A voiceprint monitoring method for substation main equipment based on speech recognition

Abstract

The present invention discloses a voiceprint monitoring method for a substation main equipment based on speech recognition, comprising: when the substation main equipment is not in operation, taking the position of the substation main equipment as a collection point, collecting sound information of multiple time periods in the surrounding environment, and processing the sound information to obtain a voiced fundamental tone periodic signal of the surrounding environment; when the substation main equipment is operating normally, the voiceprint monitoring method for the substation main equipment based on speech recognition collects and processes the surrounding environment of the substation main equipment, then collects the sound information of the substation main equipment when it is operating in combination with the surrounding environment, and uses the collected sound information of the surrounding environment to filter and remove the sound information of the substation main equipment when it is operating, which can effectively improve the authenticity of the sound information of the substation main equipment when it is operating, and greatly reduce the influence of the surrounding environment on the voiceprint monitoring of the substation main equipment when it is operating.

Description

Technical Field

[0001] The present invention relates to the technical field of substation equipment monitoring, and in particular to a voiceprint monitoring method for substation main equipment based on speech recognition. Background Art

[0002] Substation equipment is one of the distribution equipment in the power system. When the substation equipment is in operation, the operation process of the substation equipment needs to be monitored in real time. Most of the existing monitoring methods for substation equipment are visual monitoring, that is, real-time monitoring of the substation equipment is achieved by collecting the operation process images of the substation equipment in real time. However, the visual monitoring method requires the substation equipment to be in a bright environment, so lighting equipment is required to increase the brightness of the monitoring equipment at all times.

[0003] The voiceprint monitoring method based on speech recognition achieves the purpose of monitoring through voice feature recognition. The currently disclosed patent document: A partial discharge monitoring method based on voiceprint (application number: 202110900753.8) discloses a monitoring method for partial discharge of high-voltage equipment. The voiceprint-based monitoring method is relatively simple and does not perform environmental soundprint processing on the environment where the high-voltage equipment is located. The monitoring results of partial discharge are easily affected by the external environment, and the monitoring results of partial discharge of high-voltage equipment are not accurate. Summary of the invention

[0004] In order to overcome the problems existing in the prior art, the purpose of the present invention is to provide a voiceprint monitoring method for substation main equipment based on speech recognition, which uses the collected sound information of the surrounding environment to filter and remove the sound information of the substation main equipment during operation, which can effectively improve the authenticity of the sound information of the substation main equipment during operation and greatly reduce the impact of the surrounding environment on the voiceprint monitoring of the substation main equipment during operation.

[0005] The purpose of the present invention is achieved through the following technical solutions:

[0006] A voiceprint monitoring method for a substation main device based on voice recognition comprises the following specific steps:

[0007] S1. When the main substation equipment is not in operation, the location of the main substation equipment is used as the collection point to collect sound information of the surrounding environment in multiple time periods, and the sound information is processed to obtain the voiced fundamental tone periodic signal of the surrounding environment;

[0008] S2. When the main substation equipment is operating normally, collect the sound information of multiple operating cycles of the main substation equipment during normal operation, and process the sound information to obtain the periodic sound signal of the main substation equipment during normal operation;

[0009] S3, obtaining a periodic sound signal when the main substation equipment is operating normally without ambient sound;

[0010] S4. Establish a database of standard operation sound signals of main substation equipment;

[0011] S5, collecting sound information of multiple operation cycles of multiple different abnormal operation states of the main substation equipment, and processing the sound information to obtain periodic sound signals of multiple different abnormal operation states of the main substation equipment;

[0012] S6. Acquire periodic sound signals of multiple abnormal operation states of the main substation equipment without ambient sound;

[0013] S7. Establish a database of abnormal operation sound signals of main substation equipment;

[0014] S8. Establishing a database of sound signals of abnormal operation of main substation equipment;

[0015] S9. When the main substation equipment is in operation, the main substation equipment standard operation sound signal database, the main substation equipment abnormal operation starting sound signal database and the main substation equipment abnormal operation sound signal database are used to determine whether the main substation equipment is in abnormal operation.

