A distributed sensor fault-tolerant cooperative control method and system based on spatial interpolation reconstruction

Abstract

This invention provides a distributed sensor fault-tolerant collaborative control method and system based on spatial interpolation reconstruction. The method includes: intelligent actuator nodes maintaining a list of neighboring sensors and receiving the measured values ​​and self-confidence levels of each sensor; the actuator nodes dynamically calculating the confidence level of each neighboring sensor; when the confidence level of the master sensor is lower than a fault threshold, the actuator nodes automatically select a set of reliable sensors, reconstruct the environmental parameters of the control point using inverse distance weighted or Kriging interpolation, and use the reconstructed values ​​as feedback for control output; simultaneously, the system continuously monitors failed sensors and automatically switches back to the measured values ​​after the sensor recovers. This invention eliminates the need for redundant hardware, completes fault detection and parameter reconstruction at the edge level in milliseconds, achieves reconstruction accuracy that meets control requirements, significantly reduces system costs, improves fault tolerance, real-time performance, and control continuity, and is suitable for environments such as laboratories and cleanrooms.

Description

Technical Field

[0001] This invention relates to the fields of industrial automation and laboratory environment control technology, and in particular to a distributed sensor fault-tolerant collaborative control method and system based on spatial interpolation reconstruction. Background Technology

[0002] In environments with stringent requirements for environmental parameters (such as differential pressure, temperature, humidity, and toxic gas concentration), such as laboratories, cleanrooms, and data centers, sensor failure is one of the main causes of control system failure. Traditional solutions suffer from the following problems: High hardware redundancy costs: Configuring dual or triple sensors for key measurement points significantly increases hardware costs and installation space requirements.

[0003] The central fault tolerance delay is large: the central controller detects sensor faults and switches to the backup strategy. The fault detection, communication and decision-making delay can reach the second level, which is difficult to meet the millisecond-level requirements of negative pressure laboratories for maintaining pressure gradients.

[0004] The fault-tolerant mode is crude: after a failure, a fixed opening degree or historical average value is usually used as a substitute, which cannot respond to real dynamic changes in the environment and may lead to safety loss of control or energy waste.

[0005] Existing technologies lack a low-cost, high-real-time fault-tolerant solution that does not rely on redundant hardware, can perform sensor fault detection and parameter reconstruction in real time at the edge, and whose reconstruction accuracy meets control requirements. Summary of the Invention

[0006] The purpose of this invention is to address the shortcomings of existing technologies by proposing a distributed sensor fault-tolerant collaborative control method and system based on spatial interpolation reconstruction.

[0007] To achieve the above objectives, the present invention adopts the following technical solution: a distributed sensor fault-tolerant cooperative control method based on spatial interpolation reconstruction, comprising the following sub-steps: S1: The actuator receives data from various sensors. coordinate According to Euclidean distance Generate a list of nearby sensors It also pre-stores coordinate information and outputs a list of nearby sensors. For use in subsequent trust assessments; Furthermore, S1 includes the following sub-steps: S11: Based on the preset communication radius or the number of nearest neighbors Calculate its relationship with each sensor Euclidean distance: ;in, These are the coordinates of the control points for each actuator, typically the actuator's own installation location or a virtual control point; S12: Will satisfy The sensors are sorted in ascending order of distance, and the top ones are selected. Each sensor generates a list of nearby sensors: ; S13: The actuator simultaneously stores a list of nearby sensors. Coordinates of each sensor and the corresponding Euclidean distance ; Furthermore, the list of proximity sensors It can be statically configured or dynamically updated via heartbeat packets, for example, by triggering a rearrangement when the signal strength RSSI changes.

[0008] S2: The sensor collects environmental parameters and calculates the self-confidence level by integrating the self-diagnostic status. It then packages and broadcasts the measured value, self-confidence level, and timestamp for nearby actuators to receive. Furthermore, the environmental parameters include wind speed, pressure difference, and gas concentration; the self-diagnostic status includes power supply voltage. Signal amplitude and rate of change ; Furthermore, sensors At a fixed frequency Perform the following sub-steps: S21: Collect environmental parameters and obtain measured values. ; The sensor's sensing element converts the environmental parameters into analog electrical signals (voltage or current). After signal conditioning circuitry (amplification and filtering), the signals are sampled and quantized by an analog-to-digital converter (ADC), and then converted into measured values ​​with physical units according to pre-stored calibration coefficients (linear or polynomial mapping). .

