Pipeline identification method and device, air conditioner, storage medium and computer program product

Abstract

The invention discloses a pipeline identification method and device, an air conditioner, a storage medium and a computer program product, and the method is used for identifying whether a to-be-detected pipeline of equipment is cracked or not. The pipeline identification method comprises the following steps: for a to-be-detected pipeline in pipelines of the equipment, acquiring an acoustic emission signal of the to-be-detected pipeline; extracting characteristic parameters from the acoustic emission signals of the to-be-detected pipeline as acoustic emission characteristic parameters of the to-be-detected pipeline; and based on the acoustic emission characteristic parameters of the to-be-detected pipeline, identifying whether the to-be-detected pipeline is cracked by using a pre-trained pipeline identification model, so as to initiate a reminding message of cracking of the to-be-detected pipeline under the condition of identifying that the to-be-detected pipeline is cracked. According to the scheme, the acoustic emission technology and the neural network technology are adopted, whether the pipeline is cracked or not is recognized based on the acoustic emission signal of the pipeline of the equipment (such as an air conditioner), and the reminding message is initiated when the pipeline is recognized to be cracked, so that measures are taken, the service life of the pipeline is prolonged, and normal use of the equipment is guaranteed.

Description

Technical Field

[0001] The present invention belongs to the technical field of pipeline identification, and specifically relates to a pipeline identification method, device, air conditioner, storage medium and computer program product, and in particular to a fatigue crack identification method, device, air conditioner, storage medium and computer program product of an air conditioning pipeline based on acoustic emission technology. Background Art

[0002] The vibration of equipment (such as air conditioners) pipelines has a great impact on the reliability and noise level of the unit. The vibration of air conditioning pipelines is usually caused by the vibration excitation of the compressor. During the long-term operation of the air conditioning pipelines, due to the influence of multiple factors such as compressor vibration, temperature changes, pressure fluctuations, etc., the material of the air conditioning pipelines gradually fatigues, increases the stress of the air conditioning pipelines, and cracks appear at the welding positions of the air conditioning pipelines, which may eventually cause the air conditioning pipelines to rupture or be damaged. However, the relevant scheme cannot detect whether the air conditioning pipelines are cracked. It is not discovered until the air conditioning pipelines are ruptured or damaged due to cracks. This not only affects the service life of the air conditioning pipelines, but also affects the normal use of the air conditioner.

[0003] The above contents are only used to assist in understanding the technical solution of the present invention and do not constitute an admission that the above contents are prior art. Summary of the invention

[0004] The object of the present invention is to provide a pipeline identification method, device, air conditioner, storage medium and computer program product to solve the problem that the excessive stress of the pipeline of equipment (such as air conditioner) cannot be detected, resulting in fatigue cracks in the pipeline, affecting the service life of the pipeline and the normal use of the equipment, so as to achieve the effect of identifying whether the pipeline is cracked based on the acoustic emission signal of the pipeline of the equipment (such as air conditioner) by adopting acoustic emission technology and neural network technology, and initiating a reminder message when the pipeline is cracked, so as to take measures to extend the service life of the pipeline and ensure the normal use of the equipment.

[0005] The present invention provides a pipeline identification method, which is applied to identifying whether a pipeline to be tested of a device is cracked; the pipeline identification method comprises: obtaining an acoustic emission signal of the pipeline to be tested in the pipelines of the device; extracting characteristic parameters from the acoustic emission signal of the pipeline to be tested as acoustic emission characteristic parameters of the pipeline to be tested; based on the acoustic emission characteristic parameters of the pipeline to be tested, using a pipeline identification model obtained by pre-training, identifying whether the pipeline to be tested is cracked, so as to initiate a reminder message of the cracking of the pipeline to be tested when it is identified that the pipeline to be tested is cracked.

[0006] In some embodiments, the training process of the pre-trained pipeline identification model includes: in the experimental process of detecting cracks in the experimental pipeline, collecting the acoustic emission signal of the experimental pipeline and collecting the cracking parameters of the experimental pipeline; extracting characteristic parameters from the acoustic emission signal of the experimental pipeline as the acoustic emission characteristic parameters of the experimental pipeline; using the acoustic emission characteristic parameters of the experimental pipeline and the cracking parameters of the experimental pipeline as training samples, and using a preset neural network model as a basic model; using the acoustic emission characteristic parameters of the experimental pipeline as input parameters of the basic model, and using the cracking parameters of the experimental pipeline as output parameters of the basic model, and performing neural network training to obtain the required pipeline identification model.

[0007] In some embodiments, extracting characteristic parameters from the acoustic emission signal of the pipeline to be tested as the acoustic emission characteristic parameters of the pipeline to be tested includes: converting the acoustic emission signal of the pipeline to be tested into an electrical signal to obtain the acoustic emission electrical signal of the pipeline to be tested; using wavelet packet analysis to extract features from the acoustic emission electrical signal of the pipeline to be tested to obtain characteristic parameters, which are recorded as the acoustic emission characteristic parameters of the pipeline to be tested; and / or, in the case where the training process of the pre-trained pipeline recognition model includes extracting characteristic parameters from the acoustic emission signal of the experimental pipeline, extracting characteristic parameters from the acoustic emission signal of the experimental pipeline as the acoustic emission characteristic parameters of the experimental pipeline includes: converting the acoustic emission signal of the experimental pipeline into an electrical signal to obtain the acoustic emission electrical signal of the experimental pipeline; using wavelet packet analysis to extract features from the acoustic emission electrical signal of the experimental pipeline to obtain characteristic parameters, which are recorded as the acoustic emission characteristic parameters of the experimental pipeline.

[0008] In some embodiments, the experimental process for detecting cracking of an experimental pipeline includes: building a finite element model for the experimental pipeline in the pipeline of the device; using the finite element model to perform simulation analysis on the experimental pipeline to determine the location where a preset degree of fracture or slip occurs in the experimental pipeline, which is recorded as a dangerous location of the experimental pipeline; collecting acoustic emission signals of the experimental pipeline from the dangerous location of the experimental pipeline and within a set area of ​​the dangerous location of the experimental pipeline, and collecting cracking parameters of the experimental pipeline.

[0009] In some embodiments, for a pipeline to be tested in the pipeline of the device, obtaining an acoustic emission signal of the pipeline to be tested includes: obtaining an acoustic emission signal of the pipeline to be tested collected by a preset first acoustic emission signal collection system; wherein the first acoustic emission signal collection system is pre-set at a preset cracking point on the pipeline to be tested; and / or, collecting the acoustic emission signal of the experimental pipeline from a dangerous position of the experimental pipeline and a set area within the dangerous position of the experimental pipeline, including: collecting the acoustic emission signal of the experimental pipeline collected by a preset second acoustic emission signal collection system; wherein the second acoustic emission signal collection system is pre-set at the dangerous position of the experimental pipeline and a set area within the dangerous position of the experimental pipeline.

[0010] In some embodiments, when the training process of the pre-trained pipeline identification model includes extracting characteristic parameters from the acoustic emission signal of the experimental pipeline, the acoustic emission characteristic parameters of the pipeline to be tested and the acoustic emission characteristic parameters of the corresponding pipeline in the acoustic emission characteristic parameters of the experimental pipeline include: time domain parameters of the acoustic emission electrical signal of the corresponding pipeline, frequency domain parameters of the acoustic emission electrical signal of the corresponding pipeline, and energy parameters of the acoustic emission electrical signal of the corresponding pipeline; wherein the time domain parameters of the acoustic emission electrical signal of the corresponding pipeline include at least one of the following: voltage amplitude of the acoustic emission electrical signal of the corresponding pipeline, voltage rise time of the acoustic emission electrical signal of the corresponding pipeline, voltage duration of the acoustic emission electrical signal of the corresponding pipeline, and vibration within the voltage duration of the acoustic emission electrical signal of the corresponding pipeline. the count value of the bell; the energy parameters of the acoustic emission electrical signal of the corresponding pipeline, including at least one of the following: the total energy of the acoustic emission electrical signal of the corresponding pipeline, the frequency band energy of the acoustic emission electrical signal of the corresponding pipeline in a preset frequency band, and the proportion of the frequency band energy of the acoustic emission electrical signal of the corresponding pipeline in the preset frequency band in the total energy; the frequency domain parameters of the acoustic emission electrical signal of the corresponding pipeline, including at least one of the following: the frequency of the acoustic emission electrical signal of the corresponding pipeline, and the pulse factor of the acoustic emission electrical signal of the corresponding pipeline; wherein, the frequency of the acoustic emission electrical signal of the corresponding pipeline includes at least one of the following: the initial frequency of the acoustic emission electrical signal of the corresponding pipeline, the peak frequency of the acoustic emission electrical signal of the corresponding pipeline, the average frequency of the acoustic emission electrical signal of the corresponding pipeline, and the inverse calculated frequency of the acoustic emission electrical signal of the corresponding pipeline.

[0011] Matching the above method, the present invention provides a pipeline identification device on the other hand, which is used to identify whether a pipeline to be tested of a device is cracked; the pipeline identification device includes: an acquisition unit, configured to acquire an acoustic emission signal of the pipeline to be tested for the pipeline to be tested in the pipeline of the device; a control unit, configured to extract characteristic parameters from the acoustic emission signal of the pipeline to be tested as acoustic emission characteristic parameters of the pipeline to be tested; the control unit is also configured to identify whether the pipeline to be tested is cracked based on the acoustic emission characteristic parameters of the pipeline to be tested and using a pipeline identification model obtained by pre-training, so as to initiate a reminder message of the cracking of the pipeline to be tested when it is identified that the pipeline to be tested is cracked.

[0012] In some embodiments, the control unit, the training process of the pipeline identification model obtained by pre-training, includes: in the experimental process of detecting the cracking of the experimental pipeline, collecting the acoustic emission signal of the experimental pipeline, and collecting the cracking parameters of the experimental pipeline; extracting characteristic parameters from the acoustic emission signal of the experimental pipeline as the acoustic emission characteristic parameters of the experimental pipeline; using the acoustic emission characteristic parameters of the experimental pipeline and the cracking parameters of the experimental pipeline as training samples, and using a preset neural network model as a basic model; using the acoustic emission characteristic parameters of the experimental pipeline as input parameters of the basic model, and using the cracking parameters of the experimental pipeline as output parameters of the basic model, and performing neural network training to obtain the required pipeline identification model.

