Fault detection algorithm for sensor of suspension system

Abstract

The invention discloses a suspension system sensor fault detection algorithm, which comprises the following steps: establishing a suspension system model according to the motion of a real part of a suspension system; designing a state observer based on the suspension system model; acquiring and processing a sensor signal to obtain a residual value; and comparing the residual value with a threshold value to obtain fault information of the suspension system sensor. According to the fault detection method based on the algorithm, the change of the system state can be sensed earlier, and an early warning signal is sent out at the initial stage of sensor fault occurrence. The magnetic levitation system is widely applied to scenes of rail transit, air transportation, medical care and the like, the algorithm provides guarantee for normal operation of the system, meanwhile, maintenance personnel provide sufficient time for checking and repairing a fault sensor, further deterioration of faults is effectively prevented, safe and stable operation of the scenes where the magnetic levitation system is applied is ensured, and the service life of the magnetic levitation system is prolonged. And the safety and the reliability are improved.

Description

Technical Field

[0001] The invention relates to the technical field of electromagnetic suspension, and in particular to a suspension system sensor fault detection algorithm. Background Art

[0002] The maglev system is a technology that uses electromagnetic force to suspend and drive. It is designed to achieve contactless, frictionless, high-speed relative motion. The maglev system is widely used in rail transit, air transportation, health care and other scenarios. The maglev system is mainly composed of key components such as sensors, controllers and actuators (including electromagnets). In the maglev system, sensors play a vital role. They can detect the suspension state and position information of objects in relative motion in real time and provide accurate data support for the controller. Rail transit and air transportation have significant advantages such as high speed, low noise, low vibration, environmental protection and energy saving, and are one of the important directions for future transportation development.

[0003] The operation of the maglev system depends on the stability and reliability of the maglev system. Fault detection can detect potential problems in the system in a timely manner, thereby avoiding serious safety accidents. Fault detection helps to detect abnormal conditions of the maglev system in advance, so that preventive measures can be taken to avoid further deterioration of the fault and reduce maintenance costs. Therefore, fault detection is crucial for the operation of rail transportation and air transportation.

[0004] At present, the technology of fault detection of sensors in suspension systems based on observers is in a stage of rapid development but still needs to be improved. This technology builds an observer model, monitors the system state variables in real time, obtains residual values ​​based on errors and compares them with thresholds to determine whether the sensor is faulty. However, in practical applications, this technology still faces some technical challenges and problems that need to be solved.

[0005] The ability to detect multiple faults is a major drawback of current technology. In complex systems, there are often situations where multiple sensors fail at the same time, and most existing fault detection methods can only detect a single fault. For situations where multiple faults occur at the same time, the detection effect may not be good. Therefore, how to design an algorithm that can detect multiple faults at the same time is a problem to be solved. Secondly, the robustness problem is also a technical problem that needs to be solved urgently. In actual environments, in the suspension systems of rail transit and air transportation, the dynamic equations are usually described as second-order systems. However, factors such as system modeling errors, external disturbances and noise may affect the accuracy of fault detection results. Therefore, improving the robustness of the fault detection method so that it can resist the influence of these unknown input factors and reduce the false alarm rate and missed alarm rate is also a problem to be solved by technical personnel in this field. Summary of the invention

[0006] According to an embodiment of the present invention, to address the above deficiencies of the prior art, a fault detection algorithm for a suspension system sensor is provided, which includes the following steps: Step 1: Based on the actual motion of rail transit and air transportation, establish a suspension system model for rail transit and air transportation; Step 2: Based on the suspension system model, design a third-order linear extended state observer; Step 3: Collect and process sensor signals to obtain a residual value; Step 4: Compare the residual with a threshold to obtain the fault information of the suspension system sensor.

[0007] Preferably, Step 3 further includes the following steps: Based on the third-order linear extended state observer, establish a third-order linear fault detection observer model and an observer error model; Establish a fault model of the suspension system and subtract it from the fault detection observer model to obtain a residual model; Based on the observer error model and the residual model, obtain the steady-state solution as the residual value.

[0008] Preferably, the sensor signals include a gap signal and an acceleration signal.

[0009] Preferably, it further includes Step 5: Based on the residual value, reconstruct the fault information of the sensor.

[0010] Preferably, in Step 4, when the residual value is less than the threshold, the sensor has no fault; when the residual value is greater than the threshold, the sensor has a fault.

[0011] Preferably, the threshold is an adaptive threshold.

