes over the control process and executes the output. Multi-level dynamic fault-tolerant filtering steps: The redundant sensor data collected by the first heterogeneous CPU and the second heterogeneous CPU are summarized, arbitrated and filtered. After removing invalid data, the remaining valid data are averaged to obtain the final collected value.
[0012] Furthermore, after each output is completed, the first heterogeneous CPU notifies the second heterogeneous CPU of the output success via SRIO, and the second heterogeneous CPU monitors the output result of the first heterogeneous CPU via SRIO in each control cycle.
[0013] Furthermore, the output permissions of the first heterogeneous CPU and the second heterogeneous CPU are controlled by the FPGA on the interface board via a token.
[0014] Furthermore, the token control includes: The output token is by default assigned to the first heterogeneous CPU of the host. If the first heterogeneous CPU fails to complete the output control within the preset control period, the token ownership is changed to the second heterogeneous CPU of the backup machine, and the second heterogeneous CPU takes over the output control. Starting from the next control cycle, the token will be assigned to the second heterogeneous CPU by default. If the second heterogeneous CPU does not output within the preset control cycle, the token will be assigned to the first heterogeneous CPU, which will then take over the token and control the output.
[0015] Furthermore, if the value collected by any sensor is determined to be invalid by both the first heterogeneous CPU and the second heterogeneous CPU for three consecutive control cycles, the sensor is set to a failed state and the system switches to a backup sensor.
[0016] Compared with the prior art, the present invention has the following beneficial technical effects: This invention solves the common-mode problem that easily occurs under homogeneous redundancy by using two different CPUs, each running different software and using the same algorithm. By using SRIO networking, the primary and backup CPUs of heterogeneous redundancy can simultaneously collect information from all interface boards, realizing "primary and backup machines collecting from the same source" on the basis of redundancy, and solving the problem of information synchronization difficulties between heterogeneous redundancies.
[0017] This invention achieves dual-CPU output status synchronization through the SRIO bus. When the main CPU fails and cannot output, the backup CPU can immediately take over control and output during the same fault cycle, ensuring the continuity of control. By using the same source for the main and backup CPUs, it ensures that the inputs used for calculation by the main and backup CPUs are the same in each control cycle, effectively solving the control jitter caused by the switching between the main and backup CPUs.
[0018] This invention improves the reliability of acquisition results by using a synchronous acquisition and arbitration method between the primary and backup units for homogeneous dual redundant sensors. Furthermore, it achieves validity judgment of acquisition results through software arbitration. When both the primary and backup sensors are working normally, it improves acquisition accuracy through multi-value averaging. The multi-level dynamic fault-tolerant filtering algorithm module in this invention can realize any working combination of primary CPU + primary sensor, primary CPU + backup sensor, backup CPU + primary sensor, and backup CPU + backup sensor. This solves the problem in traditional redundancy mode where only some functions of the primary and backup units are abnormal, but the product cannot work normally due to the limitations of the switching method. Attached Figure Description
[0019] Figure 1 This is a hardware topology diagram of a heterogeneous dual-redundant cooperative control system in an embodiment of the present invention; Figure 2 This is a flowchart of the transient hot-switching module's workflow. Figure 3 This is a schematic diagram of the token status of the transient hot-switching module; Figure 4 This is a schematic diagram of a multi-level dynamic fault-tolerant filtering algorithm. Detailed Implementation
[0020] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.
[0021] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.
[0022] This invention provides a heterogeneous dual-redundant cooperative control system, such as... Figure 1As shown, it includes a primary and a backup unit with the same architecture. The architecture of the primary or backup unit consists of: an SRIO switch, multiple interface boards, and a CPU. The input terminals of the SRIO switch are connected to the CPU and each interface board via an SRIO bus. Specifically: The host includes a first SRIO switch, whose input is connected to a first heterogeneous CPU and multiple interface boards via an SRIO bus. Each interface board is equipped with an FPGA, and sensors are connected to the interface boards via the FPGAs. Each interface board is interconnected with the first heterogeneous CPU via an SRIO bus. The standby includes a second SRIO switch, whose input is connected to a second heterogeneous CPU and multiple interface boards via an SRIO bus. Each interface board is connected to a sensor, and each interface board is interconnected with the second heterogeneous CPU via an SRIO bus.