[0013] In some embodiments, the control unit extracts characteristic parameters from the acoustic emission signal of the pipeline to be tested as the acoustic emission characteristic parameters of the pipeline to be tested, including: converting the acoustic emission signal of the pipeline to be tested into an electrical signal to obtain the acoustic emission electrical signal of the pipeline to be tested; using wavelet packet analysis to perform feature extraction on the acoustic emission electrical signal of the pipeline to be tested to obtain characteristic parameters, which are recorded as the acoustic emission characteristic parameters of the pipeline to be tested; and / or the control unit extracts characteristic parameters from the acoustic emission signal of the experimental pipeline as the acoustic emission characteristic parameters of the experimental pipeline when the training process of the pre-trained pipeline recognition model includes extracting characteristic parameters from the acoustic emission signal of the experimental pipeline, including: converting the acoustic emission signal of the experimental pipeline into an electrical signal to obtain the acoustic emission electrical signal of the experimental pipeline; using wavelet packet analysis to perform feature extraction on the acoustic emission electrical signal of the experimental pipeline to obtain characteristic parameters, which are recorded as the acoustic emission characteristic parameters of the experimental pipeline.

[0014] In some embodiments, the control unit performs an experimental process for detecting cracking of an experimental pipeline, including: building a finite element model for the experimental pipeline in the pipeline of the device; using the finite element model to perform simulation analysis on the experimental pipeline to determine the location where a preset degree of fracture or slip occurs in the experimental pipeline, and record it as a dangerous location of the experimental pipeline; collecting acoustic emission signals of the experimental pipeline from the dangerous location of the experimental pipeline and within a set area of ​​the dangerous location of the experimental pipeline, and collecting cracking parameters of the experimental pipeline.

[0015] In some embodiments, the acquisition unit acquires the acoustic emission signal of the pipeline to be tested in the pipeline of the device, including: acquiring the acoustic emission signal of the pipeline to be tested acquired by a preset first acoustic emission signal acquisition system; wherein the first acoustic emission signal acquisition system is pre-set at a preset cracking point on the pipeline to be tested; and / or, the control unit collects the acoustic emission signal of the experimental pipeline from a dangerous position of the experimental pipeline and a set area within the dangerous position of the experimental pipeline, including: collecting the acoustic emission signal of the experimental pipeline acquired by a preset second acoustic emission signal acquisition system; wherein the second acoustic emission signal acquisition system is pre-set at the dangerous position of the experimental pipeline and a set area within the dangerous position of the experimental pipeline.

[0016] In some embodiments, the control unit, when the training process of the pre-trained pipeline recognition model includes extracting characteristic parameters from the acoustic emission signal of the experimental pipeline, the acoustic emission characteristic parameters of the pipeline to be tested and the acoustic emission characteristic parameters of the experimental pipeline include: time domain parameters of the acoustic emission electrical signal of the corresponding pipeline, frequency domain parameters of the acoustic emission electrical signal of the corresponding pipeline, and energy parameters of the acoustic emission electrical signal of the corresponding pipeline; wherein the time domain parameters of the acoustic emission electrical signal of the corresponding pipeline include at least one of the following: voltage amplitude of the acoustic emission electrical signal of the corresponding pipeline, voltage rise time of the acoustic emission electrical signal of the corresponding pipeline, voltage duration of the acoustic emission electrical signal of the corresponding pipeline, and voltage duration of the acoustic emission electrical signal of the corresponding pipeline. the count value of the ringing in the interval; the energy parameters of the acoustic emission electrical signal of the corresponding pipeline, including at least one of the following: the total energy of the acoustic emission electrical signal of the corresponding pipeline, the frequency band energy of the acoustic emission electrical signal of the corresponding pipeline in a preset frequency band, and the proportion of the frequency band energy of the acoustic emission electrical signal of the corresponding pipeline in the preset frequency band in the total energy; the frequency domain parameters of the acoustic emission electrical signal of the corresponding pipeline, including at least one of the following: the frequency of the acoustic emission electrical signal of the corresponding pipeline, and the pulse factor of the acoustic emission electrical signal of the corresponding pipeline; wherein, the frequency of the acoustic emission electrical signal of the corresponding pipeline includes at least one of the following: the initial frequency of the acoustic emission electrical signal of the corresponding pipeline, the peak frequency of the acoustic emission electrical signal of the corresponding pipeline, the average frequency of the acoustic emission electrical signal of the corresponding pipeline, and the inverse calculated frequency of the acoustic emission electrical signal of the corresponding pipeline.

[0017] Matching the above-mentioned device, the present invention further provides an air conditioner on the other hand, including: the pipeline identification device mentioned above.

[0018] In accordance with the above method, the present invention provides a storage medium on another aspect, wherein the storage medium includes a stored program, wherein when the program is executed, the device where the storage medium is located is controlled to execute the steps of the pipeline identification method described above.

[0019] In accordance with the above method, the present invention provides a computer program product on another aspect, including a computer program, which implements the steps of the pipeline identification method described above when executed by a processor.

[0020] Therefore, the scheme of the present invention is to build an air-conditioning pipeline model (such as an air-conditioning pipeline finite element model) for simulation analysis of equipment pipelines (such as air-conditioning pipelines), select an area where cracks (small fractures or slips) appear in the air-conditioning pipeline, arrange an acoustic emission signal acquisition system (such as an acoustic emission sensor) in the selected area, and collect the acoustic emission signal of the air-conditioning pipeline by the acoustic emission signal acquisition system; use the wavelet packet analysis method to extract the characteristic parameters (such as eigenvalues) of the acoustic emission signal, and use the characteristic parameters of the acoustic emission signal and the air-conditioning pipeline parameters (such as parameters indicating whether the air-conditioning pipeline is cracked represented by the corresponding characteristic parameters) as samples, based on For this sample (with characteristic parameters of acoustic emission signals as input and air-conditioning pipeline parameters as output), BP neural network is used to perform neural network training to obtain an air-conditioning pipeline recognition model for identifying whether the air-conditioning pipeline is cracked; then, the air-conditioning pipeline recognition model is used to identify whether the air-conditioning pipeline is cracked based on the acoustic emission signals of the air-conditioning pipeline; thus, by adopting acoustic emission technology and neural network technology, whether the air-conditioning pipeline is cracked is identified based on the acoustic emission signals of equipment pipelines (such as air-conditioning pipelines), and a reminder message is initiated when the air-conditioning pipeline is identified to take measures to extend the service life of the air-conditioning pipeline and ensure the normal use of the air-conditioning.

[0021] Other features and advantages of the present invention will be set forth in the description which follows, and in part will be apparent from the description, or may be learned by practice of the present invention.

[0022] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 A schematic diagram of a flow chart of an embodiment of a pipeline identification method of the present invention;

[0024] Figure 2 A flow chart of an embodiment of a training process of a pipeline identification model obtained by pre-training in the method of the present invention;

[0025] Figure 3 It is a flow chart of an embodiment of extracting characteristic parameters from the acoustic emission signal of the pipeline to be tested in the method of the present invention;

[0026] Figure 4 It is a flow chart of an embodiment of extracting characteristic parameters from the acoustic emission signal of the experimental pipeline in the method of the present invention;

[0027] Figure 5 It is a flow chart of an embodiment of an experimental process for detecting cracking of an experimental pipeline in the method of the present invention;

[0028] Figure 6 It is a structural schematic diagram of an embodiment of a pipeline identification device of the present invention;

[0029] Figure 7 It is a schematic diagram of the process of identifying cracks in air-conditioning pipes based on acoustic emission technology and neural network technology;

[0030] Figure 8 This is the stress cloud diagram of the air conditioning pipeline;

[0031] Fig. 9 It is the displacement cloud diagram of the air conditioning pipeline;

[0032] Fig.10 A schematic diagram of a curve of frequency band energy numbered for each frequency band;

[0033] Fig.11 A schematic diagram showing the ratio of the frequency band energy numbered for each frequency band to the total energy;

[0034] Fig.12 is a curve diagram of acoustic emission parameters;

[0035] Fig.13 Schematic diagram of the distribution of pulse damage factors in different stages (such as initiation stage, extension stage and crack stage).

[0036] In conjunction with the accompanying drawings, the reference numerals in the embodiments of the present invention are as follows:

[0037] 102 - acquisition unit; 104 - control unit. DETAILED DESCRIPTION

[0038] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the specific embodiments of the present invention and the corresponding drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0039] According to an embodiment of the present invention, a pipeline identification method is provided, such as Figure 1 The pipeline identification method is applied to identify whether the pipeline to be tested of a device (such as an air conditioner) is cracked. In the scheme of the present invention, Figure 1 As shown, the pipeline identification method includes: steps S110 to S130.

[0040] In step S110, for a pipeline to be tested in the pipelines of the device, an acoustic emission signal of the pipeline to be tested is obtained.

[0041] In step S120, characteristic parameters are extracted from the acoustic emission signal of the pipeline to be tested as the acoustic emission characteristic parameters of the pipeline to be tested.

[0042] At step S130, based on the acoustic emission characteristic parameters of the pipeline to be tested, a pipeline identification model obtained by pre-training is used to identify whether the pipeline to be tested is cracked, so that when it is identified that the pipeline to be tested is cracked, a reminder message of the cracking of the pipeline to be tested is initiated to remind relevant personnel to take measures on the pipeline to be tested in time to avoid the cracking of the pipeline to be tested, which may cause the pipeline to be ruptured or damaged and unable to be used normally, which is beneficial to extending the service life of the pipeline to be tested and ensuring the normal use of the equipment; wherein, the pre-trained pipeline identification model is a pipeline identification model to be tested that is pre-trained based on acoustic emission technology, wavelet packet analysis method and neural network technology.

[0043] In order to solve the problem of fatigue cracks in air-conditioning pipelines due to excessive stress, the solution of the present invention applies acoustic emission technology and neural network technology to preventive maintenance of air-conditioning pipelines. By early detection of fatigue damage in air-conditioning pipelines and real-time detection to identify whether cracks occur in air-conditioning pipelines, cracks in air-conditioning pipelines can be identified and detected, and timely measures can be taken to extend the service life of air-conditioning pipelines. For example, smaller cracks can be repaired by welding or sealant, and larger cracks can be replaced in time to extend the service life of air-conditioning pipelines.

[0044] In some embodiments, the specific process of the training process of the pipeline identification model pre-trained in step S130 is shown in the following exemplary description.

[0045] Combine the following Figure 2 The flowchart of an embodiment of the training process of the pipeline identification model pre-trained in the method of the present invention further illustrates the specific process of the training process of the pipeline identification model pre-trained in step S130, including: step S210 to step S240.

[0046] Step S210, during the experimental process of detecting cracks in the experimental pipeline, collect the acoustic emission signal of the experimental pipeline and collect the cracking parameters of the experimental pipeline. The experimental pipeline is the pipeline used for the experiment in the equipment. The cracking parameters of the experimental pipeline are used to characterize whether the experimental pipeline is actually cracked; for example: if the experimental pipeline is not cracked, the cracking parameters of the experimental pipeline are values ​​within the set range; if the experimental pipeline is not cracked, the cracking parameters of the experimental pipeline are values ​​outside the set range.

