Substation equipment operation state monitoring method and system based on voiceprint recognition

Abstract

The invention discloses a transformer substation equipment operation state monitoring method and system based on voiceprint recognition, and belongs to the technical field of dynamic monitoring. Operation voiceprint data and operation time periods of all power transformation equipment in the transformer substation during historical operation are obtained, and an operation time point set is constructed; constructing a voiceprint drift model based on the operation time point and the operation voiceprint data; obtaining a current operation time point of the power transformation equipment and operation voiceprint data corresponding to the current operation time point; acquiring operation voiceprint data corresponding to all operation time points matched with the current operation time point, and calculating a comprehensive voiceprint difference degree; the comprehensive difference degree of all the power transformation devices in the transformer substation at the current operation time point is obtained, the comprehensive difference similarity between different power transformation devices is calculated, a threshold value is preset, and linkage early warning between the power transformation devices is analyzed, so that the accuracy and the real-time performance of monitoring the operation state of the power transformation devices are improved; and the intelligent level of abnormal linkage identification and early warning response is also enhanced.

Description

Technical Field

[0001] The present invention relates to the technical field of dynamic monitoring, and in particular to a method and system for monitoring the operating status of substation equipment based on voiceprint recognition. Background Technique

[0002] As the core hub of the power grid, the real-time monitoring and intelligent diagnosis of the operating status of substation equipment have become important means to ensure the safe and stable operation of the power grid. Traditional condition monitoring methods mostly rely on electrical parameters (such as current, voltage, temperature, etc.) and visual recognition means. Although these methods have formed a certain system in actual operation, they still have problems such as coarse perception granularity, high response delay, and lagging abnormal recognition when facing complex and changeable operating environments, the linkage effects between equipment, and sudden abnormal events. In recent years, with the development of voiceprint recognition, acoustic feature extraction, and multi-source data fusion technologies, equipment status recognition based on voiceprint has gradually become a research hotspot. This type of method analyzes based on the acoustic features generated by the equipment during operation, and has advantages such as strong non-invasiveness, convenient data collection, and wide equipment adaptability, and has gradually been explored and applied in fields such as wind power, hydropower, and rail transit.

[0003] However, currently in the substation field, the voiceprint-based status recognition method is still in the initial research stage. Existing technologies mostly focus on the abnormal detection of single equipment or instantaneous judgment based on short-time sound features, lacking systematic modeling of the operating voiceprint of equipment throughout its life cycle and correlation analysis between multiple equipment. On the one hand, existing technologies usually ignore the "voiceprint drift" phenomenon generated by the evolution of the operating voiceprint of equipment over time, resulting in insufficient accuracy and robustness of abnormal detection; on the other hand, most substation equipment operates in coordination. When a potential abnormality occurs in a certain equipment, its adjacent or associated equipment often also shows certain changes in voiceprint response. Existing solutions are difficult to effectively capture and model this linkage behavior between equipment. In addition, the deep fusion of voiceprint data and time information, the dynamic comparison of multi-round historical data, and the quantitative evaluation method of multi-dimensional voiceprint differences are not yet mature, restricting the practical and intelligent development of voiceprint recognition in substation operating status monitoring. Summary of the Invention

[0004] The purpose of the present invention is to provide a method and system for monitoring the operating status of substation equipment based on voiceprint recognition to solve the problems raised in the above background technique.

[0005] To solve the above technical problems, the present invention provides the following technical solutions: Method for monitoring the operating state of substation equipment based on voiceprint recognition. This method includes the following steps: Step S1: Obtain the operating voiceprint data and operating time periods of all substation equipment during historical operation, and construct an operating time point set; Step S2: Based on the operating time points and the corresponding operating voiceprint data, construct a voiceprint drift model; Step S3: Obtain the current operating time point of the substation equipment and the corresponding operating voiceprint data; Obtain the operating voiceprint data corresponding to all operating time points that match the current operating time point, and calculate the comprehensive voiceprint difference degree; Step S4: Obtain the comprehensive difference degrees of all substation equipment within the current operating time point, calculate the comprehensive difference similarity between different substation equipment, preset a threshold, and analyze and conduct linkage warnings between substation equipment.