[0016] Furthermore, the specific method of collecting the ambient sound information in S1 is:

[0017] S101, when the main substation equipment is not in operation, taking the location of the main substation equipment as a collection point, collecting sound information at different times of the day;

[0018] S102, continuously collecting sound information at different time periods over multiple days;

[0019] S103, acquiring periodic signals of voiced sounds in the surrounding environment based on the collected environmental sound information for multiple days, and acquiring non-periodic specific sound information emitted by a single object, wherein the specific sound information includes vehicle sounds, wind sounds, rain sounds, and thunder sounds;

[0020] The method for obtaining the voiced pitch periodic signal of the surrounding environment in S1 is:

[0021] Based on Matlab, using the short-time autocorrelation function R n (k), obtain the ambient voice pitch period signal x n (m):

[0022]

[0023] Wherein, k is the maximum delay point number, N is the period of the pitch period signal, ω(n) is the window function, and n and m are the starting sampling point and the ending sampling point of the speech frame in the sound information, respectively.

[0024] Furthermore, the location for collecting the sound information of the main substation equipment during operation in S2, S5 and S9 is the same as the location for collecting the sound information of the surrounding environment in multiple time periods in S1.

[0025] Furthermore, the specific steps of obtaining the periodic sound signal of the main substation equipment without ambient sound during normal operation in S3 are:

[0026] According to the periodic signal of the voiced fundamental tone of the surrounding environment obtained in S1, the periodic sound signal of the voiced fundamental tone of the surrounding environment in the periodic sound signal of the substation main equipment during normal operation is filtered out and removed, and at the same time, specific sound information is removed to obtain the periodic sound signal of the substation main equipment during normal operation without ambient sound.

[0027] Furthermore, the specific steps of establishing the standard operation sound signal database of the substation main equipment in S4 are:

[0028] Extract the sound signal of each period of the periodic sound signal of the main substation equipment when it is operating normally without ambient sound, overlap the sound signals of multiple periods, obtain the standard operating sound signal range of the main substation equipment when it is operating normally, and establish a standard operating sound signal database for the main substation equipment.

[0029] Furthermore, the specific steps of obtaining the periodic sound signals of the multiple abnormal operation states of the substation main equipment without ambient sound in S6 are:

[0030] According to the periodic signal of the ambient voiced fundamental tone obtained in S1, the periodic sound signal of the ambient voiced fundamental tone in the periodic sound signals of multiple different abnormal operating states of the substation main equipment is filtered out to obtain the periodic sound signals of multiple abnormal operating states of the substation main equipment without ambient sound.

[0031] Furthermore, the specific steps of establishing the abnormal operation sound signal database of the main substation equipment in S7 are:

[0032] Extract the sound signal of each period of the periodic sound signals of multiple abnormal operating states of the substation main equipment without ambient sound, overlap the sound signals of multiple periods of multiple abnormal operating states of the substation main equipment, obtain the range of abnormal operation sound signals under multiple abnormal operating states of the substation main equipment, and establish a database of abnormal operation sound signals of the substation main equipment.

[0033] Furthermore, the specific steps of establishing the abnormal operation start sound signal database of the main substation equipment in S8 are:

[0034] The abnormal operation sound signals of multiple substation main equipment in abnormal operation state are compared with the standard operation sound signals of the substation main equipment in normal operation in the same period, the sound signal range of the starting point of abnormal operation of multiple substation main equipment is obtained, and a database of abnormal operation starting sound signals of substation main equipment is established.

[0035] Further, the specific method for determining the standard operation sound signal range of the substation main equipment in normal operation and the abnormal operation sound signal range of multiple substation main equipment in abnormal operation state in S4 and S7 and the sound signal range of the starting point of abnormal operation of the substation main equipment in S6 is:

[0036] Using the short-time average amplitude function M n Get the energy of a frame of speech signal in the standard running sound signal:

[0037]

[0038] The energy of the speech signal of a sound signal of one cycle is obtained by using the short-time average amplitude function;

[0039] The sound signals of multiple cycles during the normal operation of the main substation equipment are overlapped to obtain the standard operating sound signal range (M n1 ,M n2 );

[0040] The sound signals of multiple cycles under the abnormal operation state of the main substation equipment are overlapped to obtain the abnormal operation sound signal range (M n3 ,M n4 )、(M n5 ,M n6 )、(M n7 ,M n8 ),···;

[0041] The sound signal range of the starting point of abnormal operation of the main substation equipment is obtained by using the short-time average amplitude function (E n3 ,E n4 )、(E n5 ,E n6 )、(E n7 ,E n8 ),···.