[0023] The first SRIO switch and the second SRIO switch are interconnected via an SRIO bus. The first heterogeneous CPU is used for output, and the second heterogeneous CPU is used for computation and backup output.
[0024] The first heterogeneous CPU and the second heterogeneous CPU have different models.
[0025] See Figure 1 The dual-redundancy collaborative control module includes a primary and a backup unit with identical hardware architecture. Each unit contains one SRIO switch, one CPU board (the primary and backup units use different CPUs), and multiple interface boards (A0 / B0 are redundant, A1 / B1 are redundant, and Ax / Bx are redundant). SRIO switch A is interconnected with SRIO switch B, and both CPUA and CPUB are interconnected with SRIO switches A and B. During operation, CPUA communicates with the primary interface board through SRIO switch A and with CPUB and the backup interface board through SRIO switch B; CPUB communicates with the backup interface board through SRIO switch B and with CPUA and the primary interface board through SRIO switch A. If one switch fails during operation, the primary and backup units can still communicate unaffected through the other switch.
[0026] One embodiment of this invention provides a heterogeneous dual-redundancy collaborative control system, addressing the challenge of implementing heterogeneous redundancy in embedded systems with high reliability requirements. It achieves "zero-error" collaborative control between the heterogeneous dual redundancies by networking dual CPUs and multiple interface boards via an SRIO bus. Each CPU can collect information from all interface boards and share the final information via the SRIO bus. Each CPU interprets the data from all the results collected from both sides to obtain the final control information. Transient hot-switching is implemented, with both CPUs performing control-related calculations simultaneously. The primary CPU is assumed to complete the control output; if a fault occurs before the primary CPU's output, the hot-standby CPU immediately takes over and completes the current and subsequent control outputs. A multi-level dynamic fault-tolerant filtering algorithm is implemented. The primary and standby CPUs collect all interface board information and aggregate it via SRIO. The primary and standby CPUs then separately determine the validity of the aggregated information. When all data is valid, the final result is obtained by averaging all the data. If some data is determined to be invalid, it is discarded, and an adaptive degradation strategy is adopted, averaging only the valid data.
[0027] The system mainly consists of three modules: a dual-redundancy collaborative control module, a transient hot-switching module, and a multi-level dynamic fault-tolerant filtering algorithm module.
[0028] The dual-redundant collaborative control module uses two different CPUs interconnected via an SRIO bus. The sensors are homogeneous and dual-redundant, with primary and backup sensors connected to primary and backup interface boards respectively. Data acquisition is controlled by an FPGA on the interface boards, which in turn connects to the CPU board via the SRIO bus. The FPGA shares sensor data with the CPU board through SRIO, achieving SRIO interconnection between all sensors and the CPU. After power-on, both CPUs run their respective programs to receive external commands and simultaneously initiate the control flow. During data acquisition, both CPUs acquire all sensor data via SRIO. The primary CPU outputs the data, while the backup CPU only performs calculations and does not output. When the primary CPU fails to output, the backup CPU takes over. Multiple interface boards can be connected depending on the application scenario, each implementing different functions.
[0029] The transient hot-switching module synchronizes the state machine output between the two CPUs via SRIO. The default working mode is that the main CPU calculates and outputs. After the output is completed, the backup CPU is notified of the output success via SRIO. After the backup CPU completes its calculation, it monitors the output result of the main CPU via SRIO. If the main CPU has not output before the end of the control cycle (which can be adjusted according to the actual control cycle), the backup CPU takes over the control flow and outputs the result to ensure uninterrupted control. At the same time, since the main and backup CPUs have the same data acquisition source, the calculation results will be the same, effectively avoiding control jitter caused by switching.
[0030] The multi-level dynamic fault-tolerant filtering algorithm module uses both the primary and backup units to simultaneously collect information from all interface boards via SRIO. The collected information is then aggregated and arbitrated using SRIO. Taking angle sensors as an example, the primary and backup units collect values from angle sensors A and B respectively, obtaining a total of four mechanical angles. The primary and backup units set thresholds based on the dynamic changes in physical quantities, discarding values exceeding the thresholds and averaging the remaining valid values as the final collected value for calculation. If a sensor's collected values are deemed invalid by both the primary and backup units for three consecutive cycles, the sensor is set to a failed state and will not be collected again thereafter.
[0031] In this embodiment, the main CPU is the first heterogeneous CPU in the host machine, and the backup CPU is the second heterogeneous CPU in the backup machine. The two are of different models and are heterogeneous.
