A belt roasting production line equipment fault detection method and device

Abstract

The present invention discloses a method and device for detecting faults in a belt-type roasting production line. Several preset positions on the belt-type roasting equipment are respectively provided with audio acquisition components, wherein the method comprises: obtaining a real-time audio signal of a current audio sampling period at a preset position based on the audio acquisition component; extracting audio signals from the real-time audio signal of the current audio sampling period according to several preset frequency bands to obtain several frequency-divided signals corresponding to the preset frequency bands; extracting a feature vector of each frequency-divided signal based on an audio signal recognition model, and comparing the feature vector with the feature vectors of several fault audio signals of the corresponding preset frequency bands in an audio feature library, and if they are similar, determining that the belt-type roasting production line equipment has a fault, and recording the feature vector of the frequency-divided signal. Analyzing equipment faults based on audio has many advantages, such as timely discovery of fault symptoms, accurate detection of fault types, and real-time and uninterrupted monitoring.

Description

Technical Field

[0001] The present invention relates to a field, and in particular to a method and a device for detecting faults of belt roasting production line equipment. Background Art

[0002] The belt roasting production line is widely used in the production of pellets in the domestic steel industry. Its promotion and application can provide the steel industry with a production line with reasonable process, low energy consumption and environmental protection, and help drive the development of my country's mining processing equipment manufacturing industry. At the same time, due to its strong adaptability to raw materials, it can reduce the risk of the steel industry restricting raw material varieties. In addition, the reasonable setting of the thermal system of the belt roasting production line and the characteristics of short pipelines create favorable conditions for realizing the recycling of roasting gas and reducing heat and electricity consumption, which helps to reduce the energy consumption level of the steel industry.

[0003] Traditional equipment status perception methods are mostly contact-based, and non-contact methods include visual analysis methods such as cameras. However, they lack audio analysis to capture subtle sound changes in equipment before a failure occurs, thereby achieving early warning of equipment failures, helping to detect potential problems in a timely manner and avoid production interruptions and losses caused by equipment failures. Summary of the invention

[0004] The purpose of the embodiment of the present invention is to provide a belt roasting production line equipment fault detection method and device, by processing the collected audio signals at preset positions, obtaining frequency-divided signals of each preset frequency band, and comparing each frequency-divided signal with the fault audio signal pre-stored in the audio feature library, to determine the similarity between the two, and when the similarity meets the preset requirements, it is determined that the equipment has a fault or fault symptom, and has many advantages such as high fault detection accuracy, early detection of fault symptoms, and no need to shut down for maintenance.

[0005] In order to solve the above technical problems, a first aspect of an embodiment of the present invention provides a method for detecting faults of belt-type roasting production line equipment, wherein audio collection components are respectively arranged at several preset positions on the belt-type roasting equipment, comprising the following steps:

[0006] Acquire a real-time audio signal of a current audio sampling period at the preset position based on the audio acquisition component;

[0007] Extracting audio signals from the real-time audio signal of the current audio sampling period according to a plurality of preset frequency bands to obtain a plurality of frequency-divided signals corresponding to the preset frequency bands;

[0008] Based on the audio signal recognition model, the feature vector of each of the frequency-divided signals is extracted, and the feature vector is compared with the feature vectors of several faulty audio signals in the corresponding preset frequency bands in the audio feature library. If they are similar, it is determined that there is a fault in the belt roasting production line equipment, and the feature vector of the frequency-divided signal is recorded.

[0009] Furthermore, extracting a feature vector of each of the frequency-divided signals and comparing the feature vector with feature vectors of a plurality of fault audio signals of corresponding preset frequency bands in the audio feature library respectively includes:

[0010] Performing windowing processing and fast Fourier transform on the frequency-divided signal to obtain a plurality of audio frames within the current audio sampling period;

[0011] Obtaining the bandwidth and spectrum centroid of each of the audio frames;

[0012] Based on the bandwidth and the spectrum centroid of the audio frame, respectively calculating the similarity between each frame of the frequency-divided signal and a plurality of corresponding frames of the fault audio signal in the corresponding preset frequency band in the audio feature library;

[0013] When the average value of the similarities is greater than a similarity threshold, it is determined that the frequency-divided signal is similar to the fault audio signal.

