Air compressor fault monitoring system based on big data

Abstract

The invention discloses an air compressor fault monitoring system based on big data. The air compressor fault monitoring system comprises a data acquisition module, an air compressor dynamic model construction module, an energy state function construction module, a state judgment module, a parameter correction module and a fault monitoring module. The invention belongs to the field of fault monitoring, and particularly relates to an air compressor fault monitoring system based on big data. According to the scheme, nonlinearity and time delay characteristics of an air compressor are captured by constructing a dynamic model based on a neural network; a delay product function is used for capturing transient and long-term trends, and then a zero value supplement item is introduced to improve the sensitivity to small changes of the air compressor; the accuracy of fault monitoring of the air compressor is further improved; introducing primary and secondary movement judgment; through two-stage movement judgment, not only can potential excellent solutions be quickly captured in the initial stage, but also local optimum can be avoided by means of global optimum information; a three-dimensional spiral track is constructed based on the overall moving unit, so that the global exploration capability is enhanced; and the fault monitoring efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the field of fault monitoring, and specifically refers to an air compressor fault monitoring system based on big data. Background Art

[0002] An air compressor fault monitoring system is a system used to monitor the operating state in real time, discover and diagnose potential faults in a timely manner. It plays an important role in ensuring the stable operation of the air compressor, improving production efficiency, and reducing maintenance costs. However, the traditional air compressor fault monitoring system has poor processing capabilities for delay effects, mutations, and transient responses, resulting in insufficiently sensitive fault warnings and prone to missing or misjudging faults; the traditional air compressor fault monitoring system has problems with improper parameter optimization, leading to poor actual operation effects of fault monitoring. Summary of the Invention

[0003] In view of the above situation, to overcome the defects of the prior art, the present invention provides an air compressor fault monitoring system based on big data. Aiming at the problem that the traditional air compressor fault monitoring system has poor processing capabilities for delay effects, mutations, and transient responses, resulting in insufficiently sensitive fault warnings and prone to missing or misjudging faults, this solution captures the non-linear and time-delay characteristics of the air compressor by constructing a dynamic model based on a neural network; uses a delay product function to capture transient and long-term trends, and then introduces a zero-value supplement term to further improve the sensitivity to small changes in the air compressor; thereby improving the accuracy of air compressor fault monitoring. Aiming at the problem that the traditional air compressor fault monitoring system has improper parameter optimization, resulting in poor actual operation effects of fault monitoring, this solution introduces primary and secondary movement determination. The primary movement determination unit uses a control factor to preliminarily adjust the individual position; the secondary movement determination unit then introduces a random factor and adjusts the individual position according to the position of the global optimal solution; through two-stage movement determination, not only can potential excellent solutions be quickly captured in the initial stage, but also the global optimal information can be used to avoid falling into local optima; and a three-dimensional spiral trajectory is constructed based on the overall movement unit to enhance the global exploration ability; thereby improving the efficiency of fault monitoring.

[0004] The technical solution adopted by the present invention is as follows: The air compressor fault monitoring system based on big data provided by the present invention includes a data acquisition module, an air compressor dynamic model construction module, an energy state function construction module, a state determination module, a parameter correction module, and a fault monitoring module;

[0005] The data acquisition module collects the historical operation data of the air compressor and obtains an air compressor training data set based on feature engineering processing;

[0006] The air compressor dynamic model construction module constructs an air compressor dynamic model based on the air compressor training data set and a neural network;

[0007] The energy state function construction module constructs and corrects the energy state function by weighting quadratic terms and integral terms;

[0008] The state determination module realizes the determination of the operating state of the air compressor by calculating the change rate of the energy state function and combining load prediction and external disturbances;

[0009] The parameter correction module optimizes the parameter settings within the parameter correction space by using a multi-stage iterative optimization method;

[0010] The fault monitoring module realizes fault monitoring for the operation data of the air compressor collected in real time.

[0011] Further, in the data acquisition module, the historical operation data of the air compressor includes sensor data, environmental parameter data, operation state data, response time data, and operation state categories; the operation state category is used as a data label; the operation state categories include normal operation and abnormal operation; the operation state data includes load changes and operation records; and the collected data is subjected to feature engineering processing; to obtain an air compressor training data set.

