Clean room air hammer fluctuation source positioning method and system

Abstract

The present invention belongs to the technical field of fault location, and discloses a method and system for locating the fluctuation source of a pneumatic hammer in a clean room; the method comprises: collecting n kinds of real-time pneumatic hammer operating data in the clean room; analyzing the n kinds of parameters collected in real time, judging whether the n kinds of parameters are abnormal, putting the first characteristic parameter performance value into the input data set as the characteristic parameter, and obtaining the collection time corresponding to the first characteristic parameter, and marking it as the first collection time; continuing to collect m kinds of parameters other than the first characteristic parameter in a unit time after the first collection time; obtaining a complete input data set, and obtaining the collection time of each characteristic parameter in the input data set; inputting the complete input data set and the collection time of each characteristic parameter in the input data set into a trained fault type recognition model in a fixed format, and obtaining a pneumatic hammer fault type label. The accuracy of fault source location is effectively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of fault location, and more specifically, to a method and system for locating a pneumatic hammer fluctuation source in a clean room. Background Art

[0002] A clean room is a special environment with strict control over air quality, temperature, humidity, pressure, etc. It is widely used in semiconductor, biomedicine, food processing and other fields. In a clean room, it is crucial to accurately locate and troubleshoot sources of fluctuations that may cause environmental pollution.

[0003] Prior art, such as the Chinese patent with authorization announcement number CN114544188B, discloses a method for identifying and eliminating vibration fluctuation faults caused by multi-source beat vibrations of aircraft engines. For the situation where the vibration value monitored by the aircraft engine is a periodic sinusoidal fluctuation, the method includes: step one, dividing the vibration value fluctuation mode of the periodic sinusoidal fluctuation into a first mode and a second mode; step two, calculating the speed ratio of the high-pressure and low-pressure rotors when the vibration fluctuation occurs; step three, identifying the beat vibration fault mode according to the speed ratio R; step four, after confirming the beat vibration fault mode, by adjusting the relationship between the high-pressure and low-pressure rotor speeds, changing the speed ratio to eliminate the multi-source coupled beat vibration. The method for identifying and eliminating vibration fluctuation faults caused by multi-source beat vibrations of aircraft engines in this application can realize the rapid identification and accurate positioning of the three modes of multi-source beat vibrations by combining the actual working characteristics of aircraft engines and calculating the size of the high-pressure and low-pressure speed ratios and the vibration fluctuations, and can eliminate the multi-source beat vibration faults of aircraft engines.

[0004] The existing technology is the rapid identification and accurate positioning of the three modes of multi-source vibration; however, it ignores the fact that one fault-affecting parameter abnormality will gradually lead to multiple fault parameter abnormalities, that is, a chain reaction of multiple different types of faults caused by one fault. When identifying and eliminating the fault, it is difficult to accurately locate the source of the fault, resulting in low efficiency in eliminating the fault source. In addition, in a clean room environment, for example, when the air hammer fluctuation source is abnormal, not only will the air hammer be unable to operate normally, but the gas environment in the clean room will also be destroyed. Therefore, for the positioning of this type of fault source, it is necessary to have the characteristic of accurate positioning.

[0005] In view of this, the present invention proposes a method and system for locating a clean room air hammer fluctuation source to solve the above-mentioned problem. Summary of the invention

[0006] In order to overcome the above-mentioned defects of the prior art and to achieve the above-mentioned purpose, the present invention provides the following technical solutions: a method for locating the wave source of a pneumatic hammer in a clean room;

[0007] Collect real-time air hammer operation data in the clean room, which includes n parameters;

[0008] Analyze the n parameters collected in real time to determine whether they are abnormal. If any of the n parameters is an abnormal parameter, mark the parameter as the first characteristic parameter.

[0009] Put the first characteristic parameter performance value into the input data set as the characteristic parameter, obtain the collection time corresponding to the first characteristic parameter, and mark it as the first collection time;

[0010] Continue to collect m parameters other than the first characteristic parameter within a unit time after the first collection time, where m=n-1, and continue to determine the remaining m characteristic parameters in the input data set, and determine the collection time of the m characteristic parameters; obtain a complete input data set, and obtain the collection time of each characteristic parameter in the input data set;

[0011] The complete input data set and the acquisition time of each characteristic parameter in the input data set are input into the trained fault type recognition model in a fixed format to obtain the air hammer fault type label.

