A distributed sensor fault diagnosis method and system based on multi-moving body networking system

Abstract

The present application provides a distributed sensor fault diagnosis method and system based on a multi-moving body networking system, the method comprising: establishing a mathematical model of the dynamics of the multi-moving body networking system; establishing a fault speed sensor model and a healthy speed sensor model affected by measurement interference based on the sensor principle; selecting appropriate residual variables based on the dynamics model to establish an observer with residual variables and selecting appropriate observation thresholds; exploring the applicable conditions of the residual variable fault observer through simulation experiments; designing and establishing an auxiliary state observer based on the residual variable observer to assist in diagnosing sensor faults; and obtaining a fault diagnosis method through simulation analysis. The sensor fault diagnosis method proposed in the present application can confirm sensor faults in the case of multiple sensor faults and find the sensor fault location.

Description

Technical Field

[0001] The present invention relates to a distributed sensor fault diagnosis method, and in particular to a distributed sensor fault diagnosis method and system based on a multi-moving body networking system. Background Art

[0002] Sensors are usually designed for specific working conditions. However, in actual measurements, they may be affected by factors such as temperature, humidity, vibration, and dust, which may exceed the working conditions within their design range. Sensors may fail under conditions such as excessively high or low pressure, excessive vibration, and extreme temperature. For multi-moving body networking systems, since information related to motion posture is collected, processed, and transmitted through various sensors of the information control system, the effect and performance of sensor fault diagnosis are of great significance to ensuring the safety and reliability of the operation of multi-moving body networking systems.

[0003] There are many different sensor fault diagnosis methods in use today, and statistical methods are more widely used and more mature. However, this approach also has great limitations: first, the statistical results must be accurate; second, the threshold selection must be reasonable, otherwise misjudgment may occur. Therefore, in order to better meet the safety needs of increasingly complex and intelligent sensors, it is necessary to further explore and develop sensor fault diagnosis methods based on statistical models and combined with model-based methods. Summary of the invention

[0004] In order to solve the problem that each sensor in a distributed speed sensor network system of a multi-moving body networking system is affected by measurement interference, resulting in erroneous measurement results, the present application provides a distributed sensor fault diagnosis method and system based on a multi-moving body networking system, which can effectively diagnose sensor faults and find the location of the faulty sensor in the event of multiple sensor failures.

[0005] In order to achieve the above object, the present invention adopts the following technical solution:

[0006] A distributed sensor fault diagnosis method based on a multi-moving body networking system, comprising:

[0007] Step 1: Establish a mathematical model of the dynamics of the multi-moving body network system;

[0008] Step 2: Establish a faulty speed sensor model and a healthy speed sensor model affected by measurement interference based on the sensor principle;

[0009] Step 3: Based on the mathematical model of the dynamics of the multi-moving body network system, an observer with residual variables of the faulty speed sensor model and the healthy speed sensor model is established, and an observation error threshold is determined;

[0010] Step 4: Based on the observation error threshold, determine the applicable conditions of the observer with residual variables through simulation;

[0011] Step 5: Establish an auxiliary state observer based on the observer with residual variables to assist in diagnosing sensor faults;

[0012] Step 6: Based on the observation error threshold, the applicable conditions of the observer and the auxiliary state observer, the fault diagnosis strategy is obtained through simulation analysis to determine the faulty sensor.

[0013] Furthermore, the mathematical model of the multi-moving body networking system dynamics is:

[0014] ;

[0015] in, , for The derivative of Represents the mass of the main moving body in the same moving body in the multi-moving body networking system. Represents the mass of the auxiliary moving body connected to the main moving body; , Respectively represent their displacement relative to the ground; , Respectively , The first-order derivative of ; F is positive, indicating the traction force output by the main moving body in the multi-moving body network system during the acceleration phase, F is negative, indicating the braking force output by the main moving body in the multi-moving body network system during the deceleration phase, b indicates the adhesion coefficient between the main moving body and the moving carrier in the multi-moving body network system, k indicates the elastic coefficient between the main moving body and the auxiliary moving body, , , respectively and The speed is measured by the corresponding speed sensor, and the output Y is the speed , the input is the traction force F.

[0016] Furthermore, the sensor principle is:

[0017] ;

[0018] Where D represents the wheel diameter, represents the power shaft speed, Represents the sensor output.

