A method for non-stop detection of the acoustic fingerprint of a transformer and status warning

Abstract

The present invention provides a method for non-power-off detection of the acoustic fingerprint of a transformer and status warning, which includes the following steps: collecting the acoustic fingerprint during the operation of the transformer; collecting a plurality of historical current values and the corresponding first historical noise values during the operation of the transformer; fitting the plurality of historical current values and the corresponding first historical noise values to obtain a current noise function curve; collecting the real-time operating current and the corresponding first real-time noise value during the operation of the transformer; calculating a first predicted noise value according to the real-time operating current and the current noise function curve; performing noise cancellation processing on the acoustic fingerprint according to the first predicted noise value and the first real-time noise value; performing non-power-off detection on the real-time status of the transformer and warning of the fault status according to the acoustic fingerprint after the noise cancellation processing. The present invention can perform non-power-off detection on the real-time status of the transformer and warning of the fault status.

Description

Technical Field

[0001] The present invention relates to the technical field of electrical equipment detection, and more particularly to a method for non-power-off detection of transformer acoustic fingerprints and status warning. Background Art

[0002] During the operation of a transformer, mechanical deformation occurs under the combined action of various factors such as internal current and magnetic field. Through its own structure conduction, it is manifested as a vibration signal, and this signal is transmitted through the surrounding air medium, generating the sound signal of the transformer operation. These signals can largely reflect the operating conditions of the transformer. In on-site inspection operations, experienced technicians can often judge the operating conditions of the transformer, discover the types of faults and even roughly locate the faults by closely attaching an industrial stethoscope to the transformer box and carefully listening to the sound inside the transformer. This widely used diagnostic method relies heavily on the subjective judgment and personal experience of technicians and has great uncertainty.

[0003] To solve the above problems, Chinese Patent Invention No. 202210002728.2 discloses a transformer diagnosis method and device based on acoustic wave interference. The method includes: collecting the acoustic signal of the transformer; sequentially performing signal enhancement and signal filtering on the acoustic signal to obtain a filtered acoustic signal; inputting the filtered acoustic signal into a fault model to output a fault state and annotation corresponding to the filtered acoustic signal; and realizing fault diagnosis of the transformer based on the fault state and the annotation. This method is convenient for obtaining acoustic signals without contacting the equipment, does not generate electromagnetic signals during signal acquisition, does not interfere with the normal operation of the equipment, and is simple and convenient; in addition, the device is simple and the sensor installation is relatively flexible.

[0004] The above-mentioned transformer diagnosis method based on acoustic wave interference can only diagnose the transformer through acoustic signals when the transformer fails, and cannot warn of the transformer fault state, so it needs to be improved. Summary of the Invention

[0005] Based on this, in order to solve the problem that the existing transformer diagnosis method based on acoustic wave interference cannot warn of the transformer fault state, the present invention provides a method for non-power-off detection of transformer acoustic fingerprints and status warning, and its specific technical solution is as follows:

[0006] A method for non-power-off detection of transformer acoustic fingerprints and status warning, which includes the following steps:

[0007] Collect the acoustic fingerprints during the operation of the transformer;

[0008] Collect multiple historical current values during the operation of the transformer and the first historical noise values corresponding to the multiple historical current values;

[0009] Fit multiple historical current values and the corresponding first historical noise values of the multiple historical current values to obtain a current noise function curve;

[0010] Collect the real-time operating current during the operation of the transformer and the corresponding first real-time noise value;

[0011] Calculate the first predicted noise value according to the real-time operating current and the current noise function curve;

[0012] Perform noise cancellation processing on the acoustic fingerprint according to the first predicted noise value and the first real-time noise value;

[0013] Perform live detection on the real-time state of the transformer and give an early warning of the fault state according to the acoustic fingerprint after noise cancellation processing.

[0014] The method for live detection and state warning of the transformer acoustic fingerprint obtains a current noise function curve by collecting multiple historical current values during the operation of the transformer and the corresponding first historical noise values of the multiple historical current values and fitting the multiple historical current values and the corresponding first historical noise values. By performing noise cancellation processing on the acoustic fingerprint according to the first predicted noise value and the first real-time noise value, the acoustic fingerprint during the operation of the transformer can be obtained more accurately, and the accuracy of monitoring the operation state of the transformer through the acoustic fingerprint is improved.