[0009] S22: Based on the measured values ​​in S21 Calculate the rate of change ,in Take the initial value or 0; S23: Calculate self-confidence level ; Self-confidence is based solely on the sensor's own diagnostic information, reflecting the sensor's own health status, without considering spatial consistency with surrounding sensors. In this invention, self-confidence will be used as one of the inputs for subsequent comprehensive confidence assessment (S3), and the calculation formula is as follows: ,in Power supply voltage confidence: Deviation of the sensor's internal power supply voltage from its rated value will affect measurement accuracy. This is designed to be: That is, when the voltage deviation is within ±10% of the rated value, the weighting is linearly reduced; otherwise, it is set to 0. Rated voltage; For range confidence: If the measured value exceeds the sensor's specified physical range, the sensor will inevitably fail. ,in For sensor range; Confidence level for rate of change: The rate of change of environmental parameters is limited by the inertia of the physical system. If the rate of change of the measured value exceeds the maximum possible rate of change... This indicates that the sensor may be affected by noise or malfunction, and is defined as: ,in Preset based on the physical characteristics of the controlled object.

[0010] The three factors mentioned above cover the most common sensor failure modes: power supply anomaly, over-range measurement, and signal jump. It should be noted that self-confidence level is insufficient to determine sensor reliability on its own (e.g., slow drift faults cannot be detected by the above factors). Therefore, this invention further introduces spatial prediction consistency in S3 to perform a comprehensive reliability assessment of the sensor.

[0011] S24: Generate data packet The timestamp This is the count value of the local clock of the sensor node.

[0012] S3: The actuator receives data packets. Combined with pre-stored coordinates, the dynamic reliability of each sensor is calculated in each control cycle, and the sensor status is determined based on the reliability changes in consecutive cycles, and the average reliability and status flag are output. Furthermore, each actuator Perform the following sub-steps in each control cycle T (e.g., 50ms): S31: Receive neighbor list Various sensors data packets Simultaneously acquire pre-stored sensor coordinates Set preset parameters: maximum permissible deviation R, fault threshold. Trust threshold Filter coefficient α = 0.4, sampling frequency ; S32: Perform timestamp verification and outdated data processing; If the difference between the current time t and the data packet timestamp is greater than If the data is deemed outdated, the self-confidence level will be temporarily downgraded. If no valid data is available for three consecutive periods, the current original confidence level will be reset. The sensor was then marked as "communication failure".

[0013] S33: Calculate spatial prediction values ; For each sensor , using Other than, the average confidence level of the previous period For other nearby sensors, inverse distance-weighted interpolation is used:

[0014] ; If the number of neighbors that meet the condition is less than 2, then let and the consistency items in S34 The value is directly assigned to 0.5.

[0015] S34: Calculate the original credibility of the current period: The min function ensures that the expression within the parentheses is non-negative.

[0016] S35: Apply a first-order low-pass filter to the original confidence level to obtain the average confidence level. Setting initial ; S36: Perform status determination and suspicious item count; Maintain a suspicious counter for each sensor (Initial value set to 0), if Mark as "normal" and clear the suspicious counter; if If marked as "suspicious", the suspicious counter is incremented by 1; if Or suspicious counter 3. Mark as "Failed". Output the average confidence level for each neighboring sensor. and status indicators (normal / suspicious / invalid).

[0017] S4: The actuator acquires the coordinates of the control point and the set of trusted sensors. When the main sensor fails, it uses inverse distance weighted interpolation to reconstruct the parameters and outputs the reconstructed value as a feedback quantity. Furthermore, each actuator The following sub-steps are executed in each control cycle T: S41: Obtain control point coordinates Main sensor identifier and a list of nearby sensors The status and coordinates of each sensor Measured values Average credibility .

[0018] Furthermore, set the maximum permissible interpolation standard deviation. and backup safety value Main sensor identifier Pre-configured as the sensor closest to or specified to the control point.

[0019] S42: Check main sensor status: If the main sensor is marked If the status is "normal", then its measured value is used directly as the control feedback quantity. If the current condition is met, the system will jump to S5; otherwise, it will enter fault-tolerant refactoring mode.