[0042] Furthermore, the specific steps for determining the operating status of the main substation equipment in S9 are:

[0043] S901, using a voice recognition device to collect sound information of the main substation equipment in operation in real time, and process the sound information to obtain a real-time sound signal of the main substation equipment in operation;

[0044] S902, filtering and removing the real-time sound signal of the substation main equipment when it is running according to the voiced fundamental tone periodic signal and specific sound information of the surrounding environment obtained in S1, and obtaining the real-time sound signal of the substation main equipment when it is running without the ambient sound;

[0045] S903, comparing the real-time sound signal of the substation main equipment when it is running without ambient sound with the starting sound signal in the database of the starting sound signal of abnormal operation of the substation main equipment;

[0046] S9031. When the sound signal of the real-time sound signal at a certain moment is within the sound signal range of the starting point of abnormal operation of a certain substation main equipment, the sound signal subsequent to the real-time sound signal is compared with the abnormal operation sound signal range of the corresponding substation main equipment in the abnormal operation state in the substation main equipment abnormal operation sound signal database. When the real-time sound signal falls into the abnormal operation sound signal range of the corresponding substation main equipment in the abnormal operation state, it is determined that the substation main equipment is operating abnormally. Otherwise, proceed to S9032.

[0047] S9032. Compare the subsequent sound signals of the real-time sound signal at a certain moment in S9031 with the standard operation sound signals in the standard operation sound signal database of the substation main equipment. When the real-time sound signal falls into the standard operation sound signal range when the substation main equipment is operating normally, it is determined that the substation main equipment is operating normally. Otherwise, it is determined that the operating status of the substation main equipment is not identified, and an early warning prompt is issued.

[0048] Compared with the prior art, the present invention has the following beneficial effects:

[0049] 1. The voiceprint monitoring method for substation main equipment based on speech recognition collects and processes the surrounding environment of the substation main equipment, then collects the sound information of the substation main equipment when it is running in combination with the surrounding environment, and uses the collected sound information of the surrounding environment to filter and remove the sound information of the substation main equipment when it is running. It can effectively improve the authenticity of the sound information of the substation main equipment when it is running, greatly reduce the impact of the surrounding environment on the voiceprint monitoring of the substation main equipment when it is running, and is conducive to improving the accuracy of the monitoring results of the operating status of the substation main equipment.

[0050] 2. The voiceprint monitoring method for substation main equipment based on speech recognition can obtain the sound signal of the starting point of abnormal operation of the substation main equipment after collecting the sound information of the substation main equipment during normal operation and abnormal operation, and processing the sound information. When monitoring the substation main equipment, the abnormal operation of the substation main equipment can be identified and monitored in a targeted manner, which is beneficial to reducing the amount of data for comparative identification of abnormal operation of the substation main equipment, and is also beneficial to improving the efficiency of monitoring and identification of abnormal operation of the substation main equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 This is a flow chart of the voiceprint monitoring method of the present invention;

[0052] Figure 2 This is a flow chart for determining whether the main substation equipment is operating abnormally in the present invention;

[0053] Figure 3 It is a schematic diagram of the simulation of the sound signal when the main power transformation equipment of the present invention is running;

[0054] Figure 4 It is a schematic diagram of the simulation structure of multiple periodic sound signals when the main power transformation equipment of the present invention is running. DETAILED DESCRIPTION

[0055] 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.

[0056] The embodiment of the voiceprint monitoring method of the main substation equipment based on speech recognition is as follows:

[0057] See also Figure 1-Figure 4 , a voiceprint monitoring method for a substation main equipment based on speech recognition, comprising the following specific steps:

[0058] S1. When the main substation equipment is not in operation, the location of the main substation equipment is used as the collection point to collect the sound information of the surrounding environment in multiple periods, and the sound information is processed to obtain the periodic signal of the voiced fundamental tone of the surrounding environment:

[0059] S101, when the main substation equipment is not in operation, taking the location of the main substation equipment as a collection point, collecting sound information at different times of the day;

[0060] S102, continuously collecting sound information at different time periods over multiple days;

[0061] S103, acquiring periodic signals of voiced sounds in the surrounding environment based on the collected environmental sound information for multiple days, and acquiring non-periodic specific sound information emitted by a single object, wherein the specific sound information includes vehicle sounds, wind sounds, rain sounds, and thunder sounds;

[0062] Based on Matlab, using the short-time autocorrelation function R n (k), obtain the ambient voice pitch period signal x n (m):

[0063]

[0064] Wherein, k is the maximum delay point number, N is the period of the pitch period signal, ω(n) is the window function, and n and m are the starting sampling point and the ending sampling point of the speech frame in the sound information, respectively.