[0015] After performing noise cancellation processing on the acoustic fingerprint according to the first predicted noise value and the first real-time noise value, perform live detection on the real-time state of the transformer and give an early warning of the fault state using the acoustic fingerprint after noise cancellation processing. The method for live detection and state warning of the transformer acoustic fingerprint can not only perform live detection on the real-time state of the transformer, but also give an early warning of the fault state of the transformer, send the early warning information to the technical personnel in advance, facilitate the technical personnel to perform timely maintenance and repair on the transformer, and ensure the normal operation of the transformer.

[0016] Further, the multiple historical current values include no-load current values, and the method further includes the following steps:

[0017] Obtain the first historical noise value N' corresponding to the no-load current value;

[0018] According to the formula N = N' + λlgI' - λlgI max Calculate the second predicted noise value corresponding to the real-time operating current I';

[0019] where I max represents the rated current of the transformer, and λ represents the noise adjustment coefficient.

[0020] Further, the noise cancellation process for the acoustic fingerprint based on the first predicted noise value and the first real-time noise value is specifically as follows: The noise cancellation process for the acoustic fingerprint is performed according to the first predicted noise value, the second predicted noise value, and the first real-time noise value.

[0021] Further, the method further includes the following steps: Calibrating the current noise function curve according to the second predicted noise value.

[0022] Further, the specific method for calibrating the current noise function curve according to the second predicted noise value includes the following steps:

[0023] Calculating a first noise difference between the first predicted noise value and the second predicted noise value corresponding to the current noise function curve at the same real-time operating current;

[0024] Determining whether the first noise difference is greater than a first preset threshold. If so, calculating the noise mean value of the first historical noise values corresponding to multiple historical current values equal to the real-time operating current. If not, refitting the multiple historical current values, the second predicted noise value, and the first historical noise values corresponding to the multiple historical current values to obtain a new current noise function curve, thereby completing the calibration of the current noise function curve;

[0025] Determining whether a second noise difference between the noise mean value and the second predicted noise value is greater than a second preset threshold. If so, excluding the second predicted noise value.

[0026] Further, the method further includes the following steps:

[0027] When the second noise difference is less than or equal to the second preset threshold, refitting the multiple historical current values, the second predicted noise value, and the first historical noise values corresponding to the multiple historical current values to obtain a new current noise function curve, thereby completing the calibration of the current noise function curve.

[0028] Further, the method further includes the following steps:

[0029] Obtaining the fault category of the transformer and the fault sound wave corresponding to the fault category;

[0030] Obtaining a fault sample set according to the fault sound wave.

[0031] Further, the noise adjustment coefficient λ is equal to 10.

[0032] Further, a computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the transformer acoustic fingerprint non-power-off detection and status warning method is implemented. Description of the Drawings

[0033] The present invention can be further understood from the following description in conjunction with the accompanying drawings. The components in the drawings are not necessarily drawn to scale, but the emphasis is placed on showing the principles of the embodiments. In different views, the same reference numerals designate corresponding parts.

[0034] Figure 1 It is a schematic diagram of the overall process of a method for non - power - off detection and status warning of the acoustic fingerprint of a transformer in an embodiment of the present invention. Detailed implementation manners

[0035] In order to make the objectives, technical solutions and advantages of the present invention clearer and more understandable, the present invention will be further described in detail below in conjunction with its embodiments. It should be understood that the specific implementation manners described herein are only used to explain the present invention and do not limit the protection scope of the present invention.

[0036] It should be noted that when an element is referred to as being "fixed to" another element, it can be directly on the other element or there can also be an intermediate element. When an element is considered to be "connected" to another element, it can be directly connected to the other element or there may be an intermediate element at the same time. The terms "vertical", "horizontal", "left", "right" and similar expressions used herein are only for the purpose of illustration and do not represent the only implementation manners.