[0006] As a preferred solution of the method for monitoring the operating state of substation equipment based on voiceprint recognition according to the present invention, all substation equipment in the substation is obtained, and voiceprint vibration sensors are arranged inside the substation equipment. One voiceprint vibration sensor is arranged corresponding to one substation equipment; The operating voiceprint data of the substation equipment collected by the voiceprint vibration sensor is stored in the substation background monitoring system; Retrieve the historical operation data log of the substation equipment from the substation background monitoring system. The historical operation data log includes the operating voiceprint data and operating time periods of the substation equipment during historical operation. The operating time period is evenly divided into several operating time points to construct an operating time point set. Denote the operating time point set corresponding to the nth historical operation as , where represents the ith operating time point corresponding to the nth historical operation, and I represents the total number of operating time points.

[0007] As a preferred solution of the method for monitoring the operating state of substation equipment based on voiceprint recognition according to the present invention, based on the ith operating time point corresponding to the nth historical operation and the operating voiceprint data corresponding to the operating time point , construct the voiceprint drift model corresponding to the nth historical operation, specifically as follows: Traverse the operating time point set to obtain the operating voiceprint data corresponding to each operating time point, and arrange them in the order of the operating time points; Perform normalization processing on the arranged operating voiceprint data. Based on the normalized operating voiceprint data, construct a voiceprint drift model, denoted as , where represents the voiceprint drift model corresponding to the nth historical operation, represents the operating voiceprint data corresponding to the normalized operating time point , and I represents the total number of operating time points.

[0008] As a preferred solution of the substation equipment operation status monitoring method based on voiceprint recognition according to the present invention, obtain the current operation time point of the substation equipment and the operation voiceprint data corresponding to the current operation time point, and record them as and respectively; obtain the voiceprint drift model corresponding to a total of N historical operations, preset an operation time point matching rule, and based on the operation time point matching rule, extract from the voiceprint drift model the operation time point that matches the current operation time point of the substation equipment ; the operation time point matching rule is specifically: if , then it is determined that the operation time point matches the current operation time point , where represents a preset operation duration tolerance threshold, and represents the m-th operation time point corresponding to the n-th historical operation; Obtain the operation voiceprint data corresponding to all operation time points that match the current operation time point , and compare it with the operation voiceprint data corresponding to the current operation time point to calculate the comprehensive voiceprint difference degree. The calculation formula is as follows: ; Among them, represents the comprehensive difference degree, M represents the total number of operation voiceprint data corresponding to all operation time points that match the current operation time point , represents the operation voiceprint data corresponding to the operation time point , and represents a preset exponential decay factor.

[0009] It should be noted that this formula screens historical time points close to the current time point through the threshold to ensure that the compared voiceprint data is in a similar operation condition (such as the same load period), and is used to quantify the difference degree between the voiceprint data at the current operation time point and the voiceprint data at the historical matching time point . Its core design includes time decay weighting and difference accumulation and normalization. Among them, through the exponential function , higher weights are given to historical data closer in time. The smaller the time difference , the greater the weight, and vice versa for attenuation; by accumulating and weighting the voiceprint differences in the numerator part, the denominator normalizes the sum of weights to ensure that the result is not affected by the amount of matching data (M); this formula combines time decay with voiceprint differences, solving the problem that traditional methods ignore "voiceprint drift" (the gradual change of voiceprint caused by equipment aging and environmental changes). For example, due to internal wear of a transformer, the voiceprint characteristics may change slowly. The formula can distinguish normal drift from sudden anomalies by weighting historical data.

[0010] As a preferred solution of the substation equipment operation status monitoring method based on voiceprint recognition described in the present invention, obtain the current operation time point within, the comprehensive difference degree of all substation equipment in the substation, use the cosine similarity to calculate the comprehensive difference similarity between different substation equipment, preset the comprehensive difference similarity threshold, and mark all substation equipment with a comprehensive difference similarity greater than or equal to the comprehensive difference similarity threshold as similar equipment; Based on the comprehensive difference degree , preset the comprehensive difference degree threshold. If the comprehensive difference degree is greater than or equal to the comprehensive difference degree threshold, it is determined that there is a voiceprint drift anomaly in the substation equipment at the current operation time point, that is, there is an operation status anomaly in the substation equipment at the current operation time point, and then unified early warning is carried out for the equipment similar to the substation equipment; Real-time obtain the operation voiceprint data of the substation equipment at the operation time point, and dynamically monitor the operation status of all substation equipment in the substation.