[0020] S43: Filtering a set of trustworthy sensors: From a list of nearby sensors Selected if the status is "normal" and it is not the main sensor identifier The sensors constitute the candidate set. .like Then a backup strategy is adopted: the most recent valid historical value is used first. (If it exists), otherwise use the backup safety value, issue an alarm, and then jump to S5.

[0021] S44: Calculate the reconstructed values ​​at the control points using inverse distance-weighted interpolation with a trusted set. ; S45: Perform reconstruction quality assessment and calculate the interpolation standard deviation: If the denominator is 0 (which theoretically will not happen), then directly set... ;like If the reconstruction is deemed unreliable, switch to secure backup mode: If the value is not specified, an alarm will be issued; otherwise, the reconstructed value will be used as the control feedback value. .

[0022] S5: The actuator provides feedback with the reconstructed value or the measured value, calculates the control command through PID and broadcasts the fault-tolerant status, and automatically switches back to the measured value after the sensor recovers. Furthermore, each actuator performs the following sub-steps in each control cycle T: S51: Obtain control feedback quantity Target setpoint, PID parameters Output limiting Maximum fault tolerance time (e.g., 1800s), number of consecutive normal cycles N required for recovery (e.g., 10). Maintenance mode variable Mode (0=normal, 1=fault-tolerant, 2=safe backup) and the start time for entering fault-tolerant mode. .

[0023] S52: Calculation of Deviation ; S53: Perform a seamless handover (only when the mode changes); When switching from normal mode to fault-tolerant mode, or vice versa, the PID integral term is reset to prevent output jumps; assuming the PID form is: The new integral term is then set as follows: And use it to update the integrator.

[0024] S54: The control output is calculated using an incremental PID algorithm. , Initial time Limit the output: And drive the actuator to move.

[0025] S55: Fault-tolerant status broadcast: If currently in fault-tolerant mode A status message is broadcast every 10 cycles: ; S56: Self-healing recovery monitoring (executed only in fault-tolerant mode): Receives a failed master sensor identifier every cycle. The data packets are recalculated for their average credibility according to the method in S3. .

[0026] Maintenance counter :like ,but Add 1, otherwise reset to zero; when When the sensor has stabilized and recovered, proceed with the execution. Perform a seamless handover again (S53) and broadcast a recovery message. ; S57: Timeout Protection: If the fault-tolerant mode duration is... Then it will automatically switch to the safe backup mode: (Lock the output) and send a "sensor permanently failed" alarm to the central monitoring system.

[0027] Compared with existing technologies, the advantages of this invention are as follows: This invention utilizes the spatial interpolation reconstruction capability of distributed actuator nodes to reconstruct environmental parameters at the failure point in real time using measurement data from nearby trusted sensors when a sensor fails, eliminating the need for any redundant sensors. Taking fume hood surface wind speed control as an example, traditional solutions require 2-3 wind speed sensors for voting, while this invention only requires one main sensor in conjunction with nearby auxiliary sensors to achieve fault tolerance, reducing hardware costs by more than 50% and avoiding implementation difficulties caused by insufficient installation space for redundant sensors. This invention decentralizes fault detection, trust assessment, and spatial interpolation reconstruction to the edge of the actuator node, eliminating the need for a central controller. Experimental data shows that the total time from sensor failure to actuator completion of fault-tolerant switching can be controlled within 50ms (measured fume hood surface wind speed fluctuation does not exceed ±0.03m / s), which is 1-2 orders of magnitude better than traditional central fault-tolerant solutions, fully meeting the stringent real-time requirements of highly dynamic environments. This invention employs inverse distance weighted interpolation, fully utilizing the spatial correlation of environmental parameters to dynamically reconstruct the parameter values ​​at the failure point based on measured values ​​from reliable sensors. For scenarios with strong spatial correlation, Kriging interpolation can also be used to obtain the optimal linear unbiased estimate. Experimental results show that the reconstruction error of inverse distance weighted interpolation is ≤5% of the range, and the reconstruction error of Kriging is ≤2% of the range. The reconstruction accuracy is far superior to simple historical value substitution or fixed value substitution, enabling the maintenance of control quality and avoiding safety risks during sensor failure. Attached Figure Description

[0028] Figure 1 This is a flowchart illustrating the implementation of Embodiment 1 of the present invention. Detailed Implementation