[0037] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the technical field to which the present invention belongs. The terms used herein in the description of the present invention are only for the purpose of describing specific implementation manners and are not intended to limit the present invention. The term "and / or" used herein includes any and all combinations of one or more of the related listed items.

[0038] The "first" and "second" described in the present invention do not represent specific quantities and sequences, but are only used for name distinction.

[0039] As Figure 1 shown, a method for non - power - off detection and status warning of the acoustic fingerprint of a transformer in an embodiment of the present invention includes the following steps:

[0040] S1. Collect the acoustic fingerprint during the operation of the transformer. The acoustic fingerprint here refers to the sound signal formed by the vibration signal of the transformer during operation, which is propagated through the air medium under the action of various factors such as current and magnetic field.

[0041] S2. Collect multiple historical current values during the operation of the transformer and the corresponding first historical noise values of the multiple historical current values.

[0042] The historical current values are collected by a current sensor, while the first historical noise values can be collected by a noise sensor.

[0043] The historical current values and the first historical noise values correspond one by one and are stored in a server or a controller.

[0044] S3. Fit multiple historical current values and the corresponding first historical noise values of the multiple historical current values to obtain a current noise function curve. Here, since the data fitting method belongs to conventional technical means, it will not be elaborated here.

[0045] S4. Collect the real-time operating current during the operation of the transformer and the corresponding first real-time noise value.

[0046] S5. Calculate the first predicted noise value according to the real-time operating current and the current noise function curve.

[0047] S6. Perform noise cancellation processing on the acoustic fingerprint according to the first predicted noise value and the first real-time noise value. Here, the noise cancellation processing can be achieved by superimposing the first predicted noise value and the first real-time noise value on the acoustic fingerprint.

[0048] Of course, it is also possible to set a corresponding filter according to the amplitudes of the first predicted noise value and the first real-time noise value, and then filter the acoustic fingerprint through the set filter to achieve noise cancellation processing.

[0049] S7. Perform live detection on the real-time state of the transformer and give an early warning for the fault state according to the acoustic fingerprint after noise cancellation processing.

[0050] Specifically, the fault category of the transformer and the fault sound wave corresponding to the fault category can be obtained first, and then the fault sample set can be obtained according to the fault sound wave. After obtaining the fault sample set, the real-time state of the transformer is detected without power-off according to the matching degree between the acoustic fingerprint after noise cancellation processing and the fault sample set. At the same time, according to the similarity between the acoustic fingerprint after noise cancellation processing and the fault sample set, an early warning is given for the fault state of the transformer according to the similarity.

[0051] That is to say, when the similarity is greater than the preset similarity threshold, it is judged that the transformer may or is about to have a fault, and an early warning is given in time and the warning information is fed back to the technical personnel or the manager.

[0052] The method for non-power-off detection and status warning of the transformer acoustic fingerprint collects multiple historical current values and the corresponding first historical noise values during the operation of the transformer, fits the multiple historical current values and the corresponding first historical noise values, obtains the current noise function curve, and performs noise cancellation processing on the acoustic fingerprint according to the first predicted noise value and the first real-time noise value, so as to more accurately obtain the acoustic fingerprint during the operation of the transformer and improve the accuracy of monitoring the operation status of the transformer through the acoustic fingerprint.

[0053] After performing noise cancellation processing on the acoustic fingerprint according to the first predicted noise value and the first real-time noise value, the non-power-off detection of the real-time status of the transformer and the warning of the fault status are carried out by using the acoustic fingerprint after noise cancellation processing. The method for non-power-off detection and status warning of the transformer acoustic fingerprint can not only perform non-power-off detection on the real-time status of the transformer, but also warn of the fault status of the transformer, send the warning information to the technical personnel in advance, facilitate the technical personnel to perform timely maintenance and repair on the transformer, and ensure the normal operation of the transformer.

[0054] In one embodiment, the multiple historical current values include no-load current values, and the method further includes the following steps:

[0055] Obtain the first historical noise value N' corresponding to the no-load current value;

[0056] According to the formula N = N' + λlgI' - λlgI max Calculate the second predicted noise value corresponding to the real-time operating current I';

[0057] Wherein, I max represents the rated current of the transformer, and λ represents the noise adjustment coefficient. Preferably, the noise adjustment coefficient λ is equal to 10.