[0011] A substation equipment operation status monitoring system based on voiceprint recognition. This system includes: a data acquisition module, a voiceprint drift model construction module, a data matching and difference degree calculation module, and a similarity calculation and analysis early warning module; The data acquisition module: obtain the operation voiceprint data and operation time period of all substation equipment in the substation during historical operation, and construct an operation time point set; The voiceprint drift model construction module: construct a voiceprint drift model based on the operation time point and the operation voiceprint data corresponding to the operation time point; The data matching and difference degree calculation module: obtain the current operation time point of the substation equipment and the operation voiceprint data corresponding to the current operation time point; obtain the operation voiceprint data corresponding to all operation time points matching the current operation time point, and calculate the comprehensive voiceprint difference degree; The similarity calculation and analysis early warning module: obtain the comprehensive difference degree of all substation equipment in the substation within the current operation time point, calculate the comprehensive difference similarity between different substation equipment, preset the threshold, and analyze and carry out linkage early warning between substation equipment.

[0012] Further, the data acquisition module includes a data acquisition unit; The data acquisition unit: acquires all the substation equipment in the substation, and arranges acoustic fingerprint vibration sensors inside the substation equipment, where one acoustic fingerprint vibration sensor is arranged corresponding to one substation equipment; stores the operation acoustic fingerprint data of the substation equipment collected by the acoustic fingerprint vibration sensors into the substation background monitoring system; retrieves the historical operation data log of the substation equipment from the substation background monitoring system, where the historical operation data log includes the operation acoustic fingerprint data and the operation time period during the historical operation of the substation equipment, evenly divides the operation time period into several operation time points, constructs an operation time point set, and records the operation time point set corresponding to the nth historical operation as , where represents the ith operation time point corresponding to the nth historical operation, and I represents the total number of operation time points.

[0013] Furthermore, the acoustic fingerprint drift model construction module includes an acoustic fingerprint drift model construction unit; The acoustic fingerprint drift model construction unit: based on the ith operation time point corresponding to the nth historical operation and the operation acoustic fingerprint data corresponding to the operation time point , constructs the acoustic fingerprint drift model corresponding to the nth historical operation, specifically as follows: traverses the operation time point set , obtains the operation acoustic fingerprint data corresponding to each operation time point, and arranges them in the order of the operation time points; performs normalization processing on the arranged operation acoustic fingerprint data, and constructs an acoustic fingerprint drift model based on the normalized operation acoustic fingerprint data, denoted as , where represents the acoustic fingerprint drift model corresponding to the nth historical operation, represents the operation acoustic fingerprint data corresponding to the normalized operation time point , and I represents the total number of operation time points.

[0014] Furthermore, the data matching and difference degree calculation module includes a data matching unit and a difference degree calculation unit; The data matching unit: obtains the current operation time point of the substation equipment and the operation acoustic fingerprint data corresponding to the current operation time point, denoted as and respectively; obtains the acoustic fingerprint drift models corresponding to a total of N historical operations, presets an operation time point matching rule, and based on the operation time point matching rule, extracts the operation time point that matches the current operation time point of the substation equipment from the acoustic fingerprint drift models; the operation time point matching rule is specifically: if , then it is determined that the operation time point matches the current operation time point , where Indicates a preset tolerance threshold for the operating duration. Indicates the m-th operating time point corresponding to the n-th historical operation. The difference degree calculation unit: obtains the operating voiceprint data corresponding to all operating time points that match the current operating time point and compares it with the operating voiceprint data corresponding to the current operating time point to calculate the comprehensive voiceprint difference degree. Specifically, the similarity calculation and analysis warning module includes a similarity calculation unit and an analysis warning unit;

[0015] Furthermore, the similarity calculation unit: obtains the comprehensive difference degrees of all substation equipment within the current operating time point and uses the cosine similarity to calculate the comprehensive difference similarity between different substation equipment. A preset comprehensive difference similarity threshold is set, and all substation equipment with a comprehensive difference similarity greater than or equal to the comprehensive difference similarity threshold is marked as similar equipment; The analysis warning unit: based on the comprehensive difference degree a preset comprehensive difference degree threshold is set. If the comprehensive difference degree is greater than or equal to the comprehensive difference degree threshold, it is determined that there is an abnormal voiceprint drift in the substation equipment at the current operating time point, that is, there is an abnormal operating state in the substation equipment at the current operating time point, and unified warnings are given to the equipment similar to the substation equipment; the operating voiceprint data of the substation equipment at the operating time point is obtained in real time to dynamically monitor the operating states of all substation equipment in the substation.

