Condition monitoring and diagnosis method of centrifugal pump based on cavitation noise

Abstract

The present invention discloses a centrifugal pump condition monitoring and diagnosis method based on cavitation noise, which belongs to the technical field of intelligent monitoring of centrifugal pumps. The monitoring and diagnosis method obtains the distribution of cavitation bubbles on the centrifugal pump blades at different speeds through a high-speed image acquisition camera to verify the distribution of cavitation bubbles on the centrifugal pump blades at different speeds obtained through modeling simulation, and adjusts the simulation model according to actual conditions. By introducing acoustic information during the cavitation occurrence process into the adjusted simulation model, cavitation noise information corresponding to different cavitation stages is obtained. When the same batch of centrifugal pumps is subjected to production inspection, the corresponding cavitation noise information at different speeds is obtained. By comparing the actual cavitation noise spectrum of each monitoring point obtained with the standard cavitation noise spectrum, it is determined whether the condition of the centrifugal pump being inspected is normal. This method can simplify the inspection process of the centrifugal pump and quickly and accurately monitor the condition of the finished centrifugal pump.

Description

Technical Field

[0001] The present invention belongs to the technical field of intelligent monitoring of centrifugal pumps, and in particular relates to a centrifugal pump state monitoring and diagnosis method based on cavitation noise. Background Art

[0002] When the centrifugal pump is working, cavitation will occur when the centrifugal pump blades rotate at high speed. Cavitation is the process of formation, development and collapse of vapor or gas bubbles inside the liquid or on the liquid-solid interface when the local pressure in the liquid is reduced. The pulse effect generated when the cavitation collapses increases the vibration of the structure, which will cause great damage to the centrifugal pump blades and also cause the centrifugal pump to generate strong noise during operation.

[0003] Cavitation noise is one of the main noises generated by centrifugal pumps during operation, and the sound intensity of cavitation noise is closely related to the volume change and distribution of cavitation bubbles, which in turn are related to the rotation of the centrifugal pump blades.

[0004] In the prior art, when monitoring and diagnosing the working status of a centrifugal pump, many items need to be monitored, and the underwater part cannot be monitored intuitively, resulting in the inability to ensure efficient and accurate status monitoring of the centrifugal pump. In order to solve the above problems, the present invention provides the following technical solutions. Summary of the Invention

[0005] The purpose of the present invention is to provide a centrifugal pump status monitoring and diagnosis method based on cavitation noise, so as to solve the problem in the prior art that when monitoring and diagnosing the working status of a centrifugal pump, there are many items that need to be monitored, and the underwater part cannot be monitored intuitively, resulting in the inability to ensure efficient and accurate status monitoring of the centrifugal pump.

[0006] The purpose of the present invention can be achieved through the following technical solutions:

[0007] The centrifugal pump condition monitoring and diagnosis method based on cavitation noise includes:

[0008] S1. Obtain the changes in cavitation position and morphology on centrifugal pump blades at different speeds through simulation model;

[0009] S2, using a PIV system combined with a high-speed image acquisition camera to obtain the changes in cavitation position and morphology on centrifugal pump blades at different speeds;

[0010] S3. Within the preset speed range, obtain n speed points by arithmetic progression, and mark these n speed points as reference points;

[0011] Obtain the A1 morphology map and A2 morphology map corresponding to each control point;

[0012] Obtain the similarity β of vacuolar distribution in the A1 morphology map and the A2 morphology map corresponding to each control point;

[0013] A1 morphology diagram refers to the changes in the position and morphology of cavitation on the centrifugal pump blade at different speeds obtained in step S1, and the cavitation position and morphology diagram on the centrifugal pump blade corresponding to the control point is obtained;

[0014] A2 morphology diagram is a diagram showing the changes in the position and morphology of cavitation bubbles on the centrifugal pump blades at different speeds obtained in step S2, and a morphology diagram of the position and morphology of cavitation bubbles on the centrifugal pump blades corresponding to the control point is obtained;

[0015] S4. When the average value of the similarity β between the cavitation distribution in the A1 morphology map and the A2 morphology map corresponding to each control point is greater than the preset value βy, the model established in step S1 is considered accurate. Otherwise, the corresponding parameters are adjusted until the average value of the similarity β between the cavitation distribution in the A1 morphology map and the A2 morphology map corresponding to each control point is greater than the preset value βy, and the model is considered accurate, and the next step is performed.