[0058] Here, the first historical noise value N' corresponding to the no-load current value of the transformer can be obtained through a noise sensor. Of course, the first historical noise value N' corresponding to the no-load current value of the transformer can also be obtained through the current noise function curve.

[0059] As a preferred technical solution, the specific method for obtaining the first historical noise value N' corresponding to the no-load current value is: first obtain the predicted no-load current noise value corresponding to the no-load current value of the transformer through the current noise function curve, and obtain the real-time no-load current noise values corresponding to multiple no-load current values of the transformer in real time through a noise sensor, and then calculate the no-load current noise mean value of the predicted no-load current noise value and the real-time no-load current noise values corresponding to multiple no-load current values of the transformer, and use the no-load current noise mean value as the first historical noise value N'.

[0060] Through the above solution, by combining the predicted no-load current noise value and the real-time no-load current noise values corresponding to multiple transformer no-load current values, the first historical noise value N' can be calculated more accurately.

[0061] After calculating the first historical noise value, according to the formula N = N'+λlgI'-λlgI max , the second predicted noise value corresponding to the real-time operating current I' can also be calculated more conveniently and accurately.

[0062] In one embodiment, the noise cancellation processing of the acoustic fingerprint according to the first predicted noise value and the first real-time noise value is specifically as follows: the noise cancellation processing of the acoustic fingerprint is performed according to the first predicted noise value, the second predicted noise value, and the first real-time noise value.

[0063] Specifically, multiple corresponding filters can be constructed according to the amplitudes of the first predicted noise value, the second predicted noise value, and the first real-time noise value, and then the noise cancellation processing of the acoustic fingerprint is performed through the filters.

[0064] Performing the noise cancellation processing of the acoustic fingerprint according to the first predicted noise value, the second predicted noise value, and the first real-time noise value can make the acoustic fingerprint more consistent with the sound signal generated during the real-time operation of the transformer.

[0065] In one embodiment, the method further includes the following steps: S8, correcting the current noise function curve according to the second predicted noise value.

[0066] Specifically, in step S8, the specific method for correcting the current noise function curve according to the second predicted noise value includes the following steps:

[0067] S80, calculating the first noise difference between the first predicted noise value and the second predicted noise value corresponding to the current noise function curve under the same real-time operating current.

[0068] S81, determining whether the first noise difference is greater than the first preset threshold. If so, calculating the noise mean value of the first historical noise values corresponding to multiple historical current values equal to the real-time operating current. If not, re-fitting the multiple historical current values, the second predicted noise value, and the first historical noise values corresponding to the multiple historical current values to obtain a new current noise function curve, and completing the correction of the current noise function curve.

[0069] S82, determining whether the second noise difference between the noise mean value and the second predicted noise value is greater than the second preset threshold. If so, removing the second predicted noise value.

[0070] S83. When the second noise difference is less than or equal to the second preset threshold, re-fit the multiple historical current values, the second predicted noise value, and the first historical noise values corresponding to the multiple historical current values to obtain a new current noise function curve, thereby completing the correction of the current noise function curve.

[0071] Based on the first predicted noise value and the second predicted noise value, a current noise function curve that better fits the actual situation can be obtained. When the second noise difference between the noise mean value and the second predicted noise value is greater than the second preset threshold, the second predicted noise value is excluded. And when the first noise difference is less than or equal to the first preset threshold, re-fit the multiple historical current values, the second predicted noise value, and the first historical noise values corresponding to the multiple historical current values, and the current noise function curve can be double-corrected according to the first predicted noise value and the second predicted noise value.

[0072] After obtaining the new current noise function curve, first re-calculate the first predicted noise value according to the new current noise function curve and the real-time operating current, then perform noise cancellation processing on the acoustic fingerprint according to the first predicted noise value and the first real-time noise value, and finally perform live detection on the real-time state of the transformer and give an early warning of the fault state according to the acoustic fingerprint after the noise processing. In this way, after double-correcting the current noise function curve, the live detection of the real-time state of the transformer and the early warning of the fault state can be more accurate.