[0016] ​Compared with the prior art, the beneficial effects achieved by the present invention are as follows: In the method and system for monitoring the operation state of substation equipment based on voiceprint recognition provided by the present invention, by arranging voiceprint vibration sensors inside the substation equipment, voiceprint data of each equipment during historical operation is obtained, and the operation time point set is divided in combination with the operation time period, providing structured and high-resolution basic data for subsequent modeling of the voiceprint evolution trend and enhancing the feasibility of time series comparative analysis. Based on the normalized historical voiceprint data, a voiceprint drift model is constructed to depict the regular characteristics of the voiceprint of the equipment operation changing with time, realizing accurate modeling of the historical operation state mode of the equipment and laying a model foundation for anomaly recognition. By introducing the operation time point matching rule, the corresponding relationship between the current operation moment and the historical state is obtained, and the comprehensive voiceprint difference degree is calculated through exponential decay weighting, highlighting the dynamic difference between the current state and the historical stable state, and improving the sensitivity and robustness of anomaly detection. On the basis of extracting the comprehensive difference degree of all equipment, the mutual correlation degree between equipment is calculated by means of cosine similarity, and a dual-threshold mechanism is set for linkage warning, which not only realizes the active identification of abnormal equipment, but also constructs a risk perception network for associated equipment, enhancing the system's early intervention ability for potential fault propagation among multiple equipment. Through the whole-process modeling and analysis process driven by voiceprint data, the present invention not only improves the accuracy and real-time performance of substation equipment operation state monitoring, but also enhances the intelligent level of anomaly linkage recognition and warning response. Brief Description of the Drawings

[0017] The drawings are used to provide a further understanding of the present invention, and constitute a part of the specification. They are used together with the embodiments of the present invention to explain the present invention, and do not constitute a limitation to the present invention.

[0018] Figure 1 It is a schematic diagram of the steps of the method for monitoring the operation state of substation equipment based on voiceprint recognition of the present invention; Figure 2 It is a schematic diagram of the structure of the system for monitoring the operation state of substation equipment based on voiceprint recognition of the present invention. Detailed Embodiments

[0019] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0020] Please refer to Figure 1 , in the first embodiment: A method for monitoring the operation state of substation equipment based on voiceprint recognition is provided, and the method includes the following steps: Step S1: Obtain the running voiceprint data and running time period of all substation equipment during historical operation, and construct a set of running time points.

[0021] Specifically, obtain all the substation equipment in the substation, and deploy voiceprint vibration sensors inside the substation equipment. One voiceprint vibration sensor is deployed corresponding to one substation equipment; store the running voiceprint data of the substation equipment collected by the voiceprint vibration sensor into the substation background monitoring system. Furthermore, retrieve the historical operation data log of the substation equipment from the substation background monitoring system. The historical operation data log includes the running voiceprint data and running time period of the substation equipment during historical operation. Divide the running time period evenly into several running time points to construct a set of running time points. Denote the set of running time points corresponding to the nth historical operation as where represents the ith running time point corresponding to the nth historical operation, and I represents the total number of running time points.

[0022] Step S2: Construct a voiceprint drift model based on the running time point and the running voiceprint data corresponding to the running time point.

[0023] Specifically, based on the ith running time point corresponding to the nth historical operation and the running voiceprint data corresponding to the running time point construct the voiceprint drift model corresponding to the nth historical operation as follows: Traverse the set of running time points to obtain the running voiceprint data corresponding to each running time point, and arrange them in the order of the running time points. Perform normalization processing on the arranged running voiceprint data. Based on the normalized running voiceprint data, construct a voiceprint drift model, denoted as where represents the voiceprint drift model corresponding to the nth historical operation, represents the running voiceprint data corresponding to the normalized running time point and I represents the total number of running time points.

[0024] In the present invention, each time point is bound to the normalized voiceprint feature to form a "time - voiceprint" trajectory, intuitively depicting the dynamic evolution of the equipment operation state. If the current voiceprint is abnormal, the data of the same historical time point in the model can be traced back to analyze whether the abnormality is a periodic fluctuation (such as a normal change caused by the start and stop of a fan).

