Airport luggage transmission system fault diagnosis method and system based on machine learning

Abstract

The invention discloses an airport luggage transmission system fault diagnosis method and system based on machine learning, and the method comprises the steps: setting a time window and the number of windows, collecting a working state data feature value in the time window, and updating the working state data feature value in the time window in a first-in first-out manner, training the original fault model according to the working state data characteristic values in the time window to generate an updated fault model, and fitting parameters of the updated fault model; setting a model evaluation function, calculating an error value between the updated fault model and the original fault model, evaluating the fitted updated fault model, and when the error value between the updated fault model and the original fault model exceeds a preset error threshold value, continuously fitting parameters of the updated fault model to obtain an updated fault model; and taking the corresponding updated fault model as a final fault model until the error is smaller than a preset error threshold value, and performing fault diagnosis on the airport luggage transmission system.

Description

Technical Field

[0001] The present invention belongs to the technical field of conveyor belt health monitoring, and more specifically, relates to a machine learning-based fault diagnosis method and system for airport baggage conveying systems. Background Art

[0002] To ensure the safety of passengers and civil aircraft, passengers must undergo multiple levels of airport security before boarding. Airport security checks include both personal and luggage checks. These tasks are complex and tedious, and even the slightest inadvertent missteps can lead to security incidents. Therefore, it is crucial that security personnel conduct compliant checks on passengers and their luggage during airport security checks. Furthermore, during the actual security check process, passengers' luggage items may be lost. In such cases, due to a lack of comprehensive monitoring and records of the connection between passengers and luggage, it is impossible to promptly track the location of the lost luggage and determine whether it was caused by the passenger themselves or by the actions of security personnel. This can cause inconvenience for passengers and pose challenges for airport management.

[0003] To ensure the reliability and stability of baggage handling systems, health monitoring technology has gradually become a key means of maintaining and optimizing these systems. Airport baggage handling systems (BHSs) are a crucial component of airport operations, and their operational efficiency and stability directly impact passenger travel experiences and the overall efficiency of the airport. However, existing fault prediction technologies typically rely on fault models to predict and monitor the system based on sensor data. These fault models mostly rely on historical data and specific mathematical algorithms. However, due to the diversity and complexity of sensor data, these models often suffer from high error rates and inaccuracies, resulting in insufficient fault prediction accuracy. For example, some existing fault models rely solely on a single signal threshold for judgment, ignoring the interactivity between multi-dimensional and multi-level sensor data in the system, which can easily lead to false positives or missed negatives. Existing fault prediction methods typically focus on a single type of sensor data, such as vibration or temperature signals. However, baggage handling systems involve multiple devices and complex operating environments, and equipment fault states are often manifested in multiple dimensions (such as vibration, noise, temperature, and pressure). In addition, these models are often unable to adapt to the dynamic changes of equipment and operating environments, resulting in poor prediction results when facing new types of faults or new operating environments, and cannot effectively improve the accuracy of fault detection. Summary of the Invention

[0004] To solve the above technical problems, the present invention proposes a machine learning-based fault diagnosis method for airport baggage transfer systems. It adopts multi-dimensional and multi-level sensor data acquisition and fusion technology. Through sensors installed on the airport baggage transfer system, the working status data characteristic values ​​of the equipment on the airport baggage transfer system are collected in real time. Among them, the working status data characteristic values ​​include: equipment working temperature, equipment working pressure, equipment working vibration, and equipment working speed. It can comprehensively capture the operating status of the equipment, achieve more accurate fault diagnosis, and reduce false alarms and missed alarms.

[0005] According to a first aspect of the present invention, a method for diagnosing a fault in an airport baggage handling system based on machine learning is provided, comprising:

[0006] Using sensors installed on the airport baggage handling system, the system collects in real time the characteristic values ​​of the operating status data of the equipment on the airport baggage handling system, wherein the characteristic values ​​of the operating status data include: equipment operating temperature, equipment operating pressure, equipment operating vibration, and equipment operating speed;

[0007] Setting an original fault model, and calculating a fault value of the airport baggage transfer system according to the characteristic value of the working status data, and determining whether the airport baggage transfer system has a fault according to the fault value;

[0008] Setting a time window and the number of windows, collecting characteristic values ​​of working status data within the time window, and updating the characteristic values ​​of the working status data within the time window in a first-in-first-out manner, training the original fault model based on the characteristic values ​​of the working status data within the time window, generating an updated fault model, and fitting the parameters of the updated fault model;

[0009] A model evaluation function is set to calculate the error value between the updated fault model and the original fault model, thereby evaluating the fitted updated fault model. When the error value between the updated fault model and the original fault model exceeds a preset error threshold, the parameters of the updated fault model are continuously fitted until it is less than the preset error threshold. The corresponding updated fault model is used as the final fault model, and fault diagnosis of the airport baggage transfer system is performed.