[0073] In one of the embodiments, a computer-readable storage medium stores a computer program, which implements the method for live detection and state early warning of the transformer acoustic fingerprint when the computer program is executed by a processor.

[0074] The technical features of the above-described embodiments can be combined arbitrarily. For the sake of brevity of description, not all possible combinations of the technical features in the above-described embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as the scope described in this specification.

[0075] The above-described embodiments merely represent several implementation manners of the present invention. The description is relatively specific and detailed, but it should not be construed as a limitation on the scope of the invention patent. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present invention, several modifications and improvements can still be made, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the present invention patent shall be subject to the appended claims.

Claims

1. A method for non-stop detection of the acoustic fingerprint of a transformer and status warning, characterized in that, The steps include: Collect the acoustic fingerprint of the transformer during operation; Collecting a plurality of historical current values during the operation of the transformer and first historical noise values corresponding to the plurality of historical current values; Fitting a plurality of historical current values and first historical noise values corresponding to the plurality of historical current values to obtain a current noise function curve; Collecting the real-time operating current and the corresponding first real-time noise value during the operation of the transformer; Calculate a first predicted noise value according to the real-time operating current and the current noise function curve; Performing noise elimination processing on the acoustic fingerprint according to the first predicted noise value and the first real-time noise value; The real-time status of the transformer is detected without power outage and fault status is warned based on the acoustic fingerprint after noise elimination. The plurality of historical current values include a no-load current value, and the method further comprises the following steps: Acquire a first historical noise value N' corresponding to the no-load current value; According to the formula N = N' + λlgI' - λlgI max calculate the second predicted noise value corresponding to the real-time operating current I'; where I max represents the rated current of the transformer, and λ represents the noise adjustment coefficient.

2. The method for non-power-off detection of the acoustic fingerprint of a transformer and status early warning according to claim 1, characterized in that, Performing noise elimination processing on the acoustic fingerprint according to the first predicted noise value and the first real-time noise value specifically includes: performing noise elimination processing on the acoustic fingerprint according to the first predicted noise value, the second predicted noise value and the first real-time noise value.

3. The method for non-power-off detection of the acoustic fingerprint of a transformer and status warning according to claim 2, characterized in that The method further comprises the following step: correcting the current noise function curve according to the second predicted noise value.

4. The method for non-power-off detection of the acoustic fingerprint of a transformer and status warning according to claim 3, characterized in that, The specific method for correcting the current noise function curve according to the second predicted noise value comprises the following steps: Calculating a first noise difference between a first predicted noise value and a second predicted noise value corresponding to a current noise function curve under the same real-time operating current; Determine whether the first noise difference is greater than a first preset threshold value; if so, calculate a noise mean of the first historical noise value corresponding to a plurality of historical current values equal to the real-time operating current; if not, refit the plurality of historical current values, the second predicted noise value, and the first historical noise value corresponding to the plurality of historical current values to obtain a new current noise function curve, and complete the correction of the current noise function curve; It is determined whether a second noise difference between the noise mean and the second predicted noise value is greater than a second preset threshold; if so, the second predicted noise value is discarded.

5. The method for non-power-off detection of the acoustic fingerprint of a transformer and status warning according to claim 4, characterized in that, The method further comprises the steps of: When the second noise difference is less than or equal to the second preset threshold, the multiple historical current values, the second predicted noise value and the first historical noise values corresponding to the multiple historical current values are refitted to obtain a new current noise function curve to complete the correction of the current noise function curve.

6. The method for non-power-off detection of the acoustic fingerprint of a transformer and status warning according to claim 5, characterized in that The method further comprises the steps of: Obtaining a fault type of the transformer and a fault sound wave corresponding to the fault type; A fault sample set is obtained according to the fault sound wave.

7. The method for non-power-off detection of the acoustic fingerprint of a transformer and status warning according to claim 6, characterized in that, The noise adjustment coefficient λ is equal to 10.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the transformer acoustic fingerprint non-stop power-off detection and status warning method according to any one of claims 1 to 7 is implemented.

Images