Rudder feedback sensor fault diagnosis method based on data and knowledge fusion under dual-redundancy sensing

Abstract

The invention discloses a data and knowledge fused rudder feedback sensor fault diagnosis method under dual-redundancy sensing. The method comprises the steps of obtaining a first detection statistic according to a difference value between dual-redundancy rudder feedback sensors and a preset scaling scale; obtaining a second detection statistical magnitude according to a difference value between the instruction and the first rudder feedback sensor and a preset scaling scale; obtaining a third detection statistical magnitude according to a difference value between the instruction and the second rudder feedback sensor and a preset scaling scale; and judging whether to enter fault alarm decision logic and fault positioning decision logic or not according to the relationship between the first detection statistic and the first detection threshold. According to the invention, reliable alarm and positioning of the fault rudder feedback sensor under dual-redundancy sensing are realized.

Description

Technical Field

[0001] This invention belongs to the field of steering system and fault diagnosis technology, and in particular relates to a fault diagnosis method for steering feedback sensors that integrates data and knowledge under dual-redundant sensing. Background Technology

[0002] As a key sensing and control element, the rudder feedback sensor monitors and feeds back information such as the position and angle of the rudder surface in real time, enabling high-precision tracking control of the rudder system. Affected by a combination of external and internal factors such as mechanical vibration, temperature changes, and component wear, the rudder feedback sensor is a vulnerable link in the rudder system prone to failure. However, due to design complexity, cost constraints, and space limitations, the redundancy configuration of the rudder feedback sensor is often limited, requiring a trade-off between performance, reliability, and cost. On the one hand, excessive sensor redundancy increases system weight and size, leading to higher maintenance costs; the optimal sensor redundancy configuration should be as streamlined as possible while meeting reliability requirements. On the other hand, when the rudder feedback sensor fails, the limited redundancy configuration may not provide sufficient information to accurately locate the fault, posing a challenge to fault diagnosis and fault-tolerant design of the rudder system.

[0003] For rudder systems, the configuration of dual-redundant rudder feedback sensors is a relatively streamlined design architecture. However, the redundancy information provided by dual-redundant feedback sensor data is limited, making it difficult to accurately identify complex fault modes and faulty sensors. In fact, the mechanism of fault propagation can often be characterized by the temporal changes in data, but current fault diagnosis methods do not fully utilize the knowledge of fault propagation mechanisms. Summary of the Invention

[0004] The technical problem solved by this invention is to overcome the shortcomings of the prior art and provide a fault diagnosis method for rudder feedback sensors based on data and knowledge fusion under dual-redundant sensing, so as to realize the reliability alarm and location of faulty rudder feedback sensors under dual-redundant sensing.

[0005] The objective of this invention is achieved through the following technical solution: a fault diagnosis method for a rudder feedback sensor based on data and knowledge fusion under dual-redundant sensing, comprising: obtaining a first detection statistic based on the difference between dual-redundant rudder feedback sensors and a preset scaling scale; obtaining a second detection statistic based on the difference between the command and the first rudder feedback sensor and a preset scaling scale; obtaining a third detection statistic based on the difference between the command and the second rudder feedback sensor and a preset scaling scale; wherein, the first rudder feedback sensor participates in closed-loop control, and the second rudder feedback sensor serves as a redundant reference; and determining whether to enter the fault alarm decision logic and fault location decision logic based on the relationship between the first detection statistic and a first detection threshold.

[0006] In the above-mentioned fault diagnosis method for rudder feedback sensors based on data and knowledge fusion under dual-redundant sensing, the first detection statistic... It can be obtained through the following formula: ; in, This represents the difference between the dual-redundant rudder feedback sensors. This is for scaling up or down.

[0007] In the above-mentioned fault diagnosis method for rudder feedback sensors based on data and knowledge fusion under dual-redundant sensing, the second detection statistic... It can be obtained through the following formula: ; in, The difference between the command and the first rudder feedback sensor. This is for scaling up or down.

[0008] In the above-mentioned fault diagnosis method for rudder feedback sensors based on data and knowledge fusion under dual-redundant sensing, the third detection statistic... It can be obtained through the following formula: ; in, The difference between the command and the second rudder feedback sensor. This is for scaling up or down.

