Bearing fault diagnosis method, device and system for natural energy water lifting equipment

Abstract

The invention discloses a bearing fault diagnosis method, device and system for natural energy water lifting equipment, and the method comprises the steps: obtaining original vibration data, carrying out the cepstrum transformation, and obtaining target fault information; performing filtering of different scales on the cepstrum, fusing spatial information of different scales, and adaptively adjusting the weight; adopting a clustering algorithm to adaptively determine a centroid, and inputting the centroid into a classification algorithm; training data to obtain a corresponding weight coefficient matrix and a fuzzy normalization factor matrix; in the testing process, after unknown original vibration data is subjected to cepstrum transformation, filtering and classification, the unknown original vibration data, the weight coefficient matrix and the fuzzy normalization factor matrix are subjected to Hadamard product, and a final diagnosis result is obtained. According to the bearing fault detection method, accurate fault detection can be carried out under the condition that part of faults are unknown, and the bearing fault detection method is combined with an artificial intelligence algorithm, has very high expandability and is very suitable for bearing fault diagnosis.

Description

Technical Field

[0001] The present invention belongs to the technical field of bearing fault diagnosis, and in particular relates to a bearing fault diagnosis method, device and system for natural energy water pumping equipment. Background Art

[0002] With the continuous improvement of science and technology, mechanical equipment is increasingly tending to be large-scale, systematized, high-speed, precise and automated. As the "joint" of mechanical equipment, rolling bearings are one of the most basic components that are used in large quantities and are most easily damaged. Therefore, it is necessary to monitor the condition of bearings, diagnose faults in a timely manner, and even identify potential faults in advance. Compared with the traditional fault diagnosis method based on signal analysis and processing, the data-driven fault diagnosis method is the most widely used in the current fault diagnosis field.

[0003] Patent CN113723546B performs fault detection on the bearing to be detected based on the feature vectors of different fault types of the bearing to be detected and the trained discrete hidden Markov model, which can reduce the dependence on the initial value, improve the search ability, and thus improve the accuracy of fault detection. Patent CN117723300A integrates the channel attention mechanism, BiLSTM and BN into the WDCNN model, and realizes the fault diagnosis and classification of the bearing by directly processing the original noise signal of the rolling bearing and analyzing it. However, these data-driven bearing fault diagnosis methods must have complete fault samples. The number of fault data can be more or less, but the data of each fault category must be available, which is difficult to meet in the industry. Considering that many faults will cause huge losses, few factories are allowed to run to the fault and collect fault samples to train the fault diagnosis system. Typical machine failures are the result of a gradual degradation process from normal operation to failure, which makes it time-consuming and expensive to obtain sufficient fault samples. Traditional fault diagnosis methods are not effective in practical applications due to the lack of fault data. Therefore, bearing fault diagnosis under incomplete training data has become a very classic fault diagnosis problem encountered in real applications and practice. Summary of the invention

[0004] In order to solve the above technical problems, the present invention provides a bearing fault diagnosis method, device and system for natural energy water pumping equipment, so as to realize accurate bearing fault diagnosis even when fault data is missing, improve the durability of the equipment and reduce maintenance costs.

[0005] To achieve the above objectives, the present invention provides a bearing fault diagnosis method, device and system for natural energy water pumping equipment. A bearing fault diagnosis method for natural energy water pumping equipment includes:

[0006] Acquire original vibration data, and perform cepstrum transformation on the original vibration data to obtain target fault information;

[0007] According to the target fault information, the cepstrum is filtered at different scales, spatial information at different scales is integrated, and weights are adaptively adjusted;

[0008] Adaptively determining a centroid using a clustering algorithm, and inputting the centroid into a classification algorithm;

[0009] The training data obtains the corresponding weight coefficient matrix and fuzzy normalization factor matrix;

[0010] During the test, the unknown original vibration data is subjected to cepstrum transformation, filtering and classification, and then Hadamard product is performed with the weight coefficient matrix and the fuzzy normalization factor matrix to obtain the final diagnosis result.

