Operation machinery intelligent identification method for external force damage prevention scene

Abstract

The invention relates to the technical field of engineering construction safety monitoring, in particular to a working machinery intelligent identification method for an external damage prevention scene, and the method comprises the steps: obtaining a sound spectrum characteristic quantity through a voiceprint sensor, and generating a voiceprint recognition model; importing the voiceprint recognition model into an original model database for comparison; whether the voiceprint recognition model belongs to an early warning model or a safety model is judged according to the comparison result, and the judgment result is updated to the original model database; according to the vibration induction value, whether a first sound-light alarm is sent to a machine corresponding to the early warning model is judged; judging whether to send a second sound-light alarm or not according to the laser receiving condition of the laser sensor; the voiceprint sensor is used for obtaining the sound spectrum characteristic quantity and generating the voiceprint recognition model according to the sound spectrum characteristic quantity, the sound characteristics are effectively extracted, the sound comparison accuracy is improved, and meanwhile the accuracy and stability of intelligent recognition of the operation machine are effectively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of engineering construction safety monitoring, and in particular to an intelligent identification method for operating machinery in scenarios of protection against external force damage. Background Art

[0002] As the urban process continues to accelerate, the contradiction between the construction of urban upgrading and reconstruction projects and the requirements for the distribution network to prevent external force damage has become increasingly prominent. At present, around the task of preventing external force damage, urban construction sites are inspected and monitored through manual inspections, safety briefings, installation of surveillance cameras, and setting up tilt monitoring identification piles. At present, the monitoring method for external force damage is inefficient and labor-intensive, and it is impossible to quickly identify and warn of the risk of external force damage within the power corridor. Based on the voiceprint extraction algorithm, this project extracts the sound spectrum characteristics of different operating machinery from the collected mixed sounds on site. The original voiceprint recognition model is trained using the sound spectrum characteristics to generate voiceprint recognition models for different operating machinery, and a typical database of external force damage is constructed to store the generated voiceprint recognition model. Accurate identification of operating machinery is achieved, and external force damage risk warning signals are promptly issued to operation and maintenance personnel for the identified operating machinery.

[0003] Chinese Patent Authorization Announcement No.: CN117672266B discloses a voiceprint recognition method based on a DCN deep cross network. The invention provides a voiceprint recognition method based on a DCN deep cross network in the field of engineering construction safety monitoring technology, including: step S1, obtaining a large amount of historical voiceprint data during the engineering construction process, building a data set based on each historical voiceprint data, and dividing the data set into a training set, a verification set and a test set; step S2, creating a voiceprint recognition model based on DCN; step S3, training the voiceprint recognition model through the training set, and verifying the trained voiceprint recognition model through the verification set; step S4, testing the verified voiceprint recognition model through the test set, and continuously optimizing the loss function and hyperparameters of the voiceprint recognition model; step S5, collecting the voiceprint analog signal emitted by the construction machinery from the engineering construction site through the voiceprint sensor array, converting the voiceprint analog signal into a voiceprint digital signal and inputting it into the tested voiceprint recognition model to perform voiceprint recognition. The advantage of this invention is that it greatly improves the accuracy of voiceprint recognition.

[0004] Chinese Patent Publication No.: CN118173101A discloses a method for identifying the risk of external force damage based on voiceprint. The invention discloses a method for identifying the risk of external force damage based on voiceprint, comprising the following steps: Step S1, collecting the voiceprint data generated during the operation of the transformer, and marking the risk type of the collected voiceprint data; Step S2, expanding the voiceprint data collected in step S1 based on the adversarial generative network model of the optimal architecture search; Step S3, extracting the original feature representation of the voiceprint data in step S2 based on the neural network model of the multi-granularity honeypot optimization algorithm; Step S4, reconstructing the voiceprint data representation in step S3 based on the autoencoder of the potential Dirichlet allocation, and outputting the reconstructed feature representation; Step S5, classifying the reconstructed feature representation in step S4 based on the classifier of the difference branch random forest, and determining the risk type of external force risk. The invention is applied to the identification of the risk of external force damage to the transformer. By analyzing the sound data of the construction machinery, it can effectively identify activities that may pose a threat to the power facilities.