[0014] Furthermore, the number of frames of the frequency-divided signal is the same as the number of frames of the fault audio signal corresponding to the corresponding preset frequency band;

[0015] The average value calculation formula of the similarity is:

[0016]

[0017]

[0018] Where n is the number of audio frames in the frequency-divided signal, cos(θ i ) is the similarity value between the i-th audio frame of the frequency-divided signal and the i-th audio frame of the fault audio signal, BW Ri is the bandwidth value of the i-th audio frame in the frequency-divided signal, BW Ti is the bandwidth value of the i-th audio frame in the faulty audio signal, SC Ri is the spectrum centroid value of the i-th audio frame in the frequency-divided signal, SC Ti is the spectrum centroid value of the i-th audio frame in the fault audio signal;

[0019] The similarity value cos(θ i) has a setting range of (-1,1). The larger the value, the higher the similarity between the two; the smaller the value, the lower the similarity between the two.

[0020] Furthermore, after comparing the feature vector with the feature vectors of several fault audio signals of corresponding preset frequency bands in the audio feature library, the method further includes:

[0021] When the feature vector is not similar to a plurality of fault audio signals of a corresponding preset frequency band in the audio feature library, comparing the feature vector with a plurality of normal audio signals of a corresponding preset frequency band in the audio feature library;

[0022] If the characteristic vector of each frequency-divided signal in the real-time audio signal is similar to a characteristic vector of the corresponding normal audio signal, it is determined that the belt roasting production line equipment is operating normally;

[0023] If the characteristic vector of each frequency-divided signal in the real-time audio signal is not similar to a corresponding characteristic vector of the normal audio signal, it is determined that the belt roasting production line equipment has a fault, and the characteristic vector of the frequency-divided signal is recorded.

[0024] Furthermore, after recording the characteristic vector of the frequency-divided signal, the method further includes:

[0025] Obtaining the number of times the characteristic vector of the frequency-divided signal is recorded;

[0026] When the number of records is greater than a first preset number, a first alarm signal of equipment failure is sent to the control system of the belt roasting production line;

[0027] When the number of records is greater than a second preset number, an equipment emergency shutdown signal is sent to the control system of the belt roasting production line.

[0028] Furthermore, the plurality of preset positions are respectively provided with image acquisition components, each of the preset positions is provided with an image recognition base point, and the surrounding environment of the preset positions is provided with a plurality of image recognition reference points at the same distance as the image recognition base point;

[0029] After determining that the belt roasting production line equipment has a fault, the method further includes:

[0030] Based on the image acquisition component, the distance value between each image recognition base point in the preset position and the corresponding plurality of image recognition reference points is obtained, and each of the distance values ​​is determined to be less than or equal to a first distance threshold;

[0031] If yes, it is determined that the image recognition base point at the preset position is at the initial position;

[0032] If not, it is determined that the image recognition base point of the preset position has deviated from the initial position, and a second equipment failure alarm signal is sent to the belt roasting production line equipment control system.

[0033] Furthermore, after determining that each of the distance values ​​is less than or equal to the first distance threshold, the method further includes:

[0034] Determining whether the distance value is greater than a second distance threshold;

[0035] If so, an equipment emergency shutdown signal is sent to the control system of the belt roasting production line.

[0036] Accordingly, a first aspect of an embodiment of the present invention provides a belt roasting production line equipment fault detection device, wherein several preset positions on the belt roasting equipment are respectively provided with audio collection components, including:

[0037] An audio receiving module, which is used to obtain a real-time audio signal of a current audio sampling period at the preset position based on the audio acquisition component;

[0038] A signal extraction module, which is used to extract audio signals from the real-time audio signal of the current audio sampling period according to a plurality of preset frequency bands, and obtain a plurality of frequency-divided signals corresponding to the preset frequency bands respectively;

[0039] The first feature comparison module is used to extract the feature vector of each of the frequency-divided signals based on the audio signal recognition model, and compare the feature vector with the feature vectors of several faulty audio signals in the corresponding preset frequency bands in the audio feature library. If they are similar, it is determined that there is a fault in the belt roasting production line equipment, and the feature vector of the frequency-divided signal is recorded.