[0011] Preferably, the process of obtaining raw vibration data includes:

[0012] Obtain the vibration data of the bearing under normal operation and the vibration data generated by the bearing at the fault location, and classify the vibration data generated by the bearing at the fault location and the missing diameter into a known fault class F∈{ };

[0013] For the total fault set F∈{ , , , …}, d∈(I, II, III, …), the data includes slice data with partial diameters, described as ∈{ , , , …}, da∈d, where da represents the available data with diameter d; missing data are described as ∈{ , , , …}, dm∈d, where dm represents missing data with diameter d;

[0014] Available training data For incomplete fault data, is missing data, satisfying:

[0015] .

[0016] Preferably, the process of performing cepstrum transformation on the original vibration data comprises:

[0017] A fast Fourier transform is performed on the original vibration data, and the absolute value and logarithm of the fast Fourier transform result are taken; then an inverse fast Fourier transform is performed, and the sideband spectral line family on the original spectrum is reduced to a single spectral line.

[0018] Preferably, the formula expression of the cepstrum transform is:

[0019]

[0020] Where Ceps(n) represents the signal after cepstrum transformation, FFT[·] represents Fourier transform calculation, IFFT[·] represents inverse Fourier transform calculation, and Re(·) represents real part calculation. Pick To avoid NAN values ​​during calculation.

[0021] Preferably, according to the target fault information, a multi-scale transformation formula for filtering the cepstrum at different scales is expressed as:

[0022]

[0023] Where Scalings(t) represents the signal after filtering, t = 0, 1, ..., N / L-1, N is the data length, L = 2S-1, and S represents the scaling factor.

[0024] Preferably, the clustering algorithm adopts MMA-Clustering algorithm, and the MMA-Clustering algorithm includes Euclidean distance, Manhattan distance, and Chebyshev distance;

[0025] The classification algorithm adopts MMW-Classify algorithm;

[0026] The centroid is defined as , where S represents scale, M represents measurement, and O represents category.

[0027] Preferably, the formula expression of the fuzzy normalization factor matrix is:

[0028]

[0029] in, represents the fuzzy normalized matrix, is the average of the training data above the median, is the average value of the training data below the median, and is called the normalization factor.

[0030] Preferably, after the unknown original vibration data is subjected to cepstrum transformation, filtering and classification, the formula for making a Hadamard product with the weight coefficient matrix and the fuzzy normalization factor matrix is ​​as follows:

[0031]

[0032] in, is the Hadamard product.

[0033] The present invention provides a bearing fault diagnosis device for natural energy water pumping equipment, comprising:

[0034] Bearing condition monitoring part, bearing fault diagnosis part;

[0035] The bearing status monitoring part includes a bearing temperature monitoring module and a bearing oil monitoring module;

[0036] Wherein, the bearing temperature monitoring module is installed on both sides of the bearing to monitor the bearing temperature in real time;

[0037] The bearing oil monitoring module is installed on the bearing oil supply side and is used to monitor the lubricating oil content and the impurity content of the lubricating oil;

[0038] The bearing fault diagnosis part includes an original data acquisition module, a vibration data preprocessing module and a diagnosis network module;

[0039] Wherein, the raw data acquisition module is installed on the bearing housing and is used to collect the raw vibration signal when the bearing is in operation;

[0040] The vibration data preprocessing module is installed on the main control board and is used to perform filtering and cepstrum preprocessing operations on the original vibration data;

[0041] The diagnostic network module is installed on the main control board and is used to diagnose bearing faults.