[0005] However, the above method has the following problems: it processes and identifies voiceprints in a single way, cannot make accurate judgments on the complex noise at the construction site, and does not use a variety of detection instruments to effectively confirm the sounds emitted by suspected destructive machinery. Summary of the invention

[0006] To this end, the present invention provides an intelligent identification method for operating machinery in scenarios of preventing external force damage, so as to overcome the problems in the prior art of single processing and identification of voiceprints, inability to accurately judge the complex noise at the construction site, and failure to use multiple detection instruments to effectively confirm the sounds emitted by suspected destructive machinery.

[0007] To achieve the above object, the present invention provides an intelligent identification method for operating machinery in a scenario of preventing external force damage, comprising:

[0008] Determine whether to turn on the voiceprint sensor and ground vibration sensor based on the ambient sound intensity value;

[0009] Using the voiceprint sensor to obtain a sound spectrum feature quantity and generating a voiceprint recognition model according to the sound spectrum feature quantity;

[0010] Importing the voiceprint recognition model into the original model database and comparing it with the warning model and the security model in turn to obtain a warning voiceprint feature similarity value and a security voiceprint feature similarity value;

[0011] Determine whether the voiceprint recognition model belongs to the warning model or the first comparison model according to the warning voiceprint feature similarity value, and determine whether the first comparison model belongs to the safety model or the second comparison model according to the safety voiceprint feature similarity value;

[0012] Determine whether the second comparison model belongs to the warning model or the safety model according to the laser sensor receiving laser light, and update the determination result to the original model database;

[0013] Using the ground vibration sensor to obtain a vibration sensing value, determining whether to issue a first sound and light alarm to the machine corresponding to the early warning model according to the vibration sensing value, determining whether to issue a second sound and light alarm according to the laser receiving condition of the laser sensor, and capturing an image of the sound source object through a plurality of cameras;

[0014] Among them, the sound spectrum characteristics include frequency, period, wavelength and sound speed; the original model database includes the early warning model and the safety model; the early warning model is the sound model of a machine that can cause damage to overhead cables; the safety model is the sound model of an object that will not cause damage to overhead cables; the voiceprint sensor, the ground vibration sensor and the laser sensor are set one by one.

[0015] Further, whether to turn on the voiceprint sensor and the ground vibration sensor is determined according to the ambient sound intensity value, wherein:

[0016] If the ambient sound intensity value is greater than or equal to the ambient sound intensity preset value, turning on the voiceprint sensor and the ground vibration sensor;

[0017] Among them, the preset value of the ambient sound intensity is negatively correlated with the straight-line distance from the working area of ​​the operating machinery to the voiceprint sensor.

[0018] Furthermore, the steps of using the voiceprint sensor to obtain the sound spectrum feature quantity and generating a voiceprint recognition model according to the sound spectrum feature quantity include:

[0019] The voiceprint sensor extracts the sound spectrum characteristics of different operating machines from the acquired environmental sounds based on the voiceprint extraction algorithm;

[0020] The voiceprint sensor generates voiceprint recognition models of different operating machines using the sound spectrum feature quantity.

[0021] Further, judging whether the voiceprint recognition model belongs to the early warning model according to the early warning voiceprint feature similarity value includes:

[0022] If the warning voiceprint feature similarity value is greater than or equal to a preset warning similarity value, the voiceprint recognition model is determined to be a warning model;

[0023] If the warning voiceprint feature similarity value is less than the preset warning similarity value, determining that the voiceprint recognition model is the first comparison model;

[0024] Among them, the preset warning similarity value is negatively correlated with the straight-line distance from the working area of ​​the operating machinery to the voiceprint sensor.