[0012] Further, the air compressor dynamic model construction module is based on the air compressor training data set and a neural network, describes the multi-dimensional state of the air compressor as a state vector, and assumes that the state vector of the air compressor is x(·), and the air compressor dynamic model is expressed as: ; where, is the state change rate; t is the time variable; A1 is the basic control matrix, reflecting the mechanical damping characteristics; W0, W1, and W2 are all neural network weight matrices, which are used to represent the direct influence term, the delayed influence term, and the non-linear mapping respectively; f(·) is the activation function; τ(·) is the internal transfer delay function of the air compressor, reflecting the response delay; is the external disturbance; the original sample is mapped through a Gaussian kernel and expressed as: ; where, G(·) is the Gaussian kernel mapping function; σ is the bandwidth parameter; is the Gaussian kernel center; is the jth sample data.

[0013] Further, the energy state function construction module uses the energy function to quantify the healthy energy of the operation state of the air compressor, specifically including the following contents:

[0014] Initial energy state function construction unit, initial energy state function is defined as: ; where, T is the transpose operation; P and Q are positive definite matrices, P is used to weight the contribution of the state vector in the energy function, and Q is used to weight the contribution of the state vector in the integral term; s is the integral variable; T is the transpose operation;

[0015] Delay product function construction unit; constructing short-term trend term , expressed as: ; constructing long-term trend term , expressed as: ; constructing dynamic change characteristic term , expressed as: ; where is the delay margin at the current moment; is the longest allowable delay during the response process of the air compressor;

[0016] Zero-value supplementary term introduction unit; expressed as: ; where Q2 and Q3 are positive definite matrices used to adjust the influence of the zero-value term; is the combined vector containing the original state of the air compressor and the information after non-linear mapping;

[0017] Energy state function correction unit; fusing various characteristics to form the corrected energy state function, expressed as: ; where R is a positive definite matrix used to balance the contributions of the delay term and the zero-value supplementary term to the energy.

[0018] Furthermore, the state determination module determines the state by using the change rate of the energy function, expressed as: , when the condition is met, it is determined that the air compressor is in normal operation; if not, it is determined to be in abnormal operation; where is the time derivative of the energy state function, reflecting the change rate of energy with time; is the load change predicted by the neural network; sym{·} is the symmetric change of energy under perturbation; γ is the response parameter.

[0019] Furthermore, the parameter correction module specifically includes the following:

[0020] Initialization unit; constructing a parameter correction space based on the basic control matrix, direct and delay influence terms, non-linear mapping, maximum delay, response parameter, and bandwidth parameter; randomly initializing the individual positions of the optimization population, and using the determination accuracy rate of the training model on the test set as the fitness evaluation;

[0021] Initial movement determination unit; obtaining the individual initial movement determination using the control factor A, expressed as: ; ; where C s (·) is the distance of the individual initial movement determination; i is the search number index; P s (·) is the individual position; α and β are both adjustment parameters used to control the non-linear change of A; Max is the maximum search number;

[0022] Quadratic movement determination unit; obtaining the individual quadratic movement determination using the random factor B, expressed as: ; ; thereby initially updating the individual position, expressed as: ; where M s (·) is the distance of the individual quadratic movement determination; P Bbt is the position of the global optimal solution; f c is the individual fitness value; D s (·) is the position after the individual's initial update; rand is a random number between 0 and 1;

[0023] Overall movement unit; calculating the three-dimensional helical motion trajectory, expressed as: ; performing overall movement on the optimized individual to obtain the final iterative position, expressed as: ; where x, y, and z are the three-dimensional coordinates of the helical motion trajectory respectively; r is the motion radius; is the random angle;

[0024] Iterative determination unit; when there is an individual fitness value higher than the determination threshold, the iteration ends, and finally the global optimal individual parameters are used to set the module parameters, and the parameter correction module training is completed; if the maximum number of iterations is reached, return to the initialization unit; otherwise, return to the initial movement determination unit to continue the iteration.