[0012] Furthermore, the method of determining the remaining m characteristic parameters in the input data set and determining the acquisition time of the m characteristic parameters includes:

[0013] If there is an abnormal parameter among the m parameters, the abnormal parameter performance value is put into the input data set as a feature parameter, and the collection time corresponding to the abnormal parameter is obtained;

[0014] If no abnormal parameters appear in the m parameters, mark the normal parameters, put the normal parameter performance values ​​collected at the end time node of the unit time into the input data set as feature parameters, and use the end time node of the unit time as the collection time of the normal parameters.

[0015] Furthermore, the method for determining whether a parameter is abnormal includes:

[0016] Preset each parameter threshold interval, and compare and analyze the parameter with the corresponding preset parameter threshold interval;

[0017] If the parameter is within the corresponding preset parameter threshold range, the corresponding parameter is determined to be a normal parameter;

[0018] If the parameter is not within the corresponding preset parameter threshold range, the corresponding parameter is determined to be an abnormal parameter.

[0019] Furthermore, the training method of the fault type identification model includes:

[0020] Collect x groups of historical pneumatic hammer sample data, where x is an integer greater than 1, and the historical pneumatic hammer sample data includes pneumatic hammer characteristic data and pneumatic hammer fault types corresponding to the pneumatic hammer characteristic data; the pneumatic hammer characteristic data includes an input data set and the acquisition time of each characteristic parameter in the input data set, and numerically number each characteristic parameter in the input data set in sequence according to the acquisition time, and set a label for the pneumatic hammer fault type;

[0021] The air hammer feature data is used as the input of the fault type recognition model, the predicted air hammer fault type label corresponding to each group of air hammer feature data is used as the output, and the actual air hammer fault type label corresponding to each group of air hammer feature data is used as the prediction target; minimizing the sum of the prediction errors of all air hammer feature data is used as the training target; the fault type recognition model is trained until the sum of the prediction errors reaches convergence and the training is stopped to obtain the fault type recognition model.

[0022] Furthermore, the fault type identification model is a regression analysis model, a Bayesian network model, a support vector machine model or other suitable models.

[0023] Furthermore, the calculation formula for the prediction error is R x =(α x -μ x ) 2 , where Rx is the prediction error, x is the group number of the feature vector, α x is the actual air hammer fault type label of the xth group of air hammer characteristic data, μ x Predict the air hammer fault type label for the xth set of air hammer feature data.

[0024] Furthermore, the n parameters include sound data, vibration values ​​and temperature values, and the sound data includes sound vibration frequency.

[0025] Furthermore, the sound source filtering module directs the main lobe direction of the directional microphone toward the location of the corresponding air hammer sound source, calculates the incident angle of the sound source through the acoustic collection formula, compares and analyzes the incident angle of the sound source with the preset incident angle range, and determines whether to retain the acquired sound data.

[0026] Further, The acoustic collection formula of a directional microphone is as follows:

[0027] V = K*cos(θ)*P;

[0028] In the formula, V represents the output voltage of the directional microphone sensor, which is collected by the directional microphone sensor; K represents the voltage output corresponding to the unit sound pressure, which is obtained by looking at the directional microphone parameters; θ represents the incident angle of the sound source, that is, the angle between the sound and the main lobe direction of the microphone; P represents the sound pressure, that is, the intensity of the sound pressure, which is collected by the directional microphone sensor;

[0029] The cos(θ) value is obtained through the acoustic acquisition formula to obtain the corresponding incident angle θ of the sound source;

[0030] Compare the incident angle θ of the sound source collected in real time with the preset incident angle range of the directional microphone.

[0031] If the incident angle θ of the sound source is within the preset incident angle range of the directional microphone, the sound data is retained;

[0032] If the incident angle θ of the sound source is not within the preset angle range of the directional microphone, it is an interference sound source and the sound data is discarded.

[0033] A clean room air hammer fluctuation source positioning system implements the clean room air hammer fluctuation source positioning method, comprising:

[0034] A data acquisition module is used to collect real-time air hammer operation data in the clean room, and the real-time air hammer operation data includes n parameters;

[0035] The data processing module analyzes the n parameters collected in real time to determine whether the n parameters are abnormal. If any of the n parameters is an abnormal parameter, the parameter is marked as the first characteristic parameter.