[0009] In the aforementioned fault diagnosis method for rudder feedback sensors based on data and knowledge fusion under dual-redundant sensing, determining whether to enter the fault alarm decision logic and fault location decision logic according to the relationship between the first detection statistic and the first detection threshold includes: if a fault is detected... And alarm decision signals Then set the diagnostic decision initiation flag. If the value is "1", the system will sequentially enter the fault alarm decision logic and the fault location decision logic; otherwise, it will output an alarm decision signal. Fault location decision signal ;in, This is the first detection statistic. This is the first detection threshold.

[0010] In the above-mentioned fault diagnosis method for rudder feedback sensors based on data and knowledge fusion under dual-redundant sensing, the fault alarm decision logic includes: Step B1: If Then set the fault alarm flag. Otherwise, set a fault alarm flag. ; This is the first detection statistic. The first detection threshold is set; Step B2: If the value of the first counting register is... Less than the maximum count limit for delayed alarm time Then place and will The value is pushed onto the first stack. The length of the first stack is equal to the upper limit of the delay alarm count; Step B3: If the value of the first count register... Equal to the upper limit of delayed alarm time count Then place Calculate the first stack alarm numbers That is, the number of elements that are "1", based on the number of alarms. and the upper limit of delayed alarm time count Obtain the alarm count ratio within the delayed alarm time. Step B4: If Then set an alarm decision signal. Otherwise, set an alarm decision signal. ;in, The fault alarm ratio threshold; Step B5: If the value of the first counter register... Greater than the maximum count limit for delayed alarm time Then reset the diagnostic decision initiation flag. Clear the first stack and the value of the first counter register .

[0011] In the above-mentioned fault diagnosis method for rudder feedback sensors based on data and knowledge fusion under dual-redundant sensing, the fault location decision logic includes: Step C1: If and Then set the fault location auxiliary label. ;like and Then place ;like and Then place Otherwise, place ;in, This is the second detection statistic. The second detection threshold, This is the third detection statistic. The third detection threshold; Step C2: If the value of the second counter register... Less than the upper limit of fault location time count Then place and will The value is pushed onto the second stack. The length of the second stack is equal to the upper limit of the fault location time count; Step C3: If the value of the second count register... Equal to the upper limit of fault location time count Then place Traverse the second stack In Value, if it exists Then set the first auxiliary identifier to exist flag bit. Otherwise If it exists Then set the second auxiliary identifier to exist flag bit. Otherwise If it exists Then set the third auxiliary identifier to exist flag bit. Otherwise Step C4: Calculate the values ​​in the stack respectively. The smallest sequence number is denoted as the first auxiliary identifier existence flag index. ; calculate the stack respectively The smallest sequence number is denoted as the index of the second auxiliary identifier existence flag. ; calculate the stack respectively The smallest sequence number is denoted as the index of the third auxiliary identifier existence flag. ;like and and Then let , and Step C5: If Then set the fault location decision signal If the first rudder feedback sensor malfunctions, proceed to step C6; otherwise, proceed to step C6. Step C6: If... If yes, proceed to step C7; otherwise, proceed to step C8; Step C7: If Then place If the second rudder feedback sensor malfunctions, otherwise determine whether the condition is met. ;like Then place Otherwise Step C8: If Then place Otherwise, effective fault location cannot be given; Step C9: If the value of the second counter register is... Greater than the upper limit of fault location time count Waiting for the value of the first counter register Greater than the maximum count limit for delayed alarm time Clear the second stack and the value of the second counter register .

[0012] In the aforementioned fault diagnosis method for rudder feedback sensors based on data and knowledge fusion under dual-redundant sensing, the alarm count ratio within the delayed alarm time is... It can be obtained through the following formula: .

[0013] A fault diagnosis system for a rudder feedback sensor based on data and knowledge fusion under dual-redundant sensing includes: a first module, used to diagnose faults based on the difference between dual-redundant rudder feedback sensors and a preset scaling scale. Obtain the first detection statistic Based on the difference between the command and the first rudder feedback sensor and the preset scaling scale. Obtain the second detection statistic Based on the difference between the command and the second rudder feedback sensor and the preset scaling scale. Obtain the third detection statistic The second module is used to determine whether to enter the fault alarm decision logic and fault location decision logic based on the relationship between the first detection statistic and the first detection threshold.