[0047] Step S220, extracting characteristic parameters from the acoustic emission signal of the experimental pipeline as the acoustic emission characteristic parameters of the experimental pipeline.

[0048] Step S230: Using the acoustic emission characteristic parameters of the experimental pipeline and the cracking parameters of the experimental pipeline as training samples, and using the preset neural network model as the basic model. The preset neural network model is, for example, a BP neural network model.

[0049] Step S240, using the acoustic emission characteristic parameters of the experimental pipeline as input parameters of the basic model and the cracking parameters of the experimental pipeline as output parameters of the basic model, performing neural network training to obtain the required pipeline identification model.

[0050] In the scheme of the present invention, an artificial neural network is adopted to establish the relationship between the cracks in the air-conditioning pipeline and the acoustic emission signal by utilizing the characteristic value parameters in the acoustic emission signal, thereby realizing real-time detection of whether cracks have occurred in the air-conditioning pipeline, so that timely measures can be taken when cracks have been detected in the air-conditioning pipeline, thereby extending the service life of the air-conditioning pipeline.

[0051] In some embodiments, the specific process of extracting characteristic parameters from the acoustic emission signal of the pipeline to be tested in step S120 as the acoustic emission characteristic parameters of the pipeline to be tested is described in the following exemplary embodiments.

[0052] Combine the following Figure 3 The flowchart of an embodiment of extracting characteristic parameters from the acoustic emission signal of the pipeline to be tested in the method of the present invention further illustrates the specific process of extracting characteristic parameters from the acoustic emission signal of the pipeline to be tested in step S120, including: steps S310 to S320.

[0053] Step S310, converting the acoustic emission signal of the pipeline to be tested into an electrical signal to obtain the acoustic emission electrical signal of the pipeline to be tested.

[0054] Step S320, using the wavelet packet analysis method, extracting features from the acoustic emission electrical signal of the pipeline to be tested, obtaining feature parameters, and recording them as acoustic emission feature parameters of the pipeline to be tested, such as feature values ​​of the acoustic emission signal of the pipeline to be tested. Specifically, using the wavelet packet analysis method, extracting feature parameters that meet preset parameter requirements from the acoustic emission electrical signal of the pipeline to be tested, and recording them as acoustic emission feature parameters of the pipeline to be tested, such as feature values ​​of the acoustic emission signal of the pipeline to be tested.

[0055] In some embodiments, when the training process of the pipeline identification model pre-trained in step S130 includes extracting characteristic parameters from the acoustic emission signal of the experimental pipeline, the specific process of extracting characteristic parameters from the acoustic emission signal of the experimental pipeline in step S220 as the acoustic emission characteristic parameters of the experimental pipeline is described in the following exemplary embodiment.

[0056] Combine the following Figure 4 The flowchart of an embodiment of extracting characteristic parameters from the acoustic emission signal of the experimental pipeline in the method of the present invention further illustrates the specific process of extracting characteristic parameters from the acoustic emission signal of the experimental pipeline in step S220, including: steps S410 to S420.

[0057] Step S410, converting the acoustic emission signal of the experimental pipeline into an electrical signal to obtain the acoustic emission electrical signal of the experimental pipeline.

[0058] Step S420, using the wavelet packet analysis method, extracting features from the acoustic emission electrical signal of the experimental pipeline to obtain characteristic parameters, which are recorded as the acoustic emission characteristic parameters of the experimental pipeline, such as the characteristic values ​​of the acoustic emission signal of the experimental pipeline. Specifically, using the wavelet packet analysis method, characteristic parameters that meet the preset parameter requirements are extracted from the acoustic emission electrical signal of the experimental pipeline, and are used as the acoustic emission characteristic parameters of the experimental pipeline, such as the characteristic values ​​of the acoustic emission signal of the experimental pipeline.

[0059] In the scheme of the present invention, the wavelet packet analysis method is adopted to extract characteristic values ​​from the acoustic emission signals of the corresponding pipelines (such as the pipeline to be tested and the experimental pipeline) in the time domain and the frequency domain, as the acoustic emission characteristic parameters of the corresponding pipelines, and then the BP neural network is trained based on the acoustic emission characteristic parameters of the experimental pipeline to obtain a pipeline recognition model, and the pipeline recognition model is used to identify whether the pipeline to be tested is cracked based on the acoustic emission characteristic parameters of the pipeline to be tested, so as to take timely measures when cracks are detected in the air-conditioning pipeline, thereby extending the service life of the air-conditioning pipeline.

[0060] In some embodiments, the specific process of the experimental process for detecting cracks in the experimental pipeline in step S210 is described in the following exemplary embodiments.

[0061] Combine the following Figure 5The flowchart of an embodiment of the experimental process for detecting cracks in the experimental pipeline in the method of the present invention further illustrates the specific process of the experimental process for detecting cracks in the experimental pipeline in step S210, including: steps S510 to S530.

[0062] Step S510, for the experimental pipeline in the pipeline of the device, a finite element model is built. The finite element model is a model established when using the finite element analysis method, which is a group of unit combinations that are only connected at nodes, only transmit forces at nodes, and only constrained at nodes.

[0063] Step S520, using the finite element model to perform simulation analysis on the experimental pipeline, determine the location where a preset degree of fracture or slip occurs in the experimental pipeline, and record it as a dangerous location of the experimental pipeline. For example: determine the point where a preset degree of fracture or slip occurs in the experimental pipeline, and record it as a dangerous point of the experimental pipeline.

[0064] Step S530, collecting acoustic emission signals of the test pipeline from the dangerous position of the test pipeline and within a set area of ​​the dangerous position of the test pipeline, and collecting cracking parameters of the test pipeline.

[0065] Figure 7 The figure is a flow chart of crack identification of air conditioning pipelines based on acoustic emission technology and neural network technology. Figure 7 As shown in the figure, the process of identifying cracks in air-conditioning pipes based on acoustic emission technology and neural network technology includes:

[0066] Step 1: Build a finite element model of the air conditioning pipeline, and use the finite element model to simulate and analyze the air conditioning pipeline to obtain the dangerous points of the air conditioning pipeline, and then perform step 2. The dangerous points of the air conditioning pipeline are points where tiny fractures or slippage occur in the air conditioning pipeline.

[0067] In step 1, a finite element model of the air-conditioning pipeline is constructed and the finite element model of the air-conditioning pipeline is used to simulate, analyze and calculate the air-conditioning pipeline to obtain the dangerous points of the air-conditioning pipeline. Specifically, a real three-dimensional model of the pipeline is constructed in a drawing software (such as Creo) at a ratio of 1:1, and the three-dimensional model of the pipeline is imported into the finite element simulation software. The pipeline is constrained according to the installation condition of the pipeline in the air conditioner, and according to the power spectral density (PSD) spectrum (that is, the vibration excitation or acceleration excitation received by the pipeline during actual working process), the dangerous points of pipeline stress and displacement are finally obtained.

[0068] Figure 8 is the stress cloud diagram of the air conditioning pipeline. Fig. 9 is the displacement cloud diagram of the air conditioning pipeline. Figure 8 and Fig. 9The bottom of the figure is the scaled size of the figure. In step 1, the air conditioning pipeline is simulated and analyzed using the air conditioning pipeline finite element model, and the stress cloud diagram of the air conditioning pipeline can be obtained (such as Figure 8 ) and displacement cloud diagram of air conditioning pipeline (as shown in Fig. 9 As shown in Figure 1, the dangerous points of the air conditioning pipeline can be obtained by calculating the stress cloud map and displacement cloud map of the air conditioning pipeline. Figure 8 In the stress cloud diagram of the air-conditioning pipeline shown, from the red area (such as the Max area) to the blue area (such as the Min area), the air-conditioning pipeline in the blue area (such as the Min area) is subjected to less stress, and the air-conditioning pipeline in the red area (such as the Max area) is subjected to greater stress. It can be considered that the point on the air-conditioning pipeline corresponding to the red area is the dangerous point of the air-conditioning pipeline.

[0069] exist Fig. 9 In the displacement cloud map of the air-conditioning pipeline shown, from the red area (such as the Max area) to the blue area (such as the Min area), the displacement deformation of the air-conditioning pipeline in the blue area (such as the Min area) is smaller, and the displacement deformation of the air-conditioning pipeline in the red area (such as the Max area) is larger. It can be considered that the point on the air-conditioning pipeline corresponding to the red area is the dangerous point of the air-conditioning pipeline.

[0070] Step 2: Select the dangerous points of the air-conditioning pipeline obtained in step 1 and the areas near the dangerous points as the layout points of the acoustic emission sensor signals; then use the acoustic emission signal acquisition system (i.e., the acquisition system composed of acoustic emission sensors) to collect the acoustic emission signals of the air-conditioning pipeline in the fatigue test, and then execute step 3.

[0071] In the scheme of the present invention, fatigue cracks in air-conditioning pipelines are identified and detected by using neural network technology through acoustic emission. Specifically, acoustic emission signals of air-conditioning pipelines are collected through acoustic emission technology, characteristic values ​​of the signals are selected, and characteristic value parameters and energy frequency bands in the acoustic emission signals are input into a BP neural network (Backpropagation Neural Network) for neural network training, so as to use the pipeline identification model obtained by training to identify whether the pipeline to be tested is cracked, thereby realizing the recognition and detection of cracks in the air-conditioning pipelines by the neural network, and taking timely measures when cracks are detected in the air-conditioning pipelines to extend the service life of the air-conditioning pipelines.

[0072] In some embodiments, in step S110, for the pipeline to be tested in the pipeline of the device, obtaining the acoustic emission signal of the pipeline to be tested includes: obtaining the acoustic emission signal of the pipeline to be tested collected by a pre-set first acoustic emission signal collection system; wherein the first acoustic emission signal collection system is pre-set at a preset cracking point on the pipeline to be tested. In other words, the first acoustic emission signal collection system is pre-arranged at the preset cracking point of the pipeline to be tested (such as a point where excessive stress may occur), and the acoustic emission signal of the pipeline to be tested collected by the first acoustic emission signal collection system is obtained.