[0019] Furthermore, the observer with residual variables is:

[0020] ;

[0021] in, , , and , n, i, j respectively refer to the i-th sensor and j-th sensor of the main moving body, i=1,2,3,...,n, j=1,2,3,...,n, for The derivative of; n represents the number of sensors, , They refer to the measured values ​​of the main moving body's No. i speed sensor and No. j speed sensor respectively; , They refer to the displacements derived from the corresponding velocity sensor measurements.

[0022] Furthermore, the observation error threshold is:

[0023] ;

[0024] in, represents the observation error threshold, It is the daily wear and tear of the wheels without affecting the normal operation of the main moving body.

[0025] Furthermore, the applicable conditions of the observer with residual variables are: , Indicates the number of speed sensors that are not faulty on the same main moving body. The specific faulty sensor cannot be found.

[0026] Furthermore, the auxiliary state observer is:

[0027] ;

[0028] in, represents the state space variable set of the main moving body corresponding to the i-th velocity sensor, , , i represents the number of the same main moving body speed sensor, i=1,2,3,...,n, Indicates the measurement result of the main moving body's No. i sensor, represents the displacement of the main moving body derived from the measurement results of the corresponding velocity sensor, Indicates the speed of the moving body by the accurate attachment Derived displacement of the attached moving body.

[0029] Furthermore, the fault diagnosis strategy obtained in step 6 is:

[0030] Step 6-1, substitute the measurement data of each velocity sensor of the main moving body into the corresponding observer with residual variables to obtain the observer result;

[0031] Step 6-2, determine whether the results of each observer exceed the observation error threshold, that is, whether there is a fault. If there is a fault, execute step 6-3;

[0032] Step 6-3, determine whether the number of faulty sensors meets the applicable conditions. If the applicable conditions are met and the sensor fault location can be directly observed, the diagnosis ends. Otherwise, determine whether the attached moving body has accurate observation data. If the attached moving body has accurate observation data, execute step 6-4. Otherwise, execute step 6-5.

[0033] Step 6-4, substitute the correct observation data of the corresponding auxiliary moving body and the measurement data of the main moving body into the corresponding auxiliary measurement state observer and observe the results, comprehensively observe the observation results of the two observers, and determine the position of the faulty speed sensor;

[0034] Step 6-5, obtain the correct observation data of the adjacent moving body by means of information transmission, and execute step 6-4.

[0035] A distributed sensor fault diagnosis system based on a multi-moving body networking system, comprising:

[0036] System dynamics model building unit, used to establish a mathematical model of the dynamics of a multi-moving body network system;

[0037] A fault speed sensor model and a healthy speed sensor model building unit is used to build a fault speed sensor model and a healthy speed sensor model affected by measurement interference according to sensor principles;

[0038] The observer construction unit establishes observers with residual variables of the faulty speed sensor model and the healthy speed sensor model based on the mathematical model of the multi-moving body network system dynamics, and determines the observation error threshold;

[0039] An applicable condition determination unit, which determines the applicable condition of the observer with residual variables through simulation based on the observation error threshold;

[0040] An auxiliary diagnosis sensor establishment unit is used to establish an auxiliary state observer based on an observer with residual variables to assist in diagnosing sensor faults;

[0041] The fault diagnosis strategy determination unit obtains the fault diagnosis strategy through simulation analysis based on the observation error threshold, the applicable condition of the observer and the auxiliary state observer, and is used to determine the faulty sensor.

[0042] A computer storage medium stores an executable program, and the executable program is executed by a processor to implement the distributed sensor fault diagnosis method.

[0043] Compared with the prior art, the beneficial effects of the present invention are as follows: the physical system studied in the present application is a distributed speed sensor network system based on a multi-moving body networking system, and is described by a general continuous state space model. In order to observe the changes in sensor measurement values ​​over time and thereby find out the location of the faulty sensor, the present application designs a fault state observer with residual variables based on the dynamics model of the multi-moving body networking system, and determines the definition of the faulty sensor in the present application by selecting a suitable error threshold; the present application designs an auxiliary observer based on the fault state observer with residual variables to assist in observing the sensor state, thereby solving the problem that the fault state observer with residual variables cannot draw conclusions in some cases; finally, the present application designs a fault sensor model, and verifies the correctness of the present application through simulation experiments to derive the final distributed sensor fault diagnosis method and verify the correctness of the design. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 It is a flow chart of the method provided by this application.