[0025] Further, determining whether the first comparison model belongs to the security model according to the security voiceprint feature similarity value includes:

[0026] If the security voiceprint feature similarity value is greater than or equal to a preset security similarity value, determining that the first comparison model is the security model;

[0027] If the security voiceprint feature similarity value is less than a preset security similarity value, determining that the first comparison model is the second comparison model;

[0028] Among them, the preset safety similarity value is negatively correlated with the straight-line distance from the working area of ​​the operating machinery to the voiceprint sensor.

[0029] Further, when it is determined that the first comparison model is the second comparison model, it is determined that the second comparison model belongs to the warning model or the safety model according to the laser sensor receiving laser, wherein:

[0030] If the laser sensor normally receives the laser emitted by the adjacent laser sensor, it is determined that the second comparison model belongs to the safety model;

[0031] If the laser sensor does not receive the laser emitted by the adjacent laser sensor, it is determined that the second comparison model belongs to the early warning model.

[0032] Furthermore, the determination result based on the laser receiving condition of the laser sensor is updated to the original model database.

[0033] Further, when the voiceprint recognition model is determined to be an early warning model, the ground vibration sensor is used to obtain the vibration sensing value, and it is determined according to the vibration sensing value whether to issue a first sound and light alarm to the machine corresponding to the early warning model, wherein:

[0034] If the vibration sensing value is less than the preset vibration sensing value, it is determined that the first sound and light alarm is not issued;

[0035] If the vibration sensing value is not less than the preset vibration sensing value, it is determined that the first sound and light alarm is issued to the machine corresponding to the early warning model;

[0036] The preset vibration sensing value is positively correlated with the straight-line distance from the working area of ​​the operating machinery to the soundprint sensor.

[0037] Further, determining whether to issue a second sound and light alarm according to the laser sensor receiving laser light includes:

[0038] If the laser sensor normally receives the laser emitted by the adjacent laser sensor, the second sound and light alarm will not be issued;

[0039] If the laser sensor does not receive the laser emitted by the adjacent laser sensor, a second sound and light alarm is issued.

[0040] Furthermore, when the first sound and light alarm and / or the second sound and light alarm is issued, the camera captures the action of shooting the sound source object to generate image information and transmits the image information to the central control room.

[0041] Compared with the prior art, the beneficial effect of the present invention lies in that a voiceprint sensor is used to obtain sound spectrum characteristics and a voiceprint recognition model is generated based on the sound spectrum characteristics, thereby effectively extracting the characteristics of the sound and clearly distinguishing different sounds. This improves the accuracy of sound comparison while effectively improving the accuracy and stability of intelligent recognition of operating machinery.

[0042] Furthermore, by comparing the voiceprint recognition model with the early warning model and the safety model, the voiceprint recognition model is judged and classified in a targeted manner. While accurately determining whether the operating machinery will cause harm to the overhead cables, the accuracy and stability of the intelligent identification of the operating machinery are further improved.

[0043] Furthermore, by using laser sensors to detect the movements of operating machinery, better early warning can be given before hazards occur. While effectively improving the safety of overhead cables, it further improves the accuracy and stability of intelligent identification of operating machinery.

[0044] Furthermore, whether to turn on the voiceprint sensor and the ground vibration sensor is determined based on the ambient sound intensity value, which effectively reduces the equipment utilization rate and effectively increases the service life of the equipment. While effectively reducing energy consumption, it further improves the accuracy and stability of intelligent identification of operating machinery. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 It is a flow chart of the intelligent identification method of working machinery for the scenario of preventing external force damage according to the present invention;

[0046] Figure 2 A flowchart of early warning model and safety model determination according to an embodiment of the present invention;

[0047] Figure 3 This is a flow chart of the first sound and light alarm determination according to an embodiment of the present invention;

[0048] Figure 4 This is a flow chart of the second sound and light alarm determination according to an embodiment of the present invention. DETAILED DESCRIPTION

[0049] In order to make the objects and advantages of the present invention more clearly understood, the present invention is further described below in conjunction with embodiments; it should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0050] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only used to explain the technical principles of the present invention and are not intended to limit the protection scope of the present invention.