[0036] Put the first characteristic parameter performance value into the input data set as the characteristic parameter, obtain the collection time corresponding to the first characteristic parameter, and mark it as the first collection time;

[0037] Continue to collect m parameters other than the first characteristic parameter within a unit time after the first collection time, where m=n-1, and continue to determine the remaining m characteristic parameters in the input data set, and determine the collection time of the m characteristic parameters; obtain a complete input data set, and obtain the collection time of each characteristic parameter in the input data set;

[0038] The pneumatic hammer fault type identification module inputs the complete input data set and the acquisition time of each characteristic parameter in the input data set into the trained fault type identification model in a fixed format to obtain the pneumatic hammer fault type label.

[0039] Technical effects and advantages of the clean room air hammer wave source positioning method and system of the present invention:

[0040] By determining the first abnormal parameter and the first collection time of the first abnormal parameter, and then determining multiple abnormal parameters according to the unit time after the first collection time, considering the occurrence time of multiple abnormal parameters, and taking multiple abnormal parameters and the corresponding occurrence time as the influencing parameters of the specific fault type, the chain reaction of multiple different types of faults caused by one fault is considered, which effectively improves the accuracy of fault source locating; it is possible to take corresponding maintenance measures in time according to the fault type, improve maintenance efficiency, and avoid serious consequences such as damage to the clean room environment quality, production line interruption or equipment damage. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 This is a schematic diagram of a clean room air hammer wave source positioning system in Example 1 of the present invention;

[0042] Figure 2 Schematic diagram of a clean room air hammer fluctuation source positioning system in Example 2 of the present invention;

[0043] Figure 3 This is a flow chart of the method for locating the air hammer fluctuation source in a clean room in Example 3 of the present invention;

[0044] Figure 4 is a schematic diagram of an electronic device in Embodiment 4 of the present invention;

[0045] Figure 5 This is a schematic diagram of the storage medium in Example 5 of the present invention. DETAILED DESCRIPTION

[0046] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. 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 creative work are within the scope of protection of the present invention.

[0047] Example 1

[0048] A data acquisition module is used to collect real-time air hammer operation data in the clean room, where the real-time air hammer operation data includes n parameters, for example, the n parameters include sound data, vibration value and temperature value, where the sound data includes sound vibration frequency;

[0049] Under normal circumstances, the air hammer will emit a stable and uniform sound. When the air hammer fails or wears out, the noise of the air hammer increases, producing collision sounds or other abnormal sounds, and the vibration frequency of the sound will change. Use a highly sensitive microphone placed close to the moving parts of the air hammer to collect the sound vibration frequency of the air hammer under different operating conditions.

[0050] When the pneumatic hammer is in operation, it usually generates vibrations of specific frequencies and vibration modes. The vibration characteristics of the pneumatic hammer may be different under different operating conditions. A vibration sensor is used to collect the vibration values ​​of the pneumatic hammer under different operating conditions.

[0051] When the hammer parts fail or are damaged, it will cause abnormal temperature rise or fall. By monitoring temperature changes, potential problems or failures can be detected early. Use temperature sensors to collect temperature values ​​under different operating conditions of the hammer.

[0052] The data processing module analyzes the n parameters collected in real time to determine whether the n parameters are abnormal. If any of the n parameters is an abnormal parameter, the parameter is marked as the first characteristic parameter.

[0053] Put the first characteristic parameter performance value into the input data set as the characteristic parameter, obtain the collection time corresponding to the first characteristic parameter, and mark it as the first collection time;

[0054] Continue to collect m parameters other than the first characteristic parameter within a unit time after the first collection time, where m=n-1, and continue to determine the remaining m characteristic parameters in the input data set, and determine the collection time of the m characteristic parameters; obtain a complete input data set, and obtain the collection time of each characteristic parameter in the input data set;

[0055] The method for determining the remaining m feature parameters in the input data set and determining the acquisition time of the m feature parameters includes:

[0056] If there is an abnormal parameter among the m parameters, the abnormal parameter performance value is put into the input data set as a feature parameter, and the collection time corresponding to the abnormal parameter is obtained;

[0057] If there is no abnormal parameter among the m parameters, mark the normal parameter, put the normal parameter performance value collected at the end time node of the unit time into the input data set as the feature parameter, and take the end time node of the unit time as the collection time of the normal parameter;

[0058] To facilitate understanding of the above technical solution, take n parameters including sound vibration frequency, vibration value and temperature value as an example, that is, n is 3:

[0059] When the sound vibration frequency is the first abnormal parameter among the sound vibration frequency, vibration value and temperature value collected in real time, the sound vibration frequency is marked as the first characteristic parameter, the sound vibration frequency performance value is put into the input data set as the characteristic parameter, and the collection time corresponding to the sound vibration frequency is obtained and marked as the first collection time;

[0060] If there is an abnormal parameter among the m parameters, that is, the vibration value or temperature value per unit time is abnormal, the abnormal vibration value or temperature value is put into the input data set as a feature parameter, and the collection time corresponding to the abnormal vibration value or temperature value is obtained;

[0061] If no abnormal parameter appears in the m parameters, mark the normal parameter, for example, the temperature value or vibration value has never been abnormal, then the temperature value or vibration value collected at the end time node of the unit time is put into the input data set as the characteristic parameter, and the end time node of the unit time is used as the collection time of the temperature value or vibration value;

[0062] Methods for determining whether a parameter is abnormal include:

[0063] Preset each parameter threshold interval, and compare and analyze the parameter with the corresponding preset parameter threshold interval;

[0064] If the parameter is within the corresponding preset parameter threshold range, the corresponding parameter is determined to be a normal parameter;

[0065] If the parameter is not within the corresponding preset parameter threshold range, the corresponding parameter is determined to be an abnormal parameter;

[0066] Examples of preset threshold ranges for each parameter are as follows:

[0067] For example, if the n parameters are sound vibration frequency, vibration value and temperature value, then the vibration frequency threshold interval, vibration threshold interval and temperature threshold interval are preset. When comparing, the sound vibration frequency is compared and analyzed with the preset vibration frequency threshold, the vibration value is compared and analyzed with the preset vibration threshold interval, and the temperature value is compared and analyzed with the preset temperature threshold interval;

[0068] The pneumatic hammer fault type identification module inputs the complete input data set and the acquisition time of each characteristic parameter in the input data set into the trained fault type identification model in a fixed format to obtain the pneumatic hammer fault type label;

[0069] This embodiment considers n types of parameter anomalies and the order of their occurrence times as inputs to the fault type identification model. The reason is that:

[0070] Taking the n parameters as sound vibration frequency, vibration value and temperature value, and the pneumatic hammer failure types including wear of pneumatic hammer internal parts, loose pneumatic hammer parts and pneumatic hammer overload as an example;

[0071] For example, the internal parts of the air hammer are worn out. The internal parts of the air hammer include pistons, sealing rings and other parts. If the pistons, sealing rings and other parts are worn out due to long-term use or poor working conditions, the vibration value of the air hammer will increase first, then the sound value will increase abnormally, and finally the temperature will increase abnormally;

[0072] The air hammer parts are loose. The air hammer parts include the pipe connectors, brackets and fixings of the air hammer. Pipe connectors such as threaded joints or flange connections are loose or fall off during operation, which will cause the air hammer to first have abnormal sound values ​​and then abnormally increase vibration values. It has little effect on the temperature in a short time, that is, the temperature is normal;

[0073] Pneumatic hammer overload refers to the overload of the pneumatic hammer due to long-term or frequent use of the pneumatic hammer, which will cause the temperature of the pneumatic hammer to rise abnormally first, then the vibration value to increase abnormally, and finally the sound value to increase abnormally;

[0074] The training method of the fault type recognition model includes:

[0075] Collect x groups of historical pneumatic hammer sample data, where x is an integer greater than 1, and the historical pneumatic hammer sample data includes pneumatic hammer characteristic data and pneumatic hammer fault types corresponding to the pneumatic hammer characteristic data; the pneumatic hammer characteristic data includes an input data set and the acquisition time of each characteristic parameter in the input data set, and each characteristic parameter in the input data set is numerically numbered in sequence according to the acquisition time, and labels are set for the pneumatic hammer fault types, and the pneumatic hammer fault types include pneumatic hammer internal component wear, pneumatic hammer component loosening, pneumatic hammer overload, and pneumatic hammer seal failure; the pneumatic hammer internal component wear label is set to 1, the pneumatic hammer component loosening label is set to 2, the pneumatic hammer overload label is set to 3, and the pneumatic hammer seal failure label is set to 4;

[0076] The air hammer feature data is used as the input of the fault type recognition model, the predicted air hammer fault type label corresponding to each group of air hammer feature data is used as the output, and the actual air hammer fault type label corresponding to each group of air hammer feature data is used as the prediction target; minimizing the sum of the prediction errors of all air hammer feature data is used as the training target; the fault type recognition model is trained until the sum of the prediction errors reaches convergence and the training is stopped.