[0045] Figure 2 This is a diagram of the dynamic physical model of the multi-moving body networking system provided by this application.

[0046] Figure 3 It is the schematic diagram of the speed sensor provided in this application.

[0047] Figure 4 It is a sensor fault modeling diagram provided by this application.

[0048] Figure 5 The residual variable fault observer provided in the embodiment of the present application is V12 and X12 observation results at .

[0049] Figure 6 The residual variable fault observer provided in the embodiment of the present application is The V13 and X13 observation results at .

[0050] Figure 7 The residual variable fault observer provided in the embodiment of the present application is V14 and X14 observation results at .

[0051] Figure 8 The residual variable fault observer provided in the embodiment of the present application is The V23 and X23 observation results at .

[0052] Fig. 9 The residual variable fault observer provided in the embodiment of the present application is V24 and X24 observation results at .

[0053] Fig.10 The residual variable fault observer provided in the embodiment of the present application is V34 and X34 observation results at .

[0054] Fig.11 The residual variable fault observer provided in the embodiment of the present application is V12 and X12 observation results at .

[0055] Fig.12 The residual variable fault observer provided in the embodiment of the present application is The V13 and X13 observation results at .

[0056] Fig.13 The residual variable fault observer provided in the embodiment of the present application is V14 and X14 observation results at .

[0057] Fig.14 The residual variable fault observer provided in the embodiment of the present application is The V23 and X23 observation results at .

[0058] Fig.15 The residual variable fault observer provided in the embodiment of the present application is V24 and X24 observation results at .

[0059] Fig.16 The residual variable fault observer provided in the embodiment of the present application is V34 and X34 observation results at 1.10 pm.

[0060] Fig.17 The residual variable fault observer provided in the embodiment of the present application is V12 and X12 observation results at .

[0061] Fig.18 The residual variable fault observer provided in the embodiment of the present application is The V13 and X13 observation results at .

[0062] Fig.19 The residual variable fault observer provided in the embodiment of the present application is V14 and X14 observation results at .

[0063] Fig. 20 The residual variable fault observer provided in the embodiment of the present application is The V23 and X23 observation results at .

[0064] Fig.21 The residual variable fault observer provided in the embodiment of the present application is V24 and X24 observation results at .

[0065] Fig. 22 The residual variable fault observer provided in the embodiment of the present application is V34 and X34 observation results at 1.10 pm.

[0066] Fig.23 The residual variable fault observer provided in the embodiment of the present application is Observation results of V12 and X12 with the influence of rotation repair.

[0067] Fig.24 The residual variable fault observer provided in the embodiment of the present application is Observation results of V13 and X13 with the influence of rotation repair.

[0068] Fig.25 The residual variable fault observer provided in the embodiment of the present application is Observation results of V14 and X14 with the influence of rotation repair.

[0069] Fig.26 The residual variable fault observer provided in the embodiment of the present application is Observation results of V23 and X23 with the influence of rotation repair.

[0070] Fig. 27 The residual variable fault observer provided in the embodiment of the present application is Observation results of V24 and X24 with the influence of rotation repair.

[0071] Fig.28 The residual variable fault observer provided in the embodiment of the present application is Observation results of V34 and X34 with the influence of rotation repair.

[0072] Fig.29 The auxiliary observer provided in this application example is Observation results of VTD1 and XTD1 at .

[0073] Fig.30 The auxiliary observer provided in this application example is VTD2 and XTD2 observation results at .

[0074] Fig.31 The auxiliary observer provided in this application example is VTD3 and XTD3 observation results at .

[0075] Fig.32 The auxiliary observer provided in this application example is VTD4 and XTD4 observation results at .

[0076] Fig.33 The auxiliary observer provided in this application example is Observation results of VTD1 and XTD1 with the influence of rotation repair.

[0077] Fig.34 The auxiliary observer provided in this application example is Observation results of VTD2 and XTD2 with the influence of rotation repair. DETAILED DESCRIPTION

[0078] In order to make the objectives, technical solutions and advantages of the present application clearer, the implementation methods of the present application will be further described in detail below with reference to the accompanying drawings.