[0073] In some embodiments, collecting the acoustic emission signal of the experimental pipeline from the dangerous position of the experimental pipeline and the set area of ​​the dangerous position of the experimental pipeline in step S530 includes: collecting the acoustic emission signal of the experimental pipeline collected by a pre-set second acoustic emission signal collection system; wherein the second acoustic emission signal collection system is pre-set at the dangerous position of the experimental pipeline and the set area of ​​the dangerous position of the experimental pipeline. That is, the acoustic emission signal collection system is arranged at the dangerous position of the experimental pipeline and the set area of ​​the dangerous position of the experimental pipeline to collect the acoustic emission signal of the experimental pipeline collected by the acoustic emission signal collection system; and based on the dangerous position of the experimental pipeline, the cracking parameters of the experimental pipeline are collected.

[0074] In the scheme of the present invention, it is considered that when cracks or other damages occur in the air-conditioning pipeline, the stress state inside the material changes, resulting in tiny fractures or slips in local areas. These tiny fractures or slips will release the characteristics of the sound wave signal; therefore, the sound wave signal is converted into an electrical signal by acoustic emission technology, and the electrical signal is collected and processed to obtain the characteristic value of the classification signal. The parameter list analysis method is used to determine whether the air-conditioning pipeline has cracks, so that timely measures can be taken when cracks are detected in the air-conditioning pipeline to extend the service life of the air-conditioning pipeline.

[0075] In some embodiments, the training process of the pipeline identification model pre-trained in step S130 includes extracting characteristic parameters from the acoustic emission signal of the experimental pipeline. The acoustic emission characteristic parameters of the pipeline to be tested in step S120 and the acoustic emission characteristic parameters of the experimental pipeline in step S220 are the acoustic emission characteristic parameters of the corresponding pipeline, that is, the characteristic parameters obtained by extracting the acoustic emission electrical signals of the corresponding pipeline, including: time domain parameters of the acoustic emission electrical signals of the corresponding pipeline, frequency domain parameters of the acoustic emission electrical signals of the corresponding pipeline, and energy parameters of the acoustic emission electrical signals of the corresponding pipeline.

[0076] Among them, the time domain parameters of the acoustic emission electrical signal of the corresponding pipeline include at least one of the following: the voltage amplitude of the acoustic emission electrical signal of the corresponding pipeline, the voltage rise time of the acoustic emission electrical signal of the corresponding pipeline, the voltage duration of the acoustic emission electrical signal of the corresponding pipeline, and the count value of ringing within the voltage duration of the acoustic emission electrical signal of the corresponding pipeline.

[0077] The energy parameters of the acoustic emission electrical signal of the corresponding pipeline include at least one of the following: the total energy of the acoustic emission electrical signal of the corresponding pipeline, the frequency band energy of the acoustic emission electrical signal of the corresponding pipeline in a preset frequency band, and the proportion of the frequency band energy of the acoustic emission electrical signal of the corresponding pipeline in the preset frequency band in the total energy.

[0078] The frequency domain parameters of the acoustic emission electrical signal of the corresponding pipeline include at least one of the following: the frequency of the acoustic emission electrical signal of the corresponding pipeline, and the pulse factor of the acoustic emission electrical signal of the corresponding pipeline; wherein the frequency of the acoustic emission electrical signal of the corresponding pipeline includes at least one of the following: the initial frequency of the acoustic emission electrical signal of the corresponding pipeline, the peak frequency of the acoustic emission electrical signal of the corresponding pipeline, the average frequency of the acoustic emission electrical signal of the corresponding pipeline, and the inverse calculated frequency of the acoustic emission electrical signal of the corresponding pipeline.

[0079] Specifically, the characteristic parameters obtained by feature extraction of the acoustic emission electrical signal of the pipeline to be tested include: time domain parameters of the acoustic emission electrical signal of the pipeline to be tested, and frequency domain parameters of the acoustic emission electrical signal of the pipeline to be tested; wherein the time domain parameters of the acoustic emission electrical signal of the pipeline to be tested include at least one of the following: voltage amplitude of the acoustic emission electrical signal of the pipeline to be tested, voltage rise time of the acoustic emission electrical signal of the pipeline to be tested, voltage duration of the acoustic emission electrical signal of the pipeline to be tested, and count value of ringing within the voltage duration of the acoustic emission electrical signal of the pipeline to be tested; the frequency domain parameters of the acoustic emission electrical signal of the pipeline to be tested include the following At least one of: the total energy of the acoustic emission signal of the pipeline to be tested, the frequency of the acoustic emission signal of the pipeline to be tested, the pulse factor of the acoustic emission signal of the pipeline to be tested, the frequency band energy of the acoustic emission signal of the pipeline to be tested in a preset frequency band, and the proportion of the frequency band energy of the acoustic emission signal of the corresponding pipeline in the preset frequency band in the total energy; wherein, the frequency of the acoustic emission signal of the pipeline to be tested includes at least one of the following: the initial frequency of the acoustic emission signal of the pipeline to be tested, the peak frequency of the acoustic emission signal of the pipeline to be tested, the average frequency of the acoustic emission signal of the pipeline to be tested, and the inverse calculated frequency of the acoustic emission signal of the pipeline to be tested.

[0080] The characteristic parameters obtained by feature extraction of the acoustic emission electrical signal of the experimental pipeline include: time domain parameters of the acoustic emission electrical signal of the experimental pipeline, and frequency domain parameters of the acoustic emission electrical signal of the experimental pipeline; wherein the time domain parameters of the acoustic emission electrical signal of the experimental pipeline include at least one of the following: voltage amplitude of the acoustic emission electrical signal of the experimental pipeline, voltage rise time of the acoustic emission electrical signal of the experimental pipeline, voltage duration of the acoustic emission electrical signal of the experimental pipeline, and count value of ringing within the voltage duration of the acoustic emission electrical signal of the experimental pipeline; the frequency domain parameters of the acoustic emission electrical signal of the experimental pipeline include at least the following: One of: the total energy of the acoustic emission signal of the experimental pipeline, the frequency of the acoustic emission signal of the experimental pipeline, the pulse factor of the acoustic emission signal of the experimental pipeline, the frequency band energy of the acoustic emission signal of the experimental pipeline in a preset frequency band, and the proportion of the frequency band energy of the acoustic emission signal of the corresponding pipeline in the preset frequency band in the total energy; wherein, the frequency of the acoustic emission signal of the experimental pipeline includes at least one of the following: the initial frequency of the acoustic emission signal of the experimental pipeline, the peak frequency of the acoustic emission signal of the experimental pipeline, the average frequency of the acoustic emission signal of the experimental pipeline, and the inverse calculated frequency of the acoustic emission signal of the experimental pipeline.

[0081] like Figure 7 As shown in the figure, the process of identifying cracks in air-conditioning pipes based on acoustic emission technology and neural network technology also includes:

[0082] Step 3: Based on the acoustic emission signal of the air-conditioning pipeline in the fatigue test, the acoustic emission characteristic value is selected: the characteristic value of the acoustic emission signal is selected through two different dimensions of time-energy and frequency-energy to obtain the characteristic value of the acoustic emission signal of the air-conditioning pipeline, and then step 4 is executed.

[0083] In step 3, the characteristic values ​​of the acoustic emission signal are selected through two different dimensions of time-energy and frequency-energy. Specifically, the wavelet packet transform method can be used to characterize the acoustic emission signal to obtain the characteristic values ​​of the acoustic emission signal. Among them, the wavelet packet transform (WPT) is a signal processing technology that inherits the time-frequency analysis characteristics of the wavelet transform and further decomposes the undecomposed high-frequency band signal to improve the frequency resolution.

[0084] Among them, the signal collected by acoustic emission is a set of voltage amplitudes with equal time intervals. If it is analyzed directly, it is very complicated, and the acoustic emission signal needs to be described by various characteristic values. The parameters describing the characteristics of the acoustic emission waveform include amplitude, rise time, duration, ring count, etc.; the parameters describing the frequency characteristics include initial frequency, peak frequency, back-calculated frequency, average frequency, etc.; the parameters describing the energy characteristics include energy, signal strength, ASL, etc.; these parameter characteristics can be used to describe and analyze the acoustic emission signal, and the acoustic emission signal obtained by using time-energy and frequency-energy to describe it at the same time is more complete and accurate.

[0085] Specifically, the process of characterizing the acoustic emission signal using the wavelet packet transform method may include: first determining the spectral characteristics of the acoustic emission signal, selecting an appropriate decomposition level n (n is a positive integer), decomposing the original signal into 2 n The bandwidth of each node is 1 / 2 of the bandwidth of the original signal. n Finally, the wavelet coefficient of each node is inversely transformed, and the signal component corresponding to the node is calculated to obtain the signal energy of the node. The energy of different frequency bands is selected from the calculated signal energy as the characteristic value of the acoustic emission signal.

[0086] For example: select the acoustic emission signal of one of the air conditioning pipes for wavelet packet transform, select the db4 series wavelet function, perform 3-layer wavelet packet decomposition, and obtain the energy values ​​of 8 frequency bands of the signal, so as to select the energy of different frequency bands in the acoustic emission signal. Among them, the db4 series wavelet function is a discrete wavelet transform function. Among them, in the process of selecting the characteristic value of the acoustic emission signal, the energy frequency band needs to be extracted by wavelet packet transform, and the characteristic value of the energy of each frequency band can be obtained based on the acoustic emission signal.

[0087] When wavelet packet transform is used to characterize the acoustic emission signal, the energy of the signal and the energy of the signal components in each frequency band can be calculated by equations (1) and (2):

[0088]

[0089]

[0090] Among them, formula (1) can calculate the energy in the signal, where: E i represents the energy of frequency band i; represents the signal component of frequency band i, and A is the conversion coefficient, which is related to the model of the acoustic emission sensor. Formula (2) can calculate the energy of the signal component in each frequency band. The calculated frequency band energy distribution is as follows: Fig.10 and Fig.11 shown. Fig.10A schematic diagram of the frequency band energy curve for each frequency band number. Fig.11 A schematic diagram showing the ratio of the energy of each frequency band number to the total energy. Fig.10 and Fig.11 It can be seen that the frequency band energy is the highest in the frequency band numbered 2. Both formula (1) and formula (2) calculate the energy of the acoustic emission signal in each frequency band. The difference between the two lies in A (conversion coefficient). The coefficient of A is related to the sensor type. The acoustic emission sensors used in the test are the same, so A can be omitted, that is, formula (2) is used.

[0091] Step 4: Take the characteristic value of the acoustic emission signal of the air-conditioning pipeline obtained in step 3 as a sample, use the BP neural network to train the neural network, and obtain the air-conditioning pipeline recognition model; then, for the characteristic value of the acoustic emission signal of the air-conditioning pipeline obtained on site, use the air-conditioning pipeline recognition model to identify whether the air-conditioning pipeline on site is cracked.