[0025] Furthermore, the fault monitoring model collects the operation data of the air compressor in real time. After feature engineering processing, it is processed by the air compressor dynamic model construction module and the energy state function construction module, and the operation state is output by the state determination module. If the state determination module determines that the operation is abnormal, warning processing is performed.

[0026] The beneficial effects achieved by the present invention using the above solution are as follows:

[0027] (1) Aiming at the problem that the traditional air compressor fault monitoring system has poor processing capabilities for delay effects, mutations, and transient responses, resulting in insensitive fault warnings and prone to missing or misjudging faults, this solution captures the non-linear and time-delay characteristics of the air compressor by constructing a neural network-based dynamic model; uses the delay product function to capture transient and long-term trends, and then introduces a zero-value supplement term to further improve the sensitivity to small changes in the air compressor; thereby improving the accuracy of air compressor fault monitoring.

[0028] (2) Aiming at the problem that the traditional air compressor fault monitoring system has poor actual operation effect of fault monitoring due to improper parameter optimization, this solution introduces primary and secondary movement determination. The primary movement determination unit preliminarily adjusts the individual position using a control factor; the secondary movement determination unit introduces a random factor and adjusts the individual position according to the position of the global optimal solution. Through the two-stage movement determination, not only can potential excellent solutions be quickly captured in the initial stage, but also the global optimal information can be used to avoid falling into local optima; and a three-dimensional spiral trajectory is constructed based on the overall movement unit to enhance the global exploration ability; thereby improving the efficiency of fault monitoring. Brief Description of the Drawings

[0029] Figure 1 It is a schematic flowchart of the air compressor fault monitoring system based on big data provided by the present invention;

[0030] Figure 2 It is a schematic flowchart of the energy state function construction module.

[0031] The drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation to the present invention. Detailed Embodiments

[0032] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments; based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the scope of protection of the present invention.

[0033] In the description of the present invention, it should be understood that the terms "upper", "lower", "front", "rear", "left", "right", "top", "bottom", "inner", "outer", etc. indicate the orientation or position relationship based on the orientation or position relationship shown in the drawings, and are 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 understood as a limitation to the present invention.

[0034] Embodiment 1, referring to Figure 1 The air compressor fault monitoring system based on big data provided by the present invention includes a data acquisition module, an air compressor dynamic model construction module, an energy state function construction module, a state determination module, a parameter correction module, and a fault monitoring module;

[0035] The data acquisition module collects the historical operation data of the air compressor, processes it based on feature engineering to obtain the air compressor training data set; and sends the data to the air compressor dynamic model construction module;

[0036] The air compressor dynamic model construction module constructs an air compressor dynamic model based on the air compressor training data set and a neural network; and sends the data to the energy state function construction module;

[0037] The energy state function construction module constructs and corrects the energy state function through weighted quadratic terms and integral terms; and sends the data to the state determination module;

[0038] The state determination module realizes the determination of the operating state of the air compressor by calculating the change rate of the energy state function and combining load prediction and external disturbances; and sends the data to the parameter correction module;

[0039] The parameter correction module uses a multi-stage iterative optimization method to optimize the parameter settings in the parameter correction space; and sends the data to the fault monitoring module;

[0040] The fault monitoring module realizes fault monitoring of the real-time collected operating data of the air compressor.

[0041] Example 2, refer to Figure 1 , this example is based on the above example. In the data acquisition module, the historical operating data of the air compressor includes sensor data, environmental parameter data, operating state data, response time data, and operating state categories; the operating state category is used as the data label; the operating state categories include normal operation and abnormal operation; the sensor data includes pressure, flow rate, temperature, wear degree, oil temperature, vibration, and rotational speed; the environmental parameter data includes environmental temperature, humidity, power supply voltage, and current; the operating state data includes load changes and operation records; and feature engineering processing is performed on the collected data; an air compressor training data set is obtained.