[0010] Furthermore, the model evaluation function includes:

[0011]

[0012] Among them, M is the error value between the updated fault model and the original fault model, N is the number of characteristic values ​​of the working state data, and x ft is the updated prediction result output after the updated new working status data feature value at time t is input into the updated fault model, x aiis the original prediction result output after the i-th original working state data characteristic value is input into the original fault model, and α′ is the adjustment factor of the model evaluation function.

[0013] Furthermore, the parameters of the updated fault model are fitted, specifically:

[0014] β(k+1)=β(k)-[J T J+λ·diag(J T J)] -1 J T (yf(β(k)))

[0015] Where β(k+1) is the parameter of the k+1th iteration, β(k) is the parameter of the kth iteration, J is the Jacobian matrix, T is the matrix transpose, λ is the adjustment factor of the fitting function, diag is, y is the observation value, representing the actual state of the airport baggage transfer system, and f(β(k)) is the model output of the current fault model based on the parameters of the kth iteration.

[0016] Furthermore, a deep transfer learning function L is set. By minimizing the deep transfer learning function L, the source domain representing the characteristic values ​​of the historical working status data is transferred to the target domain of the characteristic values ​​of the current working status data, thereby improving the generalization ability of the fault model. Specifically:

[0017] L=L s (θ s )+λ′·L t (θ t )+μ·MMD(P s , P t )

[0018] Among them, L s (θ s ) is the source domain θ s The loss function, λ′ is the first adjustment factor of the deep transfer learning function L, L t (θ t ) is the target domain θ t The loss function, μ is the second adjustment factor of the deep transfer learning function L, MMD(P s , P t ) is the source domain θ s The distribution of working status data eigenvalues ​​and target domain θ t The maximum mean difference of the characteristic value distribution of the working status data.

[0019] Furthermore, the original fault model is specifically:

[0020] X(t)=[x1(t),x2(t),...,x n (t)] T

[0021]

[0022] Among them, x n (t) is the nth characteristic value of the working status data at time t, T is the transpose of the matrix, is the fault identification spectrum corresponding to the set X(t) of the working status data eigenvalues ​​at time t, W is the transformation matrix used to control the mapping of the working status data eigenvalues, σ is sigmoid or ReLU, A is the weight matrix, b is the bias term, and c is the constant term.

[0023] Furthermore, the original fault model also includes:

[0024]

[0025] Where ∈(t) is the dynamic adaptive tolerance at time t, which is used to reduce false alarms, and α″ is the first adjustment factor of the dynamic adaptive tolerance. is the fault identification spectrum corresponding to the set X(t-1) of the characteristic values ​​of the working status data at time t-1, β″ is the second adjustment factor of the dynamic adaptive tolerance, and F(t) is the fault value of the airport baggage handling system at time t.

[0026] The present invention also proposes a machine learning-based airport baggage transfer system fault diagnosis system, comprising:

[0027] A collection module is used to collect operating status data characteristic values ​​of equipment on the airport baggage transfer system in real time through sensors installed on the airport baggage transfer system, wherein the operating status data characteristic values ​​include: equipment operating temperature, equipment operating pressure, equipment operating vibration, and equipment operating speed;

[0028] Setting a fault model module, used for setting an original fault model, and calculating a fault value of the airport baggage transfer system according to the characteristic value of the working status data, and judging whether the airport baggage transfer system has a fault according to the fault value;

[0029] A fitting module is used to set a time window and the number of windows, collect characteristic values ​​of the working status data within the time window, and update the characteristic values ​​of the working status data within the time window using a first-in-first-out method, train the original fault model based on the characteristic values ​​of the working status data within the time window, generate an updated fault model, and fit the parameters of the updated fault model;

[0030] The model evaluation module is used to set a model evaluation function and calculate the error value between the updated fault model and the original fault model, thereby evaluating the fitted updated fault model. When the error value between the updated fault model and the original fault model exceeds a preset error threshold, the parameters of the updated fault model are continuously fitted until the error value is less than the preset error threshold. The corresponding updated fault model is used as the final fault model, and fault diagnosis of the airport baggage transfer system is performed.

[0031] Furthermore, the model evaluation function includes:

[0032]

[0033] Among them, M is the error value between the updated fault model and the original fault model, N is the number of characteristic values ​​of the working state data, and x ft is the updated prediction result output after the updated new working status data feature value at time t is input into the updated fault model, x ai is the original prediction result output after the i-th original working state data characteristic value is input into the original fault model, and α′ is the adjustment factor of the model evaluation function.