[0079] Combination Figure 1 , the present application provides a distributed sensor fault diagnosis method based on a multi-moving body networking system, and the specific implementation method is:

[0080] Step 1: Establish a mathematical model of the dynamics of the multi-moving body network system. Figure 2 Take the CRH3 short-marshaling train as an example: the CRH3 short-marshaling train is an eight-carriage train consisting of six motors and two tractors, which can be divided into two traction units, each with four carriages. Figure 2 The modeling method is more flexible: from a macroscopic point of view, the two traction units can be regarded as the structure of two carriages when the force analysis is carried out as a whole; from a microscopic point of view, each traction unit can also be divided into two parts for force analysis in pairs, and the two parts just constitute Figure 2 So just follow Figure 2 The modeling is performed in a manner, and then the modeling of the entire train can be completed by simply replacing it. In this application, the motor car of the train is regarded as the main moving body part of a moving body, and the trailer car is regarded as the corresponding auxiliary moving body.

[0081] Specifically, the following mathematical model of train dynamics is established:

[0082]

[0083] in . Represents the quality of the vehicle, Represents the mass of the trailer; , Represent the ground displacement of the motor vehicle and trailer respectively; , Respectively , The first-order derivative of ; F represents the traction force output by the EMU during the acceleration phase, which is always positive (its value is negative during the deceleration phase, indicating the braking force output by the EMU), b represents the adhesion coefficient between the train wheelset and the track, and k represents the elastic coefficient of the coupler and buffer between the two carriages. , Respectively represent motor vehicles and trailer The speed of the train is measured by the speed sensor of the corresponding carriage. Output Y is selected as the speed of the train. , the input is selected as the traction force F of the motor vehicle.

[0084] Step 2: designing a faulty speed sensor model and a healthy speed sensor model affected by measurement interference according to the speed sensor principle, that is, introducing measurement interference into the existing modeling method of the faulty speed sensor model and the healthy speed sensor model;

[0085] There are many types of speed sensors, including magnetoelectric, photoelectric, centrifugal, Hall and other speed sensors. Among them, pulse speed sensors are widely used in high-speed railways. The basic working principle of the wheel axle pulse speed sensor speed measurement is as follows: Figure 3 As shown in the figure, the circumference of the wheel is used as a "ruler" to measure the distance traveled by the train, and the train running speed is calculated based on the measured distance. The basic formula is as follows:

[0086]

[0087] Where D represents the diameter of the train wheel, and s represents the wheel speed. The pulse speed sensor is installed on the wheel axle. The number of pulses output by the wheel axle every time the wheel rotates is fixed. Therefore, we only need to measure the number of pulses output by the pulse speed sensor per unit time to convert the train wheel speed and then find the train speed.

[0088] And follow Figure 4 The train speed sensor is modeled, and the sensor failure is modeled by introducing measurement disturbance.

[0089] Step 3: Design a fault state observer based on residual variables according to the mathematical model of the multi-moving body network system dynamics and define the fault sensor by selecting a suitable error threshold;

[0090] In actual high-speed railway operation lines, sensors are usually used to monitor various parameters of the train, such as speed, position, temperature, etc., to ensure the normal operation of the train system and the safety of passengers. In order to reduce the failure rate of sensors, the railway transportation department usually takes the following measures:

[0091] (1) Regular inspection and maintenance: Regularly inspect and maintain the sensor to ensure that it is working properly and to detect and repair potential problems in a timely manner;

[0092] (2) Environmental adaptability design: When designing and installing sensors, the characteristics of the train operating environment need to be considered, and appropriate sensor types and materials should be selected to improve their stability and durability.

[0093] (3) Redundant design: Redundant sensors are used at key locations to back up and verify data. Once a sensor fails, the system can automatically switch to the backup sensor to ensure safe operation of the train.

[0094] Therefore, at the same time, the probability that all speed sensors on a carriage have measurement failures is very low. Therefore, before providing the detection method, the present invention makes the following assumptions.

[0095] Assumption 1: Assume that there are at least two non-faulty speed sensors on the same carriage at the same time.