[0051] It should be noted that, in the description of the present invention, terms such as "up", "down", "left", "right", "inside" and "outside" indicating directions or positional relationships are based on the directions or positional relationships shown in the drawings. This is merely for the convenience of description and does not indicate or imply that the device or element must have a specific orientation, be constructed and operated in a specific orientation. Therefore, it cannot be understood as a limitation on the present invention.

[0052] In addition, it should be noted that in the description of the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For those skilled in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0053] See also Figure 1 As shown, it is a flow chart of the method for intelligently identifying a working machine for a scene of preventing external force damage according to the present invention. The method for intelligently identifying a working machine for a scene of preventing external force damage comprises:

[0054] Step S1, judging whether to turn on the voiceprint sensor and the ground vibration sensor according to the ambient sound intensity value;

[0055] Step S2, using the voiceprint sensor to obtain a sound spectrum feature quantity and generating a voiceprint recognition model according to the sound spectrum feature quantity;

[0056] Step S3, importing the voiceprint recognition model into the original model database and comparing it with the warning model and the security model in turn to obtain the warning voiceprint feature similarity value and the security voiceprint feature similarity value;

[0057] Step S4, determining whether the voiceprint recognition model belongs to the warning model or the first comparison model according to the warning voiceprint feature similarity value, and determining whether the first comparison model belongs to the safety model or the second comparison model according to the safety voiceprint feature similarity value;

[0058] Step S5, determining whether the second comparison model belongs to the warning model or the safety model according to the laser receiving situation of the laser sensor and updating the determination result to the original model database;

[0059] Step S6, using the ground vibration sensor to obtain a vibration sensing value, determining whether to issue a first sound and light alarm to the machine corresponding to the early warning model according to the vibration sensing value, determining whether to issue a second sound and light alarm according to the laser receiving condition of the laser sensor, and capturing an image of the sound source object by using a plurality of cameras;

[0060] Among them, the sound spectrum characteristics include frequency, period, wavelength and sound speed; the original model database includes early warning model and safety model; the early warning model is the sound model of machinery that can cause damage to overhead cables; the safety model is the sound model of objects that will not cause damage to overhead cables; the voiceprint sensor, ground vibration sensor and laser sensor are set one by one.

[0061] It is understandable that, for those skilled in the art, it is obvious to use a sound level meter or other equipment to measure the ambient sound intensity value, which will not be described in detail here.

[0062] Specifically, in step S1, it is determined whether to turn on the voiceprint sensor and the ground vibration sensor according to the ambient sound intensity value, wherein:

[0063] If the ambient sound intensity value is greater than or equal to the ambient sound intensity preset value, the voiceprint sensor and the ground vibration sensor are turned on;

[0064] Example 1: In this example, the preset value of the ambient sound intensity is 80 decibels. If the measured ambient sound intensity is 100 decibels, which is greater than the preset value of 80 decibels, the voiceprint sensor and the ground vibration sensor are turned on.

[0065] Among them, the preset value of the ambient sound intensity is negatively correlated with the straight-line distance from the working area of ​​the operating machinery to the voiceprint sensor.

[0066] In one embodiment, the straight-line distance from the working area of ​​the operating machine to the soundprint sensor is 50 meters, and the preset value of the ambient sound intensity is 80 decibels;

[0067] The straight-line distance from the working area of ​​the operating machinery to the soundprint sensor is 100 meters, and the preset value of the ambient sound intensity is 70 decibels;

[0068] The straight-line distance from the working area of ​​the operating machinery to the soundprint sensor is 150 meters, and the preset value of the ambient sound intensity is 60 decibels.

[0069] It can be understood that the closer the straight-line distance from the working area of ​​the operating machinery to the voiceprint sensor, the greater the sound intensity of the mechanical construction received by the voiceprint sensor, and the greater the preset value of the ambient sound intensity.