[0040] Furthermore, the first feature comparison module includes:

[0041] An audio frame acquisition unit, which is used to perform windowing processing and fast Fourier transform on the frequency-divided signal to obtain a plurality of audio frames within the current audio sampling period;

[0042] A feature extraction unit, which is used to obtain the bandwidth and spectrum centroid of each of the audio frames;

[0043] A similarity calculation unit, which is used to calculate the similarity between each frame of the frequency-divided signal and a plurality of corresponding frames of the fault audio signal in the corresponding preset frequency band in the audio feature library based on the bandwidth and the spectrum centroid of the audio frame;

[0044] The first similarity determination unit is configured to determine that the frequency-divided signal is similar to the fault audio signal when the average value of the similarity is greater than a similarity threshold.

[0045] Furthermore, the number of frames of the frequency-divided signal is the same as the number of frames of the fault audio signal corresponding to the corresponding preset frequency band;

[0046] The average value calculation formula of the similarity is:

[0047]

[0048]

[0049] Where n is the number of audio frames in the frequency-divided signal, cos(θ i ) is the similarity value between the i-th audio frame of the frequency-divided signal and the i-th audio frame of the fault audio signal, BW Ri is the bandwidth value of the i-th audio frame in the frequency-divided signal, BW Ti is the bandwidth value of the i-th audio frame in the faulty audio signal, SC Ri is the spectrum centroid value of the i-th audio frame in the frequency-divided signal, SC Ti is the spectrum centroid value of the i-th audio frame in the fault audio signal;

[0050] The similarity value cos(θ i ) has a setting range of (-1,1). The larger the value, the higher the similarity between the two; the smaller the value, the lower the similarity between the two.

[0051] Furthermore, it also includes: a second feature comparison module, wherein the second feature comparison module includes:

[0052] A similarity comparison unit, configured to compare the frequency-divided signal with a plurality of normal audio signals of a corresponding preset frequency band in the audio feature library when the frequency-divided signal is not similar to a plurality of faulty audio signals of a corresponding preset frequency band in the audio feature library;

[0053] A second similarity determination unit, which is used to determine that the belt roasting production line equipment is operating normally when each frequency-divided signal in the real-time audio signal is similar to a normal audio signal;

[0054] The second similarity determination unit is further configured to determine that a frequency-divided signal is an abnormal audio signal and record the frequency-divided signal when any frequency-divided signal in the real-time audio signal is not similar to the plurality of normal audio signals.

[0055] Furthermore, the second feature comparison module also includes:

[0056] An abnormal signal recording unit, which is used to abnormally obtain the number of times the frequency division signal is recorded;

[0057] An abnormal signal alarm unit, which is used to send a first alarm signal of equipment failure to the control system of the belt roasting production line when the number of records is greater than a first preset number;

[0058] The abnormal signal alarm unit is also used to send an equipment emergency shutdown signal to the control system of the belt roasting production line when the number of records is greater than a second preset number.

[0059] Furthermore, the plurality of preset positions are respectively provided with image acquisition components, each of the preset positions is provided with an image recognition base point, and the surrounding environment of the preset positions is provided with a plurality of image recognition reference points at the same distance as the image recognition base point;

[0060] The belt roasting production line equipment fault detection device also includes:

[0061] An image receiving module, which is used to obtain the distance value between each image recognition base point in the preset position and the corresponding plurality of image recognition reference points based on the image acquisition component, and determine whether each of the distance values ​​is less than or equal to a first distance threshold;

[0062] An image determination module, configured to determine that the image recognition base point at the preset position is at an initial position when the distance value is less than or equal to the first distance threshold;

[0063] The image judgment module is also used to determine that the image recognition base point of the preset position has deviated from the initial position when the distance value is greater than the first distance threshold, and send a second equipment failure alarm signal to the belt roasting production line equipment control system.

[0064] Furthermore, the image judgment module is also used to judge whether the distance value is greater than a second distance threshold, and if so, send an equipment emergency shutdown signal to the control system of the belt roasting production line.

[0065] Correspondingly, a third aspect of an embodiment of the present invention provides an electronic device, comprising: at least one processor; and a memory connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor executes the above-mentioned belt roasting production line equipment fault detection method.

[0066] Accordingly, a fourth aspect of an embodiment of the present invention provides a computer-readable storage medium having computer instructions stored thereon, which, when executed by a processor, implement the above-mentioned belt roasting production line equipment fault detection method.