[0096] Based on Assumption 1, considering the structure of the mathematical model of train dynamics, the residual value between the measured values ​​of the EMU speed sensor can be selected As system status and fault variables, When the value exceeds a certain threshold, it can be determined that there is a faulty sensor between the i-th and j-th speed sensors of the motor vehicle. At this time, the difference between the i-th and j-th sensors and other sensors can be observed to determine the specific faulty sensor. The specific observer model is as follows:

[0097]

[0098] in , , and i (=1, 2, 3, 4), j (=1, 2, 3, 4) refer to the i-th sensor and j-th sensor of the EMU respectively; , They respectively refer to the measurement value of the i-th speed sensor and the j-th speed sensor of the EMU; , They respectively refer to the displacements derived from the measurements of the corresponding vehicle speed sensors.

[0099] In actual applications, due to different track conditions and the fact that the size of train wheelsets may not be consistent due to maintenance, there will be slight differences in the measurement results of the train speed sensor. This is one of the reasons why it is not appropriate to simply use a statistical model-based method to determine the failure of the train speed sensor. Therefore, in this application, the measurement difference caused by the maintenance of the train is selected as the error threshold. It is defined that the observation result that does not exceed the error threshold is a normal observation result, and the observation result that exceeds the error threshold is a faulty observation result, and the corresponding sensor is a faulty sensor. Furthermore, the observation error threshold is selected in the following way:

[0100]

[0101] in represents the observation error threshold, The inconsistency in the size of train wheels is caused by wheel rotation during daily train maintenance.

[0102] Step 4, explore the applicable conditions of the residual variable fault observer through simulation experiments;

[0103] Assumptions (=1,2,3,4) represents the number of speed sensors that are not faulty in the same car. The simulation results are as follows: Figure 5-Figure 10 As shown, it can be directly observed that the faulty sensor is sensor No. 4; when When , the simulation results are as follows Figure 11-Figure 16 As shown, it can be directly observed that the faulty sensors are sensors 3 and 4; when When , the simulation results are as follows Figure 17-Figure 22 As shown, it is no longer possible to draw correct conclusions by observing the residual variable fault observer.

[0104] Therefore, the applicable conditions of the residual variable fault observer can be obtained: When the fault sensor position can be directly obtained by observing the residual variable fault observer, The specific faulty sensor cannot be found.

[0105] It is worth noting that when If a sensor has a low measurement result due to maintenance, it may lead to a situation where it cannot be judged, such as Figure 23-Figure 28 shown.

[0106] Step 5, design and establish an auxiliary state observer based on the residual variable observer to assist in diagnosing sensor faults;

[0107] The communication structure of the train (taking the CRH3 train as an example): Each CRH3 EMU is composed of eight cars, which can be further divided into two traction units consisting of four cars each. The traction unit contains both EMUs and trailers. The vehicle bus MVB is used in each traction unit to connect the four cars in the unit and the on-board equipment, sensors, and actuators in the same car. The topology of each car in the traction unit remains unchanged. The two traction units are connected by the train bus WTB, and the vehicle bus and the train bus communicate through the gateway. Therefore, when a train car (whether it is an EMU or a trailer) cannot determine the location of the faulty sensor through its own sensor data, the accurate observation results of other cars in the same traction unit can be used to assist in determining the location of the faulty sensor.

[0108] In order to solve the shortcomings of the residual variable fault observer, an auxiliary observer is designed based on the residual variable fault observer:

[0109]

[0110] in Represents the set of state space variables corresponding to the i-th speed sensor. , , i (=1,2,3,4) represents the number of the speed sensor of the same train carriage, It represents the measurement result of the No. i sensor of the motor vehicle. represents the vehicle displacement derived from the corresponding speed sensor measurement results, Indicates the exact trailer speed Derived trailer displacement.

[0111] when When Figure 17-Figure 22 The simulation results are shown in Figure 2. At this time, it is no longer possible to draw a correct conclusion by observing the residual variable fault observer. By bringing the accurate trailer speed and the data to be confirmed into the corresponding auxiliary observer, we can get the following: Figure 29-Figure 32 The simulation results are shown in Figure 2. By comprehensively observing the results of the two observers, it can be determined that the faulty sensors are No. 2 to No. 4. Figure 23-Figure 28 The results shown in the figure are similar. By bringing the data into the corresponding auxiliary observer, we can obtain Fig.33 , Fig.34 The results shown in the figure show that the results of the two observers are comprehensively observed, and it can be found that the faulty sensors are No. 3 and No. 4, and the measurement results of No. 2 sensor are biased due to the rotation repair. The simulation results prove that the designed auxiliary observer can achieve the expected effect.