[0042] Example 3, refer to Figure 1 , this example is based on the above example. In the actual operation of the air compressor, its operating state is affected by various factors such as environmental temperature, load fluctuations, and mechanical wear, and faults often appear in the form of state mutations or energy anomalies; based on the air compressor training data set and a neural network, the multi-dimensional state of the air compressor is described as a state vector. Let the state vector of the air compressor be x(·), and the air compressor dynamic model is expressed as: ; where is the state change rate; t is the time variable; A1 is the basic control matrix, reflecting the mechanical damping characteristics; W0, W1, and W2 are all neural network weight matrices, which are used to represent the direct influence term, the delayed influence term, and the non-linear mapping respectively; f(·) is the activation function; τ(·) is the internal transfer delay function of the air compressor, reflecting the response delay; is the external disturbance, corresponding to environmental interference and power supply fluctuations; to highlight the key fault characteristics, the original samples are mapped through a Gaussian kernel and expressed as: ; where G(·) is the Gaussian kernel mapping function; σ is the bandwidth parameter that maps the original signal to a new feature space to highlight key fault features; is the Gaussian kernel center; is the j-th sample data.

[0043] Example 4, refer to Figure 1 and Figure 2 , this example is based on the above example. The energy state function construction module uses an energy function to quantify the health energy of the air compressor operation state, specifically including the following:

[0044] Initial energy state function construction unit, the initial energy state function is defined as: ; where T is the transpose operation; P and Q are positive definite matrices. P is used to weight the contribution of the state vector in the energy function, and Q is used to weight the contribution of the state vector in the integral term; s is the integration variable; T is the transpose operation;

[0045] Delay product function construction unit; to capture the change characteristics of the air compressor state due to time delay, capture the transient and progressive changes of the air compressor respectively for easy fault location, and construct the following delay functions: construct the short-term trend term , expressed as: ; construct the long-term trend term , expressed as: ; construct the dynamic change characteristic term , expressed as: ; where is the delay margin at the current moment; is the longest delay allowed during the air compressor response process;

[0046] Zero value supplement term introduction unit; improve the sensitivity and description accuracy to small changes; expressed as: ; where Q2 and Q3 are positive definite matrices used to adjust the influence of the zero value term; is the combined vector, which contains the original state of the air compressor and the information after non-linear mapping, and is used to improve the sensitivity to small changes;

[0047] Energy state function correction unit; fuse each feature to form the corrected energy state function, expressed as: ; where R is a positive definite matrix used to balance the contributions of the delay term and the zero value supplement term in the energy.

[0048] By performing the above operations, aiming at the problems existing in the traditional air compressor fault monitoring system, such as poor ability to handle delay effects, mutations, and transient responses, resulting in insensitive fault warnings and prone to missing or misjudging faults, this solution constructs a dynamic model based on neural network to capture the nonlinear and time-delay characteristics of the air compressor; uses the delay product function to capture transient and long-term trends, and then introduces a zero-value supplementary term to further improve the sensitivity to small changes in the air compressor; thereby improving the accuracy of air compressor fault monitoring.

[0049] Embodiment Six. Refer to Figure 1 , based on the above embodiment, the state determination module determines the state by using the change rate of the energy function, which is expressed as: , when the condition is met, it is determined that the state of the air compressor is normal; if not, it is determined as abnormal operation; where is the time derivative of the energy state function, reflecting the change rate of energy over time; is the load change predicted by the neural network; sym{·} is the symmetric change of energy under perturbation; γ is the response parameter, used to adjust the response amplitude of the perturbation term to the overall energy state change.

[0050] Embodiment Seven. Refer to Figure 1 , based on the above embodiment, the parameter correction module aims at the complex operating environment and nonlinear behavior of the air compressor, and adopts parameter correction to improve the determination accuracy, specifically including the following:

[0051] Initialization unit; construct a parameter correction space based on the basic control matrix, direct and delay influence terms, nonlinear mapping, maximum delay, response parameter, and bandwidth parameter; randomly initialize the individual positions of the optimization population, and use the determination accuracy rate of the training model on the test set as the fitness evaluation;

[0052] Initial movement determination unit; use the control factor A to obtain the initial movement determination of the individual, which is expressed as: ; ; where C s (·) is the distance of the initial movement determination of the individual; i is the search times index; P s (·) is the individual position; α and β are both adjustment parameters, used to control the nonlinear change of A; Max is the maximum search times;