[0034] According to a third aspect of the present invention, there is provided a monitoring device comprising:

[0035] memory for storing computer programs;

[0036] A processor is used to implement the steps of the fault diagnosis method when executing the computer program.

[0037] According to a fourth aspect of the present invention, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps of the fault diagnosis method are implemented.

[0038] In general, the above technical solutions conceived by the present invention have the following beneficial effects compared with the prior art:

[0039] 1. The method of the present invention utilizes multi-dimensional, multi-layered sensor data acquisition and fusion technology. Sensors installed on the airport baggage handling system collect real-time characteristic data on the operating status of equipment in the system. These characteristic data include equipment operating temperature, operating pressure, operating vibration, and operating speed. This method comprehensively captures the operating status of the equipment, enabling more accurate fault diagnosis and reducing false alarms and missed alerts.

[0040] 2. The method of the present invention establishes a dynamic health screening and fault identification spectrum, combines the equipment failure mechanism, and establishes a multi-dimensional fault identification spectrum based on sensor data, namely the "fingerprint spectrum", so that each fault type has a unique identification feature. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 This is a flow chart of a method for diagnosing a fault in an airport baggage transfer system based on machine learning according to Example 1 of the present invention;

[0042] Figure 2 This is a structural diagram of an airport baggage transfer system fault diagnosis system based on machine learning according to Example 2 of the present invention. DETAILED DESCRIPTION

[0043] In order to better understand the above technical solution, the above technical solution will be described in detail below with reference to the accompanying drawings and specific implementation methods.

[0044] The method provided by the present invention can be implemented in the following terminal environment, wherein the terminal may include one or more of the following components: a processor, a storage medium, and a display screen. The storage medium stores at least one instruction, which is loaded and executed by the processor to implement the method described in the following embodiments.

[0045] A processor can include one or more processing cores. It connects various components within the terminal using various interfaces and circuits. It executes instructions, programs, code sets, or instruction sets stored in storage media, and accesses data stored in storage media to perform various terminal functions and process data.

[0046] The storage medium may include a random access memory (RAM) or a read-only memory (ROM). The storage medium may be used to store instructions, programs, codes, code sets, or instructions.

[0047] The display is used to show the user interface of each application.

[0048] In addition, those skilled in the art will appreciate that the structure of the terminal described above does not limit the terminal. The terminal may include more or fewer components, or a combination of certain components, or a different arrangement of components. For example, the terminal may also include a radio frequency circuit, an input unit, a sensor, an audio circuit, a power supply, and other components, which will not be described in detail here.

[0049] Example 1

[0050] like Figure 1As shown, an embodiment of the present invention proposes a method for diagnosing airport baggage transfer system faults based on machine learning, comprising:

[0051] Step 101: Using sensors installed on the airport baggage handling system, real-time characteristic values ​​of operating status data of equipment on the airport baggage handling system are collected, wherein the characteristic values ​​of the operating status data include: equipment operating temperature, equipment operating pressure, equipment operating vibration, and equipment operating speed;

[0052] Step 102: setting an original fault model, and calculating a fault value of the airport baggage handling system based on the characteristic value of the working status data, and determining whether the airport baggage handling system has a fault based on the fault value;

[0053] Specifically, the original fault model is:

[0054] X(t)=[x1(t),x2(t),...,x n (t)] T

[0055]

[0056] Among them, x n (t) is the nth characteristic value of the working status data at time t, T is the transpose of the matrix, is the fault identification spectrum corresponding to the set X(t) of the working status data eigenvalues ​​at time t, W is the transformation matrix used to control the mapping of the working status data eigenvalues, σ is sigmoid or ReLU, A is the weight matrix, b is the bias term, and c is the constant term.

[0057] Specifically, the original fault model also includes:

[0058]

[0059] Where ∈(t) is the dynamic adaptive tolerance at time t, which is used to reduce false alarms, and α″ is the first adjustment factor of the dynamic adaptive tolerance. is the fault identification spectrum corresponding to the set X(t-1) of the characteristic values ​​of the working status data at time t-1, β″ is the second adjustment factor of the dynamic adaptive tolerance, and F(t) is the fault value of the airport baggage handling system at time t.

[0060] Step 103: Set a time window and the number of windows, collect characteristic values ​​of the working status data within the time window, and update the characteristic values ​​of the working status data within the time window using a first-in-first-out method. Train the original fault model based on the characteristic values ​​of the working status data within the time window to generate an updated fault model, and fit the parameters of the updated fault model.