[0016] S5. While acquiring data on the position and morphology of cavitation bubbles on centrifugal pump blades under different cavitation numbers by combining the PIV system with a high-speed image acquisition camera, the cavitation acoustic signal in the pump is simultaneously acquired by a vibration noise signal acquisition device to obtain standard cavitation noise spectra at different monitoring points.

[0017] The monitoring point is a preset position corresponding to the position of the centrifugal pump as a reference point;

[0018] S6. When monitoring and diagnosing the centrifugal pump, a sound detection device is set at the corresponding monitoring point. While changing the speed of the centrifugal pump, the actual cavitation noise spectrum corresponding to each monitoring point at each control point is obtained through the sound detection device;

[0019] For the same control point, the actual cavitation noise spectrum of each monitoring point is compared with the standard cavitation noise spectrum to obtain the similarity αj between the two;

[0020] Where 1≤j≤m, m is the number of monitoring points;

[0021] When the average value of the similarity αj of all monitoring points corresponding to the same control point is greater than or equal to the preset value αy, it is considered that the state of the monitored centrifugal pump at the corresponding control point is normal; otherwise, it is considered that the state of the monitored centrifugal pump is abnormal;

[0022] Obtain whether the status of the monitored centrifugal pump is normal or abnormal at each control point;

[0023] When the status of the monitored centrifugal pump is normal at the control point with a preset ratio σ% or more, the status of the corresponding monitored centrifugal pump is considered normal; otherwise, the status of the corresponding monitored centrifugal pump is considered abnormal.

[0024] As a further solution of the present invention, a method for obtaining the changes in the position and morphology of cavitation bubbles on centrifugal pump blades at different speeds through a simulation model is as follows:

[0025] According to the design of centrifugal pump blades, the geometric model of the centrifugal pump blades is constructed using 3D modeling software. Based on the 3D model, the computational fluid dynamics pre-processor is used to establish the computational fluid domain and perform meshing.

[0026] Use finite element software to establish the finite element model of centrifugal pump blades;

[0027] Based on the hydrodynamic model and finite element model, the cavitation performance of the centrifugal pump blades was simulated and calculated, and the cavitation numerical simulation results were obtained. The information of the non-uniform flow field was imported into the fluid analysis software in the form of a compiled profile file. By setting the corresponding cavitation model, material parameters and fluid calculation parameters, the relevant boundary conditions and fluid-solid coupling interface were defined, and the changes in the cavitation position and morphology on the centrifugal pump blades at different speeds were obtained by calculation.

[0028] As a further solution of the present invention, the fluid calculation parameters include fluid type and turbulence model.

[0029] As a further solution of the present invention, a method for obtaining the similarity β of the cavitation distribution in the A1 morphology image and the A2 morphology image corresponding to each control point is as follows:

[0030] For a control point, a number of concentric rings are set in the A1 morphology diagram with the rotation axis of the centrifugal pump blade as the center, and the combination of the concentric rings completely covers the centrifugal pump blade in the A1 morphology diagram;

[0031] Label the concentric rings h11, h12, ..., h1k in a certain order;

[0032] Using the same method, obtain the corresponding concentric rings labeled h21, h22, ..., h2k on the A2 morphology diagram;

[0033] Obtain the number of cavitation bubbles in different diameter ranges in the concentric rings h1i and h2i;

[0034] Where 1≤i≤k;

[0035] The number of cavitations within the concentric ring h1i within the range of x diameters is d11, d12, ..., d1x; the number of cavitations within the concentric ring h2i within the range of x diameters is d21, d22, ..., d2x;

[0036] According to the formula The cavitation distribution similarity βi of the concentric rings h1i and h2i is calculated; where 1≤r≤x;

[0037] According to the formula The similarity β of the vacuolar distribution in the A1 morphology map and the A2 morphology map corresponding to the corresponding control point is calculated.