[0025] Step S3: Obtain the current operating time point of the power transformation equipment and the corresponding operating voiceprint data at the current operating time point; obtain the operating voiceprint data corresponding to all operating time points that match the current operating time point, and calculate the comprehensive voiceprint difference degree.

[0026] Specifically, obtain the current operating time point of the power transformation equipment and the corresponding operating voiceprint data at the current operating time point, and record them as and respectively; obtain the voiceprint drift model corresponding to the historical N operations, preset the operating time point matching rule, and based on the operating time point matching rule, extract from the voiceprint drift model the operating time point that matches the current operating time point of the power transformation equipment ; the specific operating time point matching rule is: if , then it is determined that the operating time point matches the current operating time point , where represents the preset tolerance threshold of the operating duration, represents the m-th operating time point corresponding to the n-th historical operation; Furthermore, obtain the operating voiceprint data corresponding to all operating time points that match the current operating time point , and compare it with the operating voiceprint data corresponding to the current operating time point to calculate the comprehensive voiceprint difference degree. The calculation formula is as follows: ; where represents the comprehensive difference degree, M represents the total number of operating voiceprint data corresponding to all operating time points that match the current operating time point , represents the operating voiceprint data corresponding to the operating time point , represents the preset exponential decay factor.

[0027] In the present invention, invalid historical points can be filtered through threshold screening (such as For 2 hours, only historical data within ±2 hours of the current time are retained), focusing on comparing similar operating conditions to avoid misjudgment across different operating conditions; in the complex environment of the substation (such as intermittent fan noise, lightning strikes), it can accurately identify abnormal sound patterns of the equipment itself rather than pseudo-abnormalities caused by environmental interference. By setting time matching rules (including a tolerance mechanism) to compare the current sound pattern with the matching nodes in the historical drift model, and introducing an exponential decay factor to weight the degree of difference, a comprehensive difference index reflecting the difference between the current state and the historical state is constructed. The introduction of a dynamic matching and decay mechanism makes the calculation of the difference degree not only consider the distance but also reflect the difference in historical weights, strengthening the sensitivity of the model to deviations from the latest state.

[0028] Step S4: Obtain the comprehensive difference degree of all substation equipment within the current operating time point, calculate the comprehensive difference similarity between different substation equipment, preset a threshold, and analyze and conduct linkage warnings between substation equipment.

[0029] Specifically, obtain the current operating time point within which, obtain the comprehensive difference degree of all substation equipment in the substation, use cosine similarity to calculate the comprehensive difference similarity between different substation equipment, preset a comprehensive difference similarity threshold, and mark all substation equipment with a comprehensive difference similarity greater than or equal to the comprehensive difference similarity threshold as similar equipment; Further, based on the comprehensive difference degree , preset a comprehensive difference degree threshold. If the comprehensive difference degree is greater than or equal to the comprehensive difference degree threshold, it is determined that there is an abnormal sound pattern drift of the substation equipment at the current operating time point, that is, there is an abnormal operating state of the substation equipment at the current operating time point, and a unified warning is issued for the equipment similar to the substation equipment; Real-time obtain the operating sound pattern data of the substation equipment at the operating time point, and conduct dynamic monitoring of the operating states of all substation equipment in the substation.

[0030] Please refer to Figure 2 , in the second embodiment: Provide a monitoring system for the operating state of substation equipment based on sound pattern recognition, which includes: a data acquisition module, a sound pattern drift model construction module, a data matching and difference degree calculation module, and a similarity calculation and analysis warning module; The data acquisition module: Obtain the operating sound pattern data and operating time periods of all substation equipment during historical operation in the substation, and construct an operating time point set; The sound pattern drift model construction module: Based on the operating time point and the corresponding operating sound pattern data, construct a sound pattern drift model; The data matching and difference degree calculation module: obtains the current operation time point of the substation equipment and the operation voiceprint data corresponding to the current operation time point; obtains the operation voiceprint data corresponding to all operation time points matching the current operation time point, and calculates the comprehensive voiceprint difference degree. The similarity calculation and analysis warning module: obtains the comprehensive difference degree of all substation equipment in the substation within the current operation time point, calculates the comprehensive difference similarity between different substation equipment, sets a preset threshold, and analyzes and conducts linkage warnings between substation equipment.