[0061] Step 104: Set a model evaluation function, calculate the error value between the updated fault model and the original fault model, and evaluate the updated fault model after fitting. When the error value between the updated fault model and the original fault model exceeds a preset error threshold, continue fitting the parameters of the updated fault model until it is less than the preset error threshold. The corresponding updated fault model is used as the final fault model, and fault diagnosis of the airport baggage handling system is performed.

[0062] Specifically, the model evaluation function includes:

[0063]

[0064] Among them, M is the error value between the updated fault model and the original fault model, N is the number of characteristic values ​​of the working state data, and x ft is the updated prediction result output after the updated new working status data feature value at time t is input into the updated fault model, x ai is the original prediction result output after the i-th original working state data characteristic value is input into the original fault model, and α′ is the adjustment factor of the model evaluation function.

[0065] Specifically, the parameters of the updated fault model are fitted as follows:

[0066] β(k+1)=β(k)-[J T J+λ·diag(J T J)] -1 J T (yf(β(k)))

[0067] Where β(k+1) is the parameter of the k+1th iteration, β(k) is the parameter of the kth iteration, J is the Jacobian matrix, T is the matrix transpose, λ is the adjustment factor of the fitting function, diag is, y is the observation value, representing the actual state of the airport baggage transfer system, and f(β(k)) is the model output of the current fault model based on the parameters of the kth iteration.

[0068] Specifically, a deep transfer learning function L is set. By minimizing the deep transfer learning function L, the source domain representing the characteristic values ​​of historical working status data is transferred to the target domain of the characteristic values ​​of current working status data, thereby improving the generalization ability of the fault model. Specifically:

[0069] L=L s (θ s )+λ′·L t (θ t )+μ·MMD(P s , P t )

[0070] Among them, L s (θ s ) is the source domain θ s The loss function, λ′ is the first adjustment factor of the deep transfer learning function L, L t (θ t ) is the target domain θ t The loss function, μ is the second adjustment factor of the deep transfer learning function L, MMD(P s , P t ) is the source domain θ s The distribution of working status data eigenvalues ​​and target domain θ t The maximum mean difference of the characteristic value distribution of the working status data.

[0071] Example 2

[0072] like Figure 2 As shown, an embodiment of the present invention further provides an airport baggage transfer system fault diagnosis system based on machine learning, comprising:

[0073] A collection module is used to collect operating status data characteristic values ​​of equipment on the airport baggage transfer system in real time through sensors installed on the airport baggage transfer system, wherein the operating status data characteristic values ​​include: equipment operating temperature, equipment operating pressure, equipment operating vibration, and equipment operating speed;

[0074] Setting a fault model module, used for setting an original fault model, and calculating a fault value of the airport baggage transfer system according to the characteristic value of the working status data, and judging whether the airport baggage transfer system has a fault according to the fault value;

[0075] Specifically, the original fault model is:

[0076] X(t)=[x1(t),x2(t),...,x n (t)] T

[0077]

[0078] Among them, x n (t) is the nth characteristic value of the working status data at time t, T is the transpose of the matrix, is the fault identification spectrum corresponding to the set X(t) of the working status data eigenvalues ​​at time t, W is the transformation matrix used to control the mapping of the working status data eigenvalues, σ is sigmoid or ReLU, A is the weight matrix, b is the bias term, and c is the constant term.

[0079] Specifically, the original fault model also includes:

[0080]

[0081] Where ∈(t) is the dynamic adaptive tolerance at time t, which is used to reduce false alarms, and α″ is the first adjustment factor of the dynamic adaptive tolerance. is the fault identification spectrum corresponding to the set X(t-1) of the characteristic values ​​of the working status data at time t-1, β″ is the second adjustment factor of the dynamic adaptive tolerance, and F(t) is the fault value of the airport baggage handling system at time t.

[0082] A fitting module is used to set a time window and the number of windows, collect characteristic values ​​of the working status data within the time window, and update the characteristic values ​​of the working status data within the time window using a first-in-first-out method, train the original fault model based on the characteristic values ​​of the working status data within the time window, generate an updated fault model, and fit the parameters of the updated fault model;

[0083] The model evaluation module is used to set a model evaluation function and calculate the error value between the updated fault model and the original fault model, thereby evaluating the fitted updated fault model. When the error value between the updated fault model and the original fault model exceeds a preset error threshold, the parameters of the updated fault model are continuously fitted until the error value is less than the preset error threshold. The corresponding updated fault model is used as the final fault model, and fault diagnosis of the airport baggage transfer system is performed.