[0092] The waveform of the acoustic emission signal of the air conditioning pipeline contains four characteristic values: amplitude, rise time, duration and ring count, which mainly have a good detection effect in the early stage of fatigue crack generation. The four characteristic values ​​contained in the frequency are the frequencies of the acoustic emission signal at different stages: the initial frequency is the average frequency before the acoustic emission signal reaches the peak, the peak frequency is the frequency corresponding to the maximum peak in the power spectrum, the inverse frequency is the average frequency after the peak of the acoustic emission signal, and the average frequency is the average frequency of the entire acoustic emission signal, which represent the frequencies of the acoustic emission signal at different stages. Among them, the signal obtained by acoustic emission acquisition is relatively complex and cumbersome, and it is necessary to use various characteristic values ​​to describe the acoustic emission signal. The acoustic emission signal can be described from the aspects of time, frequency, energy, etc. In terms of time and frequency, you can directly observe and record the time domain diagram (time) and frequency domain diagram (frequency) on the system according to your needs. The energy of each frequency band needs to be calculated by wavelet packet transform. Finally, the amplitude, rise time, duration, ring count, initial frequency, peak frequency, inverse frequency, average frequency, signal strength, pulse factor and the energy proportion of the first four frequency bands are selected as the characteristic value input in the neural network training. Among them, the characteristic values ​​of the acoustic emission signal (such as amplitude, rise time, duration, ring count), and the frequency band energy (i.e. the energy of the first four frequency bands) are as follows: Fig.12 shown. Fig.12The following is a curve diagram of acoustic emission parameters. Amplitude: refers to the size of the maximum peak value in the signal. Rise time: refers to the time required for the acoustic emission signal to exceed the threshold to reach the maximum peak. Duration: refers to the duration of the entire acoustic emission signal above the threshold value. Ring count: refers to the number of oscillations after exceeding the threshold value. The characteristic values ​​of the acoustic emission signal (such as amplitude, rise time, duration, and ring count) can fully describe the collective morphological characteristics of a waveform in the acoustic emission signal. The threshold value is a set critical value, and the acoustic emission signal will only be recorded if it exceeds the threshold value.

[0093] Judging whether fatigue cracks exist based on the acoustic emission signal of the air-conditioning pipeline: When cracks occur at a certain position of the air-conditioning pipeline, various acoustic emission parameters also undergo drastic changes; for example, at the position of the air-conditioning pipeline where cracks occur, the spectrum signal of the acoustic emission will increase, the energy will increase, the pulse factor will decrease, etc. By inputting the characteristic value and energy frequency band of the acoustic emission signal into the BP neural network model, the acoustic emission signal can be automatically identified after training. When the acoustic emission signal shows the above characteristics during the detection process, it indicates that fatigue cracks have occurred in the air-conditioning pipeline. Among them, the energy frequency band is obtained by performing wavelet packet transform on the acoustic emission signal and then calculating it with formula (2). The energy frequency band is part of the characteristic value; the energy frequency band is obtained by extracting the energy of each frequency band in the acoustic emission signal through the wavelet packet transform method after obtaining the acoustic emission signal. The input of the BP neural network model is the various characteristic value parameters, and the output of the BP neural network model is the occurrence of cracks in the air-conditioning pipeline. Fig.13 Shown is the distribution diagram of the pulse factor in the acoustic emission signal at different stages. Fig.13 The figure is a schematic diagram of the distribution of pulse damage factors at different stages (such as the initiation stage, the expansion stage, and the crack stage). The initiation stage refers to the initial formation process of the crack when a crack size appears on the surface of the structure, but the size is very small. Tiny cracks appear in the stress concentration area, which is the initial formation process of the crack. The expansion stage refers to the process in which the crack expands at a slower speed, the expansion rate gradually increases, and the stress concentration at the crack tip leads to local plastic deformation. The crack stage refers to the process in which the crack rapidly expands after reaching the critical size until the material breaks, and the crack expansion rate increases sharply, reaching the macro crack expansion stage.

[0094] In step 4, the cracks in the air-conditioning pipeline are detected by using a neural network: the BP neural network model is selected to establish the relationship between the air-conditioning pipeline and the acoustic emission signal, so as to realize the real-time recognition of the cracking phenomenon of the air-conditioning pipeline. i As a sample, the sample contains training feature values. The sample contains training feature values. The feature values ​​include amplitude, initial frequency, energy of the first four frequency bands, etc. A large number of training samples are input into the BP neural network to establish a nonlinear functional relationship f between the feature value and whether the air conditioning pipeline has cracked.j The feature vector for the test is input into the nonlinear function f to identify whether cracks occur in the air conditioning pipe. The input amplitude, rise time, duration, ring count, initial frequency, peak frequency, back-calculated frequency, average frequency, signal strength, ASL and energy of the first four frequency bands in the acoustic emission data are used as the input data of the initial nodes of the 14 input layers of the neural network. The output node [1, 0] indicates the presence of cracks. The training samples are randomly selected, and the training samples account for 80% of the total samples. During the training process, the Levenberg-Marquardt optimization algorithm (LM algorithm) is used, with a maximum number of iterations of 2000, a training rate of 0.01, and a minimum training error of 10 -6 ; Among them, the Levenberg-Marquardt algorithm is a numerical optimization method widely used in nonlinear least squares problems. The remaining 20% ​​of the acoustic emission data is a test sample set. Among them, rise time: refers to the time required for the acoustic emission signal to exceed the threshold value to reach the maximum peak. Duration: refers to the duration of the entire acoustic emission signal at the threshold value. Ring count: refers to the number of oscillations after exceeding the threshold value. Initial frequency: refers to the average frequency before the acoustic emission signal reaches the peak value. Peak frequency: refers to the frequency corresponding to the maximum peak in the power spectrum. Inverse frequency: refers to the average frequency after the peak value of the acoustic emission signal. Average frequency: refers to the average frequency of the entire acoustic emission signal. ASL (Average signal level): refers to the physical quantity that uses a decibel meter to describe the average value of the signal size.

[0095] In the scheme of the present invention, the acoustic emission signal in the fatigue crack of the air-conditioning pipeline is characterized and processed and then introduced into the BP neural network for neural network training and testing, and finally the cracks in the air-conditioning pipeline are identified and detected online through the neural network. The neural network is trained and tested on the air-conditioning pipeline, and the results are shown in Table 1. It can be found from Table 1 that the error ratio is 3.56%, and the neural network has a high accuracy in identifying cracks in the air-conditioning pipeline.

[0096] Table 1

[0097] Working conditions Amount of training data Training error ratio Test data volume Test error ratio Air conditioning pipeline 100000 3.56％ 30000 3.96％

[0098] In the scheme of the present invention, neural network technology is used to identify and detect fatigue cracks in air-conditioning pipelines through acoustic emission, fatigue cracks in air-conditioning pipelines are identified, and cracks are detected in real time so that timely measures can be taken to extend the service life of the air-conditioning pipelines.

[0099] By adopting the technical solution of this embodiment, an air-conditioning pipeline model (such as an air-conditioning pipeline finite element model) is built for the air-conditioning pipeline for simulation analysis, an area where cracks (small fractures or slips) appear in the air-conditioning pipeline is selected, and an acoustic emission signal acquisition system (such as an acoustic emission sensor) is arranged in the selected area. The acoustic emission signal of the air-conditioning pipeline is acquired by the acoustic emission signal acquisition system; the characteristic parameters (such as eigenvalues) of the acoustic emission signal are extracted by wavelet packet analysis, and the characteristic parameters of the acoustic emission signal and the air-conditioning pipeline parameters (such as parameters indicating whether the air-conditioning pipeline is cracked, such as the corresponding characteristic parameters) are used as samples. Based on the sample (with the characteristic parameters of the acoustic emission signal as input and the air-conditioning pipeline parameters as output), the BP neural network is used to perform neural network training to obtain an air-conditioning pipeline recognition model for identifying whether the air-conditioning pipeline is cracked; then, the air-conditioning pipeline recognition model is used to identify whether the air-conditioning pipeline is cracked based on the acoustic emission signal of the air-conditioning pipeline; thus, by adopting the acoustic emission technology and the neural network technology, whether the air-conditioning pipeline is cracked is identified based on the acoustic emission signal of the air-conditioning pipeline, and a reminder message is initiated when the air-conditioning pipeline is identified to take measures to extend the service life of the air-conditioning pipeline and ensure the normal use of the air conditioner.

[0100] According to an embodiment of the present invention, a pipeline identification device corresponding to the pipeline identification method is also provided. Figure 6 The schematic diagram of the structure of an embodiment of the device of the present invention is shown in FIG. The pipeline identification device is used to identify whether the pipeline to be tested of the equipment (such as air conditioner) is cracked. In the scheme of the present invention, Figure 6 As shown, the pipeline identification device includes: an acquisition unit 102 and a control unit 104.

[0101] The acquisition unit 102 is configured to acquire the acoustic emission signal of the pipeline to be tested in the pipelines of the device. The specific functions and processing of the acquisition unit 102 refer to step S110.

[0102] The control unit 104 is configured to extract characteristic parameters from the acoustic emission signal of the pipeline to be tested as the acoustic emission characteristic parameters of the pipeline to be tested. The specific functions and processing of the control unit 104 refer to step S120.

[0103] The control unit 104 is also configured to identify whether the pipeline to be tested is cracked based on the acoustic emission characteristic parameters of the pipeline to be tested and the pipeline identification model obtained by pre-training, so as to initiate a reminder message of the cracking of the pipeline to be tested when the pipeline to be tested is identified as cracked, so as to remind relevant personnel to take measures on the pipeline to be tested in time to avoid the cracking of the pipeline to be tested, which leads to the rupture or damage of the pipeline to be tested and cannot be used normally, which is conducive to extending the service life of the pipeline to be tested and ensuring the normal use of the equipment; wherein the pipeline identification model obtained by pre-training is a pipeline identification model to be tested obtained by pre-training based on acoustic emission technology, wavelet packet analysis method and neural network technology. The specific functions and processing of the control unit 104 are also shown in step S130.

[0104] In order to solve the problem of fatigue cracks in air-conditioning pipelines caused by excessive stress, the solution of the present invention applies acoustic emission technology and neural network technology to the preventive maintenance of air-conditioning pipelines. By early detection of fatigue damage of air-conditioning pipelines and real-time detection to identify whether cracks have occurred in the air-conditioning pipelines, cracks in the air-conditioning pipelines can be identified and detected, and measures can be taken in time to extend the service life of the air-conditioning pipelines.

[0105] In some embodiments, the control unit 104 performs a training process of the pipeline recognition model obtained by pre-training, including:

[0106] The control unit 104 is also specifically configured to collect the acoustic emission signal of the experimental pipeline and the cracking parameters of the experimental pipeline during the experimental process of detecting the cracking of the experimental pipeline. The cracking parameters of the experimental pipeline are used to characterize whether the experimental pipeline is actually cracked. The specific functions and processing of the control unit 104 are also shown in step S210.