[0029] To provide a further understanding of the purpose, structure, features, and functions of the present invention, detailed descriptions are provided below with reference to specific embodiments. The distributed sensor fault-tolerant collaborative control method based on spatial interpolation reconstruction includes the following sub-steps: S1: The actuator receives data from various sensors. coordinate According to Euclidean distance Generate a list of nearby sensors It also pre-stores coordinate information and outputs a list of nearby sensors. For use in subsequent trust assessments; Furthermore, S1 includes the following sub-steps: S11: Based on the preset communication radius or the number of nearest neighbors Calculate its relationship with each sensor Euclidean distance: ;in, These are the coordinates of the control points for each actuator, typically the actuator's own installation location or a virtual control point; S12: Will satisfy The sensors are sorted in ascending order of distance, and the top ones are selected. Each sensor generates a list of nearby sensors: ; S13: The actuator simultaneously stores a list of nearby sensors. Coordinates of each sensor and the corresponding Euclidean distance ; Furthermore, the list of proximity sensors It can be statically configured or dynamically updated via heartbeat packets, for example, by triggering a rearrangement when the signal strength RSSI changes.

[0030] By calculating the Euclidean distance to generate a neighbor list and pre-store the coordinates, the problem of self-building spatial relationships between nodes in a distributed environment is solved. This enables the actuator to autonomously identify neighboring sensors without relying on a central database, providing an efficient data foundation for subsequent collaborative fault tolerance, reducing system configuration complexity, and supporting dynamic topology updates.

[0031] S2: The sensor collects environmental parameters and calculates the self-confidence level by integrating the self-diagnostic status. It then packages and broadcasts the measured value, self-confidence level, and timestamp for nearby actuators to receive. Furthermore, the environmental parameters include wind speed, pressure difference, and gas concentration; the self-diagnostic status includes power supply voltage. Signal amplitude and rate of change ; Furthermore, sensors At a fixed frequency Perform the following sub-steps: S21: Collect environmental parameters and obtain measured values. ; The sensor's sensing element converts the environmental parameters into analog electrical signals (voltage or current). After signal conditioning circuitry (amplification and filtering), the signals are sampled and quantized by an analog-to-digital converter (ADC), and then converted into measured values ​​with physical units according to pre-stored calibration coefficients (linear or polynomial mapping). .

[0032] S22: Based on the measured values ​​in S21 Calculate the rate of change ,in Take the initial value or 0; S23: Calculate self-confidence level ; Self-confidence is based solely on the sensor's own diagnostic information, reflecting the sensor's own health status, without considering spatial consistency with surrounding sensors. In this invention, self-confidence will be used as one of the inputs for subsequent comprehensive confidence assessment (S3), and the calculation formula is as follows: ,in Power supply voltage confidence: Deviation of the sensor's internal power supply voltage from its rated value will affect measurement accuracy. This is designed to be: That is, when the voltage deviation is within ±10% of the rated value, the weighting is linearly reduced; otherwise, it is set to 0. Rated voltage; For range confidence: If the measured value exceeds the sensor's specified physical range, the sensor will inevitably fail. ,in For sensor range; Confidence level for rate of change: The rate of change of environmental parameters is limited by the inertia of the physical system. If the rate of change of the measured value exceeds the maximum possible rate of change... This indicates that the sensor may be affected by noise or malfunction, and is defined as: ,in Preset based on the physical characteristics of the controlled object.

[0033] The three factors mentioned above cover the most common failure modes of sensors: abnormal power supply, over-range, and signal jump. It should be noted that self-confidence is not sufficient to determine the reliability of a sensor on its own (for example, slow drift faults cannot be detected by the above factors). Therefore, this invention further introduces spatial prediction consistency in S3 to conduct a comprehensive reliability assessment of the sensor.

[0034] S24: Generate data packet The timestamp This is the count value of the local clock of the sensor node.

[0035] By generating and broadcasting self-confidence scores based on comprehensive self-diagnostic factors such as power supply voltage, range, and rate of change, the problem of quantifying and distributing the sensor's own health status is solved. This provides reliable self-diagnostic information for the edge side, providing basic data for subsequent dynamic confidence assessment. At the same time, it avoids misjudgments from single self-diagnoses and improves the initial screening accuracy of fault detection.