[0038] As a further solution of the present invention, the method for obtaining the standard cavitation noise spectrum at different monitoring points is:

[0039] Use computational acoustics software to build an acoustic computational model of centrifugal pump blades;

[0040] Setting multiple monitoring points in the acoustic calculation model;

[0041] The numerical simulation results of cavitation obtained from the fluid analysis software were imported into the computational acoustics software. The basic CFD quantities were converted into sound sources, which were inserted into the acoustic grid using the integration method. The sound source data were used to calculate the sound pressure signal at the monitoring point through Fourier transform, and the standard cavitation noise spectrum at each monitoring point was obtained.

[0042] Beneficial effects of the present invention:

[0043] 1. The present invention verifies the distribution of cavitation on the centrifugal pump blades at different speeds obtained through modeling simulation by using a high-speed image acquisition camera, and adjusts the simulation model according to actual conditions to ensure the accuracy of the simulation model.

[0044] 2. The present invention introduces acoustic information during the cavitation process into the adjusted simulation model to obtain cavitation noise information corresponding to different cavitation stages. When conducting production inspections on centrifugal pumps from the same batch, the rotational speed of the inspected centrifugal pumps is varied to obtain cavitation noise information corresponding to different rotational speeds. The actual cavitation noise spectrum obtained at each monitoring point is compared with the standard cavitation noise spectrum to determine whether the inspected centrifugal pump is in a normal state. This method can simplify the inspection process for centrifugal pumps and quickly and accurately monitor the status of finished centrifugal pumps.

[0045] 3. The present invention reasonably segments the cavitation distribution images obtained by the high-speed image acquisition camera and the model simulation, and compares the similarity of the segmented parts in turn, thereby obtaining the similarity of the two cavitation distribution images, and adjusting the parameter data of the simulation model in turn to ensure the accuracy of the model simulation, which is conducive to improving the accuracy of the correspondence between the cavitation distribution and cavitation noise information obtained subsequently. DETAILED DESCRIPTION

[0046] The following is a clear and complete description of the technical solutions in the embodiments of the present invention. Obviously, the embodiments described are only some of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative efforts are within the scope of protection of the present invention.

[0047] The centrifugal pump condition monitoring and diagnosis method based on cavitation noise includes:

[0048] S1. Based on the design of centrifugal pump blades, a geometric model of the centrifugal pump blades is constructed using 3D modeling software. Based on the 3D model, a computational fluid dynamics preprocessor is used to establish a computational fluid domain and perform meshing.

[0049] Use finite element software to establish the finite element model of centrifugal pump blades;

[0050] Based on the hydrodynamic model and finite element model, the cavitation performance of centrifugal pump blades was simulated and calculated, and the cavitation numerical simulation results were obtained. The information of the non-uniform flow field was imported into the fluid analysis software in the form of a compiled profile file. By setting the corresponding cavitation model, material parameters and fluid calculation parameters, the relevant boundary conditions and fluid-solid coupling interface were defined, and the changes in the cavitation position and morphology on the centrifugal pump blades at different speeds were calculated.

[0051] The fluid calculation parameters include fluid type, turbulence model, etc.

[0052] S2, using a PIV system combined with a high-speed image acquisition camera to obtain the changes in cavitation position and morphology on centrifugal pump blades at different speeds;

[0053] S3. Within the preset speed range, obtain n speed points by arithmetic progression, and mark these n speed points as reference points;

[0054] Obtain the A1 morphology map and A2 morphology map corresponding to each control point;

[0055] Obtain the similarity β of vacuolar distribution in the A1 morphology map and the A2 morphology map corresponding to each control point;

[0056] In one embodiment of the present invention, the method for obtaining the similarity β of the cavitation distribution in the A1 morphology image and the A2 morphology image corresponding to each control point is:

[0057] For a control point, a number of concentric rings are set in the A1 morphology diagram with the rotation axis of the centrifugal pump blade as the center, and the combination of the concentric rings completely covers the centrifugal pump blade in the A1 morphology diagram;

[0058] Label the concentric rings h11, h12, ..., h1k in a certain order;

[0059] Using the same method, obtain the corresponding concentric rings labeled h21, h22, ..., h2k on the A2 morphology diagram;