[0084] Specifically, the model evaluation function includes:

[0085]

[0086] Among them, M is the error value between the updated fault model and the original fault model, N is the number of characteristic values ​​of the working state data, and x ft is the updated prediction result output after the updated new working status data feature value at time t is input into the updated fault model, x ai is the original prediction result output after the i-th original working state data characteristic value is input into the original fault model, and α′ is the adjustment factor of the model evaluation function.

[0087] Specifically, the parameters of the updated fault model are fitted as follows:

[0088] β(k+1)=β(k)-[J T J+λ·diag(J T J)] -1 J T (yf(β(k)))

[0089] Where β(k+1) is the parameter of the k+1th iteration, β(k) is the parameter of the kth iteration, J is the Jacobian matrix, T is the matrix transpose, λ is the adjustment factor of the fitting function, diag is, y is the observation value, representing the actual state of the airport baggage transfer system, and f(β(k)) is the model output of the current fault model based on the parameters of the kth iteration.

[0090] Specifically, a deep transfer learning function L is set. By minimizing the deep transfer learning function L, the source domain representing the characteristic values ​​of historical working status data is transferred to the target domain of the characteristic values ​​of current working status data, thereby improving the generalization ability of the fault model. Specifically:

[0091] L=L s (θ s )+λ′·L t (θ t )+μ·MMD(P s , P t )

[0092] Among them, L s (θ s ) is the source domain θ s The loss function, λ′ is the first adjustment factor of the deep transfer learning function L, L t (θ t ) is the target domain θ t The loss function, μ is the second adjustment factor of the deep transfer learning function L, MMD(P s , P t ) is the source domain θ s The distribution of working status data eigenvalues ​​and target domain θ t The maximum mean difference of the characteristic value distribution of the working status data.

[0093] Example 3

[0094] An embodiment of the present invention further proposes a storage medium storing a plurality of instructions, wherein the instructions are used to implement the aforementioned method for diagnosing airport baggage transfer system faults based on machine learning.

[0095] Optionally, in this embodiment, the storage medium may be located in any computer terminal in a computer terminal group in a computer network, or in any mobile terminal in a mobile terminal group.

[0096] Optionally, in this embodiment, the storage medium is configured to store program codes for executing the following steps: Step 101, collecting operating status data characteristic values ​​of equipment on the airport baggage transfer system in real time via sensors installed on the airport baggage transfer system, wherein the operating status data characteristic values ​​include: equipment operating temperature, equipment operating pressure, equipment operating vibration, and equipment operating speed;

[0097] Step 102: setting an original fault model, and calculating a fault value of the airport baggage handling system based on the characteristic value of the working status data, and determining whether the airport baggage handling system has a fault based on the fault value;

[0098] Specifically, the original fault model is:

[0099] X(t)=[x1(t),x2(t),...,x n (t)] T

[0100] Among them, x n (t) is the nth characteristic value of the working status data at time t, T is the transpose of the matrix, is the fault identification spectrum corresponding to the set X(t) of the working status data eigenvalues ​​at time t, W is the transformation matrix used to control the mapping of the working status data eigenvalues, σ is sigmoid or ReLU, A is the weight matrix, b is the bias term, and c is the constant term.

[0101] Specifically, the original fault model also includes:

[0102]

[0103] Where ∈(t) is the dynamic adaptive tolerance at time t, which is used to reduce false alarms, and α″ is the first adjustment factor of the dynamic adaptive tolerance. is the fault identification spectrum corresponding to the set X(t-1) of the characteristic values ​​of the working status data at time t-1, β″ is the second adjustment factor of the dynamic adaptive tolerance, and F(t) is the fault value of the airport baggage handling system at time t.

[0104] Step 103: Set a time window and the number of windows, collect characteristic values ​​of the working status data within the time window, and update the characteristic values ​​of the working status data within the time window using a first-in-first-out method. Train the original fault model based on the characteristic values ​​of the working status data within the time window to generate an updated fault model, and fit the parameters of the updated fault model.

[0105] Step 104: Set a model evaluation function, calculate the error value between the updated fault model and the original fault model, and evaluate the updated fault model after fitting. When the error value between the updated fault model and the original fault model exceeds a preset error threshold, continue fitting the parameters of the updated fault model until it is less than the preset error threshold. The corresponding updated fault model is used as the final fault model, and fault diagnosis of the airport baggage handling system is performed.