[0031] Furthermore, the data acquisition module includes a data acquisition unit. The data acquisition unit: obtains all substation equipment in the substation, and arranges voiceprint vibration sensors inside the substation equipment, where one voiceprint vibration sensor is arranged corresponding to one substation equipment; stores the operation voiceprint data of the substation equipment collected by the voiceprint vibration sensors into the substation background monitoring system; retrieves the historical operation data log of the substation equipment from the substation background monitoring system, and the historical operation data log includes the operation voiceprint data and the operation time period during the historical operation of the substation equipment, evenly divides the operation time period into several operation time points, constructs an operation time point set, and records the operation time point set corresponding to the nth historical operation as , where represents the ith operation time point corresponding to the nth historical operation, and I represents the total number of operation time points.

[0032] Furthermore, the voiceprint drift model construction module includes a voiceprint drift model construction unit. The voiceprint drift model construction unit: based on the ith operation time point corresponding to the nth historical operation and the operation voiceprint data corresponding to the operation time point , constructs the voiceprint drift model corresponding to the nth historical operation, specifically as follows: traverses the operation time point set , obtains the operation voiceprint data corresponding to each operation time point, and arranges them in the order of the operation time points; performs normalization processing on the arranged operation voiceprint data, and constructs a voiceprint drift model based on the normalized operation voiceprint data, denoted as , where represents the voiceprint drift model corresponding to the nth historical operation, represents the operation voiceprint data corresponding to the normalized operation time point , and I represents the total number of operation time points.

[0033] Furthermore, the data matching and difference degree calculation module includes a data matching unit and a difference degree calculation unit. The data matching unit: obtains the current operation time point of the power transformation equipment and the operation voiceprint data corresponding to the current operation time point, denoted as and respectively; obtains the voiceprint drift models corresponding to the historical N operations, presets the operation time point matching rule, and based on the operation time point matching rule, extracts from the voiceprint drift models the operation time point matched with the current operation time point of the power transformation equipment; the operation time point matching rule is specifically: if , then it is determined that the operation time point is matched with the current operation time point , where represents the preset operation duration tolerance threshold, and represents the m-th operation time point corresponding to the n-th historical operation; The difference degree calculation unit: obtains the operation voiceprint data corresponding to all operation time points matched with the current operation time point , and compares it with the operation voiceprint data corresponding to the current operation time point to calculate the comprehensive voiceprint difference degree.

[0034] Furthermore, the similarity calculation and analysis warning module includes a similarity calculation unit and an analysis warning unit; The similarity calculation unit: obtains the comprehensive difference degree of all power transformation equipment in the substation within the current operation time point , uses the cosine similarity to calculate the comprehensive difference similarity between different power transformation equipment, presets the comprehensive difference similarity threshold, and marks all power transformation equipment with a comprehensive difference similarity greater than or equal to the comprehensive difference similarity threshold as similar equipment; The analysis warning unit: based on the comprehensive difference degree , presets the comprehensive difference degree threshold, if the comprehensive difference degree is greater than or equal to the comprehensive difference degree threshold, then it is determined that there is an abnormal voiceprint drift in the power transformation equipment at the current operation time point, that is, there is an abnormal operation state in the power transformation equipment at the current operation time point, and then a unified warning is given to the equipment similar to the power transformation equipment; obtains the operation voiceprint data of the power transformation equipment at the operation time point in real time, and dynamically monitors the operation states of all power transformation equipment in the substation.

[0035] It should be noted that in this text, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprising", "including" or any other variant thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements not only includes those elements, but also includes other elements not expressly listed, or elements inherent to such process, method, article or device.

[0036] Finally, it should be noted that the above are only preferred embodiments of the present invention and are not used to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A method for monitoring the operating state of substation equipment based on voiceprint recognition, characterized in that, The method includes the following steps: Step S1: Obtain the operation voiceprint data and operation time period of all substation equipment during historical operation, and construct an operation time point set; Step S2: Based on the operation time points and the operation voiceprint data corresponding to the operation time points, construct a voiceprint drift model; Step S3: Obtain the current operation time point of the substation equipment and the operation voiceprint data corresponding to the current operation time point; obtain the operation voiceprint data corresponding to all operation time points that match the current operation time point, and calculate the comprehensive voiceprint difference degree; Step S4: Obtain the comprehensive difference degree of all substation equipment in the substation at the current operation time point, calculate the comprehensive difference similarity between different substation equipment, preset a threshold, and analyze and conduct linkage early warning between substation equipment.