[0070] Specifically, whether to turn on the voiceprint sensor and ground vibration sensor is determined based on the ambient sound intensity value, which effectively reduces the equipment utilization rate and effectively increases the service life of the equipment. While effectively reducing energy consumption, it further improves the accuracy and stability of intelligent identification of operating machinery.

[0071] Specifically, in step S2, the steps of using a voiceprint sensor to obtain a sound spectrum feature quantity and generating a voiceprint recognition model according to the sound spectrum feature quantity include:

[0072] Step S21, the voiceprint sensor extracts the sound spectrum characteristics of different operating machines from the acquired environmental sounds based on the voiceprint extraction algorithm;

[0073] In step S22, the voiceprint sensor generates voiceprint recognition models of different operating machines using the sound spectrum feature quantity.

[0074] Embodiment 2: Based on Embodiment 1, this embodiment differs from it in that the frequency of the sound acquired by the voiceprint sensor is 80 Hz, the period is 0.5 seconds, the wavelength is 0.5 meters and the speed of sound is 340 meters per second.

[0075] It can be understood that the voiceprint extraction algorithm includes speech signal preprocessing, feature parameter extraction and model generation.

[0076] It can be understood that speech signal preprocessing includes filtering, A / D conversion, pre-emphasis, framing, endpoint detection, etc., and the sample amplitude is limited to a certain range through normalization processing.

[0077] It is understandable that the feature parameter extraction uses methods such as Mel Frequency Cepstral Coefficient (MFCC), including steps such as Fourier transform, Mel filter and discrete cosine transform, to extract parameters that can represent various mechanical features from the speech signal.

[0078] It can be understood that the generated model is a voiceprint recognition model generated according to the parameters of each mechanical feature.

[0079] Specifically, the beneficial effect of the present invention lies in that a voiceprint sensor is used to obtain sound spectrum characteristics and a voiceprint recognition model is generated based on the sound spectrum characteristics, thereby effectively extracting the characteristics of the sound, clearly distinguishing different sounds, and improving the accuracy of sound comparison while effectively improving the accuracy and stability of intelligent recognition of operating machinery.

[0080] See also Figure 2 As shown, it is a flowchart of the early warning model and security model determination according to an embodiment of the present invention, and whether the voiceprint recognition model belongs to the early warning model is determined according to the early warning voiceprint feature similarity value, including:

[0081] If the warning voiceprint feature similarity value is greater than or equal to the preset warning similarity value, the voiceprint recognition model is determined to be a warning model;

[0082] If the warning voiceprint feature similarity value is less than the preset warning similarity value, the voiceprint recognition model is determined to be the first comparison model;

[0083] Embodiment 3: Based on Embodiment 2, this embodiment is different from it in that the preset warning similarity value is 50%. If the warning voiceprint feature similarity value is 80% and is greater than the preset warning similarity value of 50%, the voiceprint recognition model is determined to be a warning model;

[0084] If the warning voiceprint feature similarity value is 30% and is less than the preset warning similarity value of 50%, the voiceprint recognition model is determined to be the first comparison model.

[0085] Among them, the preset warning similarity value is negatively correlated with the straight-line distance from the working area of ​​the operating machinery to the voiceprint sensor.

[0086] In one embodiment, the straight-line distance from the working area of ​​the operating machine to the voiceprint sensor is 50 meters, and the preset warning similarity value is 70%;

[0087] The straight-line distance from the working area of ​​the operating machinery to the voiceprint sensor is 100 meters, and the preset warning similarity value is 60%;

[0088] The straight-line distance from the working area of ​​the operating machinery to the voiceprint sensor is 150 meters, and the preset warning similarity value is 50%.

[0089] It is understandable that the closer the straight-line distance from the working area of ​​the operating machinery to the voiceprint sensor, the clearer the voiceprint characteristics of the sound recognized by the voiceprint sensor.