[0067] The above technical solution of the embodiment of the present invention has the following beneficial technical effects:

[0068] 1. Through audio analysis technology, it is possible to capture subtle sound changes of equipment before failure occurs, thereby achieving early warning of equipment failure, helping to discover potential problems in a timely manner and avoid production interruptions and losses caused by equipment failure;

[0069] 2. Based on audio analysis, equipment failure usually does not require the equipment to be disassembled or shut down, and will not interfere with the normal operation of the equipment; non-contact detection methods can ensure production continuity while reducing maintenance costs;

[0070] 3. Audio analysis technology can realize real-time monitoring and fault diagnosis of equipment, so that managers and operators can grasp the operating status of equipment at any time, which helps to find and solve equipment failures in time and improve production efficiency;

[0071] 4. Audio analysis of equipment failure can use advanced signal processing and artificial intelligence technology to perform high-precision identification and analysis of the equipment's sound signals, which helps to accurately determine the type and location of equipment failure and provide maintenance personnel with targeted maintenance suggestions.

[0072] 5. Through audio analysis of equipment failures, preventive maintenance of equipment can be achieved, reducing downtime and repair costs caused by equipment failures. Audio analysis technology can also help optimize the operating parameters of the equipment and improve the operating efficiency and service life of the equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0073] Figure 1 It is a flow chart of a method for detecting faults in a belt roasting production line provided by an embodiment of the present invention;

[0074] Figure 2 This is a module block diagram of a belt roasting production line equipment fault detection device provided by an embodiment of the present invention;

[0075] Figure 3 is a block diagram of a first feature comparison module provided in an embodiment of the present invention;

[0076] Figure 4 is a block diagram of a second feature comparison module provided in an embodiment of the present invention;

[0077] Figure 5 It is a block diagram of an image detection module provided by an embodiment of the present invention.

[0078] Reference numerals:

[0079] 1. Audio receiving module, 2. Signal extraction module, 3. First feature comparison module, 31. Audio frame acquisition unit, 32. Feature extraction unit, 33. Similarity calculation unit, 34. First similarity determination unit, 4. Second feature comparison module, 41. Similarity comparison unit, 42. Second similarity determination unit, 43. Abnormal signal recording unit, 44. Abnormal signal alarm unit, 5. Image detection module, 51. Image receiving unit, 52. Image determination unit. DETAILED DESCRIPTION

[0080] In order to make the purpose, technical scheme and advantages of the present invention clearer, the present invention is further described in detail below in conjunction with specific embodiments and with reference to the accompanying drawings. It should be understood that these descriptions are only exemplary and are not intended to limit the scope of the present invention. In addition, in the following description, the description of well-known structures and technologies is omitted to avoid unnecessary confusion of the concept of the present invention.

[0081] Please refer to Figure 1 A first aspect of an embodiment of the present invention provides a method for detecting faults in a belt roasting production line, wherein audio collection components are respectively arranged at several preset positions on the belt roasting equipment, comprising the following steps:

[0082] Step S100: acquiring a real-time audio signal of a current audio sampling period at a preset position based on an audio acquisition component.

[0083] Step S200: extracting audio signals from the real-time audio signal of the current audio sampling period according to a plurality of preset frequency bands to obtain a plurality of frequency-divided signals corresponding to the preset frequency bands.

[0084] Step S300, based on the audio signal recognition model, extract the feature vector of each frequency division signal, and compare the feature vector with the feature vectors of several faulty audio signals in the corresponding preset frequency bands in the audio feature library. If they are similar, it is determined that there is a fault in the belt roasting production line equipment, and the feature vector of the frequency division signal is recorded.

[0085] By processing the collected audio signals at preset positions, frequency division signals of each preset frequency band are obtained, and each frequency division signal is compared with the fault audio signal pre-stored in the audio feature library to determine the similarity between the two. When the similarity meets the preset requirements, it is determined that the equipment has a fault or fault symptom. It has many advantages such as high fault detection accuracy and early detection of fault symptoms.

[0086] Based on the frequency division signals of each preset frequency band, high-precision monitoring of fault audio signals can be achieved, which greatly improves the detection accuracy of equipment faults at various locations on the belt roasting production line. In addition, it can also be used to optimize and improve the operating status of the equipment.