[0014] An electronic device includes: a memory for storing computer-readable instructions; and a processor for executing the computer-readable instructions to perform a fault diagnosis method for a rudder feedback sensor based on data and knowledge fusion under dual-redundant sensing.

[0015] Compared with the prior art, the present invention has the following advantages: (1) This invention realizes the reliability alarm and location of faulty rudder feedback sensor under dual-redundant sensing; (2) The present invention adopts the data and knowledge fusion approach to design fault location decision logic. Based on the knowledge of the propagation mechanism of the closed loop of the rudder feedback sensor under the control action, the change sequence of the second detection statistic and the third detection statistic compared with the threshold is analyzed through the fault location auxiliary identifier and the auxiliary identifier existence flag index, so as to realize the location of the fault rudder feedback sensor under dual redundancy sensing and consider more comprehensive fault situations. Attached Figure Description

[0016] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1 This is a schematic diagram of the fault diagnosis method for rudder feedback sensor based on data and knowledge fusion under dual-redundant sensing provided in an embodiment of the present invention; Figure 2 This is a flowchart of the fault alarm decision logic provided in an embodiment of the present invention; Figure 3 This is a flowchart of the fault location decision logic provided in an embodiment of the present invention; Figure 4 This is a diagram illustrating the effect of fault detection and location decision-making provided in an embodiment of the present invention. Detailed Implementation

[0017] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the disclosure to those skilled in the art. It should be noted that, unless otherwise specified, the embodiments and features described herein can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0018] How to integrate limited redundant data information with knowledge of fault propagation mechanisms, fully extract potential information under limited sensing redundancy, and design a fault diagnosis method for rudder feedback sensors that integrates data and knowledge under dual-redundancy sensing is of great practical significance for achieving rapid and accurate diagnosis of rudder feedback sensor faults and improving the reliability and safety of rudder system operation.

[0019] like Figure 1 As shown, this embodiment provides a fault diagnosis method for a rudder feedback sensor based on data and knowledge fusion under dual-redundant sensing. The method includes: Based on the difference between the dual-redundant rudder feedback sensors and the preset scaling scale Obtain the first detection statistic Based on the difference between the command and the first rudder feedback sensor and the preset scaling scale. Obtain the second detection statistic Based on the difference between the command and the second rudder feedback sensor and the preset scaling scale. Obtain the third detection statistic Based on the relationship between the first detection statistic and the first detection threshold, determine whether to enter the fault alarm decision logic and the fault location decision logic.

[0020] This embodiment calculates a first detection statistic, a second detection statistic, and a third detection statistic based on the differences between dual-redundant rudder feedback sensors, the differences between the command and the sensors participating in the closed-loop feedback, and the differences between the command and another sensor. Based on the relationship between the first detection statistic and a first detection threshold, it calculates the alarm count ratio within the delayed alarm time, achieving reliable alarm for fault occurrence. Based on the propagation mechanism of the closed-loop circuit of rudder feedback sensing under control action, it analyzes the temporal changes of the second and third detection statistics compared to the threshold using fault location auxiliary identifiers and auxiliary identifier presence flags, achieving fault location of the rudder feedback sensor under dual-redundant sensing. It mainly includes three parts: ① fault alarm and location design framework; ② fault alarm decision logic; ③ fault location decision logic.

[0021] This embodiment calculates the difference between two redundant position sensors, the difference between the command and the position sensor participating in the closed-loop feedback, and the difference between the command and another sensor. It integrates the scaling design of the first detection statistic, the second detection statistic, and the third detection statistic, and determines whether to initiate a diagnostic decision based on the relationship between the first detection statistic and the first detection threshold, and enters the fault alarm decision logic and the fault location decision logic.

[0022] First detection statistic It can be obtained through the following formula: ; in, This represents the difference between the dual-redundant rudder feedback sensors. This is for scaling up or down.

[0023] Second detection statistic It can be obtained through the following formula: ; in, The difference between the command and the first rudder feedback sensor. This is for scaling up or down.