[0060] Obtain the number of cavitation bubbles in different diameter ranges in the concentric rings h1i and h2i;

[0061] Where 1≤i≤k;

[0062] The number of cavitation bubbles within the concentric ring h1i within the x diameter range is d11, d12, ..., d1x;

[0063] The number of cavitation bubbles within the concentric ring h2i within the x diameter range is d21, d22, ..., d2x;

[0064] According to the formula The cavitation distribution similarity βi of the concentric rings h1i and h2i is calculated; where 1≤r≤x;

[0065] According to the formula The similarity β of the vacuolar distribution in the A1 morphology map and the A2 morphology map corresponding to the control point is calculated;

[0066] This step involves performing reasonable segmentation processing on the cavitation distribution images obtained by the high-speed image acquisition camera and the model simulation, and then performing similarity comparison on the segmented parts in turn to obtain the similarity between the two cavitation distribution images. The parameter data of the simulation model is then adjusted in turn to ensure the accuracy of the model simulation, which is conducive to improving the accuracy of the subsequent correspondence between the cavitation distribution and cavitation noise information.

[0067] In one embodiment of the present invention, when the concentric rings are numbered, they are numbered from the inside out;

[0068] In one embodiment of the present invention, when calculating βi, the values ​​with large deviations corresponding to x (d1r-d2r) / d2r may be deleted first to improve the accuracy of the calculation result as much as possible;

[0069] The A1 morphology diagram refers to the changes in the position and morphology of cavitation bubbles on the centrifugal pump blades at different speeds obtained in step S1, and the cavitation position and morphology diagram on the centrifugal pump blades corresponding to the control point is obtained;

[0070] A2 morphology diagram refers to the changes in the position and morphology of cavitation on the centrifugal pump blade at different speeds obtained in step S2, and the cavitation position and morphology diagram on the centrifugal pump blade corresponding to the control point is obtained;

[0071] S4. When the average value of the similarity β between the cavitation distribution in the A1 morphology map and the A2 morphology map corresponding to each control point is greater than the preset value βy, the model established in step S1 is considered accurate. Otherwise, the corresponding parameters are adjusted until the average value of the similarity β between the cavitation distribution in the A1 morphology map and the A2 morphology map corresponding to each control point is greater than the preset value βy, and the model is considered accurate, and the next step is performed.

[0072] S5. While acquiring data on the position and morphology of cavitation bubbles on centrifugal pump blades under different cavitation numbers by combining the PIV system with a high-speed image acquisition camera, the cavitation acoustic signal in the pump is simultaneously acquired by a vibration noise signal acquisition device to obtain standard cavitation noise spectra at different monitoring points.

[0073] The monitoring point is a preset position with the position corresponding to the centrifugal pump as the reference point;

[0074] Specifically, the method for obtaining the standard cavitation noise spectrum at different monitoring points is as follows:

[0075] Computational acoustics software is used to construct an acoustic calculation model of centrifugal pump blades. The model includes two parts: acoustic finite element and acoustic infinite element.

[0076] Among them, the acoustic finite element simulates the vibration sound radiation in the near sound field, and the acoustic infinite element simulates the far field radiation;

[0077] Acoustic finite element meshes can be directly constructed in acoustic pre-processing software (such as Patran, Hypermesh, ICEM, etc.), while acoustic infinite element meshes are defined by determining the envelope surface of the finite element mesh as the infinite element basis. It is also necessary to set multiple monitoring points in the acoustic calculation model to facilitate subsequent analysis of cavitation noise changes at each monitoring point.

[0078] The numerical simulation results of cavitation obtained from the fluid analysis software were imported into the computational acoustics software. The basic CFD quantities were converted into sound sources. The sound sources were inserted into the acoustic grid using the integration method. The sound pressure signals at the monitoring points were calculated using the obtained sound source data through Fourier transform, and the standard cavitation noise spectrum at each monitoring point was obtained.