[0087] Furthermore, in step S300, the feature vector of each frequency-divided signal is extracted, and the feature vector is compared with the feature vectors of several fault audio signals of the corresponding preset frequency band in the audio feature library, including:

[0088] Step S310, windowing processing is performed on the frequency-divided signal and a fast Fourier transform is performed to obtain a plurality of audio frames in the current audio sampling period.

[0089] Step S320: Obtain the bandwidth and spectrum centroid of each audio frame.

[0090] Step S330, based on the bandwidth and spectrum centroid of the audio frame, respectively calculate the similarity between each frame of the frequency-divided signal and a number of corresponding frames of the fault audio signal in the corresponding preset frequency band in the audio feature library.

[0091] Step S340: When the average value of the similarity is greater than the similarity threshold, it is determined that the frequency-divided signal is similar to the fault audio signal.

[0092] Furthermore, the number of frames of the frequency-divided signal is the same as the number of frames of the fault audio signal corresponding to the corresponding preset frequency band, and the average value calculation formula of the frame similarity is:

[0093]

[0094]

[0095] Where n is the number of audio frames in the frequency-divided signal, cos(θ i ) is the similarity value between the i-th audio frame of the frequency-divided signal and the i-th audio frame of the faulty audio signal, BW Ri is the bandwidth value of the i-th audio frame in the frequency-divided signal, BW Ti is the bandwidth value of the i-th audio frame in the faulty audio signal, SC Ri is the spectral centroid value of the i-th audio frame in the frequency-divided signal, SC Ti is the spectral centroid value of the i-th audio frame in the faulty audio signal.

[0096] Specifically, the similarity value cos(θ i ) is set in the range of (-1,1). The larger the value, the higher the similarity between the two. The smaller the value, the lower the similarity between the two.

[0097] Furthermore, after comparing the feature vector with the feature vectors of several fault audio signals of corresponding preset frequency bands in the audio feature library, the method further includes:

[0098] Step S410: when the feature vector is not similar to a plurality of faulty audio signals of a corresponding preset frequency band in the audio feature library, the feature vector is compared with a plurality of normal audio signals of a corresponding preset frequency band in the audio feature library.

[0099] Step S420: If the feature vector of each frequency-divided signal in the real-time audio signal is similar to a feature vector of the corresponding normal audio signal, it is determined that the belt roasting production line equipment operates normally.

[0100] Step S430: if the feature vector of each frequency-divided signal in the real-time audio signal is not similar to a feature vector of the corresponding normal audio signal, it is determined that the belt roasting production line equipment has a fault, and the feature vector of the frequency-divided signal is recorded.

[0101] Furthermore, after recording the characteristic vector of the frequency-divided signal in step S430, the method further includes:

[0102] Step S431, obtaining the number of times the characteristic vector of the frequency-divided signal is recorded.

[0103] Step S432: When the number of records is greater than a first preset number, a first alarm signal of equipment failure is sent to the control system of the belt roasting production line.

[0104] Step S433: when the number of recorded times is greater than a second preset number, an equipment emergency shutdown signal is sent to the control system of the belt roasting production line.

[0105] In another implementation of the present invention, image acquisition components are respectively provided at a plurality of preset positions, an image recognition base point is provided at each preset position, and a plurality of image recognition reference points are provided in the surrounding environment of the preset position at the same distance as the image recognition base point. The image acquisition component is used to further assist in determining a part of a device or production line that is determined to be faulty based on the audio signal.

[0106] Specifically, after determining in step S300 that the belt roasting production line equipment has a fault, the method further includes:

[0107] Step S510, based on the image acquisition component, obtains the distance values ​​between the image recognition base point in each preset position and the corresponding plurality of image recognition reference points, and determines whether each distance value is less than or equal to a first distance threshold.

[0108] Step S520: If yes, it is determined that the image recognition base point at the preset position is at the initial position.

[0109] Step S530: If not, it is determined that the image recognition base point at the preset position has deviated from the initial position, and a second equipment failure alarm signal is sent to the belt roasting production line equipment control system.

[0110] Further, after determining in step S510 that each distance value is less than or equal to the first distance threshold, the method further includes:

[0111] Step S541, determining whether the distance value is greater than a second distance threshold.

[0112] Step S542: If yes, an equipment emergency shutdown signal is sent to the control system of the belt roasting production line.