[0053] Second movement determination unit; use the random factor B to obtain the second movement determination of the individual, which is expressed as: ; ; thus, the individual position is initially updated, which is expressed as: ; where M s (·) is the distance of the second movement determination of the individual; P Bbt is the position of the global optimal solution; fc is the individual fitness value; D s (·) is the position of the individual after the initial update; rand is a random number between 0 and 1;

[0054] Overall movement unit; calculate the three-dimensional spiral motion trajectory, expressed as: ; perform overall movement on the optimized individual to obtain the final iterative position, expressed as: ; where x, y, and z are the three-dimensional coordinates of the spiral motion trajectory respectively; r is the radius of motion; is the random angle;

[0055] Iterative decision-making unit; when there is an individual fitness value higher than the decision threshold, the iteration ends, and finally the global optimal individual parameters are used to set the module parameters, and the parameter correction module training is completed; if the maximum number of iterations is reached, return to the initialization unit; otherwise, return to the initial movement decision-making unit to continue the iteration.

[0056] Furthermore, the fault monitoring model collects the operation data of the air compressor in real time. After feature engineering processing, it is processed by the air compressor dynamic model construction module and the energy state function construction module, and the operation state is output by the state decision-making module. If the state decision-making module determines that the operation is abnormal, warning processing is performed.

[0057] By performing the above operations, aiming at the problem that the traditional air compressor fault monitoring system has poor actual operation effect of fault monitoring due to improper parameter optimization, this solution introduces initial and secondary movement decisions. The initial movement decision-making unit uses the control factor to preliminarily adjust the individual position; the secondary movement decision-making unit introduces the random factor and adjusts the individual position according to the position of the global optimal solution; through the two-stage movement decision, it can not only quickly capture potential excellent solutions in the initial stage, but also avoid falling into the local optimum by means of the global optimal information; and construct a three-dimensional spiral trajectory based on the overall movement unit to enhance the global exploration ability; thereby improving the efficiency of fault monitoring.

[0058] It should be noted that in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "including", "comprising" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements not only includes those elements, but also includes other elements not expressly listed, or also includes elements inherent to such process, method, article or device.

[0059] Although embodiments of the present invention have been shown and described, those of ordinary skill in the art will understand that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the present invention.

[0060] The present invention and its embodiments have been described above. Such description is not restrictive. What is shown in the drawings is only one of the embodiments of the present invention, and the actual structure is not limited thereto. Generally speaking, if those of ordinary skill in the art are inspired by it and, without departing from the gist of the present invention, creatively design a structural manner and embodiments similar to the technical solution, they shall fall within the protection scope of the present invention.

Claims

1. An air compressor fault monitoring system based on big data, characterized in that: The system includes a data acquisition module, an air compressor dynamic model construction module, an energy state function construction module, a state determination module, a parameter correction module, and a fault monitoring module; The data acquisition module collects the historical operation data of the air compressor and obtains the air compressor training data set based on feature engineering processing; The air compressor dynamic model construction module constructs an air compressor dynamic model based on the air compressor training data set and a neural network; The energy state function construction module constructs and corrects the energy state function through a weighted quadratic term and an integral term; The state determination module realizes the determination of the operation state of the air compressor by calculating the change rate of the energy state function and combining load prediction and external disturbance; The parameter correction module uses a multi-stage iterative optimization method to optimize the parameter settings within the parameter correction space; The fault monitoring module realizes fault monitoring of the real-time collected operation data of the air compressor.

2. The air compressor fault monitoring system based on big data according to claim 1, wherein: The air compressor dynamic model construction module is based on the air compressor training data set and neural network, and describes the multi-dimensional state of the air compressor as a state vector. Let the state vector of the air compressor be x(·), and the air compressor dynamic model is expressed as: ; where is the state change rate; t is the time variable; A1 is the basic control matrix, reflecting the mechanical damping characteristics; W0, W1, and W2 are all neural network weight matrices, which are used to represent the direct influence term, the delayed influence term, and the non-linear mapping respectively; f(·) is the activation function; τ(·) is the internal transfer delay function of the air compressor, reflecting the response delay; is the external disturbance; the original sample is mapped through the Gaussian kernel and expressed as: ; where G(·) is the Gaussian kernel mapping function; σ is the bandwidth parameter; is the Gaussian kernel center; is the j-th sample data.