[0036] S3: The actuator receives data packets. Combined with pre-stored coordinates, the dynamic reliability of each sensor is calculated in each control cycle, and the sensor status is determined based on the reliability changes in consecutive cycles, and the average reliability and status flag are output. Furthermore, each actuator Perform the following sub-steps in each control cycle T (e.g., 50ms): S31: Receive data from each sensor in the neighboring list data packets Simultaneously acquire pre-stored sensor coordinates Set preset parameters: maximum permissible deviation R, fault threshold. Trust threshold Filter coefficient α = 0.4, sampling frequency ; S32: Perform timestamp verification and outdated data processing; If the difference between the current time t and the data packet timestamp is greater than If the data is deemed outdated, the self-confidence level will be temporarily downgraded. If no valid data is available for three consecutive periods, the current original confidence level will be reset. The sensor was then marked as "communication failure".

[0037] S33: Calculate spatial prediction values ; For each sensor , using Other than, the average confidence level of the previous period For other nearby sensors, inverse distance-weighted interpolation is used:

[0038] ; If the number of neighbors that meet the condition is less than 2, then let and the consistency items in S34 The value is directly assigned to 0.5. S34: Calculate the original reliability for the current period: The min function ensures that the expression within the parentheses is non-negative.

[0039] S35: Apply a first-order low-pass filter to the original confidence level to obtain the average confidence level. Setting initial ; S36: Perform status determination and suspicious item count; Maintain a suspicious counter for each sensor (initial value set to 0), if... Mark as "normal" and clear the suspicious counter; if If marked as "suspicious", the suspicious counter is incremented by 1; if Or suspicious counter 3. Mark as "Failed". Output the average confidence level for each neighboring sensor. and status indicators (normal / suspicious / invalid).

[0040] By combining self-confidence and spatial prediction consistency to calculate dynamic confidence, and then using filtering and suspicious counting to determine sensor status, the problem of large delay and easy misjudgment in traditional centralized fault detection is solved. Millisecond-level accurate fault detection is achieved at the edge, reducing the false alarm rate by more than 80%, and providing a reliable basis for fault-tolerant switching.

[0041] S4: The actuator acquires the coordinates of the control point and the set of trusted sensors. When the main sensor fails, it uses inverse distance weighted interpolation to reconstruct the parameters and outputs the reconstructed value as a feedback quantity. Furthermore, each actuator The following sub-steps are executed in each control cycle T: S41: Obtain control point coordinates Main sensor identifier And the status and coordinates of each sensor in the nearby sensor list. Measured values Average credibility .

[0042] Furthermore, set the maximum permissible interpolation standard deviation. and backup safety value Main sensor identifier Pre-configured as the sensor closest to or specified to the control point.

[0043] S42: Check the status of the main sensor: If the status of the main sensor is "normal", then use its measured value directly as the control feedback quantity. If the current condition is met, the system will jump to S5; otherwise, it will enter fault-tolerant refactoring mode.

[0044] S43: Filtering a set of trustworthy sensors: From a list of nearby sensors Selected if the status is "normal" and it is not the main sensor identifier The sensors constitute the candidate set. If so, a backup strategy is adopted: the most recent valid historical value is used first (if it exists), otherwise a backup safe value is used. It will then issue an alarm and then jump to S5.

[0045] S44: Calculate the reconstructed values ​​at the control points using inverse distance-weighted interpolation with a trusted set. ; S45: Perform reconstruction quality assessment and calculate the interpolation standard deviation: If the denominator is 0 (which theoretically will not happen), then directly set it to 0; if If the reconstruction is deemed unreliable, the system switches to the safety backup mode and issues an alarm; otherwise, the reconstruction value is used as the control feedback value. .

[0046] By selecting a set from nearby reliable sensors when the main sensor fails, reconstructing control point parameters using inverse distance weighted interpolation, and performing quality assessment, the problem of relying on redundant hardware and having a coarse fault tolerance mode in traditional solutions is solved. This achieves high-precision parameter reconstruction with zero redundant hardware, a reconstruction error of ≤5% of the range, and a reduction in hardware cost of more than 50%. At the same time, the reliability of reconstruction is ensured through quality assessment.

[0047] S5: The actuator provides feedback with the reconstructed value or the measured value, calculates the control command through PID and broadcasts the fault-tolerant status, and automatically switches back to the measured value after the sensor recovers. Furthermore, each actuator performs the following sub-steps in each control cycle T: S51: Obtain control feedback quantity Target set value PID parameters Output limiting Maximum fault tolerance time (e.g., 1800s), number of consecutive normal cycles N required for recovery (e.g., 10). Maintenance mode variable Mode (0=normal, 1=fault-tolerant, 2=safe backup) and the start time for entering fault-tolerant mode. .