[0113] Accordingly, please refer to Figure 2 A first aspect of an embodiment of the present invention provides a belt roasting production line equipment fault detection device, wherein several preset positions on the belt roasting equipment are respectively provided with audio collection components, including:

[0114] An audio receiving module 1 is used to obtain a real-time audio signal of a current audio sampling period at a preset position based on an audio acquisition component;

[0115] A signal extraction module 2 is used to extract audio signals from the real-time audio signal of the current audio sampling period according to a number of preset frequency bands to obtain a number of frequency-divided signals corresponding to the preset frequency bands;

[0116] The first feature comparison module 3 is used to extract the feature vector of each frequency division signal based on the audio signal recognition model, and compare the feature vector with the feature vectors of several faulty audio signals in the corresponding preset frequency bands in the audio feature library. If they are similar, it is determined that there is a fault in the belt roasting production line equipment, and the feature vector of the frequency division signal is recorded.

[0117] For further information, please refer to Figure 4 , the first feature comparison module 3 includes:

[0118] An audio frame acquisition unit 31 is used to perform windowing processing on the frequency-divided signal and perform fast Fourier transform to obtain a number of audio frames within the current audio sampling period;

[0119] A feature extraction unit 32, which is used to obtain the bandwidth and spectrum centroid of each audio frame;

[0120] A similarity calculation unit 33 is used to calculate the similarity between each frame of the frequency-divided signal and a plurality of corresponding frames of the fault audio signal in the corresponding preset frequency band in the audio feature library based on the bandwidth and the spectrum centroid of the audio frame;

[0121] The first similarity determination unit 34 is configured to determine that the frequency-divided signal is similar to the fault audio signal when the average value of the similarity is greater than a similarity threshold.

[0122] Further, the number of frames of the frequency-divided signal is the same as the number of frames of the fault audio signal corresponding to the corresponding preset frequency band;

[0123] The average similarity calculation formula is:

[0124]

[0125]

[0126] Where n is the number of audio frames in the frequency-divided signal, cos(θ i ) is the similarity value between the i-th audio frame of the frequency-divided signal and the i-th audio frame of the faulty audio signal, BW Ri is the bandwidth value of the i-th audio frame in the frequency-divided signal, BW Ti is the bandwidth value of the i-th audio frame in the faulty audio signal, SC Ri is the spectral centroid value of the i-th audio frame in the frequency-divided signal, SC Ti is the spectral centroid value of the i-th audio frame in the faulty audio signal.

[0127] The similarity value cos(θ i ) has a setting range of (-1,1). The larger the value, the higher the similarity between the two; the smaller the value, the lower the similarity between the two.

[0128] For further information, please refer to Figure 4 The belt roasting production line equipment fault detection device also includes: a second feature comparison module 4, the second feature comparison module 4 includes:

[0129] A similarity comparison unit 41 is used to compare the frequency-divided signal with a number of normal audio signals in the corresponding preset frequency band in the audio feature library when the frequency-divided signal is not similar to a number of faulty audio signals in the corresponding preset frequency band in the audio feature library;

[0130] A second similarity determination unit 42, which is used to determine that the belt roasting production line equipment is operating normally when each frequency division signal in the real-time audio signal is similar to a normal audio signal;

[0131] The second similarity determination unit 42 is further configured to determine that a frequency-divided signal is an abnormal audio signal and record the frequency-divided signal when any frequency-divided signal in the real-time audio signal is not similar to any of the normal audio signals.

[0132] For further information, please refer to Figure 4 , the second feature comparison module 4 also includes:

[0133] An abnormal signal recording unit 43, which is used to record the number of abnormal acquisition frequency division signals;

[0134] An abnormal signal alarm unit 44, which is used to send a first alarm signal of equipment failure to the control system of the belt roasting production line when the number of recorded times is greater than a first preset number;

[0135] The abnormal signal alarm unit 44 is also used to send an equipment emergency shutdown signal to the control system of the belt roasting production line when the number of recorded times is greater than a second preset number.

[0136] Furthermore, image acquisition components are respectively provided at a plurality of preset positions, an image recognition base point is provided at each preset position, and a plurality of image recognition reference points having the same distance as the image recognition base point are provided in the surrounding environment of the preset position.