[0090] Specifically, judging whether the first comparison model belongs to a security model according to the security voiceprint feature similarity value includes:

[0091] If the security voiceprint feature similarity value is greater than or equal to the preset security similarity value, the first comparison model is determined to be a security model;

[0092] If the security voiceprint feature similarity value is less than the preset security similarity value, the first comparison model is determined to be the second comparison model;

[0093] Embodiment 4: Based on Embodiment 3, this embodiment is different from it in that the preset security similarity value is 50%. If the security voiceprint feature similarity value is 80% and is greater than the preset security similarity value of 50%, the first comparison model is determined to be a security model;

[0094] If the security voiceprint feature similarity value of 30% is less than the preset security similarity value of 50%, the first comparison model is determined to be the second comparison model.

[0095] Among them, the preset safety similarity value is negatively correlated with the straight-line distance from the working area of ​​the operating machinery to the voiceprint sensor.

[0096] In one embodiment, the straight-line distance from the working area of ​​the operating machine to the voiceprint sensor is 50 meters, and the preset safety similarity value is 70%;

[0097] The straight-line distance from the working area of ​​the operating machinery to the voiceprint sensor is 100 meters, and the preset safety similarity value is 60%;

[0098] The straight-line distance from the working area of ​​the operating machinery to the voiceprint sensor is 150 meters, and the preset safety similarity value is 50%.

[0099] It is understandable that the closer the straight-line distance from the working area of ​​the operating machinery to the voiceprint sensor, the clearer the voiceprint characteristics of the sound recognized by the voiceprint sensor.

[0100] Specifically, by comparing the voiceprint recognition model with the early warning model and the safety model, the voiceprint recognition model is judged and classified in a targeted manner. While accurately determining whether the operating machinery will cause harm to the overhead cables, the accuracy and stability of the intelligent identification of the operating machinery are further improved.

[0101] Specifically, when the first comparison model is determined to be the second comparison model, the second comparison model is determined to be a warning model or a safety model according to the laser sensor receiving laser, wherein:

[0102] If the laser sensor normally receives the laser emitted by the adjacent laser sensor, it is determined that the second comparison model is a safe model;

[0103] If the laser sensor does not receive the laser emitted by the adjacent laser sensor, it is determined that the second comparison model belongs to the early warning model.

[0104] Embodiment 5: Based on Embodiment 4, this embodiment is different from the embodiment in that, when the first comparison model is determined to be the second comparison model, if the laser sensor normally receives the laser emitted by the adjacent laser sensor, the second comparison model is determined to be a safe model;

[0105] If the laser sensor does not receive the laser emitted by the adjacent laser sensor, it is determined that the second comparison model belongs to the early warning model.

[0106] Specifically, the determination result based on the laser receiving condition of the laser sensor is updated to the original model database.

[0107] Specifically, by using laser sensors to detect the movements of operating machinery, better early warning can be given before hazards occur. While effectively improving the safety of overhead cables, it further improves the accuracy and stability of intelligent identification of operating machinery.

[0108] See also Figure 3 As shown, it is a flow chart of the first sound and light alarm determination of an embodiment of the present invention. When the voiceprint recognition model is determined to be an early warning model, a ground vibration sensor is used to obtain a vibration sensing value, and a determination is made based on the vibration sensing value whether to issue a first sound and light alarm to the machine corresponding to the early warning model, wherein:

[0109] If the vibration sensing value is less than the preset vibration sensing value, it is determined that the first sound and light alarm will not be issued;

[0110] If the vibration sensing value is not less than the preset vibration sensing value, it is determined that the first sound and light alarm is issued for the machine corresponding to the early warning model;

[0111] Embodiment 6: Based on Embodiment 5, this embodiment is different from the embodiment in that the preset vibration sensing value is 10 mm / s, and if the vibration sensing value is 8 mm / s, which is less than the preset vibration sensing value of 10 mm / s, it is determined that the first sound and light alarm is not issued;

[0112] If the vibration sensing value of 12 mm / s is greater than the preset vibration sensing value of 10 mm / s, it is determined that the first sound and light alarm is issued for the machine corresponding to the early warning model.