[0032] The dual-redundancy collaborative control module in this invention solves the common-mode problem that easily occurs under homogeneous redundancy by using two different CPUs, each running different software and using the same algorithm. Through SRIO networking, the primary and backup CPUs of heterogeneous redundancy can simultaneously collect information from all interface boards, realizing "primary and backup machines collecting information from the same source" on the basis of redundancy, and solving the problem of information synchronization difficulties between heterogeneous redundancies.
[0033] The transient hot-switching module in this invention achieves dual-CPU output status synchronization through the SRIO bus. When the main CPU fails and cannot output, the backup CPU can immediately take over control and output during the fault cycle, ensuring the continuity of control. By using the same source for the main and backup CPUs, it ensures that the inputs used for calculation by the main and backup CPUs are the same in each control cycle, effectively solving the control jitter caused by the switching between the main and backup units.
[0034] The multi-level dynamic fault-tolerant filtering algorithm module in this invention improves the reliability of the acquisition results by synchronously acquiring and arbitrating homogeneous dual redundant sensors through the main and backup units. It also achieves the validity judgment of the acquisition results through software arbitration. When both the main and backup sensors are working normally, it improves the acquisition accuracy through multi-value averaging. The multi-level dynamic fault-tolerant filtering algorithm module in this invention can realize any working combination of main CPU + main sensor, main CPU + backup sensor, backup CPU + main sensor, and backup CPU + backup sensor. This solves the problem that in the traditional redundancy mode, when both the main and backup units have only partial functional abnormalities, the product cannot work normally due to the limitations of the switching method.
[0035] This invention also provides a heterogeneous dual-redundancy cooperative control method, comprising: Dual-redundant collaborative control steps: The first heterogeneous CPU and the second heterogeneous CPU are respectively connected to multiple interface boards via SRIO bus to form a network, so that the first heterogeneous CPU and the second heterogeneous CPU can simultaneously collect information from all interface boards. Transient hot switching steps: The status is synchronously output between the first heterogeneous CPU and the second heterogeneous CPU via the SRIO bus. When the first heterogeneous CPU fails to complete the output within the preset control cycle, the second heterogeneous CPU takes over the control process and executes the output. Multi-level dynamic fault-tolerant filtering steps: The redundant sensor data collected by the first heterogeneous CPU and the second heterogeneous CPU are summarized, arbitrated and filtered. After removing invalid data, the remaining valid data are averaged to obtain the final collected value.
[0036] The detailed dual-redundancy collaborative control steps include: After the system powers on, CPU A and CPU B start their respective programs to complete the initialization of their peripherals, SRIO bus, and internal state machines. Since CPU A and CPU B are different processors with different clock frequencies and instruction execution speeds, their initialization times differ, but both must complete initialization before entering the formal control flow. The SRIO bus protocol is used to network CPU A, CPU B, and all interface boards. CPU A and CPU B identify each interface board and establish a communication link through device IDs. Multiple interface boards can be connected depending on the actual use case, each performing different functions. After the dual CPUs are powered on, each runs its own program to receive external commands and simultaneously initiates the control process. Within each control cycle, both CPU A and CPU B acquire sensor data from all interface boards via the SRIO bus. CPU A and CPU B each send all the sensor data they have collected to each other via the SRIO bus, achieving data sharing. Subsequently, CPU A and CPU B each perform a comprehensive analysis of their own collected data and the shared data to obtain the final control information. Since CPU A and CPU B use the same data acquisition source (both can access all interface boards), the final information obtained by both is consistent. CPU A executes control output by default, sending the calculation results to the actuator interface board via SRIO to complete the output; CPU B performs calculations synchronously, but does not execute output. When CPU A fails and cannot output, CPU B takes over the output.