[0112] Step 6: Obtain a fault diagnosis method through simulation analysis.

[0113] By comprehensively analyzing the results of all simulation experiments, we can get the final distributed sensor fault diagnosis method:

[0114] (1) Substitute the measurement data of each velocity sensor of the main moving body into the corresponding residual variable state observer and observe the results;

[0115] (2) Observe whether the results of each observer exceed the error threshold, that is, whether there is a fault;

[0116] (3) Number of non-faulty speed sensors on the same main moving body If the conclusion cannot be drawn directly, the corresponding auxiliary measurement state observer is enabled to assist in diagnosing the position of the faulty sensor. In the case of , the corresponding auxiliary measurement state observer is directly enabled;

[0117] (4) Substitute the correct observation data of the corresponding auxiliary moving body and the measurement data of the main moving body into the corresponding auxiliary measurement state observer and observe the results. Combine the observation results of the two observers to determine the location of the faulty speed sensor.

[0118] (5) If there is no accurate observation data of the auxiliary moving body, the correct observation data of the adjacent moving body can be brought into the observation result of the corresponding auxiliary measurement state observer by means of information transmission.

[0119] A distributed sensor fault diagnosis system based on a multi-moving body networking system, comprising:

[0120] System dynamics model building unit, used to establish a mathematical model of the dynamics of a multi-moving body network system;

[0121] A fault speed sensor model and a healthy speed sensor model building unit is used to build a fault speed sensor model and a healthy speed sensor model affected by measurement interference according to sensor principles;

[0122] The observer construction unit establishes observers with residual variables of the faulty speed sensor model and the healthy speed sensor model based on the mathematical model of the multi-moving body network system dynamics, and determines the observation error threshold;

[0123] An applicable condition determination unit, which determines the applicable condition of the observer with residual variables through simulation based on the observation error threshold;

[0124] An auxiliary diagnosis sensor establishment unit is used to establish an auxiliary state observer based on an observer with residual variables to assist in diagnosing sensor faults;

[0125] The fault diagnosis strategy determination unit obtains the fault diagnosis strategy through simulation analysis based on the observation error threshold, the applicable condition of the observer and the auxiliary state observer, and is used to determine the faulty sensor.

[0126] This embodiment also provides a computer storage medium, wherein the computer storage medium stores an executable program, and the executable program is executed by a processor to implement the distributed sensor fault diagnosis method.

[0127] The sensor fault diagnosis method proposed in the present application can confirm sensor faults in the case of multiple sensor faults and find out the sensor fault locations.

[0128] The description of the above preferred embodiments enables those skilled in the art to make or use the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be applied to other embodiments without using creative capabilities. Therefore, the present invention is not limited to the embodiments shown herein, but conforms to the broadest scope consistent with the principles and novel features disclosed herein.

Claims

1. A distributed sensor fault diagnosis method based on a multi-moving body networking system, characterized in that: include: Step 1: Establish a mathematical model of the dynamics of the multi-moving body network system; Step 2: Establish a faulty speed sensor model and a healthy speed sensor model affected by measurement interference based on the sensor principle; the sensor principle is: ； Where D represents the wheel diameter, represents the power shaft speed, represents the sensor output; Step 3: Based on the mathematical model of the dynamics of the multi-moving body network system, an observer with residual variables of the faulty speed sensor model and the healthy speed sensor model is established, and an observation error threshold is determined; Step 4: Based on the observation error threshold, determine the applicable conditions of the observer with residual variables through simulation; Step 5: Establish an auxiliary state observer based on the observer with residual variables to assist in diagnosing sensor faults; Step 6: Based on the observation error threshold, the applicable conditions of the observer and the auxiliary state observer, the fault diagnosis strategy is obtained through simulation analysis to determine the faulty sensor.