[0113] The preset vibration sensing value is positively correlated with the straight-line distance from the working area of ​​the operating machinery to the soundprint sensor.

[0114] In one embodiment, the straight-line distance from the working area of ​​the operating machine to the soundprint sensor is 50 meters, and the preset vibration sensing value is 5 millimeters per second;

[0115] The straight-line distance from the working area of ​​the operating machinery to the soundprint sensor is 100 meters, and the preset vibration sensing value is 8 mm per second;

[0116] The straight-line distance from the working area of ​​the operating machinery to the voiceprint sensor is 150 meters, and the preset vibration sensing value is 10 mm per second.

[0117] It can be understood that the voiceprint sensor and the ground vibration sensor are set correspondingly, and the closer the straight-line distance from the working area of ​​the operating machinery to the voiceprint sensor, the greater the vibration intensity sensed by the ground vibration sensor.

[0118] See also Figure 4 As shown, it is a flow chart of the second sound and light alarm determination of an embodiment of the present invention, which determines whether to issue the second sound and light alarm according to the laser receiving situation of the laser sensor, including:

[0119] If the laser sensor normally receives the laser emitted by the adjacent laser sensor, the second sound and light alarm will not be issued;

[0120] If the laser sensor does not receive the laser emitted by the adjacent laser sensor, a second sound and light alarm is issued.

[0121] Embodiment 7: Based on Embodiment 6, this embodiment is different from the embodiment 6 in that if the laser sensor normally receives the laser emitted by the adjacent laser sensor, the second sound and light alarm is not issued;

[0122] If the laser sensor does not receive the laser emitted by the adjacent laser sensor, a second sound and light alarm is issued.

[0123] Specifically, when the first sound and light alarm and / or the second sound and light alarm is issued, the camera captures the action of shooting the sound source object to generate image information and transmits the image information to the central control room.

[0124] So far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present invention.

[0125] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. An intelligent identification method for operating machinery in anti-external force damage scenarios, characterized in that: include: Determine whether to turn on the voiceprint sensor and ground vibration sensor based on the ambient sound intensity value; Using the voiceprint sensor to obtain a sound spectrum feature quantity and generating a voiceprint recognition model according to the sound spectrum feature quantity; Importing the voiceprint recognition model into the original model database and comparing it with the warning model and the security model in turn to obtain a warning voiceprint feature similarity value and a security voiceprint feature similarity value; Determine whether the voiceprint recognition model belongs to the warning model or the first comparison model according to the warning voiceprint feature similarity value, and determine whether the first comparison model belongs to the safety model or the second comparison model according to the safety voiceprint feature similarity value; Determine whether the second comparison model belongs to the warning model or the safety model according to the laser sensor receiving laser light, and update the determination result to the original model database; Using the ground vibration sensor to obtain a vibration sensing value, determining whether to issue a first sound and light alarm to the machine corresponding to the early warning model according to the vibration sensing value, determining whether to issue a second sound and light alarm according to the laser receiving condition of the laser sensor, and capturing an image of the sound source object through a plurality of cameras; Among them, the sound spectrum characteristics include frequency, period, wavelength and sound speed; the original model database includes the early warning model and the safety model; the early warning model is the sound model of a machine that can cause damage to overhead cables; the safety model is the sound model of an object that will not cause damage to overhead cables; the voiceprint sensor, the ground vibration sensor and the laser sensor are set one by one.

2. The intelligent identification method for operating machinery in the anti-external force damage scenario according to claim 1 is characterized in that: Determine whether to turn on the voiceprint sensor and the ground vibration sensor according to the ambient sound intensity value, wherein: If the ambient sound intensity value is greater than or equal to the ambient sound intensity preset value, turning on the voiceprint sensor and the ground vibration sensor; Among them, the preset value of the ambient sound intensity is negatively correlated with the straight-line distance from the working area of ​​the operating machinery to the voiceprint sensor.