[0048] S52: Calculation of Deviation ; S53: Perform a seamless handover (only when the mode changes); When switching from normal mode to fault-tolerant mode, or vice versa, the PID integral term is reset to prevent output jumps; assuming the PID form is: , then the new integral term is set to: And use it to update the integrator.

[0049] S54: The control output is calculated using an incremental PID algorithm. , Initial time Limit the output and drive the actuator to move.

[0050] S55: Fault-tolerant status broadcast: If currently in fault-tolerant mode A status message is broadcast every 10 cycles: ; S56: Self-healing recovery monitoring (executed only in fault-tolerant mode): Receives a failed master sensor identifier every cycle. The data packets are recalculated for their average credibility according to the method in S3. .

[0051] Maintain the counter: If ,but Add 1, otherwise reset to zero; when When the sensor has stabilized and recovered, proceed with the execution. Perform a seamless handover again (S53) and broadcast a recovery message. ; S57: Timeout Protection: If the fault-tolerant mode duration is... Then it will automatically switch to the safe backup mode: (Lock the output) and send a "sensor permanently failed" alarm to the central monitoring system.

[0052] By using incremental PID calculation, performing seamless switching during mode switching, and continuously monitoring failed sensors in fault-tolerant mode and automatically switching back to the measured value after recovery, the problems of output surges during mode switching and the need for manual intervention to restore sensors are solved. This achieves smooth control switching (without disturbance), self-healing recovery time ≤1 second, and no need for manual reset. At the same time, timeout protection ensures system safety.

[0053] This invention solves the problem of how nodes in a distributed environment identify spatial proximity relationships and provide a data foundation for subsequent collaborative fault tolerance by calculating Euclidean distance based on sensor coordinates and generating a list of neighboring sensors through the actuator node in S1. This achieves the beneficial effect of autonomously establishing a spatial topology without a centralized database. In S2, the sensor collects environmental parameters and calculates self-confidence based on power supply voltage, range, and rate of change before broadcasting, solving the problem of quantifying and distributing the sensor's own health status, thus providing reliable fault detection input for the edge side. In S3, the actuator node calculates dynamic reliability based on self-confidence and spatial prediction consistency, and determines the sensor status based on continuous periodic reliability changes after filtering. This solves the problems of large fault detection delays and easy misjudgments due to reliance on a central controller in traditional solutions, thus achieving millisecond-level accurate fault detection at the edge side. The system achieves beneficial effects such as fault detection and a false alarm rate reduction of over 80%. In S4, when the main sensor fails, control point parameters are reconstructed from nearby reliable sensors using inverse distance weighted interpolation, and quality assessment is performed. This solves the problems of traditional fault-tolerant schemes requiring redundant hardware and using fixed openings or historical averages as substitutes after a fault, leading to control failure or energy waste. This achieves high-precision dynamic parameter reconstruction with zero-redundancy hardware, reconstruction error ≤5% of the range, and a hardware cost reduction of over 50%. In S5, incremental PID calculation is used, seamless switching is performed during mode switching, and the system continuously monitors the failed sensor in fault-tolerant mode, automatically switching back to the measured value after recovery. This solves the problems of traditional schemes where mode switching easily generates output shocks and sensor recovery requires manual reset, resulting in poor maintainability. This achieves smooth control switching, self-healing recovery time ≤1 second, and no manual intervention required.

[0054] The present invention has been described in the above-described embodiments; however, these embodiments are merely examples for implementing the present invention. It must be noted that the disclosed embodiments do not limit the scope of the present invention. Conversely, any modifications and refinements made without departing from the spirit and scope of the present invention are within the scope of patent protection of the present invention.