[0077] The calculation formula of the prediction error is R x =(α x -μ x ) 2 , where Rx is the prediction error, x is the group number of the feature vector, α x is the actual air hammer fault type label of the xth group of air hammer characteristic data, μ x Predict the air hammer fault type label for the xth group of air hammer feature data;

[0078] The fault type identification model is a regression analysis model, a Bayesian network model, a support vector machine model or other suitable models.

[0079] This embodiment determines the first abnormal parameter and the first collection time of the first abnormal parameter, and then determines multiple abnormal parameters according to the unit time after the first collection time, considers the occurrence time of multiple abnormal parameters, and uses the multiple abnormal parameters and the corresponding occurrence time as influencing parameters of the specific fault type, taking into account the chain reaction of multiple different types of faults caused by one fault, effectively improving the accuracy of fault source locating; it can take corresponding maintenance measures according to the fault type in time, improve maintenance efficiency, and avoid serious consequences such as damage to the clean room environment quality, production line interruption or equipment damage.

[0080] Example 2

[0081] This embodiment further improves the design on the basis of implementation case 1. In actual use, there is more than one air hammer working in the clean room, and each air hammer has a fault type identification module. When one air hammer fails, the fault type identification modules of other groups can also monitor the air hammer failure sound. In order to identify which air hammer has failed, this embodiment provides a clean room air hammer fluctuation source positioning system to solve the above problem. The positioning system also includes a sound source filtering module, please refer to Figure 2 As shown;

[0082] The sound source filtering module directs the main lobe direction of the directional microphone toward the location of the corresponding air hammer sound source, calculates the incident angle of the sound source through the acoustic acquisition formula, compares and analyzes the incident angle of the sound source with the preset incident angle range, and determines whether the incident angle of the sound source is within the preset incident angle range;

[0083] Make sure the microphone is at a moderate distance from the hammer sound source to achieve the best effect; adjust the direction of the directional microphone. For example, the main lobe of the cardioid microphone is in the front, and the sound from the back and sides will be suppressed, which can better suppress the interfering sound sources caused by other hammer failures and other environmental noises.

[0084] The acoustic collection formula of a directional microphone is as follows:

[0085] V = K*cos(θ)*P;

[0086] In the formula, V represents the output voltage of the directional microphone sensor, which is collected by the directional microphone sensor; K represents the sensitivity of the microphone, which is a constant representing the voltage output corresponding to the unit sound pressure, which is obtained by looking at the directional microphone parameters; θ represents the incident angle of the sound source, that is, the angle between the sound and the main lobe direction of the microphone; P represents the sound pressure, that is, the intensity of the sound pressure, which is collected by the directional microphone sensor;

[0087] According to the formula, the response characteristics of the directional microphone can be obtained. When the sound source is located in the main lobe direction of the directional microphone, the output voltage of the sensor reaches the maximum value; as the sound source deviates from the main lobe direction, the output voltage of the sensor gradually decreases until it reaches zero.

[0088] When the sound pressure P is too large, the cos(θ) value corresponding to the incident angle is small, that is, the sound source deviates from the microphone collection direction, and the interference sound source that deviates from the microphone collection direction can also be collected. At this time, V, K and P in the formula are all known values; the cos(θ) value can be obtained through the formula to obtain the corresponding incident angle θ of the sound source.

[0089] The incident angle θ of the sound source collected in real time is compared with the preset incident angle range of the directional microphone. If the incident angle θ of the sound source is within the preset incident angle range of the directional microphone, the sound data is retained. If the incident angle θ of the sound source is not within the preset angle range of the directional microphone, it is an interference sound source and the sound data is discarded.

[0090] The directional microphone has a preset incident angle range. In the absence of other noise interference, the directional microphone collects the sound source emitted by the air hammer in this direction multiple times, and obtains a set of θ values. The minimum θ value in the θ value set is used as the minimum angle value of the incident angle range, and the maximum θ value in the θ value set is used as the maximum angle value of the incident angle range.

[0091] This embodiment uses a directional microphone to collect the sound source emitted by the air hammer in this direction to obtain a set of θ values, and uses the minimum θ value in the θ value set as the minimum angle value of the incident angle range, and uses the maximum θ value in the θ value set as the maximum angle value of the incident angle range.