2. The distributed sensor fault diagnosis method according to claim 1, characterized in that: The mathematical model of the multi-moving body networking system dynamics is: ； in, , for The derivative of Represents the mass of the main moving body in the same moving body in the multi-moving body networking system. Represents the mass of the auxiliary moving body connected to the main moving body; , Respectively represent their displacement relative to the ground; , Respectively , The first-order derivative of ; F is positive, indicating the traction force output by the main moving body in the multi-moving body network system during the acceleration phase, F is negative, indicating the braking force output by the main moving body in the multi-moving body network system during the deceleration phase, b indicates the adhesion coefficient between the main moving body and the moving carrier in the multi-moving body network system, k indicates the elastic coefficient between the main moving body and the auxiliary moving body, , , respectively and The speed is measured by the corresponding speed sensor, and the output Y is the speed , the input is the traction force F.

3. The distributed sensor fault diagnosis method according to claim 2, characterized in that: The observer with residual variables is: ； in, , , and , i and j refer to the i-th sensor and j-th sensor of the main moving body respectively, i=1,2,3,...,n, j=1,2,3,...,n, for The derivative of; n represents the number of sensors, , They refer to the measured values ​​of the main moving body's No. i speed sensor and No. j speed sensor respectively; , They refer to the displacements derived from the corresponding velocity sensor measurements.

4. The distributed sensor fault diagnosis method according to claim 3, characterized in that: The observation error threshold is: ； in, represents the observation error threshold, It is the daily wear and tear of the wheels without affecting the normal operation of the main moving body.

5. The distributed sensor fault diagnosis method according to claim 1, characterized in that: The applicable conditions of the observer with residual variables are: , Indicates the number of speed sensors that are not faulty on the same main moving body. The specific faulty sensor cannot be found.

6. The distributed sensor fault diagnosis method according to claim 2, characterized in that: The auxiliary state observer is: ； in, represents the state space variable set of the main moving body corresponding to the i-th velocity sensor, , , i represents the number of the same main moving body speed sensor, i=1,2,3,...,n, Indicates the measurement result of the main moving body's No. i sensor, represents the displacement of the main moving body derived from the measurement results of the corresponding velocity sensor, Indicates the speed of the moving body by the accurate attachment Derived displacement of the attached moving body.

7. The distributed sensor fault diagnosis method according to claim 1, characterized in that: The fault diagnosis strategy obtained in step 6 is: Step 6-1, substitute the measurement data of each velocity sensor of the main moving body into the corresponding observer with residual variables to obtain the observer result; Step 6-2, determine whether the results of each observer exceed the observation error threshold, that is, whether there is a fault. If there is a fault, execute step 6-3; Step 6-3, determine whether the number of faulty sensors meets the applicable conditions. If the applicable conditions are met and the sensor fault location can be directly observed, the diagnosis ends. Otherwise, determine whether the attached moving body has accurate observation data. If the attached moving body has accurate observation data, execute step 6-4. Otherwise, execute step 6-5. Step 6-4, substitute the correct observation data of the corresponding auxiliary moving body and the measurement data of the main moving body into the corresponding auxiliary measurement state observer and observe the results, comprehensively observe the observation results of the two observers, and determine the position of the faulty speed sensor; Step 6-5, obtain the correct observation data of the adjacent moving body by means of information transmission, and execute step 6-4.

8. A system for implementing the distributed sensor fault diagnosis method according to any one of claims 1 to 7, characterized in that: include: System dynamics model building unit, used to establish a mathematical model of the dynamics of a multi-moving body network system; A fault speed sensor model and a healthy speed sensor model building unit is used to build a fault speed sensor model and a healthy speed sensor model affected by measurement interference according to sensor principles; The observer construction unit establishes observers with residual variables of the faulty speed sensor model and the healthy speed sensor model based on the mathematical model of the multi-moving body network system dynamics, and determines the observation error threshold; An applicable condition determination unit, which determines the applicable condition of the observer with residual variables through simulation based on the observation error threshold; An auxiliary diagnosis sensor establishment unit is used to establish an auxiliary state observer based on an observer with residual variables to assist in diagnosing sensor faults; The fault diagnosis strategy determination unit obtains the fault diagnosis strategy through simulation analysis based on the observation error threshold, the applicable condition of the observer and the auxiliary state observer, and is used to determine the faulty sensor.

9. A computer storage medium, characterized in that: The computer storage medium stores an executable program, and the executable program is executed by a processor to implement the distributed sensor fault diagnosis method according to any one of claims 1 to 7.

Images