2. The method for monitoring the operation status of substation equipment based on voiceprint recognition according to claim 1, wherein, The specific implementation process of step S1 includes: Obtain all substation equipment in the substation, and deploy voiceprint vibration sensors inside the substation equipment. One voiceprint vibration sensor is deployed corresponding to one substation equipment; store the operation voiceprint data of the substation equipment collected by the voiceprint vibration sensors into the substation background monitoring system; Retrieve the historical operation data log of the substation equipment from the background monitoring system of the substation. The historical operation data log includes the operation voiceprint data and the operation time period of the substation equipment during historical operation. Divide the operation time period evenly into several operation time points to construct an operation time point set, and denote the operation time point set corresponding to the nth historical operation as , where represents the ith operation time point corresponding to the nth historical operation, and I represents the total number of operation time points.

3. The method for monitoring the operation status of substation equipment based on voiceprint recognition according to claim 2, characterized in that, The specific implementation process of step S2 includes: Based on the i-th running time point corresponding to the n-th historical run and the running time point corresponding running voiceprint data, construct the voiceprint drift model corresponding to the n-th historical run, specifically as follows: Traverse the set of running time points to obtain the running voiceprint data corresponding to each running time point, and arrange them in the order of the running time points; Normalize the arranged running voiceprint data, and construct a voiceprint drift model based on the normalized running voiceprint data, denoted as , where represents the voiceprint drift model corresponding to the nth historical run, represents the running time point after normalization corresponding running voiceprint data, and I represents the total number of running time points.

4. The method for monitoring the operation state of substation equipment based on voiceprint recognition according to claim 3, characterized in that, The specific implementation process of step S3 includes: Obtain the current operating time point of the power transformation equipment and the operating voiceprint data corresponding to the current operating time point, denoted as and ; obtain the voiceprint drift model corresponding to a total of N historical operations, preset an operating time point matching rule, and based on the operating time point matching rule, extract from the voiceprint drift model the operating time point that matches the current operating time point of the power transformation equipment ; the specific operating time point matching rule is: if , then it is determined that the operating time point matches the current operating time point , where represents the preset tolerance threshold for the operating duration, represents the m-th operating time point corresponding to the n-th historical operation; Obtain all the running voiceprint data corresponding to the running time points that match the current running time point and compare the difference with the running voiceprint data corresponding to the current running time point corresponding running voiceprint data to perform a difference comparison and calculate the comprehensive voiceprint difference degree. The calculation formula is as follows: ； Among them, represents the comprehensive difference degree, and M represents the total number of running voiceprint data corresponding to all running time points matched with the current running time point, represents the running time point corresponding running voiceprint data, represents a preset exponential decay factor.

5. The method for monitoring the operation status of substation equipment based on voiceprint recognition according to claim 4, wherein The specific implementation process of step S4 includes: Obtain the current running time point Within , for the comprehensive difference degree of all substation electrical equipment in the substation, use the cosine similarity to calculate the comprehensive difference similarity between different electrical equipment, preset the comprehensive difference similarity threshold, and mark all electrical equipment with a comprehensive difference similarity greater than or equal to the comprehensive difference similarity threshold as similar equipment; Based on the comprehensive difference degree , a preset comprehensive difference degree threshold is set. If the comprehensive difference degree is greater than or equal to the comprehensive difference degree threshold, it is determined that there is an abnormal voiceprint drift in the substation equipment at the current operation time point, that is, there is an abnormal operation state in the substation equipment at the current operation time point, and unified early warning is carried out for the equipment similar to the substation equipment; Obtain the operation voiceprint data of the substation equipment at the operation time point in real time, and dynamically monitor the operation status of all substation equipment in the substation.