[0092] The hammer sound source effectively filters out background noise, making the voice signal clearer, while also avoiding the fault identification of other groups caused by a pneumatic hammer fault noise in the clean room; it helps to improve the accuracy and stability of the clean room pneumatic hammer fluctuation source positioning system.

[0093] Example 3

[0094] See also Figure 3 As shown, the part not described in detail in this embodiment is described in Example 1, and a method for locating a source of air hammer fluctuation in a clean room is provided, the method comprising:

[0095] Collect real-time air hammer operation data in the clean room, which includes n parameters;

[0096] Analyze the n parameters collected in real time to determine whether they are abnormal. If any of the n parameters is an abnormal parameter, mark the parameter as the first characteristic parameter.

[0097] Put the first characteristic parameter performance value into the input data set as the characteristic parameter, obtain the collection time corresponding to the first characteristic parameter, and mark it as the first collection time;

[0098] Continue to collect m parameters other than the first characteristic parameter within a unit time after the first collection time, where m=n-1, and continue to determine the remaining m characteristic parameters in the input data set, and determine the collection time of the m characteristic parameters; obtain a complete input data set, and obtain the collection time of each characteristic parameter in the input data set;

[0099] The complete input data set and the acquisition time of each characteristic parameter in the input data set are input into the trained fault type recognition model in a fixed format to obtain the air hammer fault type label.

[0100] Furthermore, the method of determining the remaining m characteristic parameters in the input data set and determining the acquisition time of the m characteristic parameters includes:

[0101] If there is an abnormal parameter among the m parameters, the abnormal parameter performance value is put into the input data set as a feature parameter, and the collection time corresponding to the abnormal parameter is obtained;

[0102] If no abnormal parameters appear in the m parameters, mark the normal parameters, put the normal parameter performance values ​​collected at the end time node of the unit time into the input data set as feature parameters, and use the end time node of the unit time as the collection time of the normal parameters.

[0103] Furthermore, the method for determining whether a parameter is abnormal includes:

[0104] Preset each parameter threshold interval, and compare and analyze the parameter with the corresponding preset parameter threshold interval;

[0105] If the parameter is within the corresponding preset parameter threshold range, the corresponding parameter is determined to be a normal parameter;

[0106] If the parameter is not within the corresponding preset parameter threshold range, the corresponding parameter is determined to be an abnormal parameter.

[0107] Furthermore, the training method of the fault type identification model includes:

[0108] Collect x groups of historical pneumatic hammer sample data, where x is an integer greater than 1, and the historical pneumatic hammer sample data includes pneumatic hammer characteristic data and pneumatic hammer fault types corresponding to the pneumatic hammer characteristic data; the pneumatic hammer characteristic data includes an input data set and the acquisition time of each characteristic parameter in the input data set, and numerically number each characteristic parameter in the input data set in sequence according to the acquisition time, and set a label for the pneumatic hammer fault type;

[0109] The air hammer feature data is used as the input of the fault type recognition model, the predicted air hammer fault type label corresponding to each group of air hammer feature data is used as the output, and the actual air hammer fault type label corresponding to each group of air hammer feature data is used as the prediction target; minimizing the sum of the prediction errors of all air hammer feature data is used as the training target; the fault type recognition model is trained until the sum of the prediction errors reaches convergence and the training is stopped to obtain the fault type recognition model.

[0110] Furthermore, the fault type identification model is a regression analysis model, a Bayesian network model, a support vector machine model or other suitable models.

[0111] Furthermore, the calculation formula for the prediction error is R x =(α x -μ x ) 2 , where Rx is the prediction error, x is the group number of the feature vector, α x is the actual air hammer fault type label of the xth group of air hammer characteristic data, μ x Predict the air hammer fault type label for the xth set of air hammer feature data.

[0112] Furthermore, the n parameters include sound data, vibration values ​​and temperature values, and the sound data includes sound vibration frequency.

[0113] Collect real-time air hammer operation data in the clean room, which includes n parameters;

[0114] Analyze the n parameters collected in real time to determine whether the n parameters are abnormal. If any one of the n parameters is an abnormal parameter, mark the parameter as the first characteristic parameter.