[0012] According to the fault detection algorithm for the suspension system sensor of the embodiment of the present invention, in order to more effectively perform sensor fault detection, the present invention establishes a third-order linear extended state observer (LESO) by expanding the order of the second-order model. This observer has the following beneficial technical effects in sensor fault detection: The third-order LESO can more accurately estimate the actual state of the system, including the state deviation caused by sensor faults. The third-order LESO can simultaneously process the estimation and detection of multiple state variables, so it has stronger multi-sensor fault detection ability. The LESO has a certain robustness to system uncertainties and noises, which helps to stably perform fault detection in complex environments. By real-time monitoring the state variables and output values of the system, the fault detection algorithm based on the third-order LESO can capture the changes in the system state earlier, so it can issue a warning signal at the early stage of sensor faults. This provides more time for maintenance personnel to check and repair faulty sensors, thereby avoiding further deterioration of faults, ensuring the safe operation of the system, and guaranteeing the safety and reliability of rail transit and air transportation.

[0013] It is to be understood that both the foregoing general description and the following detailed description are exemplary and are intended to provide further explanation of the claimed technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 is a flowchart of a suspension system sensor fault detection algorithm according to an embodiment of the present invention; Figure 2 is a simulation experimental diagram of the suspension system sensor fault detection algorithm according to an embodiment of the present invention for detecting when a gap sensor fails; Figure 3 is a simulation experimental diagram of the suspension system sensor fault detection algorithm according to an embodiment of the present invention for detecting when an acceleration sensor fails. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0015] The preferred embodiments of the present invention will be described in detail below with reference to the accompanying drawings, and the present invention will be further elaborated.

[0016] In order to ensure the high stability and safety of the maglev system during operation. An embodiment of the present application provides a suspension system sensor fault detection algorithm. The maglev system is widely used in scenarios such as rail transit, air transportation, and healthcare, providing guarantee for the normal operation of the system, such as Figure 1 As shown, according to the actual motion characteristics of the maglev system, a suspension system model is constructed, providing a basis for subsequent fault detection. The suspension system model can be described by the following dynamic model: ; where is the parameter after linearization, is the input quantity, is the air gap, is the air gap change rate; is the perturbation, is the output. The suspension system model is a typical second-order system model.

[0017] Next, based on this model, the method of Embodiment 1 designs a third-order extended state observer for real-time monitoring of the system operation state.

[0018] During the execution of the algorithm, sensor signals are collected and processed to obtain a key residual value. This residual value reflects the difference between the actual state of the system and the estimated state of the observer, and is an important basis for judging whether there is a sensor fault. By comparing with a preset threshold, the algorithm can quickly and accurately identify whether the sensor has failed. This algorithm can issue an early warning at the early stage of the fault, avoid the further deterioration of the fault, and ensure the stable operation of the suspension system.

[0019] The technical solution of implementing the model construction based on the third-order linear extended state observer is specifically as follows: System modeling: First, it is necessary to establish the mathematical model of the controlled object, usually including the state equation and the output equation. The model should be able to reflect the dynamic characteristics of the system and the disturbances it receives, etc.; Determine the observer structure: For a third-order system, design a third-order linear extended state observer ESO. Its structure includes the observation and estimation of three state variables and the corresponding update equations for estimating the state and disturbances of the system; Further, select the observer parameters (specifically described as follows): Reasonably select the observer parameters to ensure the rapid convergence of the observation error and good estimation effect of the system disturbances. The parameter selection needs to be adjusted according to the characteristics of the system and the desired observation performance.

[0020] Specifically, as described in detail below: Expand the suspension system to a new third-order system through order expansion: ; Among them, matrix A: the system matrix, which is used to describe the dynamic relationship between state variables, that is, how state variables change over time; matrix B: the input matrix, which is used to represent how the input signal u affects the state variables ; matrix C: the output matrix, which is used to extract the output signal y from the state variables ; matrix D: the direct transmission matrix, which represents the direct influence of the input signal u on the output signal y without going through the dynamic process of the state variables; matrix E: which is used to represent the influence of the fault signal f on the state variables ;

[0021] Based on the expanded suspension system model, use a third-order linear extended state observer to obtain the air gap speed and interference of rail transit and air transportation. The third-order linear extended state observer is designed as: ; In the formula, is a positive real number, is the estimated value of , is 's estimated value, is the estimated value of the disturbance and is the output of the observer, . Among them, are coefficients used to adjust the performance of the observer. They play a key role in the observation error state equation and the Lyapunov equation. By selecting appropriate , the poles of the observer can be configured at the desired positions, thus ensuring that the observation error can converge quickly. The value of can be obtained by constructing the characteristic equation based on these poles. ε is a parameter used to adjust the convergence speed and robustness of the observer. It appears in the observation error dynamic equation and the update law of the observer. A smaller ε can make the observation error converge faster and enhance the robustness of the observer to system uncertainties and disturbances.