[0107] The control unit 104 is further configured to extract characteristic parameters from the acoustic emission signal of the experimental pipeline as the acoustic emission characteristic parameters of the experimental pipeline. The specific functions and processing of the control unit 104 are also shown in step S220.

[0108] The control unit 104 is further configured to use the acoustic emission characteristic parameters of the experimental pipeline and the cracking parameters of the experimental pipeline as training samples and a preset neural network model as a basic model. The specific functions and processing of the control unit 104 are also shown in step S230.

[0109] The control unit 104 is further configured to use the acoustic emission characteristic parameters of the experimental pipeline as the input parameters of the basic model and the cracking parameters of the experimental pipeline as the output parameters of the basic model to perform neural network training to obtain the required pipeline identification model. The specific functions and processing of the control unit 104 are also shown in step S240.

[0110] In the scheme of the present invention, an artificial neural network is adopted to establish the relationship between the cracks in the air-conditioning pipeline and the acoustic emission signal by utilizing the characteristic value parameters in the acoustic emission signal, thereby realizing real-time detection of whether cracks have occurred in the air-conditioning pipeline, so that timely measures can be taken when cracks have been detected in the air-conditioning pipeline, thereby extending the service life of the air-conditioning pipeline.

[0111] In some embodiments, the control unit 104 extracts characteristic parameters from the acoustic emission signal of the pipeline to be tested as the acoustic emission characteristic parameters of the pipeline to be tested, including:

[0112] The control unit 104 is also specifically configured to convert the acoustic emission signal of the pipeline to be tested into an electrical signal to obtain the acoustic emission electrical signal of the pipeline to be tested. The specific functions and processing of the control unit 104 are also shown in step S310.

[0113] The control unit 104 is also specifically configured to use the wavelet packet analysis method to extract features from the acoustic emission electrical signal of the pipeline to be tested, obtain feature parameters, and record them as the acoustic emission feature parameters of the pipeline to be tested, such as the feature values ​​of the acoustic emission signal of the pipeline to be tested. Specifically, the wavelet packet analysis method is used to extract feature parameters that meet the preset parameter requirements from the acoustic emission electrical signal of the pipeline to be tested, as the acoustic emission feature parameters of the pipeline to be tested, such as the feature values ​​of the acoustic emission signal of the pipeline to be tested. The specific functions and processing of the control unit 104 are also shown in step S320.

[0114] In some embodiments, the control unit 104, when the training process of the pre-trained pipeline recognition model includes extracting characteristic parameters from the acoustic emission signal of the experimental pipeline, extracts characteristic parameters from the acoustic emission signal of the experimental pipeline as the acoustic emission characteristic parameters of the experimental pipeline, including:

[0115] The control unit 104 is also specifically configured to convert the acoustic emission signal of the experimental pipeline into an electrical signal to obtain the acoustic emission electrical signal of the experimental pipeline. The specific functions and processing of the control unit 104 are also shown in step S410.

[0116] The control unit 104 is also specifically configured to use the wavelet packet analysis method to extract features from the acoustic emission electrical signal of the experimental pipeline, obtain characteristic parameters, and record them as the acoustic emission characteristic parameters of the experimental pipeline, such as the characteristic values ​​of the acoustic emission signal of the experimental pipeline. Specifically, the wavelet packet analysis method is used to extract characteristic parameters that meet the preset parameter requirements from the acoustic emission electrical signal of the experimental pipeline, as the acoustic emission characteristic parameters of the experimental pipeline, such as the characteristic values ​​of the acoustic emission signal of the experimental pipeline. The specific functions and processing of the control unit 104 are also shown in step S420.

[0117] In the scheme of the present invention, the wavelet packet analysis method is adopted to extract characteristic values ​​from the acoustic emission signals of the corresponding pipelines (such as the pipeline to be tested and the experimental pipeline) in the time domain and the frequency domain, as the acoustic emission characteristic parameters of the corresponding pipelines, and then the BP neural network is trained based on the acoustic emission characteristic parameters of the experimental pipeline to obtain a pipeline recognition model, and the pipeline recognition model is used to identify whether the pipeline to be tested is cracked based on the acoustic emission characteristic parameters of the pipeline to be tested, so as to take timely measures when cracks are detected in the air-conditioning pipeline, thereby extending the service life of the air-conditioning pipeline.

[0118] In some embodiments, the control unit 104 performs an experimental process for detecting cracking of the experimental pipeline, including:

[0119] The control unit 104 is further configured to build a finite element model for the experimental pipeline in the pipeline of the device. The specific functions and processing of the control unit 104 are also shown in step S510.

[0120] The control unit 104 is further configured to use the finite element model to perform simulation analysis on the experimental pipeline, determine the location of the experimental pipeline where a preset degree of fracture or slip occurs, and record it as a dangerous location of the experimental pipeline. The specific functions and processing of the control unit 104 are also shown in step S520.

[0121] The control unit 104 is further configured to collect acoustic emission signals of the test pipeline from the dangerous position of the test pipeline and the set area of ​​the dangerous position of the test pipeline, and collect cracking parameters of the test pipeline. The specific functions and processing of the control unit 104 are also shown in step S530.

[0122] Figure 7 The figure is a flow chart of crack identification of air conditioning pipelines based on acoustic emission technology and neural network technology. Figure 7 As shown in the figure, the process of identifying cracks in air-conditioning pipes based on acoustic emission technology and neural network technology includes:

[0123] Step 1: Build a finite element model of the air conditioning pipeline, and use the finite element model to simulate and analyze the air conditioning pipeline to obtain the dangerous points of the air conditioning pipeline, and then perform step 2. The dangerous points of the air conditioning pipeline are points where tiny fractures or slippage occur in the air conditioning pipeline.

[0124] Figure 8 is the stress cloud diagram of the air conditioning pipeline. Fig. 9 is the displacement cloud diagram of the air conditioning pipeline. Figure 8 and Fig. 9 The bottom of the figure is the scaled size of the figure. In step 1, the air conditioning pipeline is simulated and analyzed using the air conditioning pipeline finite element model, and the stress cloud diagram of the air conditioning pipeline can be obtained (such as Figure 8) and displacement cloud diagram of air conditioning pipeline (as shown in Fig. 9 As shown in Figure 1, the dangerous points of the air conditioning pipeline can be obtained by calculating the stress cloud map and displacement cloud map of the air conditioning pipeline. Figure 8 In the stress cloud diagram of the air-conditioning pipeline shown, from the red area (such as the Max area) to the blue area (such as the Min area), the air-conditioning pipeline in the blue area (such as the Min area) is subjected to less stress, and the air-conditioning pipeline in the red area (such as the Max area) is subjected to greater stress. It can be considered that the point on the air-conditioning pipeline corresponding to the red area is the dangerous point of the air-conditioning pipeline.

[0125] exist Fig. 9 In the displacement cloud map of the air-conditioning pipeline shown, from the red area (such as the Max area) to the blue area (such as the Min area), the displacement deformation of the air-conditioning pipeline in the blue area (such as the Min area) is smaller, and the displacement deformation of the air-conditioning pipeline in the red area (such as the Max area) is larger. It can be considered that the point on the air-conditioning pipeline corresponding to the red area is the dangerous point of the air-conditioning pipeline.

[0126] Step 2: Select the dangerous points of the air-conditioning pipeline obtained in step 1 and the areas near the dangerous points as the layout points of the acoustic emission sensor signals; then use the acoustic emission signal acquisition system (i.e., the acquisition system composed of acoustic emission sensors) to collect the acoustic emission signals of the air-conditioning pipeline in the fatigue test, and then execute step 3.

[0127] In the scheme of the present invention, fatigue cracks in air-conditioning pipelines are identified and detected by using neural network technology through acoustic emission. Specifically, acoustic emission signals of air-conditioning pipelines are collected through acoustic emission technology, characteristic values ​​of the signals are selected, and characteristic value parameters and energy frequency bands in the acoustic emission signals are input into a BP neural network (Backpropagation Neural Network) for neural network training, so as to use the pipeline identification model obtained by training to identify whether the pipeline to be tested is cracked, thereby realizing the recognition and detection of cracks in the air-conditioning pipelines by the neural network, and taking timely measures when cracks are detected in the air-conditioning pipelines to extend the service life of the air-conditioning pipelines.

[0128] In some embodiments, the acquisition unit 102 acquires the acoustic emission signal of the pipeline to be tested in the pipeline of the device, including: the acquisition unit 102 is further configured to acquire the acoustic emission signal of the pipeline to be tested acquired by a pre-set first acoustic emission signal acquisition system; wherein the first acoustic emission signal acquisition system is pre-set at a preset cracking point on the pipeline to be tested. That is, the first acoustic emission signal acquisition system is pre-arranged at the preset cracking point of the pipeline to be tested (such as a point where excessive stress may occur), and the acoustic emission signal of the pipeline to be tested acquired by the first acoustic emission signal acquisition system is acquired.

[0129] In some embodiments, the control unit 104 collects the acoustic emission signals of the experimental pipeline from the dangerous position of the experimental pipeline and the set area of ​​the dangerous position of the experimental pipeline, including: the control unit 104 is specifically configured to collect the acoustic emission signals of the experimental pipeline collected by the pre-set second acoustic emission signal collection system; wherein the second acoustic emission signal collection system is pre-set at the dangerous position of the experimental pipeline and the set area of ​​the dangerous position of the experimental pipeline. That is to say, the acoustic emission signal collection system is arranged at the dangerous position of the experimental pipeline and the set area of ​​the dangerous position of the experimental pipeline to collect the acoustic emission signals of the experimental pipeline collected by the acoustic emission signal collection system; and based on the dangerous position of the experimental pipeline, the cracking parameters of the experimental pipeline are collected.

[0130] In the scheme of the present invention, it is considered that when cracks or other damages occur in the air-conditioning pipeline, the stress state inside the material changes, resulting in tiny fractures or slips in local areas. These tiny fractures or slips will release the characteristics of the sound wave signal; therefore, the sound wave signal is converted into an electrical signal by acoustic emission technology, and the electrical signal is collected and processed to obtain the characteristic value of the classification signal. The parameter list analysis method is used to determine whether the air-conditioning pipeline has cracks, so that timely measures can be taken when cracks are detected in the air-conditioning pipeline to extend the service life of the air-conditioning pipeline.