[0137] Please refer to Figure 5 The belt roasting production line equipment fault detection device also includes: an image detection module 5, and the image detection module 5 includes:

[0138] An image receiving unit 51 is used to obtain the distance value between the image recognition base point in each preset position and the corresponding plurality of image recognition reference points based on the image acquisition component, and determine that each distance value is less than or equal to a first distance threshold;

[0139] An image determination unit 52, which is used to determine that the image recognition base point at the preset position is at an initial position when the distance value is less than or equal to a first distance threshold;

[0140] The image judgment unit 52 is also used to determine that the image recognition base point at the preset position has deviated from the initial position when the distance value is greater than the first distance threshold, and send a second equipment failure alarm signal to the belt roasting production line equipment control system.

[0141] Furthermore, the image determination unit 52 is also used to determine whether the distance value is greater than a second distance threshold value, and if so, send an equipment emergency shutdown signal to the control system of the belt roasting production line.

[0142] Accordingly, a third aspect of an embodiment of the present invention provides an electronic device, comprising: at least one processor; and a memory connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor executes the above-mentioned belt roasting production line equipment fault detection method.

[0143] Accordingly, a fourth aspect of an embodiment of the present invention provides a computer-readable storage medium having computer instructions stored thereon, which, when executed by a processor, implement the above-mentioned belt roasting production line equipment fault detection method.

[0144] The embodiment of the present invention aims to protect a belt roasting production line equipment fault detection method and device, which has the following effects:

[0145] 1. Through audio analysis technology, it is possible to capture subtle sound changes of equipment before failure occurs, thereby achieving early warning of equipment failure, helping to discover potential problems in a timely manner and avoid production interruptions and losses caused by equipment failure;

[0146] 2. Based on audio analysis, equipment failure usually does not require the equipment to be disassembled or shut down, and will not interfere with the normal operation of the equipment; non-contact detection methods can ensure production continuity while reducing maintenance costs;

[0147] 3. Audio analysis technology can realize real-time monitoring and fault diagnosis of equipment, so that managers and operators can grasp the operating status of equipment at any time, which helps to find and solve equipment failures in time and improve production efficiency;

[0148] 4. Audio analysis of equipment failure can use advanced signal processing and artificial intelligence technology to perform high-precision identification and analysis of the equipment's sound signals, which helps to accurately determine the type and location of equipment failure and provide maintenance personnel with targeted maintenance suggestions.

[0149] 5. Through audio analysis of equipment failure, preventive maintenance of equipment can be achieved, reducing downtime and repair costs caused by equipment failure. At the same time, audio analysis technology can also help optimize the operating parameters of the equipment and improve the operating efficiency and service life of the equipment.

[0150] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0151] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0152] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0153] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0154] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A belt roasting production line equipment fault detection method, characterized in that: Several preset positions on the belt roasting device are respectively provided with audio collection components, including the following steps: Acquire a real-time audio signal of a current audio sampling period at the preset position based on the audio acquisition component; Extracting audio signals from the real-time audio signal of the current audio sampling period according to a plurality of preset frequency bands to obtain a plurality of frequency-divided signals corresponding to the preset frequency bands; Based on the audio signal recognition model, extract the feature vector of each of the frequency-divided signals, and compare the feature vectors with the feature vectors of several faulty audio signals in the corresponding preset frequency bands in the audio feature library. If they are similar, it is determined that there is a fault in the belt roasting production line equipment, and the feature vector of the frequency-divided signal is recorded; The extracting of the feature vector of each of the frequency-divided signals and comparing the feature vector with the feature vectors of a plurality of fault audio signals of corresponding preset frequency bands in the audio feature library include: Performing windowing processing and fast Fourier transform on the frequency-divided signal to obtain a plurality of audio frames within the current audio sampling period; Obtaining the bandwidth and spectrum centroid of each of the audio frames; Based on the bandwidth and the spectrum centroid of the audio frame, respectively calculating the similarity between each frame of the frequency-divided signal and a plurality of corresponding frames of the fault audio signal in the corresponding preset frequency band in the audio feature library; When the average value of the similarity is greater than the similarity threshold, determining that the frequency-divided signal is similar to the fault audio signal; The number of frames of the frequency-divided signal is the same as the number of frames of the fault audio signal corresponding to the corresponding preset frequency band; The average value calculation formula of the similarity is: Where n is the number of audio frames in the frequency-divided signal, cos(θ i ) is the similarity value between the i-th audio frame of the frequency-divided signal and the i-th audio frame of the fault audio signal, BW Ri is the bandwidth value of the i-th audio frame in the frequency-divided signal, BW Ti is the bandwidth value of the i-th audio frame in the faulty audio signal, SC Ri is the spectrum centroid value of the i-th audio frame in the frequency-divided signal, SC Ti is the spectrum centroid value of the i-th audio frame in the fault audio signal; The similarity value cos(θ i ) has a setting range of (-1,1). The larger the value, the higher the similarity between the two; the smaller the value, the lower the similarity between the two. After comparing the feature vector with the feature vectors of several fault audio signals of corresponding preset frequency bands in the audio feature library, the method further includes: When the divided frequency signal is not similar to a plurality of fault audio signals of a corresponding preset frequency band in the audio feature library, comparing the divided frequency signal with a plurality of normal audio signals of a corresponding preset frequency band in the audio feature library; If each frequency-divided signal in the real-time audio signal is similar to a normal audio signal, it is determined that the belt roasting production line equipment operates normally; If any frequency-divided signal in the real-time audio signal is not similar to the plurality of normal audio signals, the frequency-divided signal is determined to be an abnormal audio signal, and the frequency-divided signal is recorded.