[0022] By adding additional state variables to comprehensively capture the dynamic behavior of the system, this feature enables the observer to more accurately reflect the true state of the system. At the same time, the third-order LESO also has a strong disturbance estimation ability, which can track unknown disturbances in the system, thereby further improving the accuracy of state estimation. In practical applications, this accurate state estimation and disturbance tracking ability enables the third-order LESO to capture small state changes caused by sensor failures at an early stage. By comparing with the actual measured values of the sensors, the algorithm can calculate the residuals, which reflect the difference between the actual state of the system and the estimated state of the observer and are an important basis for judging whether there are sensor failures. Due to the introduction of the third-order LESO, the fault detection algorithm not only realizes the accurate detection of sensor failures but also significantly improves the detection sensitivity. This means that the algorithm can issue an early warning at the initial stage of the fault, providing enough time for maintenance personnel to conduct fault troubleshooting and repair, realizing the efficient and accurate detection of sensor failures, and providing strong technical support and guarantee for the safe operation of rail transit and air transportation.

[0023] Furthermore, step three includes the data processing link in the sensor fault detection of the rail transit and air transportation suspension systems, specifically including the following sub-steps: First, based on the third-order linear extended state observer, a fault detection observer model and an observer error model are constructed. The error model is: ; Among them, is the observation error variable, representing a part of the difference between the system state and the observed state.

[0024] Secondly, a fault model of the suspension system is established as: ; In the formula, its where and respectively represent the location and mode of the fault occurrence. In addition, Represents an unknown fault. Different values can be set to meet different fault types.

[0025] The fault model is compared with the fault detection observer model to obtain the residual model; finally, combining the observer error model and the residual model, the steady-state residual value is solved. This algorithm realizes the capture and quantization of sensor fault signals through modeling and comparative analysis. Its beneficial technical effect is that it significantly improves the accuracy and sensitivity of fault detection, can detect potential faults earlier, and provides strong technical support for the safe operation of rail transit and air transportation.

[0026] Preferably, in the data processing stage, a parallel processing mechanism is adopted to synchronously process the signals of the acceleration sensor and the gap sensor. The third-order linear extended state observer (LESO) has strong state estimation and disturbance tracking capabilities, and can capture the dynamic behavior of the system in real time and accurately. Among them, the residual model of the gap sensor is: ; Specifically, matrix A: the system matrix, which is used to describe the dynamic relationship between state variables, that is, how the state variables change over time; matrix L: the gain matrix, which is used for the gain of the state observer, that is ; matrix : the fault matrix, which represents the location and manner of the fault occurrence; e is the residual value of the gap; f(t) : unknown fault; The residual value corresponding to the corresponding steady-state solution is: ; Similarly, the residual value of the acceleration sensor is: ; Specifically, is to obtain the speed signal by using the observer; is to obtain the speed signal by directly integrating the acceleration signal; the speed signal under the fault-free condition; the residual of the speed signal; acceleration fault; the time when the acceleration fault occurs.

[0027] By constructing a fault detection observer model based on LESO, synchronous capture and processing of the signals of the acceleration sensor and the gap sensor can be achieved. Synchronously capturing and processing the signals of the acceleration sensor and the gap sensor can more comprehensively reflect the state changes of the suspension systems of rail transit and air transportation. The signals of the two sensors are processed and analyzed at the same time, enabling the estimation and detection of multiple state variables to be achieved simultaneously. Therefore, it has stronger multi-sensor fault detection capabilities, thereby improving the efficiency and accuracy of fault detection.

[0028] Preferably, in step four, a threshold is set to determine whether the sensor has failed. Under normal circumstances, due to the accuracy of the observer and the stability of the system, this difference should be kept within a small range, that is, the residual value will be less than the threshold. However, when the sensor fails, its output signal will deviate from the normal state, resulting in an increase in the difference between the actual state of the system and the estimated state of the observer, thus causing the residual value to exceed the threshold. Therefore, specifically, it is performed according to the following method: when the calculated residual value is less than this preset threshold, it is considered that the sensor in question is in a normal state and there is no fault; while when the residual value is greater than the threshold, it is determined that the sensor has failed.

[0029] Preferably, the threshold is set using an adaptive threshold, and the selection of the adaptive threshold is based on Bayesian decision theory. It includes: based on the residual value, characteristic calculations are carried out; the prior probability of the suspension system is obtained, and based on Bayesian decision theory, a decision variable of the suspension system is established. Among them, the characteristic calculation includes determining the mean and variance of the difference between the integral signal and the actual signal under normal conditions based on the residual value. Specifically, the adaptive threshold is: , where N w is the data length in the operating state, and P t is the prior probability of the system being fault-free. Bayesian decision theory can take into account historical information and changes in the system state. Therefore, the adaptive threshold method based on this theory can better adapt to the fault detection requirements under different working conditions, can be dynamically adjusted according to the real-time state of the system, and the adaptive technology helps to maintain the stability and balance of the system, preventing system crashes or performance degradation caused by improper threshold settings, thereby better adapting to various complex and changeable situations. At the same time, since the need for manual threshold setting and the possible errors are reduced, the automation level of the system is improved.