[0042] The present invention also provides a bearing fault diagnosis system for natural energy water pumping equipment, comprising:

[0043] Photovoltaic power generation modules are used to directly convert solar radiation into electrical energy to power equipment;

[0044] The motor module is used to drive the water pump;

[0045] A water pump module, used for pumping groundwater into a water storage tank under the drive of a motor;

[0046] The filter module comprises a coarse filtration unit and a fine filtration unit; wherein the coarse filtration unit is used to preliminarily treat the groundwater through an absorption pipe network; and the fine filtration unit is used to perform multi-layer purification and disinfection operations on the groundwater after the coarse filtration;

[0047] The sensor and detector module has a built-in vibration sensor, a temperature sensor, an oil detector and a turbidity detector; wherein the vibration sensor is used to collect the original vibration data of the water pump bearing in real time; the temperature sensor and the oil detector are used to monitor the temperature and oil state of the water pump bearing in real time; the turbidity detector is used to detect the water impurity content during groundwater filtration;

[0048] The pulse pump and liquid-gas device module comprises a pulse pump and a liquid-gas device; wherein the pulse pump is used to generate water flow by utilizing the drop in height, periodically stopping the water flow instantaneously, generating impact force to press the water to a high place; the liquid-gas device is used to generate high-pressure air by utilizing the drop in water flow, and pressurizing the water to a high place by means of the high-pressure air.

[0049] Compared with the prior art, the present invention has the following advantages and technical effects:

[0050] The bearing fault detection method provided by the present invention can perform accurate fault detection when some faults are unknown. By combining with an artificial intelligence algorithm, it has strong scalability and is very suitable for bearing fault diagnosis.

[0051] The bearing fault detection device provided by the present invention can both monitor the bearing status in real time and perform bearing fault diagnosis.

[0052] The bearing fault detection system provided by the present invention can utilize solar energy to generate electricity, thereby saving electricity costs; it can perform coarse and fine filtration on water to purify the water quality; it utilizes a pulse pump and a liquid-gas device to pump water, thereby effectively solving the problem of a large water supply stroke from low to high. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] The drawings constituting a part of the present application are used to provide a further understanding of the present application. The illustrative embodiments and descriptions of the present application are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0054] Figure 1 A schematic diagram of the overall process of a bearing fault diagnosis method for natural energy water pumping equipment provided in Embodiment 1 of the present invention;

[0055] Figure 2 A schematic diagram of a diagnostic process flow of a method for diagnosing bearing faults of natural energy water pumping equipment provided in Embodiment 1 of the present invention;

[0056] Figure 3 A schematic diagram of the overall process of a bearing fault diagnosis device for natural energy water pumping equipment provided in the second embodiment of the present invention;

[0057] Figure 4 This is a schematic diagram of the overall process of a bearing fault diagnosis system for natural energy water pumping equipment provided in Example 3 of the present invention. DETAILED DESCRIPTION

[0058] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0059] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0060] Embodiment 1

[0061] like Figure 1-2 As shown, this embodiment provides a bearing fault diagnosis method for natural energy water pumping equipment, including the following steps:

[0062] S1: Obtaining original vibration data, including normal bearing data and some fault data;

[0063] S2: Perform cepstrum transformation on the original vibration data to obtain meaningful fault information;

[0064] S3: Filter the cepstrum at different scales, fuse the spatial information at different scales, and adaptively adjust the appropriate weights;

[0065] S4: The clustering algorithm is used to adaptively determine the final centroid and input it into the classification algorithm;

[0066] S5: training data to obtain the corresponding weight coefficient matrix and fuzzy normalization factor matrix;

[0067] S6: During the test, the unknown raw vibration data is transformed, filtered and classified, and then Hadamard product is performed with the weight coefficient matrix and the fuzzy normalization factor matrix to finally diagnose the result.

[0068] As a preferred solution of the bearing fault diagnosis method of this embodiment, step S1 obtains original vibration data, including vibration data of the bearing under normal operation and vibration data generated by the bearing at certain faults;

[0069] Furthermore, partially missing fault data and missing diameters should be classified into known fault classes F∈{ }.

[0070] Furthermore, for the total fault set F∈{ , , , …}, d∈(I, II, III, …). The available data contain slice data with partial diameters, which can be described as ∈{ , , , …}, da∈d, where da represents the available data with diameter d. Similarly, the partial missing data can be described as ∈{ , , , …}, dm∈d, where dm represents missing data with diameter d.