[0106] Specifically, the model evaluation function includes:

[0107]

[0108] Among them, M is the error value between the updated fault model and the original fault model, N is the number of characteristic values ​​of the working state data, and x ft is the updated prediction result output after the updated new working status data feature value at time t is input into the updated fault model, x ai is the original prediction result output after the i-th original working state data characteristic value is input into the original fault model, and α′ is the adjustment factor of the model evaluation function.

[0109] Specifically, the parameters of the updated fault model are fitted as follows:

[0110] β(k+1)=β(k)-[J T J+λ·diag(J T J)] -1 J T (yf(β(k)))

[0111] Where β(k+1) is the parameter of the k+1th iteration, β(k) is the parameter of the kth iteration, J is the Jacobian matrix, T is the matrix transpose, λ is the adjustment factor of the fitting function, diag is, y is the observation value, representing the actual state of the airport baggage transfer system, and f(β(k)) is the model output of the current fault model based on the parameters of the kth iteration.

[0112] Specifically, a deep transfer learning function L is set. By minimizing the deep transfer learning function L, the source domain representing the characteristic values ​​of historical working status data is transferred to the target domain of the characteristic values ​​of current working status data, thereby improving the generalization ability of the fault model. Specifically:

[0113] L=L s (θ s )+λ′·L t (θ t )+μ·MMD(P s , P t )

[0114] Among them, L s (θ s ) is the source domain θ s The loss function, λ′ is the first adjustment factor of the deep transfer learning function L, L t (θ t ) is the target domain θ t The loss function, μ is the second adjustment factor of the deep transfer learning function L, MMD(P s , P t ) is the source domain θ s The distribution of working status data eigenvalues ​​and target domain θ t The maximum mean difference of the characteristic value distribution of the working status data.

[0115] Example 4

[0116] An embodiment of the present invention also proposes an electronic device, including a processor and a storage medium connected to the processor, wherein the storage medium stores multiple instructions, which can be loaded and executed by the processor to enable the processor to execute the airport baggage transfer system fault diagnosis method based on machine learning.

[0117] Specifically, the electronic device of this embodiment may be a computer terminal, which may include: one or more processors, and a storage medium.

[0118] The storage medium can be used to store software programs and modules, such as the corresponding program instructions / modules for the machine learning-based airport baggage handling system fault diagnosis method in an embodiment of the present invention. The processor executes the software programs and modules stored on the storage medium to perform various functional applications and data processing, thereby implementing the machine learning-based airport baggage handling system fault diagnosis method. The storage medium can include high-speed random access memory (RAM) and non-volatile storage media, such as one or more magnetic storage systems, flash memory, or other non-volatile solid-state storage media. In some instances, the storage medium may further include storage media located remotely from the processor, which can be connected to the terminal via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0119] The processor can call information and applications stored in the storage medium through the transmission system to perform the following steps: Step 101, using sensors installed on the airport baggage transfer system, collect operating status data characteristic values ​​of equipment on the airport baggage transfer system in real time, wherein the operating status data characteristic values ​​include: equipment operating temperature, equipment operating pressure, equipment operating vibration, and equipment operating speed;

[0120] Step 102: setting an original fault model, and calculating a fault value of the airport baggage handling system based on the characteristic value of the working status data, and determining whether the airport baggage handling system has a fault based on the fault value;

[0121] Specifically, the original fault model is:

[0122] X(t)=[x1(t),x2(t),...,x n (t)] T

[0123]

[0124] Among them, x n (t) is the nth characteristic value of the working status data at time t, T is the transpose of the matrix, is the fault identification spectrum corresponding to the set X(t) of the working status data eigenvalues ​​at time t, W is the transformation matrix used to control the mapping of the working status data eigenvalues, σ is sigmoid or ReLU, A is the weight matrix, b is the bias term, and c is the constant term.

[0125] Specifically, the original fault model also includes:

[0126]

[0127] Where ∈(t) is the dynamic adaptive tolerance at time t, which is used to reduce false alarms, and α″ is the first adjustment factor of the dynamic adaptive tolerance. is the fault identification spectrum corresponding to the set X(t-1) of the characteristic values ​​of the working status data at time t-1, β″ is the second adjustment factor of the dynamic adaptive tolerance, and F(t) is the fault value of the airport baggage handling system at time t.

[0128] Step 103: Set a time window and the number of windows, collect characteristic values ​​of the working status data within the time window, and update the characteristic values ​​of the working status data within the time window using a first-in-first-out method. Train the original fault model based on the characteristic values ​​of the working status data within the time window to generate an updated fault model, and fit the parameters of the updated fault model.