[0024] Third detection statistic It can be obtained through the following formula: ; in, The difference between the command and the second rudder feedback sensor. This is for scaling up or down.

[0025] Based on the relationship between the first detection statistic and the first detection threshold, the determination of whether to enter the fault alarm decision logic and fault location decision logic includes: if a fault is detected... And alarm decision signals Then set the diagnostic decision initiation flag. If the value is "1", the system will sequentially enter the fault alarm decision logic and the fault location decision logic; otherwise, it will output an alarm decision signal. Fault location decision signal ;in, This is the first detection statistic. This is the first detection threshold.

[0026] The fault alarm decision logic includes: Step B1: If Then set the fault alarm flag. Otherwise, set a fault alarm flag. ; This is the first detection statistic. The first detection threshold is set; Step B2: If the value of the first counting register is... Less than the maximum count limit for delayed alarm time Then place and will The value is pushed onto the first stack. The length of the first stack is equal to the upper limit of the delay alarm count; Step B3: If the value of the first count register... Equal to the upper limit of delayed alarm time count Then place Calculate the first stack alarm numbers That is, the number of elements that are "1", based on the number of alarms. and the upper limit of delayed alarm time count Obtain the alarm count ratio within the delayed alarm time. Step B4: If Then set an alarm decision signal. Otherwise, set an alarm decision signal. ;in, The fault alarm ratio threshold; Step B5: If the value of the first counter register... Greater than the maximum count limit for delayed alarm time Then reset the diagnostic decision initiation flag. Clear the first stack and the value of the first counter register .

[0027] The fault alarm decision logic adopts a delayed alarm method. A first counting register is designed. If the value of the first counting register is less than the upper limit of the delayed alarm time count, the fault alarm flag is stored in the first stack. The length of the first stack is equal to the value of the upper limit of the delayed alarm time count. If the alarm count ratio within the delayed alarm time exceeds the set fault alarm ratio threshold, the alarm decision signal is set to "1" to achieve reliable fault alarm.

[0028] Fault location decision logic includes: Step C1: If and Then set the fault location auxiliary label. ;like and Then place ;like and Then place Otherwise, place ;in, This is the second detection statistic. The second detection threshold, This is the third detection statistic. The third detection threshold; Step C2: If the value of the second counter register... Less than the upper limit of fault location time count Then place and will The value is pushed onto the second stack. The length of the second stack is equal to the upper limit of the fault location time count; Step C3: If the value of the second count register... Equal to the upper limit of fault location time count Then place Traverse the second stack In Value, if it exists Then set the first auxiliary identifier to exist flag bit. Otherwise If it exists Then set the second auxiliary identifier to exist flag bit. Otherwise If it exists Then set the third auxiliary identifier to exist flag bit. Otherwise Step C4: Calculate the values ​​in the stack respectively. The smallest sequence number is denoted as the first auxiliary identifier existence flag index. ; calculate the stack respectively The smallest sequence number is denoted as the index of the second auxiliary identifier existence flag. ; calculate the stack respectively The smallest sequence number is denoted as the index of the third auxiliary identifier existence flag. ;like and and Then let , and Step C5: If Then set the fault location decision signal If the first rudder feedback sensor malfunctions, otherwise (i.e.) Proceed to step C6; Step C6: If If yes, proceed to step C7; otherwise ( Proceed to step C8; Step C7: If Then place If the second rudder feedback sensor malfunctions, otherwise determine whether the condition is met. ;like Then place Otherwise ; where & represents "and", a relationship that is satisfied simultaneously; Step C8: If Then place ,otherwise( Unable to provide a valid fault location; Step C9: If the value of the second counter register... Greater than the upper limit of fault location time count Waiting for the value of the first counter register Greater than the maximum count limit for delayed alarm time Clear the second stack and the value of the second counter register .

[0029] The fault location decision logic integrates data with the mechanism of fault propagation. Based on the closed-loop propagation characteristics of the rudder feedback sensor under control, it designs an auxiliary fault location identifier and an auxiliary identifier presence index. It analyzes the time sequence of changes of the second and third detection statistics compared with the threshold, realizes the location of the faulty rudder feedback sensor under dual-redundant sensing, and has a certain robustness to the case where the detection statistics exceed the limit due to the decline in tracking performance or the conservative threshold design.