[0071] It should be noted that the available training data is incomplete, is missing, the result of fault diagnosis is F∈{ }, instead of F∈{ }. The problem satisfies the following relationship:

[0072]

[0073] As a preferred solution of the bearing fault diagnosis method of this embodiment, in step S2, during the training process, the original vibration data used for training is subjected to cepstrum transformation, and the redundant periodic signals are converted into signals with significant characteristics;

[0074] Furthermore, for the vibration signal x(n), its cepstrum can be divided into the following three steps:

[0075] S21: Perform Fast Fourier Transform (FFT) on x(n) to keep the signal length constant;

[0076] S22: Get the absolute value and logarithm of the FFT result;

[0077] S23: Perform Inverse Fast Fourier Transform (IFFT).

[0078] After these steps, the family of sideband lines on the original spectrum is reduced to a single line using the following conversion formula:

[0079]

[0080] Where Ceps(n) represents the signal after cepstrum transformation, FFT[·] represents Fourier transform calculation, IFFT[·] represents inverse Fourier transform calculation, and Re(·) represents real part calculation. Pick This is to avoid NAN values ​​during calculation.

[0081] As a preferred solution of the bearing fault diagnosis method of this embodiment, step S3 filters the cepstrum at different scales, fuses spatial information at different scales, and adaptively adjusts appropriate weights;

[0082] Furthermore, the multi-scale transformation formula is as follows:

[0083]

[0084] Where Scalings(t) represents the signal after filtering, t=0, 1, ..., N / L-1, N is the data length, L = 2S-1, and S represents the scaling factor. The selection of the scaling factor S is similar to the selection of the number of layers in artificial neural networks (ANN). The default value of S is 3, but it is not fixed and can be appropriately increased according to the actual test accuracy and device computing power during the training process.

[0085] Furthermore, after multi-scale transformation, the data at each scale is filtered, which can eliminate interfering data points, retain prominent features, and improve robustness. In addition, obtaining data at multiple scales will make the amount of information more sufficient.

[0086] As a preferred solution of the bearing fault diagnosis method of this embodiment, step S4 uses a clustering algorithm to calculate the centroid of the training data, performs pre-training and iteration, adaptively determines the final centroid, and inputs it into the classification algorithm;

[0087] Furthermore, the MMA-Clustering algorithm is specifically used to adaptively determine the final centroid, and input it into the MMW-Classify algorithm.

[0088] The training data set X consists of N samples x, and the data set can be expressed as:

[0089]

[0090] Where n is the Hadamard product, and the initialized w is a vector of 1s with a length of N.

[0091] Furthermore, the distance between the data and the centroid is measured in different metric spaces, and the sample with the smallest distance to a cluster centroid is assigned to the cluster, while the sample assigned to the wrong cluster indicates that the sample is far away from the cluster centroid in the metric space.

[0092] Furthermore, the sample is added to the training data again. After several rounds of iterative calculations, this sample will get a new weight w. The cluster centroid of the training data is calculated based on this result. The cluster centroid will move closer to this sample, and the cluster centroid position with the best result will be determined as the final centroid.

[0093] Furthermore, for scale S, metric M, and category O, the centroid is defined as , where S represents scale, M represents measurement, and O represents category.

[0094] The MMA-Clustering algorithm includes three distances:

[0095] Euclidean distance:

[0096]

[0097] Manhattan distance:

[0098]

[0099] Chebyshev distance:

[0100]

[0101] Wherein, D represents the distance.

[0102] As a preferred solution of the bearing fault diagnosis method of this embodiment, step S5 uses a classification algorithm to calculate the distance between the training data and the centroids in different metric spaces, and classifies the training data into the closest data. The corresponding weight coefficient matrix Q (S, M, O) and fuzzy normalization factor matrix D are obtained through the training data;

[0103] Furthermore, during the training process, the distances between the training data and the centroids in different metric spaces are calculated by the MMW-Classify algorithm, and the training data are classified into the closest data.