6. A substation equipment operation status monitoring system based on voiceprint recognition, which executes the method for monitoring the operation status of substation equipment based on voiceprint recognition according to any one of claims 1-5, characterized in that, The system includes: a data acquisition module, a voiceprint drift model construction module, a data matching and difference degree calculation module, and a similarity calculation and analysis early warning module; The data acquisition module: Obtain the operation voiceprint data and operation time period of all substation equipment during historical operation, and construct an operation time point set; The voiceprint drift model construction module: Based on the operation time points and the operation voiceprint data corresponding to the operation time points, construct a voiceprint drift model; The data matching and difference degree calculation module: Obtain the current operation time point of the substation equipment and the operation voiceprint data corresponding to the current operation time point; obtain the operation voiceprint data corresponding to all operation time points that match the current operation time point, and calculate the comprehensive voiceprint difference degree; The similarity calculation and analysis early warning module: Obtain the comprehensive difference degree of all substation equipment in the substation at the current operation time point, calculate the comprehensive difference similarity between different substation equipment, preset a threshold, and analyze and conduct linkage early warning between substation equipment.

7. The substation equipment operation status monitoring system based on voiceprint recognition according to claim 6, characterized in that: The data acquisition module includes a data acquisition unit; The data acquisition unit: acquires all substation equipment within the substation, and arranges acoustic fingerprint vibration sensors inside the substation equipment, where one acoustic fingerprint vibration sensor is arranged corresponding to one substation equipment; stores the operation acoustic fingerprint data of the substation equipment collected by the acoustic fingerprint vibration sensors into the substation background monitoring system; retrieves the historical operation data log of the substation equipment from the substation background monitoring system, where the historical operation data log includes the operation acoustic fingerprint data and the operation time period during the historical operation of the substation equipment, evenly divides the operation time period into several operation time points, constructs an operation time point set, and records the operation time point set corresponding to the nth historical operation as , where represents the ith operation time point corresponding to the nth historical operation, and I represents the total number of operation time points.

8. The substation equipment operation status monitoring system based on voiceprint recognition according to claim 7, characterized in that: The voiceprint drift model construction module includes a voiceprint drift model construction unit; The voiceprint drift model construction unit: Based on the i-th running time point corresponding to the n-th historical run and the running time point corresponding running voiceprint data, construct a voiceprint drift model corresponding to the n-th historical run, specifically as follows: Traverse the running time point set to obtain the running voiceprint data corresponding to each running time point, and arrange them in the order of the running time points; Normalize the arranged running voiceprint data, and construct a voiceprint drift model based on the normalized running voiceprint data, denoted as , where represents the voiceprint drift model corresponding to the nth historical run, represents the running time point after normalization corresponding running voiceprint data, and I represents the total number of running time points.

9. The substation equipment operation status monitoring system based on voiceprint recognition according to claim 8, wherein: The data matching and difference degree calculation module includes a data matching unit and a difference degree calculation unit; The data matching unit: obtains the current operating time point of the substation equipment and the operating voiceprint data corresponding to the current operating time point, denoted as and ; Obtain the voiceprint drift models corresponding to a total of N historical runs, preset a running time point matching rule, and based on the running time point matching rule, extract from the voiceprint drift models the running time point that matches the current running time point of the substation equipment that matches; the running time point matching rule is specifically: if , then it is determined that the running time point matches the current running time point , where represents the preset running duration tolerance threshold, represents the m-th running time point corresponding to the n-th historical run; The difference calculation unit: obtains the running voiceprint data corresponding to all running time points that match the current running time point and compares it with the running voiceprint data corresponding to the current running time point to calculate the comprehensive voiceprint difference by comparing the difference 10. The substation equipment operation status monitoring system based on voiceprint recognition according to claim 9, wherein: The similarity calculation and analysis early warning module includes a similarity calculation unit and an analysis early warning unit; The similarity calculation unit: obtains the comprehensive difference degree of all substation equipment within the current running time point and uses the cosine similarity to calculate the comprehensive difference similarity between different substation equipment. A preset comprehensive difference similarity threshold is set, and all substation equipment with a comprehensive difference similarity greater than or equal to the comprehensive difference similarity threshold is marked as similar equipment; The analysis and early warning unit: Based on the comprehensive difference degree , a comprehensive difference degree threshold is preset. If the comprehensive difference degree is greater than or equal to the comprehensive difference degree threshold, it is determined that there is a voiceprint drift anomaly in the substation equipment at the current operation time point, that is, there is an abnormal operation state in the substation equipment at the current operation time point, and unified early warning is carried out for the equipment similar to the substation equipment; Obtain the operation voiceprint data of the substation equipment at the operation time point in real time, and dynamically monitor the operation status of all substation equipment in the substation.

Images