[0079] S6. When monitoring and diagnosing the centrifugal pump, a sound detection device is set at the corresponding monitoring point. While changing the speed of the centrifugal pump, the actual cavitation noise spectrum corresponding to each monitoring point at each control point is obtained through the sound detection device;

[0080] For the same control point, the actual cavitation noise spectrum of each monitoring point is compared with the standard cavitation noise spectrum to obtain the similarity αj between the two;

[0081] Where 1≤j≤m, m is the number of monitoring points;

[0082] When the average value of the similarity αj of all monitoring points corresponding to the same control point is greater than or equal to the preset value αy, it is considered that the state of the monitored centrifugal pump at the corresponding control point is normal; otherwise, it is considered that the state of the monitored centrifugal pump is abnormal;

[0083] Obtain in sequence at each control point whether the status of the monitored centrifugal pump is normal or abnormal;

[0084] When the status of the monitored centrifugal pump is normal at the control point with a preset ratio σ% or more, the status of the corresponding monitored centrifugal pump is considered normal; otherwise, the status of the corresponding monitored centrifugal pump is considered abnormal.

[0085] The present invention obtains the distribution of cavitation bubbles on the blades of a centrifugal pump at different speeds through a high-speed image acquisition camera to verify the distribution of cavitation bubbles on the blades of a centrifugal pump at different speeds obtained through modeling simulation, adjusts the simulation model according to actual conditions, and introduces acoustic information during the cavitation occurrence process into the adjusted simulation model to obtain cavitation noise information corresponding to different cavitation stages. When performing production inspection on centrifugal pumps of the same batch, the speed of the inspected centrifugal pump is changed to obtain the cavitation noise information corresponding to the different speeds. The actual cavitation noise spectrum of each monitoring point obtained is compared with the standard cavitation noise spectrum to determine whether the state of the inspected centrifugal pump is normal. This method can simplify the inspection process of the centrifugal pump and quickly and accurately monitor the state of the finished centrifugal pump.

[0086] Throughout the specification, references to terms such as "one embodiment," "example," or "specific example" indicate that the specific features, structures, materials, or characteristics described in conjunction with that embodiment or example are included in at least one embodiment or example of the present invention. In this specification, schematic representations of these terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

[0087] The above contents are merely examples and explanations of the present invention. Those skilled in the art may make various modifications or additions to the described specific embodiments or replace them in similar ways. As long as they do not deviate from the invention or exceed the scope defined by the claims, they should all fall within the scope of protection of the present invention.

Claims

1. A centrifugal pump condition monitoring and diagnosis method based on cavitation noise, characterized in that: include: S1. Obtain the changes in cavitation position and morphology on centrifugal pump blades at different speeds through simulation model; S2, using a PIV system combined with a high-speed image acquisition camera to obtain the changes in cavitation position and morphology on centrifugal pump blades at different speeds; S3. Within the preset speed range, obtain n speed points by arithmetic progression, and mark these n speed points as reference points; Obtain the A1 morphology map and A2 morphology map corresponding to each control point; Obtain the similarity β of vacuolar distribution in the A1 morphology map and the A2 morphology map corresponding to each control point; A1 morphology diagram refers to the changes in the position and morphology of cavitation on the centrifugal pump blade at different speeds obtained in step S1, and the cavitation position and morphology diagram on the centrifugal pump blade corresponding to the control point is obtained; A2 morphology diagram is a diagram showing the changes in the position and morphology of cavitation bubbles on the centrifugal pump blades at different speeds obtained in step S2, and a morphology diagram of the position and morphology of cavitation bubbles on the centrifugal pump blades corresponding to the control point is obtained; S4. When the average value of the similarity β between the cavitation distribution in the A1 morphology map and the A2 morphology map corresponding to each control point is greater than the preset value βy, the model established in step S1 is considered accurate. Otherwise, the corresponding parameters are adjusted until the average value of the similarity β between the cavitation distribution in the A1 morphology map and the A2 morphology map corresponding to each control point is greater than the preset value βy, and the model is considered accurate, and the next step is performed. S5. While acquiring data on the position and morphology of cavitation bubbles on centrifugal pump blades under different cavitation numbers by combining the PIV system with a high-speed image acquisition camera, the cavitation acoustic signal in the pump is simultaneously acquired by a vibration noise signal acquisition device to obtain standard cavitation noise spectra at different monitoring points. The monitoring point is a preset position corresponding to the position of the centrifugal pump as a reference point; S6. When monitoring and diagnosing the centrifugal pump, a sound detection device is set at the corresponding monitoring point. While changing the speed of the centrifugal pump, the actual cavitation noise spectrum corresponding to each monitoring point at each control point is obtained through the sound detection device; For the same control point, the actual cavitation noise spectrum of each monitoring point is compared with the standard cavitation noise spectrum to obtain the similarity αj between the two; Where 1≤j≤m, m is the number of monitoring points; When the average value of the similarity αj of all monitoring points corresponding to the same control point is greater than or equal to the preset value αy, it is considered that the state of the monitored centrifugal pump at the corresponding control point is normal; otherwise, it is considered that the state of the monitored centrifugal pump is abnormal; Obtain whether the status of the monitored centrifugal pump is normal or abnormal at each control point; When the status of the monitored centrifugal pump is normal at the control point with a preset ratio σ% or more, the status of the corresponding monitored centrifugal pump is considered normal; otherwise, the status of the corresponding monitored centrifugal pump is considered abnormal.