[0115] Put the first characteristic parameter performance value into the input data set as the characteristic parameter, obtain the collection time corresponding to the first characteristic parameter, and mark it as the first collection time;

[0116] Continue to collect m parameters other than the first characteristic parameter within a unit time after the first collection time, where m=n-1, and continue to determine the remaining m characteristic parameters in the input data set, and determine the collection time of the m characteristic parameters; obtain a complete input data set, and obtain the collection time of each characteristic parameter in the input data set;

[0117] The complete input data set and the acquisition time of each characteristic parameter in the input data set are input into the trained fault type recognition model in a fixed format to obtain the air hammer fault type label.

[0118] Example 4

[0119] See also Figure 4As shown, according to another aspect of the present application, an electronic device 500 is further provided. The electronic device 500 may include one or more processors and one or more memories. The memories store computer readable codes, and when the computer readable codes are executed by one or more processors, the clean room air hammer fluctuation source locating method as described above can be executed.

[0120] The method or system according to the embodiment of the present application can also be used by Figure 4 The electronic device architecture shown in FIG. Figure 4 As shown, the electronic device 500 may include a bus 501, one or more CPUs 502, a read-only memory (ROM) 503, a random access memory (RAM) 504, a communication port 505 connected to a network, an input / output 506, a hard disk 507, etc. The storage device in the electronic device 500, such as the ROM 503 or the hard disk 507, may store the clean room air hammer fluctuation source positioning method provided in the present application. Furthermore, the electronic device 500 may also include a user interface 508. Of course, Figure 4 The architecture shown is only exemplary and can be omitted according to actual needs when implementing different devices. Figure 4 One or more components of an electronic device are shown.

[0121] Example 5

[0122] See also Figure 5 As shown, a computer-readable storage medium 600 according to an embodiment of the present application is shown. Computer-readable instructions are stored on the computer-readable storage medium 600. When the computer-readable instructions are executed by the processor, the method for locating the source of air hammer fluctuations in a clean room according to an embodiment of the present application described with reference to the above figures can be executed. The storage medium 600 includes, but is not limited to, for example, volatile memory and / or non-volatile memory. Volatile memory may include, for example, random access memory (RAM) and cache memory (cache), etc. Non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc.

[0123] In addition, according to the implementation of the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the present application provides a non-transitory machine-readable storage medium, which stores machine-readable instructions, and the machine-readable instructions can be executed by a processor to execute instructions corresponding to the method steps provided in the present application, for example: an industrial control visual motion control method based on AI visual analysis. When the computer program is executed by a central processing unit (CPU), the above functions defined in the method of the present application are executed.

[0124] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present invention is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website site, computer, server or data center to another website site, computer, server or data center through a wired network or a wireless network. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (for example, a floppy disk, a hard disk, a tape), an optical medium (for example, a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.

[0125] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed in the present invention can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0126] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0127] In the several embodiments provided by the present invention, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of the units is only one, and there may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0128] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0129] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0130] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

[0131] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for locating the source of air hammer fluctuations in a clean room, characterized by: Collect real-time air hammer operation data in the clean room, which includes n parameters; the n parameters include sound data, vibration value and temperature value, and the sound data includes sound vibration frequency; Analyze the n parameters collected in real time to determine whether the n parameters are abnormal. If any one of the n parameters is an abnormal parameter, mark the parameter as the first characteristic parameter. Put the first characteristic parameter performance value into the input data set as the characteristic parameter, obtain the collection time corresponding to the first characteristic parameter, and mark it as the first collection time; Continue to collect m parameters other than the first characteristic parameter within a unit time after the first collection time, m=n-1, and continue to determine the remaining m characteristic parameters in the input data set, as well as the collection time of the m characteristic parameters; Obtain a complete input data set, and obtain the acquisition time of each characteristic parameter in the input data set; The method for determining the remaining m feature parameters in the input data set and determining the acquisition time of the m feature parameters includes: If there is an abnormal parameter among the m parameters, the abnormal parameter performance value is put into the input data set as a feature parameter, and the collection time corresponding to the abnormal parameter is obtained; If there is no abnormal parameter among the m parameters, mark the normal parameter, put the normal parameter performance value collected at the end time node of the unit time into the input data set as the feature parameter, and take the end time node of the unit time as the collection time of the normal parameter; The complete input data set and the acquisition time of each characteristic parameter in the input data set are input into the trained fault type recognition model in a fixed format to obtain the air hammer fault type label. The air hammer fault types include wear of the air hammer internal parts, loose air hammer parts and air hammer overload. The training method of the fault type identification model includes: Collect x groups of historical pneumatic hammer sample data, where x is an integer greater than 1, and the historical pneumatic hammer sample data includes pneumatic hammer characteristic data and pneumatic hammer fault types corresponding to the pneumatic hammer characteristic data; the pneumatic hammer characteristic data includes an input data set and the acquisition time of each characteristic parameter in the input data set, and numerically number each characteristic parameter in the input data set in sequence according to the acquisition time, and set a label for the pneumatic hammer fault type; The air hammer feature data is used as the input of the fault type recognition model, the predicted air hammer fault type label corresponding to each group of air hammer feature data is used as the output, and the actual air hammer fault type label corresponding to each group of air hammer feature data is used as the prediction target; minimizing the sum of the prediction errors of all air hammer feature data is used as the training target; the fault type recognition model is trained until the sum of the prediction errors reaches convergence and the training is stopped to obtain the fault type recognition model.