[0129] Step 104: Set a model evaluation function, calculate the error value between the updated fault model and the original fault model, and evaluate the updated fault model after fitting. When the error value between the updated fault model and the original fault model exceeds a preset error threshold, continue fitting the parameters of the updated fault model until it is less than the preset error threshold. The corresponding updated fault model is used as the final fault model, and fault diagnosis of the airport baggage handling system is performed.

[0130] Specifically, the model evaluation function includes:

[0131]

[0132] Among them, M is the error value between the updated fault model and the original fault model, N is the number of characteristic values ​​of the working state data, and x ft is the updated prediction result output after the updated new working status data feature value at time t is input into the updated fault model, x ai is the original prediction result output after the i-th original working state data characteristic value is input into the original fault model, and α′ is the adjustment factor of the model evaluation function.

[0133] Specifically, the parameters of the updated fault model are fitted as follows:

[0134] β(k+1)=β(k)-[J T J+λ·diag(J T J)] -1 J T (yf(β(k)))

[0135] Where β(k+1) is the parameter of the k+1th iteration, β(k) is the parameter of the jth iteration, J is the Jacobian matrix, T is the matrix transpose, λ is the adjustment factor of the fitting function, diag is, y is the observation value, representing the actual state of the airport baggage transfer system, and f(β(k)) is the model output of the current fault model based on the parameters of the kth iteration.

[0136] Specifically, a deep transfer learning function L is set. By minimizing the deep transfer learning function L, the source domain representing the characteristic values ​​of historical working status data is transferred to the target domain of the characteristic values ​​of current working status data, thereby improving the generalization ability of the fault model. Specifically:

[0137] L=L s (θ s )+λ′·L t (θ t )+μ·MMD(P s , P t )

[0138] Among them, L s (θ s ) is the source domain θ s The loss function, λ′ is the first adjustment factor of the deep transfer learning function L, L t (θ t ) is the target domain θ t The loss function, μ is the second adjustment factor of the deep transfer learning function L, MMD(P s , Pt ) is the source domain θ s The distribution of working status data eigenvalues ​​and target domain θ t The maximum mean difference of the characteristic value distribution of the working status data.

[0139] The serial numbers of the above embodiments of the present invention are for description only and do not represent the advantages or disadvantages of the embodiments.

[0140] In the above embodiments of the present invention, the description of each embodiment has its own focus. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0141] In the several embodiments provided by the present invention, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the system embodiments described above are merely illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, and can be electrical or other forms.

[0142] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0143] In addition, the functional units in the various embodiments of the present invention may be integrated into a single processing unit, each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0144] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, read-only storage medium (ROM, Read-Only Memory), random access storage medium (RAM, Random Access Memory), mobile hard disk, magnetic disk or optical disk, etc. Various media that can store program codes.

[0145] Obviously, the above embodiments are merely examples for clarity of explanation and are not intended to limit the implementation methods. Those skilled in the art will readily appreciate that other variations or modifications based on the above descriptions are possible. It is not necessary and impossible to enumerate all implementation methods here. Obvious variations or modifications arising therefrom remain within the scope of protection of the present invention.

Claims

1. A machine learning-based fault diagnosis method for airport baggage transfer systems, characterized in that: include: Using sensors installed on the airport baggage handling system, the system collects in real time the characteristic values ​​of the operating status data of the equipment on the airport baggage handling system, wherein the characteristic values ​​of the operating status data include: equipment operating temperature, equipment operating pressure, equipment operating vibration, and equipment operating speed; Setting an original fault model, and calculating a fault value of the airport baggage transfer system according to the characteristic value of the working status data, and determining whether the airport baggage transfer system has a fault according to the fault value; Setting a time window and the number of windows, collecting characteristic values ​​of working status data within the time window, and updating the characteristic values ​​of the working status data within the time window in a first-in-first-out manner, training the original fault model based on the characteristic values ​​of the working status data within the time window, generating an updated fault model, and fitting the parameters of the updated fault model; A model evaluation function is set to calculate the error value between the updated fault model and the original fault model, thereby evaluating the fitted updated fault model. When the error value between the updated fault model and the original fault model exceeds a preset error threshold, the parameters of the updated fault model are continuously fitted until it is less than the preset error threshold. The corresponding updated fault model is used as the final fault model, and fault diagnosis of the airport baggage transfer system is performed.

2. The method for diagnosing airport baggage handling system faults based on machine learning according to claim 1, characterized in that: The model evaluation function includes: Among them, M is the error value between the updated fault model and the original fault model, N is the number of characteristic values ​​of the working state data, and x ft is the updated prediction result output after the updated new working status data feature value at time t is input into the updated fault model, x ai is the original prediction result output after the i-th original working state data characteristic value is input into the original fault model, and α′ is the adjustment factor of the model evaluation function.