2. The centrifugal pump condition monitoring and diagnosis method based on cavitation noise according to claim 1 is characterized in that: The method for obtaining the similarity β of the cavitation distribution in the A1 morphology map and the A2 morphology map corresponding to each control point is: For a control point, a number of concentric rings are set in the A1 morphology diagram with the rotation axis of the centrifugal pump blade as the center, and the combination of the concentric rings completely covers the centrifugal pump blade in the A1 morphology diagram; Label the concentric rings h11, h12, ..., h1k in a certain order; Using the same method, obtain the corresponding concentric rings labeled h21, h22, ..., h2k on the A2 morphology diagram; Obtain the number of cavitation bubbles in different diameter ranges in the concentric rings h1i and h2i; Where 1≤i≤k; The number of cavitations within the concentric ring h1i within the range of x diameters is d11, d12, ..., d1x; the number of cavitations within the concentric ring h2i within the range of x diameters is d21, d22, ..., d2x; According to the formula The cavitation distribution similarity βi of the concentric rings h1i and h2i is calculated; where 1≤r≤x; According to the formula The similarity β of the vacuolar distribution in the A1 morphology map and the A2 morphology map corresponding to the corresponding control point is calculated.

3. The centrifugal pump condition monitoring and diagnosis method based on cavitation noise according to claim 1, characterized in that: The method of obtaining the changes in cavitation position and morphology on the centrifugal pump blades at different speeds through the simulation model is as follows: According to the design of centrifugal pump blades, the geometric model of the centrifugal pump blades is constructed using 3D modeling software. Based on the 3D model, the computational fluid dynamics pre-processor is used to establish the computational fluid domain and perform meshing. Use finite element software to establish the finite element model of centrifugal pump blades; Based on the hydrodynamic model and finite element model, the cavitation performance of the centrifugal pump blades was simulated and calculated, and the cavitation numerical simulation results were obtained. The information of the non-uniform flow field was imported into the fluid analysis software in the form of a compiled profile file. By setting the corresponding cavitation model, material parameters and fluid calculation parameters, the relevant boundary conditions and fluid-solid coupling interface were defined, and the changes in the cavitation position and morphology on the centrifugal pump blades at different speeds were obtained by calculation.

4. The centrifugal pump condition monitoring and diagnosis method based on cavitation noise according to claim 3 is characterized in that: The fluid calculation parameters include fluid type and turbulence model.

5. The centrifugal pump condition monitoring and diagnosis method based on cavitation noise according to claim 3 is characterized in that: The method for obtaining the standard cavitation noise spectrum at different monitoring points is: Use computational acoustics software to build an acoustic computational model of centrifugal pump blades; Setting multiple monitoring points in the acoustic calculation model; The numerical simulation results of cavitation obtained from the fluid analysis software were imported into the computational acoustics software. The basic CFD quantities were converted into sound sources, which were inserted into the acoustic grid using the integration method. The sound source data were used to calculate the sound pressure signal at the monitoring point through Fourier transform, and the standard cavitation noise spectrum at each monitoring point was obtained.

Images