[0030] The fault location decision logic is designed with a second stack to store the values ​​of fault location auxiliary identifiers. The length of the second stack is equal to the upper limit of the fault location time count. The values ​​of the fault location auxiliary identifiers in the second stack are traversed to obtain the presence flags of the first, second, and third auxiliary identifiers. The smallest sequence number of the fault location auxiliary identifiers in the second stack that is "1", "2", or "3" is taken and recorded as the index of the presence flag of the first, second, and third auxiliary identifiers.

[0031] Alarm count ratio during delayed alarm time It can be obtained through the following formula: .

[0032] This embodiment can be divided into three layers in terms of design logic. The first layer is the fault alarm and location design framework located at the outermost layer (see step A), the second layer is the fault alarm decision logic (see step B), and the third layer is the fault location decision logic (see step C).

[0033] Step A specifically includes the following steps: Step A1: Given the first detection threshold, the second detection threshold, and the third detection threshold, as well as the upper limit of the delayed alarm time count, the fault alarm ratio threshold, the upper limit of the fault location time count, and the scaling scale. Step A2: Initialize the fault location auxiliary identifier, the fault alarm identifier, the alarm decision signal, and the fault location decision signal. Step A3: Calculate the difference between the two redundant steering feedback sensors, the difference between the command and the sensor participating in the closed-loop feedback, and the difference between the command and another sensor. Step A4: Calculate the first detection statistic, the second detection statistic, and the third detection statistic based on the differences calculated in Step A3 and the scaling scale, respectively. Step A5: Assign a value to the diagnostic decision start flag based on the relationship between the first detection statistic and the first detection threshold. Step A6: If the diagnostic decision start flag is detected as "1", then sequentially enter the fault alarm decision logic and the fault location decision logic, and output the alarm decision signal and the fault location decision signal.

[0034] Specifically, step A1: Given a first detection threshold Second detection threshold Third detection threshold Maximum count of delayed alarm time Fault alarm ratio threshold Fault location time count limit and scaling .

[0035] Step A2: Initialize fault location auxiliary markers Fault alarm labels Alarm decision signals Fault location decision signals .

[0036] Step A3: Calculate the difference between the two redundant rudder feedback sensors, denoted as The difference between the calculation command and the sensor participating in the closed-loop feedback is denoted as... Calculate the difference between the command and another sensor, and record it as . .

[0037] Step A4: Calculate the first detection statistic as follows. Second detection statistics and the third detection statistics ,

[0038] Step A5: If detected And alarm decision signals Then set the diagnostic decision initiation flag. Set the value to "1"; otherwise, output an alarm decision signal. Fault location decision signal .

[0039] Step A6: If detected If the value is "1", the fault alarm decision logic and fault location decision logic will be entered sequentially; otherwise, an alarm decision signal will be output. Fault location decision signal .

[0040] Step B specifically includes the following steps: Step B1: Assign a value to the fault alarm identifier based on the relationship between the first detection statistic and the first detection threshold. Step B2: If the value of the first counter register is less than the upper limit of the delayed alarm time count, send the value of the fault alarm identifier to the first stack. Step B3: If the value of the first counter register is equal to the upper limit of the delayed alarm time count, calculate the number of alarms in the first stack and calculate the alarm count ratio within the delayed alarm time. Step B4: If the alarm count ratio within the delayed alarm time exceeds the set fault alarm ratio threshold, set the alarm decision signal to "1". Step B5: If the value of the counter register is greater than the upper limit of the delayed alarm time count, reset the diagnostic decision start flag and clear the first stack and the value of the first counter register.

[0041] Specifically, step B1: If Then set the fault alarm flag. Otherwise, set a fault alarm flag. .

[0042] Step B2: If the value of the first counter register... Less than the maximum count limit for delayed alarm time Then place and will The value is pushed onto the first stack. The length of the first stack is equal to the upper limit of the delay alarm count.

[0043] Step B3: If the value of the first counter register... Equal to the upper limit of delayed alarm time count Then place Calculate the first stack The alarm count, i.e., the number of elements that are "1", is denoted as The alarm count ratio during the delay alarm time is calculated as follows: .