2. The method for locating the source of air hammer fluctuation in a clean room according to claim 1, characterized in that: The method for determining whether the parameter is abnormal includes: Preset each parameter threshold interval, and compare and analyze the parameter with the corresponding preset parameter threshold interval; If the parameter is within the corresponding preset parameter threshold range, the corresponding parameter is determined to be a normal parameter; If the parameter is not within the corresponding preset parameter threshold range, the corresponding parameter is determined to be an abnormal parameter.

3. The method for locating the source of air hammer fluctuation in a clean room according to claim 2, characterized in that: The fault type identification model is a regression analysis model, a Bayesian network model or a support vector machine model.

4. The method for locating the source of air hammer fluctuation in a clean room according to claim 3, characterized in that: The calculation formula for the prediction error is: ,in is the prediction error, x is the group number of the feature vector, The actual air hammer fault type label of the xth group of air hammer characteristic data, Predict the air hammer fault type label for the xth set of air hammer feature data.

5. The method for locating the source of air hammer fluctuation in a clean room according to claim 4, characterized in that: Point the main lobe of the directional microphone toward the location of the corresponding air hammer sound source, calculate the incident angle of the sound source through the acoustic collection formula, compare and analyze the incident angle of the sound source with the preset incident angle range, and determine whether to retain the acquired sound data.

6. The method for locating the source of air hammer fluctuation in a clean room according to claim 5, characterized in that: The acoustic collection formula of a directional microphone is as follows: V = K * cos(θ) * P; In the formula, V represents the output voltage of the directional microphone sensor, which is collected by the directional microphone sensor; K represents the voltage output corresponding to the unit sound pressure, which is obtained by looking at the directional microphone parameters; θ represents the incident angle of the sound source, that is, the angle between the sound and the main lobe direction of the microphone; P represents the sound pressure, that is, the intensity of the sound pressure, which is collected by the directional microphone sensor; The cos(θ) value is obtained through the acoustic acquisition formula, thereby obtaining the corresponding incident angle θ of the sound source; The incident angle θ of the sound source collected in real time is compared with the incident angle range preset by the directional microphone. If the incident angle θ of the sound source is within the incident angle range preset by the directional microphone, the sound data is retained; If the incident angle of the sound source If the sound is not within the preset angle range of the directional microphone, it is an interference sound source and the sound data is discarded.

7. A clean room air hammer fluctuation source positioning system, implementing the clean room air hammer fluctuation source positioning method according to any one of claims 1 to 6, characterized in that: include: A data acquisition module is used to collect real-time air hammer operation data in the clean room, and the real-time air hammer operation data includes n parameters; The n parameters include sound data, vibration values ​​and temperature values, and the sound data includes the sound vibration frequency; The first data processing module analyzes the n parameters collected in real time to determine whether the n parameters are abnormal. If any one of the n parameters is an abnormal parameter, the parameter is marked as the first characteristic parameter. Put the first characteristic parameter performance value into the input data set as the characteristic parameter, obtain the collection time corresponding to the first characteristic parameter, and mark it as the first collection time; Continue to collect m parameters other than the first characteristic parameter within a unit time after the first collection time, m=n-1, and continue to determine the remaining m characteristic parameters in the input data set, as well as the collection time of the m characteristic parameters; Obtain a complete input data set, and obtain the acquisition time of each characteristic parameter in the input data set; The pneumatic hammer fault type identification module inputs the complete input data set and the acquisition time of each characteristic parameter in the input data set into the trained fault type identification model in a fixed format to obtain the pneumatic hammer fault type label.

Images