[0131] In some embodiments, the control unit 104, when the training process of the pre-trained pipeline identification model includes extracting characteristic parameters from the acoustic emission signal of the experimental pipeline, the acoustic emission characteristic parameters of the pipeline to be tested and the acoustic emission characteristic parameters of the experimental pipeline, that is, the characteristic parameters obtained by extracting the characteristics of the acoustic emission electrical signals of the corresponding pipeline, include: time domain parameters of the acoustic emission electrical signals of the corresponding pipeline, frequency domain parameters of the acoustic emission electrical signals of the corresponding pipeline, and energy parameters of the acoustic emission electrical signals of the corresponding pipeline.

[0132] Among them, the time domain parameters of the acoustic emission electrical signal of the corresponding pipeline include at least one of the following: the voltage amplitude of the acoustic emission electrical signal of the corresponding pipeline, the voltage rise time of the acoustic emission electrical signal of the corresponding pipeline, the voltage duration of the acoustic emission electrical signal of the corresponding pipeline, and the count value of ringing within the voltage duration of the acoustic emission electrical signal of the corresponding pipeline.

[0133] The energy parameters of the acoustic emission electrical signal of the corresponding pipeline include at least one of the following: the total energy of the acoustic emission electrical signal of the corresponding pipeline, the frequency band energy of the acoustic emission electrical signal of the corresponding pipeline in a preset frequency band, and the proportion of the frequency band energy of the acoustic emission electrical signal of the corresponding pipeline in the preset frequency band in the total energy.

[0134] The frequency domain parameters of the acoustic emission electrical signal of the corresponding pipeline include at least one of the following: the frequency of the acoustic emission electrical signal of the corresponding pipeline, and the pulse factor of the acoustic emission electrical signal of the corresponding pipeline; wherein the frequency of the acoustic emission electrical signal of the corresponding pipeline includes at least one of the following: the initial frequency of the acoustic emission electrical signal of the corresponding pipeline, the peak frequency of the acoustic emission electrical signal of the corresponding pipeline, the average frequency of the acoustic emission electrical signal of the corresponding pipeline, and the inverse calculated frequency of the acoustic emission electrical signal of the corresponding pipeline.

[0135] Specifically, the characteristic parameters obtained by feature extraction of the acoustic emission electrical signal of the pipeline to be tested include: time domain parameters of the acoustic emission electrical signal of the pipeline to be tested, and frequency domain parameters of the acoustic emission electrical signal of the pipeline to be tested; wherein the time domain parameters of the acoustic emission electrical signal of the pipeline to be tested include at least one of the following: voltage amplitude of the acoustic emission electrical signal of the pipeline to be tested, voltage rise time of the acoustic emission electrical signal of the pipeline to be tested, voltage duration of the acoustic emission electrical signal of the pipeline to be tested, and count value of ringing within the voltage duration of the acoustic emission electrical signal of the pipeline to be tested; the frequency domain parameters of the acoustic emission electrical signal of the pipeline to be tested include the following At least one of: the total energy of the acoustic emission signal of the pipeline to be tested, the frequency of the acoustic emission signal of the pipeline to be tested, the pulse factor of the acoustic emission signal of the pipeline to be tested, the frequency band energy of the acoustic emission signal of the pipeline to be tested in a preset frequency band, and the proportion of the frequency band energy of the acoustic emission signal of the corresponding pipeline in the preset frequency band in the total energy; wherein, the frequency of the acoustic emission signal of the pipeline to be tested includes at least one of the following: the initial frequency of the acoustic emission signal of the pipeline to be tested, the peak frequency of the acoustic emission signal of the pipeline to be tested, the average frequency of the acoustic emission signal of the pipeline to be tested, and the inverse calculated frequency of the acoustic emission signal of the pipeline to be tested.

[0136] The characteristic parameters obtained by feature extraction of the acoustic emission electrical signal of the experimental pipeline include: time domain parameters of the acoustic emission electrical signal of the experimental pipeline, and frequency domain parameters of the acoustic emission electrical signal of the experimental pipeline; wherein the time domain parameters of the acoustic emission electrical signal of the experimental pipeline include at least one of the following: voltage amplitude of the acoustic emission electrical signal of the experimental pipeline, voltage rise time of the acoustic emission electrical signal of the experimental pipeline, voltage duration of the acoustic emission electrical signal of the experimental pipeline, and count value of ringing within the voltage duration of the acoustic emission electrical signal of the experimental pipeline; the frequency domain parameters of the acoustic emission electrical signal of the experimental pipeline include at least the following: One of: the total energy of the acoustic emission signal of the experimental pipeline, the frequency of the acoustic emission signal of the experimental pipeline, the pulse factor of the acoustic emission signal of the experimental pipeline, the frequency band energy of the acoustic emission signal of the experimental pipeline in a preset frequency band, and the proportion of the frequency band energy of the acoustic emission signal of the corresponding pipeline in the preset frequency band in the total energy; wherein, the frequency of the acoustic emission signal of the experimental pipeline includes at least one of the following: the initial frequency of the acoustic emission signal of the experimental pipeline, the peak frequency of the acoustic emission signal of the experimental pipeline, the average frequency of the acoustic emission signal of the experimental pipeline, and the inverse calculated frequency of the acoustic emission signal of the experimental pipeline.

[0137] like Figure 7 As shown in the figure, the process of identifying cracks in air-conditioning pipes based on acoustic emission technology and neural network technology also includes:

[0138] Step 3: Based on the acoustic emission signal of the air-conditioning pipeline in the fatigue test, the acoustic emission characteristic value is selected: the characteristic value of the acoustic emission signal is selected through two different dimensions of time-energy and frequency-energy to obtain the characteristic value of the acoustic emission signal of the air-conditioning pipeline, and then step 4 is executed.

[0139] In step 3, the characteristic values ​​of the acoustic emission signal are selected through two different dimensions of time-energy and frequency-energy. Specifically, the wavelet packet transform method can be used to characterize the acoustic emission signal to obtain the characteristic values ​​of the acoustic emission signal. Among them, the wavelet packet transform (WPT) is a signal processing technology that inherits the time-frequency analysis characteristics of the wavelet transform and further decomposes the undecomposed high-frequency band signal to improve the frequency resolution.

[0140] Specifically, the process of characterizing the acoustic emission signal using the wavelet packet transform method may include: first determining the spectral characteristics of the acoustic emission signal, selecting an appropriate decomposition level n (n is a positive integer), decomposing the original signal into 2 n The bandwidth of each node is 1 / 2 of the bandwidth of the original signal. n Finally, the wavelet coefficient of each node is inversely transformed, and the signal component corresponding to the node is calculated to obtain the signal energy of the node. The energy of different frequency bands is selected from the calculated signal energy as the characteristic value of the acoustic emission signal.

[0141] For example, the acoustic emission signal of one of the air conditioning pipes is transformed by wavelet packet, and the db4 series wavelet function is selected to perform 3-layer wavelet packet decomposition to obtain the energy values ​​of the 8 frequency bands of the signal, and the energy of different frequency bands in the acoustic emission signal is selected by sight. Among them, the db4 series wavelet function is a discrete wavelet transform function.

[0142] When wavelet packet transform is used to characterize the acoustic emission signal, the energy of the signal and the energy of the signal components in each frequency band can be calculated by equations (1) and (2):

[0143]

[0144] Among them, formula (1) can calculate the energy in the signal, where: E i represents the energy of frequency band i; represents the signal component of frequency band i, and A is the conversion coefficient, which is related to the model of the acoustic emission sensor. Formula (2) can calculate the energy of the signal component in each frequency band. The calculated frequency band energy distribution is as follows: Fig.10 and Fig.11 shown. Fig.10 A schematic diagram of the frequency band energy curve for each frequency band number. Fig.11 A schematic diagram showing the ratio of the energy of each frequency band number to the total energy. Fig.10 and Fig.11 It can be seen that the frequency band numbered 2 has the highest frequency band energy.

[0145] Step 4: Take the characteristic value of the acoustic emission signal of the air-conditioning pipeline obtained in step 3 as a sample, use the BP neural network to train the neural network, and obtain the air-conditioning pipeline recognition model; then, for the characteristic value of the acoustic emission signal of the air-conditioning pipeline obtained on site, use the air-conditioning pipeline recognition model to identify whether the air-conditioning pipeline on site is cracked.

[0146] The waveform of the acoustic emission signal of the air-conditioning pipeline contains four characteristic values: amplitude, rise time, duration and ringing count, which mainly play a good detection role in the early stage of fatigue crack generation. The four characteristic values ​​contained in the frequency are the frequencies of the acoustic emission signal at different stages: the initial frequency is the average frequency before the acoustic emission signal reaches the peak, the peak frequency is the frequency corresponding to the maximum peak in the power spectrum, the inverse frequency is the average frequency after the peak of the acoustic emission signal, and the average frequency is the average frequency of the entire acoustic emission signal, which represent the frequencies of the acoustic emission signal at different stages. Finally, the amplitude, rise time, duration, ringing count, initial frequency, peak frequency, inverse frequency, average frequency, signal strength, pulse factor and the energy proportion of the first four frequency bands are selected as the characteristic value input in the neural network training. Among them, the characteristic values ​​of the acoustic emission signal (such as amplitude, rise time, duration, ringing count), and the frequency band energy (that is, the energy of the first four frequency bands) are as follows Fig.12 shown. Fig.12 Schematic diagram of the curve of acoustic emission parameters.

[0147] Judging whether there is fatigue crack based on the acoustic emission signal of the air-conditioning pipeline: When cracks occur at a certain position of the air-conditioning pipeline, various acoustic emission parameters also undergo drastic changes; for example, at the position of the air-conditioning pipeline where cracks occur, the spectrum signal of the acoustic emission will increase, the energy will increase, and the pulse factor will decrease. By inputting the characteristic value and energy frequency band of the acoustic emission signal into the BP neural network model, the acoustic emission signal can be automatically identified after training. When the acoustic emission signal shows the above characteristics during the detection process, it means that fatigue cracks have occurred in the air-conditioning pipeline. Fig.13 Shown is the distribution diagram of the pulse factor in the acoustic emission signal at different stages. Fig.13Schematic diagram of the distribution of pulse damage factors in different stages (such as initiation stage, extension stage and crack stage).