3. The method for diagnosing airport baggage transfer system faults based on machine learning according to claim 1, characterized in that: The parameters of the updated fault model are fitted as follows: β(k+1)=β(k)-[J T J+λ·diag(J T J)] -1 J T (yf(β(k))) Where β(k+1) is the parameter of the k+1th iteration, β(k) is the parameter of the kth iteration, J is the Jacobian matrix, T is the matrix transpose, λ is the adjustment factor of the fitting function, diag is, y is the observation value, representing the actual state of the airport baggage transfer system, and f(β(k)) is the model output of the current fault model based on the parameters of the kth iteration.

4. A machine learning-based fault diagnosis method for an airport baggage handling system according to any one of claims 1 to 3, characterized in that: Set up a deep transfer learning function L. By minimizing the deep transfer learning function L, the source domain representing the characteristic values ​​of historical working status data is transferred to the target domain representing the characteristic values ​​of current working status data, thereby improving the generalization ability of the fault model. Specifically: L=L s (θ s )+λ′·L t (θ t )+μ·MMD(P s ,P t ) Among them, L s (θ s ) is the source domain θ s The loss function, λ′ is the first adjustment factor of the deep transfer learning function L, L t (θ t ) is the target domain θ t The loss function, μ is the second adjustment factor of the deep transfer learning function L, MMD(P s , P t ) is the source domain θ s The distribution of working status data eigenvalues ​​and target domain θ t The maximum mean difference of the characteristic value distribution of the working status data.

5. A method for diagnosing airport baggage handling system faults based on machine learning according to any one of claims 1 to 3, characterized in that: The original fault model is as follows: X(t)=[x1(t),x2(t),...,x n (t)] T Among them, x n (t) is the nth characteristic value of the working status data at time t, T is the transpose of the matrix, is the fault identification spectrum corresponding to the set X(t) of the working status data eigenvalues ​​at time t, W is the transformation matrix used to control the mapping of the working status data eigenvalues, σ is sigmoid or ReLU, A is the weight matrix, b is the bias term, and c is the constant term.

6. The method for diagnosing airport baggage transfer system faults based on machine learning according to claim 5, characterized in that: The original fault model also includes: Where ∈(t) is the dynamic adaptive tolerance at time t, which is used to reduce false alarms, and α″ is the first adjustment factor of the dynamic adaptive tolerance. is the fault identification spectrum corresponding to the set X(t-1) of the characteristic values ​​of the working status data at time t-1, β″ is the second adjustment factor of the dynamic adaptive tolerance, and F(t) is the fault value of the airport baggage handling system at time t.

7. A machine learning-based airport baggage handling system fault diagnosis system, characterized in that: include: A collection module is used to collect operating status data characteristic values ​​of equipment on the airport baggage transfer system in real time through sensors installed on the airport baggage transfer system, wherein the operating status data characteristic values ​​include: equipment operating temperature, equipment operating pressure, equipment operating vibration, and equipment operating speed; Setting a fault model module, used for setting an original fault model, and calculating a fault value of the airport baggage transfer system according to the characteristic value of the working status data, and judging whether the airport baggage transfer system has a fault according to the fault value; A fitting module is used to set a time window and the number of windows, collect characteristic values ​​of the working status data within the time window, and update the characteristic values ​​of the working status data within the time window using a first-in-first-out method, train the original fault model based on the characteristic values ​​of the working status data within the time window, generate an updated fault model, and fit the parameters of the updated fault model; The model evaluation module is used to set a model evaluation function and calculate the error value between the updated fault model and the original fault model, thereby evaluating the fitted updated fault model. When the error value between the updated fault model and the original fault model exceeds a preset error threshold, the parameters of the updated fault model are continuously fitted until the error value is less than the preset error threshold. The corresponding updated fault model is used as the final fault model, and fault diagnosis of the airport baggage transfer system is performed.

8. The machine learning-based airport baggage handling system fault diagnosis system according to claim 7, characterized in that: The model evaluation function includes: Among them, M is the error value between the updated fault model and the original fault model, N is the number of characteristic values ​​of the working state data, and x ft is the updated prediction result output after the updated new working status data feature value at time t is input into the updated fault model, x ai is the original prediction result output after the i-th original working state data characteristic value is input into the original fault model, and α′ is the adjustment factor of the model evaluation function.

9. A monitoring device, characterized in that: include: memory for storing computer programs; A processor, configured to implement the steps of the fault diagnosis method according to any one of claims 1 to 6 when executing the computer program.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the fault diagnosis method according to any one of claims 1 to 6 are implemented.

Images