[0044] Step B4: If Then place Otherwise, place .

[0045] Step B5: If the value of the first counter register... Greater than the maximum count limit for delayed alarm time Then reset the diagnostic decision initiation flag. Clear the first stack and the value of the first counter register .

[0046] Step C specifically includes the following steps: Step C1: Assign fault location auxiliary identifiers based on the relationship between the second and third detection statistics and the second and third detection thresholds. Step C2: If the value of the second counter register is less than the upper limit of the fault location time count, then push the value of the fault location auxiliary identifier into the second stack. Step C3: If the value of the second counter register is equal to the upper limit of the fault location time count, then traverse the values ​​of the fault location auxiliary identifiers in the second stack, and set the first auxiliary identifier presence flag, the second auxiliary identifier presence flag, and the third auxiliary identifier presence flag according to the values ​​of the fault location auxiliary identifiers in the second stack. Step C4: Calculate the smallest sequence number of the fault location auxiliary identifiers in the second stack that are "1", "2", or "3", and record them as the indexes of the first auxiliary identifier presence flag, the second auxiliary identifier presence flag, and the third auxiliary identifier presence flag, respectively. Step C5: If the first auxiliary identifier presence flag is "1", the rudder feedback sensor participating in the closed-loop control is located to be faulty; otherwise, proceed to step C6. Step C6: Determine whether to jump to step C7 or step C8 based on the third auxiliary identifier presence flag. Step C7: Based on the presence flag of the second auxiliary identifier, the index of the presence flag of the second auxiliary identifier, and the index of the presence flag of the third auxiliary identifier, assign the fault location decision signal and locate the fault sensor. Step C8: Based on the presence flag of the second auxiliary identifier, assign the fault location decision signal and locate the fault sensor. Step C9: If the value of the second counter register is greater than the upper limit of the fault location time count, wait for the value of the first counter register to be greater than the upper limit of the delayed alarm time count, then clear the second stack and the second counter register.

[0047] Based on the propagation mechanism of the closed-loop circuit of the rudder feedback sensor under control action, a fault location decision logic with dual redundancy is designed. The specific steps are as follows: Step C1: If and Then place ;like and Then place ;like and Then place Otherwise, place .

[0048] Step C2: If the value of the second counter register Less than the upper limit of fault location time count Then place and will The value is pushed onto the second stack. The length of the second stack is equal to the upper limit of the fault location time count.

[0049] Step C3: If the value of the second counter register... Equal to the upper limit of fault location time count Then place Traverse the second stack In Value, if it exists Then set the first auxiliary identifier to exist flag bit. Otherwise If it exists Then set the second auxiliary identifier to exist flag bit. Otherwise If it exists Then set the third auxiliary identifier to exist flag bit. Otherwise .

[0050] Step C4: Calculate the values ​​in the stack separately. The smallest sequence number is denoted as the first auxiliary identifier existence flag index. ; calculate the stack respectively The smallest sequence number is denoted as the index of the second auxiliary identifier existence flag. ; calculate the stack respectively The smallest sequence number is denoted as the index of the third auxiliary identifier existence flag. ;like Then let .

[0051] Step C5: If Then place If the rudder feedback sensor involved in the closed-loop control is faulty, proceed to step C6.

[0052] Step C6: If If yes, proceed to step C7; otherwise, determine whether the condition is met. Proceed to step C8.

[0053] Step C7: If Then place If the other sensor used as a redundant reference for positioning fails, otherwise determine whether the condition is met. ;like Then place Otherwise .

[0054] Step C8: If Then place Otherwise, effective fault location cannot be provided.

[0055] Step C9: If the value of the second counter register... Greater than the upper limit of fault location time count Waiting for the value of the first counter register Greater than the maximum count limit for delayed alarm time Clear the second stack and the value of the second counter register .

[0056] This embodiment is a rudder system. Given a certain command signal, a sudden bias fault with a bias amount of 1 is injected into the rudder feedback sensor participating in the closed-loop control. The fault injection time is 1 second to meet the design verification requirements.