[0065] S2. When the main substation equipment is operating normally, the sound information of multiple operating cycles of the main substation equipment is collected at the same position in S1, and the sound information is processed to obtain the periodic sound signal of the main substation equipment when it is operating normally.

[0066] S3. Obtaining periodic sound signals when the main substation equipment is operating normally without ambient sound:

[0067] According to the periodic signal of the voiced fundamental tone of the surrounding environment obtained in S1, the periodic sound signal of the voiced fundamental tone of the surrounding environment in the periodic sound signal of the substation main equipment during normal operation is filtered out and removed, and at the same time, specific sound information is removed to obtain the periodic sound signal of the substation main equipment during normal operation without ambient sound.

[0068] S4. Establish a database of standard operation sound signals of main substation equipment:

[0069] Extract the sound signal of each period of the periodic sound signal of the main substation equipment when it is operating normally without ambient sound, overlap the sound signals of multiple periods, obtain the standard operating sound signal range of the main substation equipment when it is operating normally, and establish a standard operating sound signal database for the main substation equipment.

[0070] Using the short-time average amplitude function M n Get the energy of a frame of speech signal in the standard running sound signal:

[0071]

[0072] The short-time average amplitude function is used to obtain the energy of the voice signal of a sound signal in one cycle, and the sound signals of multiple cycles when the main substation equipment is operating normally are overlapped to obtain the standard operating sound signal range (M n1 ,M n2 ).

[0073] S5. Collect sound information of multiple operation cycles of multiple different abnormal operation states of the main substation equipment at the same position in S1, and process the sound information to obtain periodic sound signals of the main substation equipment under multiple different abnormal operation states.

[0074] S6. Obtaining periodic sound signals of multiple abnormal operating states of the main substation equipment without ambient sound:

[0075] According to the periodic signal of the ambient voiced fundamental tone obtained in S1, the periodic sound signal of the ambient voiced fundamental tone in the periodic sound signals of multiple different abnormal operating states of the substation main equipment is filtered out to obtain the periodic sound signals of multiple abnormal operating states of the substation main equipment without ambient sound.

[0076] S7. Establish a database of abnormal operation sound signals of main substation equipment:

[0077] Extract the sound signal of each period of the periodic sound signals of multiple abnormal operating states of the substation main equipment without ambient sound, overlap the sound signals of multiple periods of multiple abnormal operating states of the substation main equipment, obtain the range of abnormal operation sound signals under multiple abnormal operating states of the substation main equipment, and establish a database of abnormal operation sound signals of the substation main equipment.

[0078] Using the short-time average amplitude function M n Get the energy of a frame of speech signal in the standard running sound signal:

[0079]

[0080] The short-time average amplitude function is used to obtain the energy of the voice signal of a cycle of the sound signal, and the sound signals of multiple cycles under the abnormal operation state of the main substation equipment are overlapped to obtain the abnormal operation sound signal range (M n3 ,M n4 )、(M n5 ,M n6 )、(M n7 ,M n8 ),···.

[0081] S8. Establish a database of sound signals of abnormal operation of main substation equipment:

[0082] The abnormal operation sound signals of multiple substation main equipment in abnormal operation state are compared with the standard operation sound signals of the substation main equipment in normal operation in the same period, the sound signal range of the starting point of abnormal operation of multiple substation main equipment is obtained, and a database of abnormal operation starting sound signals of substation main equipment is established.

[0083] Using the short-time average amplitude function M n Get the energy of a frame of speech signal in the standard running sound signal:

[0084]

[0085] The short-time average amplitude function is used to obtain the energy of the voice signal of a cycle of the sound signal, and the short-time average amplitude function is used to obtain the sound signal range (E n3 ,En4 )、(E n5 ,E n6 )、(E n7 ,E n8 ),···.