[0037] Detailed Figure 2 This is a flowchart illustrating the steps of the transient hot-switching module, including: CPU A and CPU B synchronize their software output state machines via the SRIO bus. Taking a practical application in motor control as an example, the control cycle is set to 200μs. Because CPU A and CPU B use different processor models and different software implementations, the time required to run the same control algorithm differs, but neither exceeds 200μs. The two CPUs send synchronization pulses at the beginning of each control cycle via the SRIO bus to ensure that both enter the new control cycle simultaneously. In collaborative control, the backup machine with slower computation is used as the benchmark. In this embodiment, the time for CPU B to complete the acquisition and calculation is approximately 50μs. After timing 100μs from the start of the control flow (leaving a 50μs margin to ensure that the output is completed before 200μs), CPU B uses SRIO to detect whether CPU A has completed the output. If CPU B detects that CPU A has completed its output at the 100μs mark, then CPU B will normally end this control cycle, will not execute the output, and will wait for the synchronization signal for the next control cycle. If CPU B detects that CPU A has not completed its output at the 100μs mark, it determines that CPU A has malfunctioned within this control cycle. At this point, CPU B immediately takes over the control flow, sending the calculation result for this cycle to the actuator interface board via SRIO to complete the output, ensuring uninterrupted control. Since CPU A and CPU B use identical data acquisition sources and the input data used for calculation is the same for both, the calculation result output by CPU B is exactly the same as the result that CPU A should have output, effectively avoiding control jitter caused by switching. The output permissions of the primary and backup units are controlled by the FPGA on the interface board through a token mechanism. The output permissions of the first heterogeneous CPU and the second heterogeneous CPU are also controlled by the FPGA on the interface board through a token mechanism. The token mechanism specifically includes: After power-on initialization, the output token belongs to the main CPU by default; If the main CPU fails to complete the output control within the timeout period, the token ownership will be changed to the backup CPU, which will then take over the output control. From the next control cycle onwards, the token will be assigned to the backup CPU by default, and the main CPU will not be able to preempt it before the timeout period expires. If the standby CPU fails to output within the timeout period, the token ownership changes to the main CPU, which then takes over the token and controls the output.
[0038] Upon power-up and the first entry into the control flow, the output token defaults to the main CPU. If the main CPU fails to complete the output control within the timeout period, the token ownership changes to the backup CPU. At this time, the backup CPU can take over the output control and complete the subsequent work. From the next cycle onwards, the token defaults to the backup CPU. The main CPU cannot preempt the backup CPU before the timeout period expires. If the backup CPU still fails to output after the timeout, the token ownership changes to the main CPU. At this time, the main CPU can preempt the token and control the output.
[0039] Figure 4 This is a flowchart illustrating the multi-stage dynamic fault-tolerant filtering process, which specifically includes: CPU A and CPU B each acquire information from two redundant sensors. Taking the angle sensor as an example, CPU A acquires the value of angle sensor A through interface board A0 and records it as A_1, and acquires the value of angle sensor B through interface board B0 and records it as B_1; CPU B acquires the value of angle sensor A through interface board A0 and records it as A_2, and acquires the value of angle sensor B through interface board B0 and records it as B_2. Thus, a total of four mechanical angle values are obtained: A_1, A_2, B_1, and B_2. CPU A and CPU B aggregate their respective four sets of data via the SRIO bus. Each CPU obtains all four sets of data (A_1, A_2, B_1, B_2) and performs data validity checks on each. The checks are based on the dynamic changes of physical quantities; for example, the change in the angle sensor value within adjacent control cycles should not exceed a preset physical threshold. CPU A and CPU B independently perform the checks, discarding values that exceed the threshold. When all four sets of data are valid, CPU A and CPU B calculate the average of the four sets of data respectively, which is used as the final acquired value for control calculation. When some data is determined to be invalid, the invalid data is discarded, and an adaptive degradation strategy is adopted, using only the remaining valid data to calculate the average as the final acquired value. If the data collected by a certain sensor is deemed invalid by both CPU A and CPU B for three consecutive control cycles, the sensor is set to a failed state. From the fourth control cycle onwards, CPU A and CPU B no longer collect data from that sensor, but only use the data collected by the other sensor as the calculation input, that is, using the average of B_1 and B_2 as the final collected value.
[0040] Through the above-mentioned multi-level dynamic fault-tolerant filtering algorithm, any working combination of CPU A + sensor A, CPU A + sensor B, CPU B + sensor A, and CPU B + sensor B can be realized, which solves the problem that in the traditional redundancy mode, only some functions of the host and the standby are abnormal, but the product cannot work normally due to the limitation of the switching method.
[0041] The preferred embodiments of the present invention have been described in detail above; however, the present invention is not limited thereto. Within the scope of the inventive concept, various simple modifications can be made to the technical solutions of the present invention, including combinations of various technical features in any other suitable manner. These simple modifications and combinations should also be considered as the content disclosed in the present invention and are all within the protection scope of the present invention.