[0057] Figure 4 This implementation case demonstrates the effectiveness of fault detection and location decision-making. It can correctly identify faulty rudder feedback sensors and has a certain degree of robustness to conservative designs that reduce tracking performance or threshold.

[0058] This embodiment also provides a fault diagnosis system for a rudder feedback sensor based on data and knowledge fusion under dual-redundant sensing. The system includes: a first module, used to diagnose faults in rudder feedback sensors based on the difference between dual-redundant sensors and a preset scaling scale. Obtain the first detection statistic Based on the difference between the command and the first rudder feedback sensor and the preset scaling scale. Obtain the second detection statistic Based on the difference between the command and the second rudder feedback sensor and the preset scaling scale. Obtain the third detection statistic The second module is used to determine whether to enter the fault alarm decision logic and fault location decision logic based on the relationship between the first detection statistic and the first detection threshold.

[0059] This embodiment also provides an electronic device, including: a memory for storing computer-readable instructions; and a processor for running the computer-readable instructions to perform a fault diagnosis method for a rudder feedback sensor based on data and knowledge fusion under dual-redundant sensing.

[0060] This embodiment uses a data and knowledge fusion approach to design fault location decision logic. Based on the propagation mechanism of the closed loop of the rudder feedback sensor under control, it analyzes the time sequence of changes of the second and third detection statistics compared with the threshold through fault location auxiliary indicators and auxiliary indicator existence flag index, thereby realizing the location of the faulty rudder feedback sensor under dual-redundant sensing and considering more comprehensive fault scenarios.

[0061] The fault location logic in this embodiment is robust to situations where the detection statistic exceeds the limit due to decreased tracking performance or conservative threshold design. If the tracking performance is relatively stable under operating conditions, the threshold design margin is sufficient, and there is a certain tolerance for occasional false alarms, then this fault location logic can be degenerated into the following simple logic: If If the rudder feedback sensor involved in the closed-loop control is faulty, then the rudder feedback sensor is malfunctioning; if and If the other sensor used as a redundant reference fails, then the positioning sensor will malfunction.

[0062] Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make possible changes and modifications to the technical solutions of the present invention by utilizing the methods and techniques disclosed above without departing from the spirit and scope of the present invention. Therefore, any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solutions of the present invention shall fall within the protection scope of the technical solutions of the present invention.

Claims

1. A fault diagnosis method for a rudder feedback sensor based on data and knowledge fusion under dual-redundant sensing, characterized in that... include: The first detection statistic is obtained based on the difference between the dual-redundant rudder feedback sensors and the preset scaling scale; The second detection statistic is obtained based on the difference between the command and the first rudder feedback sensor and a preset scaling factor; the third detection statistic is obtained based on the difference between the command and the second rudder feedback sensor and a preset scaling factor; wherein, the first rudder feedback sensor participates in closed-loop control, and the second rudder feedback sensor serves as a redundant reference. Based on the relationship between the first detection statistic and the first detection threshold, it is determined whether to enter the fault alarm decision logic and the fault location decision logic.

2. The method for fault diagnosis of a rudder feedback sensor based on data and knowledge fusion under dual-redundant sensing as described in claim 1, characterized in that: First detection statistic It can be obtained through the following formula: ； in, This represents the difference between the dual-redundant rudder feedback sensors. This is for scaling up or down.

3. The method for fault diagnosis of a rudder feedback sensor based on data and knowledge fusion under dual-redundant sensing as described in claim 1, characterized in that: Second detection statistic It can be obtained through the following formula: ； in, The difference between the command and the first rudder feedback sensor. This is for scaling up or down.

4. The method for fault diagnosis of a rudder feedback sensor based on data and knowledge fusion under dual-redundant sensing as described in claim 1, characterized in that: Third detection statistic It can be obtained through the following formula: ； in, The difference between the command and the second rudder feedback sensor. This is for scaling up or down.

5. The method for fault diagnosis of a rudder feedback sensor based on data and knowledge fusion under dual-redundant sensing as described in claim 1, characterized in that: Based on the relationship between the first detection statistic and the first detection threshold, the determination of whether to enter the fault alarm decision logic and fault location decision logic includes: If detected And alarm decision signals Then set the diagnostic decision initiation flag. If the value is "1", the system will proceed with the fault alarm decision logic and the fault location decision logic in sequence; otherwise, an alarm decision signal will be output. Fault location decision signal ;in, This is the first detection statistic. This is the first detection threshold.