[0104] Furthermore, the weight coefficient matrix q(M, O) is constructed with the classification accuracy as the weight. The higher the accuracy, the higher the reliability of the metric space and the centroid. When MMW-Classify is applied to multi-scale space, the three-dimensional weight coefficient matrix Q(S, M, O) can be obtained by concatenating the two-dimensional matrix q.

[0105] Furthermore, in order to balance the importance of the three metric spaces, the distance results in multiple metric spaces need to be normalized. Considering the maximum-minimum normalization method, the fuzzy normalization matrix is ​​as follows:

[0106]

[0107] in, represents the fuzzy normalized matrix, is the average of the training data above the median, is the average value of the training data below the median, and is called the normalization factor.

[0108] As a preferred solution of a bearing fault diagnosis method for a natural energy water pumping equipment of this embodiment, in step S6, during the test process, the unknown original vibration data is subjected to cepstrum transformation and filtering, the MMW-Classify algorithm is used to calculate the distance in different metric spaces, the matrix d is used for fuzzy normalization, the Hadamard product of the normalized result and the weight coefficient matrix q is calculated, and then M is accumulated as follows:

[0109]

[0110] in, is the Hadamard product.

[0111] Furthermore, when it is extended to multiple scale spaces, a classification algorithm is used to calculate the distances in different metric spaces, and the fuzzy normalization factor matrix D is used for fuzzy normalization. The Hadamard product of the normalized result and the weight coefficient matrix Q is calculated, and then M is accumulated. The formula becomes:

[0112]

[0113] in, is the Hadamard product.

[0114] Furthermore, the preliminary fault diagnosis results are obtained according to the following minimum a posteriori characteristic formula:

[0115]

[0116] Among them, O is the fault set, and o is the fault diagnosis result. When the results of multiple features (O) are the same, it is necessary to learn from the Bayesian idea and use the fault category with a higher historical fault frequency as the diagnosis result.

[0117] Embodiment 2

[0118] like Figure 3 As shown, this embodiment provides a bearing fault diagnosis device for natural energy water pumping equipment, comprising:

[0119] The bearing status monitoring part includes a bearing temperature monitoring module and a bearing oil monitoring module;

[0120] Among them, the bearing temperature monitoring module is installed on both sides of the bearing to monitor the bearing temperature in real time to prevent serious wear of the bearing due to overheating.

[0121] The bearing oil monitoring module is installed on the oil supply side of the bearing to monitor the lubricating oil content and the impurity content of the lubricating oil to prevent serious wear of the bearing due to lack of oil or oil containing metal particles and metal debris.

[0122] The bearing fault diagnosis part includes the original data acquisition module, the vibration data preprocessing module and the diagnosis network module;

[0123] Among them, the original data acquisition module is installed on the bearing housing to collect the original vibration signal when the bearing is in operation.

[0124] The vibration data preprocessing module is installed on the main control board and is used to perform preprocessing operations such as filtering and cepstrum on the original vibration data to facilitate subsequent fault analysis.

[0125] The diagnostic network module is installed on the main control board to diagnose bearing faults, facilitate timely maintenance, and reduce equipment operating costs.

[0126] Embodiment 3

[0127] like Figure 4 As shown, this embodiment provides a bearing fault diagnosis system for natural energy water pumping equipment, including:

[0128] Photovoltaic power generation modules use solar radiation to directly convert into electrical energy to power subsequent devices.

[0129] The motor module is used to drive the water pump.

[0130] The water pump module, driven by the motor, pumps groundwater into the water storage tank.

[0131] The filter module is divided into coarse filtration and fine filtration. Coarse filtration is the preliminary treatment of groundwater through the absorption pipe network, and fine filtration is a multi-layer purification and disinfection operation of the groundwater after coarse filtration.