Claims

1. A distributed sensor fault-tolerant cooperative control method based on spatial interpolation reconstruction, characterized in that: S1: The actuator receives data from various sensors. coordinate According to Euclidean distance Generate a list of nearby sensors It also pre-stores coordinate information and outputs a list of nearby sensors. For use in subsequent trust assessments; S2: The sensor collects environmental parameters and calculates the self-confidence level by integrating the self-diagnostic status. It then packages and broadcasts the measured value, self-confidence level, and timestamp for nearby actuators to receive. S3: The actuator receives data packets. Combined with pre-stored coordinates, the dynamic reliability of each sensor is calculated in each control cycle, and the sensor status is determined based on the reliability changes in consecutive cycles, and the average reliability and status flag are output. S4: The actuator acquires the coordinates of the control point and the set of trusted sensors. When the main sensor fails, it uses inverse distance weighted interpolation to reconstruct the parameters and outputs the reconstructed value as a feedback quantity. S5: The actuator provides feedback based on the reconstructed or measured value, calculates the control command using PID, and broadcasts the fault-tolerant status. Once the sensor recovers, it automatically switches back to the measured value.

2. The distributed sensor fault-tolerant cooperative control method based on spatial interpolation reconstruction as described in claim 2, characterized in that: The calculation of dynamic reliability in S3 includes: For each sensor, using the exception Spatial prediction values ​​are obtained by using inverse distance-weighted interpolation with other nearby sensors whose confidence level in the previous period is not lower than the trust threshold, in addition to the sensors mentioned above. ; Calculate the original credibility ,in The confidence level is the self-confidence level after aging, and R is the maximum permissible deviation; the average confidence level is obtained by performing a first-order low-pass filter on the original confidence level. .

3. The distributed sensor fault-tolerant cooperative control method based on spatial interpolation reconstruction as described in claim 1, characterized in that: The process of determining the sensor state in S3 includes: Maintain a suspicious counter for each sensor. If the average confidence level is not lower than the trust threshold, mark it as normal and reset the counter. If the average confidence level is between the fault threshold and the trust threshold, mark it as suspicious and increment the counter by 1. If the average confidence level is lower than the fault threshold or the counter reaches 3, mark it as failed.

4. The distributed sensor fault-tolerant cooperative control method based on spatial interpolation reconstruction as described in claim 1, characterized in that: The process of selecting a set of trustworthy sensors in S4 includes: selecting sensors that are in normal condition and are not the primary sensors from the list of neighboring sensors to form a candidate set; if the size of the candidate set is less than 2, a backup strategy is adopted, using the most recent valid historical value or a preset safety value as the feedback quantity.

5. The distributed sensor fault-tolerant cooperative control method based on spatial interpolation reconstruction as described in claim 1, characterized in that: The formula for inverse distance weighted interpolation in S4 is: ;in A collection of reliable sensors, The average confidence level of the sensor, These are measured values. For sensor coordinates, These are the coordinates of the control points.

6. The distributed sensor fault-tolerant cooperative control method based on spatial interpolation reconstruction as described in claim 5, characterized in that: The reconstruction quality assessment in S4 includes: calculating the interpolation standard deviation. ,like If the interpolation standard deviation exceeds the preset maximum allowable standard deviation, the reconstruction is deemed unreliable, and the system switches to the safe backup mode.

7. The distributed sensor fault-tolerant cooperative control method based on spatial interpolation reconstruction as described in claim 1, characterized in that: The seamless switching in S5 includes: when the mode switches from normal to fault-tolerant or from fault-tolerant to normal, the PID integral term is reset to, where This is the output from the previous cycle. represents the current deviation, and represents the proportional and differential coefficients.

8. The distributed sensor fault-tolerant cooperative control method based on spatial interpolation reconstruction as described in claim 1, characterized in that: The self-healing recovery in S5 includes: in fault-tolerant mode, the dynamic reliability of the failed main sensor is recalculated in each cycle. If the reliability is not lower than the trust threshold for N consecutive cycles, the sensor is determined to have recovered and automatically switches back to normal mode.

9. The distributed sensor fault-tolerant cooperative control method based on spatial interpolation reconstruction as described in claim 1, characterized in that: The S5 also includes timeout protection: if the fault tolerance mode lasts for more than the preset maximum fault tolerance time, it will automatically switch to the safety backup mode, lock the output and send an alarm.

10. A distributed sensor fault-tolerant cooperative control system implementing the method of any one of claims 1-9, characterized in that, include: Multiple intelligent sensor nodes, each with self-diagnostic capabilities, can generate self-confidence scores in real time; Multiple intelligent actuator nodes, each node has a built-in trust evaluation module, spatial interpolation reconstruction module and PID control module; The system achieves autonomous fault-tolerant control in the event of sensor failure through edge computing of actuator nodes.

Images