6. The method for fault diagnosis of a rudder feedback sensor based on data and knowledge fusion under dual-redundant sensing as described in claim 1 or 5, characterized in that: The fault alarm decision logic includes: Step B1: If Then set the fault alarm flag. Otherwise, set a fault alarm flag. ; This is the first detection statistic. The first detection threshold; Step B2: If the value of the first counter register... Less than the maximum count limit for delayed alarm time Then place and will The value is pushed onto the first stack. The length of the first stack is equal to the upper limit of the delay alarm count; Step B3: If the value of the first counter register... Equal to the upper limit of delayed alarm time count Then place Calculate the first stack alarm numbers That is, the number of elements that are "1", based on the number of alarms. and the maximum count of delayed alarm time Obtain the alarm count ratio within the delayed alarm time. ; Step B4: If Then set an alarm decision signal. Otherwise, set an alarm decision signal. ;in, The fault alarm ratio threshold; Step B5: If the value of the first counter register... Greater than the maximum count limit for delayed alarm time Then reset the diagnostic decision initiation flag. Clear the first stack and the value of the first counter register .

7. The method for fault diagnosis of a rudder feedback sensor based on data and knowledge fusion under dual-redundant sensing as described in claim 6, characterized in that: Fault location decision logic includes: Step C1: If and Then set the fault location auxiliary label. ;like and Then place ;like and Then place Otherwise, place ;in, This is the second detection statistic. This is the second detection threshold. This is the third detection statistic. The third detection threshold; Step C2: If the value of the second counter register Less than the upper limit of fault location time count Then place and will The value is pushed onto the second stack. The length of the second stack is equal to the upper limit of the fault location time count; Step C3: If the value of the second counter register... Equal to the upper limit of fault location time count Then place Traverse the second stack In Value, if it exists Then set the first auxiliary identifier to exist flag bit. Otherwise If it exists Then set the second auxiliary identifier to exist flag bit. Otherwise If it exists Then set the third auxiliary identifier to exist flag bit. Otherwise ; Step C4: Calculate the values ​​in the stack separately. The smallest sequence number is denoted as the first auxiliary identifier existence flag index. ; calculate the stack respectively The smallest sequence number is denoted as the index of the second auxiliary identifier existence flag. ; calculate the stack respectively The smallest sequence number is denoted as the index of the third auxiliary identifier existence flag. ;like and and Then let , and ; Step C5: If Then set the fault location decision signal If the first rudder feedback sensor malfunctions, proceed to step C6; otherwise, proceed to step C6. Step C6: If If yes, proceed to step C7; otherwise, proceed to step C8. Step C7: If Then place If the second rudder feedback sensor malfunctions, otherwise determine whether the condition is met. ;like Then place Otherwise ; Step C8: If Then place Otherwise, effective fault location cannot be provided. Step C9: If the value of the second counter register... Greater than the upper limit of fault location time count Waiting for the value of the first counter register Greater than the maximum count limit for delayed alarm time Clear the second stack and the value of the second counter register .

8. The method for fault diagnosis of a rudder feedback sensor based on data and knowledge fusion under dual-redundant sensing as described in claim 6, characterized in that: Alarm count ratio during delayed alarm time It can be obtained through the following formula: 。 9. A fault diagnosis system for a rudder feedback sensor based on data and knowledge fusion under dual-redundant sensing, characterized in that... include: The first module is used to obtain the first detection statistic based on the difference between the dual-redundant rudder feedback sensors and the preset scaling scale. The second detection statistic is obtained based on the difference between the command and the first rudder feedback sensor and a preset scaling scale; the third detection statistic is obtained based on the difference between the command and the second rudder feedback sensor and a preset scaling scale. The second module is used to determine whether to enter the fault alarm decision logic and fault location decision logic based on the relationship between the first detection statistic and the first detection threshold.

10. An electronic device, characterized in that, include: Memory: Used to store computer-readable instructions; as well as Processor: for executing the computer-readable instructions to perform the method as described in any one of claims 1 to 8.

Images