[0148] In step 4, the cracks in the air-conditioning pipeline are detected by using a neural network: the BP neural network model is selected to establish the relationship between the air-conditioning pipeline and the acoustic emission signal, so as to realize the real-time recognition of the cracking phenomenon of the air-conditioning pipeline. i As a sample, the sample contains training feature values. A large number of training samples are input into the BP neural network to establish a nonlinear functional relationship f between the feature value and whether the air conditioning pipeline has cracked. j The feature vector for the test is input into the nonlinear function f to identify whether cracks occur in the air conditioning pipe. The input amplitude, rise time, duration, ring count, initial frequency, peak frequency, back-calculated frequency, average frequency, signal strength, ASL and energy of the first four frequency bands in the acoustic emission data are used as the input data of the initial nodes of the 14 input layers of the neural network. The output node [1, 0] indicates the presence of cracks. The training samples are randomly selected, and the training samples account for 80% of the total samples. During the training process, the Levenberg-Marquardt optimization algorithm (LM algorithm) is used, with a maximum number of iterations of 2000, a training rate of 0.01, and a minimum training error of 10 -6 Among them, the Levenberg-Marquardt algorithm is a numerical optimization method widely used in nonlinear least squares problems. The remaining 20% ​​of the acoustic emission data is a test sample set.

[0149] In the scheme of the present invention, the acoustic emission signal in the fatigue crack of the air-conditioning pipeline is characterized and processed and then introduced into the BP neural network for neural network training and testing, and finally the cracks in the air-conditioning pipeline are identified and detected online through the neural network. The neural network is trained and tested on the air-conditioning pipeline, and the results are shown in Table 1. It can be found from Table 1 that the error ratio is 3.56%, and the neural network has a high accuracy in identifying cracks in the air-conditioning pipeline.

[0150] Table 1

[0151] Working conditions Amount of training data Training error ratio Test data volume Test error ratio Air conditioning pipeline 100000 3.56％ 30000 3.96％

[0152] In the scheme of the present invention, neural network technology is used to identify and detect fatigue cracks in air-conditioning pipelines through acoustic emission, fatigue cracks in air-conditioning pipelines are identified, and cracks are detected in real time so that timely measures can be taken to extend the service life of the air-conditioning pipelines.

[0153] Since the processing and functions implemented by the device of this embodiment basically correspond to the embodiments, principles and examples of the aforementioned method, for the details not fully described in this embodiment, please refer to the relevant descriptions in the aforementioned embodiments, and no further elaboration will be made here.

[0154] According to an embodiment of the present invention, an air conditioner corresponding to the pipeline identification device is also provided. The air conditioner may include: the pipeline identification device described above.

[0155] Since the processing and functions implemented by the air conditioner of this embodiment basically correspond to the embodiments, principles and examples of the aforementioned devices, for the details not fully described in this embodiment, please refer to the relevant descriptions in the aforementioned embodiments and will not be repeated here.

[0156] According to an embodiment of the present invention, a computer program product corresponding to an air conditioner is also provided, comprising a computer program, and when the computer program is executed by a processor, the steps of the pipeline identification method described above are implemented.

[0157] Since the processing and functions implemented by the product of this embodiment basically correspond to the embodiments, principles and examples of the aforementioned air conditioner, for the details not fully described in this embodiment, please refer to the relevant descriptions in the aforementioned embodiments, and no further elaboration will be made here.

[0158] According to an embodiment of the present invention, a storage medium corresponding to the pipeline identification method is also provided, wherein the storage medium includes a stored program, wherein when the program is executed, the device where the storage medium is located is controlled to execute the steps of the pipeline identification method described above.

[0159] Since the processing and functions implemented by the storage medium of this embodiment basically correspond to the embodiments, principles and examples of the aforementioned method, for the details not fully described in this embodiment, please refer to the relevant descriptions in the aforementioned embodiments, and no further elaboration will be made here.

[0160] In summary, it is easy for those skilled in the art to understand that, under the premise of no conflict, the above-mentioned advantageous methods can be freely combined and superimposed.

[0161] The above description is only an embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent substitution, improvement, etc. made within the spirit and principle of the present invention shall be included in the scope of the claims of the present invention.

Claims

1. A pipeline identification method, characterized in that: The method is used to identify whether a pipeline to be tested of a device is cracked; the pipeline identification method comprises: For a pipeline to be tested in the pipelines of the device, obtaining an acoustic emission signal of the pipeline to be tested; Extracting characteristic parameters from the acoustic emission signal of the pipeline to be tested as the acoustic emission characteristic parameters of the pipeline to be tested; Based on the acoustic emission characteristic parameters of the pipeline to be tested, a pipeline identification model obtained by pre-training is used to identify whether the pipeline to be tested is cracked, so as to initiate a reminder message of the cracking of the pipeline to be tested when the pipeline to be tested is identified to be cracked.

2. The pipeline identification method according to claim 1, characterized in that: The training process of the pre-trained pipeline recognition model includes: During the experimental process of detecting cracking of the experimental pipeline, the acoustic emission signal of the experimental pipeline is collected, and the cracking parameters of the experimental pipeline are collected; Extracting characteristic parameters from the acoustic emission signal of the experimental pipeline as the acoustic emission characteristic parameters of the experimental pipeline; The acoustic emission characteristic parameters of the experimental pipeline and the cracking parameters of the experimental pipeline are used as training samples, and the preset neural network model is used as the basic model; The acoustic emission characteristic parameters of the experimental pipeline are used as input parameters of the basic model, and the cracking parameters of the experimental pipeline are used as output parameters of the basic model to perform neural network training to obtain the required pipeline identification model.

3. The pipeline identification method according to claim 1 or 2, characterized in that: in, Extracting characteristic parameters from the acoustic emission signal of the pipeline to be tested as the acoustic emission characteristic parameters of the pipeline to be tested includes: Convert the acoustic emission signal of the pipeline to be tested into an electrical signal to obtain the acoustic emission electrical signal of the pipeline to be tested; The wavelet packet analysis method is used to extract the characteristics of the acoustic emission electrical signal of the pipeline to be tested, and characteristic parameters are obtained, which are recorded as the acoustic emission characteristic parameters of the pipeline to be tested; and / or, In the case where the training process of the pre-trained pipeline recognition model includes extracting characteristic parameters from the acoustic emission signal of the experimental pipeline, extracting characteristic parameters from the acoustic emission signal of the experimental pipeline as the acoustic emission characteristic parameters of the experimental pipeline includes: Converting the acoustic emission signal of the experimental pipeline into an electrical signal to obtain the acoustic emission electrical signal of the experimental pipeline; The wavelet packet analysis method is used to extract the characteristics of the acoustic emission electrical signal of the experimental pipeline, and the characteristic parameters are obtained, which are recorded as the acoustic emission characteristic parameters of the experimental pipeline.

4. The pipeline identification method according to claim 2, characterized in that: The experimental process for detecting cracking of the experimental pipeline includes: Constructing a finite element model for an experimental pipeline in the pipeline of the device; Using the finite element model to perform simulation analysis on the experimental pipeline, determine the location where a preset degree of fracture or slip occurs in the experimental pipeline, and record it as a dangerous location of the experimental pipeline; Acoustic emission signals of the experimental pipeline are collected from the dangerous position of the experimental pipeline and within the set area of ​​the dangerous position of the experimental pipeline, and cracking parameters of the experimental pipeline are collected.

5. The pipeline identification method according to claim 4, characterized in that: in, For a pipeline to be tested in the pipeline of the device, obtaining an acoustic emission signal of the pipeline to be tested includes: Acquiring an acoustic emission signal of the pipeline to be tested collected by a preset first acoustic emission signal collection system; wherein the first acoustic emission signal collection system is pre-set at a preset cracking point on the pipeline to be tested; and / or, Collect the acoustic emission signals of the test pipeline from the dangerous position of the test pipeline and the set area of ​​the dangerous position of the test pipeline, including: The acoustic emission signals of the experimental pipeline collected by the preset second acoustic emission signal collection system are collected; wherein the second acoustic emission signal collection system is preset at the dangerous position of the experimental pipeline and in the set area of ​​the dangerous position of the experimental pipeline.

6. The pipeline identification method according to claim 1 or 2, characterized in that: In the case where the training process of the pre-trained pipeline identification model includes extracting characteristic parameters from the acoustic emission signal of the experimental pipeline, the acoustic emission characteristic parameters of the pipeline to be tested and the acoustic emission characteristic parameters of the corresponding pipeline in the acoustic emission characteristic parameters of the experimental pipeline include: time domain parameters of the acoustic emission electric signal of the corresponding pipeline, frequency domain parameters of the acoustic emission electric signal of the corresponding pipeline, and energy parameters of the acoustic emission electric signal of the corresponding pipeline; in, The time domain parameters of the acoustic emission electrical signal of the corresponding pipeline include at least one of the following: the voltage amplitude of the acoustic emission electrical signal of the corresponding pipeline, the voltage rise time of the acoustic emission electrical signal of the corresponding pipeline, the voltage duration of the acoustic emission electrical signal of the corresponding pipeline, and the count value of ringing within the voltage duration of the acoustic emission electrical signal of the corresponding pipeline; The energy parameters of the acoustic emission electric signal of the corresponding pipeline include at least one of the following: the total energy of the acoustic emission electric signal of the corresponding pipeline, the frequency band energy of the acoustic emission electric signal of the corresponding pipeline in a preset frequency band, and the proportion of the frequency band energy of the acoustic emission electric signal of the corresponding pipeline in the preset frequency band in the total energy; The frequency domain parameters of the acoustic emission electrical signal of the corresponding pipeline include at least one of the following: the frequency of the acoustic emission electrical signal of the corresponding pipeline, and the pulse factor of the acoustic emission electrical signal of the corresponding pipeline; wherein the frequency of the acoustic emission electrical signal of the corresponding pipeline includes at least one of the following: the initial frequency of the acoustic emission electrical signal of the corresponding pipeline, the peak frequency of the acoustic emission electrical signal of the corresponding pipeline, the average frequency of the acoustic emission electrical signal of the corresponding pipeline, and the inverse calculated frequency of the acoustic emission electrical signal of the corresponding pipeline.

7. A pipeline identification device, characterized in that: Used to identify whether a pipeline to be tested of a device is cracked; the pipeline identification device comprises: an acquisition unit, configured to acquire an acoustic emission signal of the pipeline to be tested in the pipelines of the device; A control unit is configured to extract characteristic parameters from the acoustic emission signal of the pipeline to be tested as the acoustic emission characteristic parameters of the pipeline to be tested; The control unit is further configured to identify whether the pipeline to be tested is cracked based on the acoustic emission characteristic parameters of the pipeline to be tested and using a pre-trained pipeline identification model, so as to initiate a reminder message of the cracking of the pipeline to be tested when the pipeline to be tested is identified to be cracked.

8. An air conditioner, characterized in that: include: The pipeline identification device as claimed in claim 7.

9. A storage medium, characterized in that: The storage medium includes a stored program, wherein when the program is executed, the device where the storage medium is located is controlled to execute the pipeline identification method according to any one of claims 1 to 6.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the pipeline identification method described in any one of claims 1 to 6 are implemented.

Images