2. The belt roasting production line equipment fault detection method according to claim 1, characterized in that: After recording the frequency-divided signal, the method further includes: Obtaining the number of times the frequency division signal is recorded; When the number of records is greater than a first preset number, a first alarm signal of equipment failure is sent to the control system of the belt roasting production line; When the number of records is greater than a second preset number, an equipment emergency shutdown signal is sent to the control system of the belt roasting production line.

3. The belt roasting production line equipment fault detection method according to claim 2, characterized in that: The plurality of preset positions are respectively provided with image acquisition components, each of the preset positions is provided with an image recognition base point, and the surrounding environment of the preset positions is provided with a plurality of image recognition reference points at the same distance as the image recognition base point; After determining that the belt roasting production line equipment has a fault, the method further includes: Based on the image acquisition component, the distance value between each image recognition base point in the preset position and the corresponding plurality of image recognition reference points is obtained, and each of the distance values ​​is determined to be less than or equal to a first distance threshold; If yes, it is determined that the image recognition base point at the preset position is at the initial position; If not, it is determined that the image recognition base point of the preset position has deviated from the initial position, and a second equipment failure alarm signal is sent to the belt roasting production line equipment control system.

4. The belt roasting production line equipment fault detection method according to claim 3, characterized in that: After determining that each of the distance values ​​is less than or equal to the first distance threshold, the method further includes: Determining whether the distance value is greater than a second distance threshold; If so, an equipment emergency shutdown signal is sent to the control system of the belt roasting production line.

5. A belt roasting production line equipment fault detection device, characterized in that: The belt roasting production line equipment fault detection method according to any one of claims 1 to 4 is used to detect the belt roasting production line equipment fault, wherein several preset positions on the belt roasting equipment are respectively provided with audio collection components, including: An audio receiving module, which is used to obtain a real-time audio signal of a current audio sampling period at the preset position based on the audio acquisition component; A signal extraction module, which is used to extract audio signals from the real-time audio signal of the current audio sampling period according to a plurality of preset frequency bands, and obtain a plurality of frequency-divided signals corresponding to the preset frequency bands respectively; A feature comparison module is used to extract a feature vector of each of the frequency-divided signals based on the audio signal recognition model, and compare the feature vector with the feature vectors of several faulty audio signals in the corresponding preset frequency bands in the audio feature library. If they are similar, it is determined that there is a fault in the belt roasting production line equipment, and the feature vector of the frequency-divided signal is recorded.

6. An electronic device, characterized in that: include: at least one processor; And a memory connected to the at least one processor; wherein the memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor executes the belt roasting production line equipment fault detection method as described in any one of claims 1-4.

7. A computer-readable storage medium, characterized in that: Computer instructions are stored thereon, and when the instructions are executed by the processor, the belt roasting production line equipment fault detection method as described in any one of claims 1-4 is implemented.

Images