[0030] An embodiment of the present invention provides a suspension system sensor fault detection algorithm, which constructs a third-order linear extended state observer (LESO) by using a second-order order expansion method. It can more accurately estimate the real-time state of the system, including the state deviation caused by sensor faults, thereby improving the accuracy of fault detection. The third-order LESO has a powerful multi-sensor fault detection ability and can simultaneously process the estimation and detection tasks of multiple state variables. The LESO has strong robustness to system uncertainties and noises, which enables it to maintain stable fault detection performance in complex environments.

[0031] According to the method of the present invention, simulation experiments are carried out accordingly, as Figure 2 shown. It can be seen that at about 4 s, the residual amplitude suddenly changes and exceeds the threshold value of 17.806, which means that the sensor fault is detected at this time and is consistent with the preset fault, indicating that the method of the present invention has a good diagnostic effect on the additive fault of the gap sensor. Similarly, as Figure 3 shown. It can be seen that at about 5 s, the residual amplitude suddenly changes and exceeds the threshold value of 25.232, which means that the sensor fault is detected at this time and is consistent with the preset fault time, indicating that the proposed method has a good diagnostic effect on the additive fault of the acceleration sensor.

[0032] As mentioned above, with reference to Figures 1 - 3 The suspension system sensor fault detection algorithm according to the embodiment of the present invention is described. It can more accurately estimate the actual state of the system, including the state deviation caused by sensor faults. The third-order LESO can simultaneously process the estimation and detection of multiple state variables, so it has a stronger multi-sensor fault detection ability. The LESO has a certain robustness to system uncertainties and noises, which helps to stably perform fault detection in complex environments. By real-time monitoring the state variables and output values of the system, the fault detection algorithm based on the third-order LESO can capture the changes in the system state earlier, so it can issue a warning signal at the early stage of sensor fault occurrence. This provides more time for maintenance personnel to check and repair the faulty sensor, thereby avoiding the further deterioration of the fault, ensuring the safe operation of the system, and guaranteeing the safety and reliability of rail transit and air transportation.

[0033] In the description of the present invention, it should be noted that unless otherwise specified, the meaning of "multiple" is two or more; the orientation or positional relationship indicated by the terms "upper", "lower", "left", "right", "inner", "outer", "top", "bottom", etc. is based on the orientation or positional relationship shown in the drawings, and is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be construed as a limitation of the present invention.

[0034] It should be noted that in this specification, the terms "comprising", "including" or any other variants thereof are intended to cover non-exclusive inclusion, such that a process, method, article or device comprising a series of elements not only includes those elements but also includes other elements not expressly listed, or further includes elements inherent to such process, method, article or device. Without further limitation, an element defined by the statement "comprising..." does not exclude the presence of additional identical elements in the process, method, article or device comprising the element.

[0035] Although the content of the present invention has been described in detail through the above preferred embodiments, it should be recognized that the above description should not be considered as a limitation of the present invention. After those skilled in the art have read the above content, various modifications and alternatives to the present invention will be obvious. Therefore, the protection scope of the present invention should be defined by the appended claims.

Claims

1. A suspension system sensor fault detection algorithm, characterized in that: Include Step 1: Establish a suspension system model based on the actual motion of the suspension system; Step 2: Based on the suspension system model, design a third-order linear extended state observer; Step 3: Collect and process the signal of the suspension system sensor to obtain a residual value; Step 4: Compare the residual value with a threshold value to obtain fault information of the suspension system sensor.

2. The suspension system sensor fault detection algorithm according to claim 1, characterized in that: Step three further comprises the following steps: Based on the third-order linear extended state observer, a third-order linear fault detection observer model and an observer error model are established; Establishing the fault model of the suspension system and subtracting it from the fault detection observer model to obtain a residual model; A steady-state solution is obtained based on the observer error model and the residual model as the residual value.

3. The suspension system sensor fault detection algorithm as claimed in claim 2, characterized in that: The signal of the sensor includes a gap signal and an acceleration signal.

4. The suspension system sensor fault detection algorithm as claimed in claim 2, characterized in that: The method further includes step five: reconstructing the fault information of the sensor based on the residual value.

5. The suspension system sensor fault detection algorithm according to claim 1, characterized in that: In step 4, when the residual value is less than the threshold, the sensor is not faulty; when the residual value is greater than the threshold, the sensor is faulty.

6. The suspension system sensor fault detection algorithm according to claim 5, characterized in that: The threshold is an adaptive threshold.

Images