3. The intelligent identification method for operating machinery in the anti-external force damage scenario according to claim 2 is characterized in that: The steps of using the voiceprint sensor to obtain the sound spectrum feature quantity and generating a voiceprint recognition model according to the sound spectrum feature quantity include: The voiceprint sensor extracts the sound spectrum characteristics of different operating machines from the acquired environmental sounds based on the voiceprint extraction algorithm; The voiceprint sensor generates voiceprint recognition models of different operating machines using the sound spectrum feature quantity.

4. The intelligent identification method for operating machinery in the anti-external force damage scenario according to claim 3 is characterized in that: Determining whether the voiceprint recognition model belongs to the early warning model according to the early warning voiceprint feature similarity value includes: If the warning voiceprint feature similarity value is greater than or equal to a preset warning similarity value, the voiceprint recognition model is determined to be a warning model; If the warning voiceprint feature similarity value is less than the preset warning similarity value, determining that the voiceprint recognition model is the first comparison model; Among them, the preset warning similarity value is negatively correlated with the straight-line distance from the working area of ​​the operating machinery to the voiceprint sensor.

5. The intelligent identification method for working machinery in the anti-external force damage scenario according to claim 4 is characterized in that: Determining whether the first comparison model belongs to the security model according to the security voiceprint feature similarity value includes: If the security voiceprint feature similarity value is greater than or equal to a preset security similarity value, determining that the first comparison model is the security model; If the security voiceprint feature similarity value is less than a preset security similarity value, determining that the first comparison model is the second comparison model; Among them, the preset safety similarity value is negatively correlated with the straight-line distance from the working area of ​​the operating machinery to the voiceprint sensor.

6. The intelligent identification method for operating machinery in the anti-external force damage scenario according to claim 5 is characterized in that: When the first comparison model is determined to be the second comparison model, the second comparison model is determined to be the warning model or the safety model according to the laser receiving condition of the laser sensor, wherein: If the laser sensor normally receives the laser emitted by the adjacent laser sensor, it is determined that the second comparison model belongs to the safety model; If the laser sensor does not receive the laser emitted by the adjacent laser sensor, it is determined that the second comparison model belongs to the early warning model.

7. The intelligent identification method for working machinery in the anti-external force damage scenario according to claim 2 is characterized in that: The determination result according to the laser receiving condition of the laser sensor is updated to the original model database.

8. The intelligent identification method for working machinery in the anti-external force damage scenario according to claim 7 is characterized in that: When the voiceprint recognition model is determined to be an early warning model, the ground vibration sensor is used to obtain the vibration sensing value, and it is determined according to the vibration sensing value whether to issue a first sound and light alarm to the machine corresponding to the early warning model, wherein: If the vibration sensing value is less than the preset vibration sensing value, it is determined that the first sound and light alarm is not issued; If the vibration sensing value is not less than the preset vibration sensing value, it is determined that the first sound and light alarm is issued to the machine corresponding to the early warning model; The preset vibration sensing value is positively correlated with the straight-line distance from the working area of ​​the operating machinery to the soundprint sensor.

9. The intelligent identification method for working machinery in the anti-external force damage scenario according to claim 8 is characterized in that: Determining whether to issue a second sound and light alarm according to the laser sensor receiving laser light includes: If the laser sensor normally receives the laser emitted by the adjacent laser sensor, the second sound and light alarm will not be issued; If the laser sensor does not receive the laser emitted by the adjacent laser sensor, a second sound and light alarm is issued.

10. The intelligent identification method for working machinery in the anti-external force damage scenario according to claim 9 is characterized in that: When the first sound and light alarm and / or the second sound and light alarm is issued, the camera captures the action of shooting the sound source object to generate image information and transmits the image information to the central control room.

Images