3. The air compressor fault monitoring system based on big data according to claim 2, characterized in that: The energy state function construction module uses the energy function to quantify the healthy energy of the operation state of the air compressor, and specifically includes the following: Initial energy state function construction unit, initial energy state function is defined as: ; where T is the transpose operation; P and Q are positive definite matrices, P is used to weight the contribution of the state vector in the energy function, and Q is used to weight the contribution of the state vector in the integral term; s is the integration variable; T is the transpose operation; Delay product function construction unit; constructing short-term trend term , expressed as: ; constructing long-term trend term , expressed as: ; constructing dynamic change characteristic term , expressed as: ; where is the delay margin at the current moment; is the longest allowable delay during the response process of the air compressor; Zero-value supplementary term introduction unit; expressed as: ; where Q2 and Q3 are positive definite matrices used to adjust the influence of zero-value terms; is a combined vector that contains the original state of the air compressor and the information after non-linear mapping; Energy state function correction unit.

4. The air compressor fault monitoring system based on big data according to claim 3, wherein: The energy state function correction unit fuses various features to form a corrected energy state function, which is expressed as: ; where R is a positive definite matrix used to balance the contributions of the delay term and the zero-value supplement term to the energy.

5. The air compressor fault monitoring system based on big data according to claim 4, wherein: The state determination module uses the rate of change of the energy function for state determination, expressed as: , when the condition is met, it is determined that the air compressor is operating normally; if not, it is determined to be operating abnormally; where is the time derivative of the energy state function, reflecting the rate of change of energy over time; is the load change predicted by the neural network; sym{·} is the symmetric change of energy under perturbation; γ is the response parameter.

6. The air compressor fault monitoring system based on big data according to claim 5, wherein: The parameter correction module specifically includes the following: Initialization unit; Construct a parameter correction space based on the basic control matrix, direct and delay impact terms, non-linear mapping, maximum delay, response parameters, and bandwidth parameters; randomly initialize the individual positions of the optimization population, and use the determination accuracy rate of the training model on the test set as the fitness evaluation; Initial movement determination unit; obtaining an individual's initial movement determination using control factor A, expressed as: ; ; where C s (·) is the distance of the individual's initial movement determination; i is the search times index; P s (·) is the individual's position; both α and β are adjustment parameters used to control the non-linear change of A; Max is the maximum number of searches; The secondary movement determination unit; obtaining the individual secondary movement determination using the random factor B, expressed as: ; ; thereby initially updating the individual position, expressed as: ; where M s (·) is the distance of the individual secondary movement determination; P Bbt is the position of the global optimal solution; f c is the individual fitness value; D s (·) is the position after the individual's initial update; rand is a random number between 0 and 1; Overall movement unit; calculating a three-dimensional helical motion trajectory, expressed as: ; performing overall movement on the optimized individual to obtain the final iterative position, expressed as: ; where x, y, and z are the three-dimensional coordinates of the helical motion trajectory respectively; r is the motion radius; is a random angle; Iterative determination unit; when there is an individual fitness value higher than the determination threshold, the iteration ends, and finally the global optimal individual parameter is used to set the module parameters, and the parameter correction module training is completed; if the maximum iteration number is reached, return to the initialization unit; otherwise, return to the initial movement determination unit to continue the iteration.

7. The air compressor fault monitoring system based on big data according to claim 6, characterized in that: In the data acquisition module, the historical operation data of the air compressor includes sensor data, environmental parameter data, operation state data, response time data, and operation state category; Use the operation state category as the data label; the operation state category includes normal operation and abnormal operation; the operation state data includes load change and operation record; and perform feature engineering processing on the collected data; Obtain the air compressor training data set.

8. The air compressor fault monitoring system based on big data according to claim 7, wherein: The fault monitoring model collects the real-time operation data of the air compressor, which is processed by the air compressor dynamic model construction module and the energy state function construction module after feature engineering processing, and the operation state is output by the state determination module. If the state determination module determines abnormal operation, warning processing is performed.

Images