Claims
1. A heterogeneous dual-redundant cooperative control system, characterized in that, Includes primary and backup servers with consistent architecture: The host includes a first SRIO switch. The input of the first SRIO switch is connected to a first heterogeneous CPU and multiple interface boards via an SRIO bus. Each interface board is connected to a sensor and is interconnected with the first heterogeneous CPU via an SRIO bus. The backup unit includes a second SRIO switch. The input of the second SRIO switch is connected to a second heterogeneous CPU and multiple interface boards via an SRIO bus. Each interface board is connected to a sensor and is interconnected with the second heterogeneous CPU via an SRIO bus. The first SRIO switch and the second SRIO switch are interconnected via an SRIO bus. The first heterogeneous CPU is used for output, and the second heterogeneous CPU is used for calculation and backup output.
2. The heterogeneous dual-redundant cooperative control system according to claim 1, characterized in that, The interface boards in the host machine and the interface boards in the standby machine are one-to-one and redundant.
3. The heterogeneous dual-redundant cooperative control system according to claim 1, characterized in that, An FPGA is installed on the interface board, and the sensor is connected to the interface board through the FPGA.
4. The heterogeneous dual-redundant cooperative control system according to claim 1, characterized in that, The first heterogeneous CPU and the second heterogeneous CPU synchronize the software output state machine through the SRIO bus; after each output is completed, the first heterogeneous CPU notifies the second heterogeneous CPU of the output success through SRIO; the second heterogeneous CPU monitors the output result of the first heterogeneous CPU through SRIO in each control cycle.
5. A heterogeneous dual-redundant cooperative control system according to claim 1, characterized in that, If the value collected by any sensor is determined to be invalid by both the first heterogeneous CPU and the second heterogeneous CPU for three consecutive control cycles, the sensor is set to a failed state and will no longer be collected from that sensor.
6. A heterogeneous dual-redundancy cooperative control method, characterized in that, The heterogeneous dual-redundant cooperative control system according to any one of claims 1-5 includes: Dual-redundant collaborative control steps: The first heterogeneous CPU and the second heterogeneous CPU are respectively connected to multiple interface boards via SRIO bus to form a network, so that the first heterogeneous CPU and the second heterogeneous CPU can simultaneously collect information from all interface boards. Transient hot switching steps: The status is synchronously output between the first heterogeneous CPU and the second heterogeneous CPU via the SRIO bus. When the first heterogeneous CPU fails to complete the output within a preset control cycle, the second heterogeneous CPU takes over the control process and executes the output. Multi-level dynamic fault-tolerant filtering steps: The redundant sensor data collected by the first heterogeneous CPU and the second heterogeneous CPU are summarized, arbitrated and filtered. After removing invalid data, the remaining valid data are averaged to obtain the final collected value.
7. The heterogeneous dual-redundancy cooperative control method according to claim 6, characterized in that, After each output is completed, the first heterogeneous CPU notifies the second heterogeneous CPU of the output success via SRIO. The second heterogeneous CPU monitors the output result of the first heterogeneous CPU via SRIO in each control cycle.
8. The heterogeneous dual-redundancy cooperative control method according to claim 6, characterized in that, The output permissions of the first heterogeneous CPU and the second heterogeneous CPU are controlled by the FPGA on the interface board via a token.
9. The heterogeneous dual-redundancy cooperative control method according to claim 8, characterized in that, The token control includes: The output token is by default assigned to the first heterogeneous CPU of the host. If the first heterogeneous CPU fails to complete the output control within the preset control period, the token ownership is changed to the second heterogeneous CPU of the backup machine, and the second heterogeneous CPU takes over the output control. Starting from the next control cycle, the token will be assigned to the second heterogeneous CPU by default. If the second heterogeneous CPU does not output within the preset control cycle, the token will be assigned to the first heterogeneous CPU, which will then take over the token and control the output.
10. The heterogeneous dual-redundancy cooperative control method according to claim 6, characterized in that, If the value collected by any sensor is determined to be invalid by both the first heterogeneous CPU and the second heterogeneous CPU for three consecutive control cycles, the sensor is set to a failed state and the system switches to a backup sensor.
Citation Information
Patent Citations
Redundancy flight control and navigation system
CN117348373A