[0086] S9, when the main substation equipment is in operation, the main substation equipment standard operation sound signal database, the main substation equipment abnormal operation start sound signal database and the main substation equipment abnormal operation sound signal database are used to determine whether the main substation equipment is in abnormal operation:

[0087] S901, using a voice recognition device to collect sound information of the main substation equipment in operation in real time, and process the sound information to obtain a real-time sound signal of the main substation equipment in operation;

[0088] S902, filtering and removing the real-time sound signal of the substation main equipment when it is running according to the voiced fundamental tone periodic signal and specific sound information of the surrounding environment obtained in S1, and obtaining the real-time sound signal of the substation main equipment when it is running without the ambient sound;

[0089] S903, comparing the real-time sound signal of the substation main equipment when it is running without ambient sound with the starting sound signal in the database of the starting sound signal of abnormal operation of the substation main equipment;

[0090] S9031. When the sound signal of the real-time sound signal at a certain moment is within the sound signal range of the starting point of abnormal operation of a certain substation main equipment, the sound signal subsequent to the real-time sound signal is compared with the abnormal operation sound signal range of the corresponding substation main equipment in the abnormal operation state in the substation main equipment abnormal operation sound signal database. When the real-time sound signal falls into the abnormal operation sound signal range of the corresponding substation main equipment in the abnormal operation state, it is determined that the substation main equipment is operating abnormally. Otherwise, proceed to S9032.

[0091] S9032. Compare the subsequent sound signals of the real-time sound signal at a certain moment in S9031 with the standard operation sound signals in the standard operation sound signal database of the substation main equipment. When the real-time sound signal falls into the standard operation sound signal range when the substation main equipment is operating normally, it is determined that the substation main equipment is operating normally. Otherwise, it is determined that the operating status of the substation main equipment is not identified, and an early warning prompt is issued.

[0092] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A voiceprint monitoring method for substation main equipment based on speech recognition, characterized in that: The specific steps include: S1. When the main substation equipment is not in operation, the location of the main substation equipment is used as the collection point to collect sound information of the surrounding environment in multiple time periods, and the sound information is processed to obtain the voiced fundamental tone periodic signal of the surrounding environment; S2. When the main substation equipment is operating normally, collect the sound information of multiple operating cycles of the main substation equipment during normal operation, and process the sound information to obtain the periodic sound signal of the main substation equipment during normal operation; S3, obtaining a periodic sound signal when the main substation equipment is operating normally without ambient sound; S4. Establish a database of standard operation sound signals of main substation equipment; S5, collecting sound information of multiple operation cycles of multiple different abnormal operation states of the main substation equipment, and processing the sound information to obtain periodic sound signals of multiple different abnormal operation states of the main substation equipment; S6. Acquire periodic sound signals of multiple abnormal operation states of the main substation equipment without ambient sound; S7. Establish a database of abnormal operation sound signals of main substation equipment; S8. Establishing a database of sound signals of abnormal operation of main substation equipment; S9, when the main substation equipment is in operation, the main substation equipment standard operation sound signal database, the main substation equipment abnormal operation start sound signal database and the main substation equipment abnormal operation sound signal database are used to determine whether the main substation equipment is in abnormal operation; The specific method of collecting ambient sound information in S1 is: S101, when the main substation equipment is not in operation, taking the location of the main substation equipment as a collection point, collecting sound information at different times of the day; S102, continuously collecting sound information at different time periods over multiple days; S103, obtaining periodic signals of voiced sounds in the surrounding environment based on the collected environmental sound information for multiple days, and simultaneously obtaining non-periodic specific sound information emitted by a single object; The method for obtaining the voiced pitch periodic signal of the surrounding environment in S1 is: Based on Matlab, using the short-time autocorrelation function R n (k), obtain the ambient voice pitch period signal x n (m): Wherein, k is the maximum delay point number, N is the period of the pitch period signal, ω(n) is the window function, and n and m are the starting sampling point and the ending sampling point of the speech frame in the sound information respectively; The specific steps of obtaining the periodic sound signal of the main substation equipment without ambient sound during normal operation in S3 are: According to the periodic signal of the voiced fundamental tone of the surrounding environment obtained in S1, the periodic signal of the voiced fundamental tone of the surrounding environment in the periodic sound signal of the substation main equipment when it is operating normally is filtered and removed, and the specific sound information is removed at the same time, so as to obtain the periodic sound signal of the substation main equipment when it is operating normally without the ambient sound; The specific steps of establishing the standard operation sound signal database of the main substation equipment in S4 are: Extract the sound signal of each period of the periodic sound signal of the main substation equipment when it is operating normally without ambient sound, overlap the sound signals of multiple periods, obtain the standard operating sound signal range of the main substation equipment when it is operating normally, and establish a database of standard operating sound signals of the main substation equipment; The specific steps of obtaining the periodic sound signals of the multiple abnormal operation states of the main substation equipment without ambient sound in S6 are: According to the periodic signal of the ambient voiced fundamental tone obtained in S1, the periodic sound signal of the ambient voiced fundamental tone in the periodic sound signals of multiple different abnormal operating states of the substation main equipment is filtered out to obtain the periodic sound signals of multiple abnormal operating states of the substation main equipment without ambient sound.

2. According to claim 1, a method for monitoring voiceprint of main substation equipment based on speech recognition, characterized in that: The location where the sound information of the main substation equipment is collected during operation in S2, S5 and S9 is the same as the location where the sound information of the surrounding environment in multiple time periods is collected in S1.

3. According to claim 1, a method for monitoring voiceprint of main substation equipment based on speech recognition, characterized in that: The specific steps of establishing the abnormal operation sound signal database of the main substation equipment in S7 are: Extract the sound signal of each period of the periodic sound signals of multiple abnormal operating states of the substation main equipment without ambient sound, overlap the sound signals of multiple periods of multiple abnormal operating states of the substation main equipment, obtain the range of abnormal operation sound signals under multiple abnormal operating states of the substation main equipment, and establish a database of abnormal operation sound signals of the substation main equipment.

4. According to claim 3, a method for monitoring voiceprint of main substation equipment based on speech recognition is characterized in that: The specific steps of establishing the abnormal operation start sound signal database of the main substation equipment in S8 are: The abnormal operation sound signals of multiple substation main equipment in abnormal operation state are compared with the standard operation sound signals of the substation main equipment in normal operation in the same period, the sound signal range of the starting point of abnormal operation of multiple substation main equipment is obtained, and a database of abnormal operation starting sound signals of substation main equipment is established.

5. According to claim 4, a method for monitoring voiceprint of main substation equipment based on speech recognition is characterized in that: The specific method for determining the standard operation sound signal range of the substation main equipment in normal operation and the abnormal operation sound signal range of multiple substation main equipment in abnormal operation states in S4 and S7 and the sound signal range of the starting point of abnormal operation of the substation main equipment in S6 is as follows: Using the short-time average amplitude function M n Get the energy of a frame of speech signal in the standard running sound signal: The energy of the speech signal of a sound signal of one cycle is obtained by using the short-time average amplitude function; The sound signals of multiple cycles during the normal operation of the main substation equipment are overlapped to obtain the standard operating sound signal range (M n1 ,M n2 ); The sound signals of multiple cycles under the abnormal operation state of the main substation equipment are overlapped to obtain the abnormal operation sound signal range (M n3 ,M n4 )、(M n5 ,M n6 )、(M n7 ,M n8 ),···; The sound signal range of the starting point of abnormal operation of the main substation equipment is obtained by using the short-time average amplitude function (E n3 ,E n4 )、(E n5 ,E n6 )、(E n7 ,E n8 ),….

6. According to claim 1, a method for voiceprint monitoring of substation main equipment based on voice recognition is characterized in that: The specific steps for determining the operating status of the main substation equipment in S9 are: S901, using a voice recognition device to collect sound information of the main substation equipment in operation in real time, and process the sound information to obtain a real-time sound signal of the main substation equipment in operation; S902, filtering and removing the real-time sound signal of the substation main equipment when it is running according to the voiced fundamental tone periodic signal and specific sound information of the surrounding environment obtained in S1, and obtaining the real-time sound signal of the substation main equipment when it is running without the ambient sound; S903, comparing the real-time sound signal of the substation main equipment when it is running without ambient sound with the starting sound signal in the database of the starting sound signal of abnormal operation of the substation main equipment; S9031. When the sound signal of the real-time sound signal at a certain moment is within the sound signal range of the starting point of abnormal operation of a certain substation main equipment, the sound signal subsequent to the real-time sound signal is compared with the abnormal operation sound signal range of the corresponding substation main equipment in the abnormal operation state in the substation main equipment abnormal operation sound signal database. When the real-time sound signal falls into the abnormal operation sound signal range of the corresponding substation main equipment in the abnormal operation state, it is determined that the substation main equipment is operating abnormally. Otherwise, proceed to S9032. S9032. Compare the subsequent sound signals of the real-time sound signal at a certain moment in S9031 with the standard operation sound signals in the standard operation sound signal database of the substation main equipment. When the real-time sound signal falls into the standard operation sound signal range when the substation main equipment is operating normally, it is determined that the substation main equipment is operating normally. Otherwise, it is determined that the operating status of the substation main equipment is not identified, and an early warning prompt is issued.

Images