[0132] The sensor and detector module has built-in vibration sensor, temperature sensor, oil detector and turbidity detector. The vibration sensor can collect the original vibration data of the water pump bearing in real time; the temperature sensor and oil detector can monitor the temperature and oil status of the water pump bearing in real time; the turbidity detector can detect the impurity content of water during groundwater filtration.

[0133] Pulse pump and liquid-gas device module, among which the pulse pump uses the drop in water flow to generate water flow, periodically stops the water flow instantly, generates huge impact force, and presses the water to a high place; the liquid-gas device uses the drop in water flow to generate high-pressure air, and presses the water to a high place through the high-pressure air.

[0134] The above are only preferred specific implementations of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed in the present application should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A method for diagnosing bearing faults of natural energy water pumping equipment, characterized in that: include: Acquire original vibration data, and perform cepstrum transformation on the original vibration data to obtain target fault information; According to the target fault information, the cepstrum is filtered at different scales, spatial information at different scales is integrated, and weights are adaptively adjusted; Adaptively determining a centroid using a clustering algorithm, and inputting the centroid into a classification algorithm; The training data obtains the corresponding weight coefficient matrix and fuzzy normalization factor matrix; During the test, the unknown original vibration data is subjected to cepstrum transformation, filtering and classification, and then Hadamard product is performed with the weight coefficient matrix and the fuzzy normalization factor matrix to obtain the final diagnosis result.

2. The method according to claim 1, characterized in that The process of obtaining raw vibration data includes: Obtain the vibration data of the bearing under normal operation and the vibration data generated by the bearing at the fault location, and classify the vibration data generated by the bearing at the fault location and the missing diameter into a known fault class F∈{ }; For the total fault set F∈{ , , , …}, d∈(I, II, III, …), the data includes slice data with partial diameters, described as ∈{ , , , …}, da∈d, where da represents the available data with diameter d; missing data are described as ∈{ , , , …}, dm∈d, where dm represents missing data with diameter d; Available training data For incomplete fault data, is missing data, satisfying: 。 3. The method according to claim 1, characterized in that The process of performing cepstrum transformation on the original vibration data includes: A fast Fourier transform is performed on the original vibration data, and the absolute value and logarithm of the fast Fourier transform result are taken; then an inverse fast Fourier transform is performed, and the sideband spectral line family on the original spectrum is reduced to a single spectral line.

4. The method according to claim 1, characterized in that: The formula expression of the cepstrum transform is: Where Ceps(n) represents the signal after cepstrum transformation, FFT [·] represents Fourier transform calculation, IFFT [·] represents inverse Fourier transform calculation, and Re (·) represents real part calculation. Pick This is to avoid NAN values ​​during calculation.

5. The method according to claim 1, characterized in that According to the target fault information, the multi-scale transformation formula for filtering the cepstrum at different scales is expressed as follows: Where Scalings(t) represents the signal after filtering, t = 0, 1, …, N / L-1, N is the data length, L = 2S-1, and S represents the scaling factor.

6. The method according to claim 1, characterized in that The clustering algorithm adopts the MMA-Clustering algorithm, and the MMA-Clustering algorithm includes Euclidean distance, Manhattan distance, and Chebyshev distance; The classification algorithm adopts MMW-Classify algorithm; The centroid is defined as , where S represents scale, M represents measurement, and O represents category.

7. The method according to claim 1, characterized in that The formula expression of the fuzzy normalization factor matrix is: in, represents the fuzzy normalized matrix, is the average of the training data above the median, is the average value of the training data below the median, and is called the normalization factor.

8. The method according to claim 1, characterized in that After the unknown original vibration data is subjected to cepstrum transformation, filtering and classification, the formula for the Hadamard product with the weight coefficient matrix and the fuzzy normalization factor matrix is ​​as follows: in, is the Hadamard product.

9. A bearing fault diagnosis device for natural energy water pumping equipment, characterized in that: include: Bearing condition monitoring part, bearing fault diagnosis part; The bearing status monitoring part includes a bearing temperature monitoring module and a bearing oil monitoring module; Wherein, the bearing temperature monitoring module is installed on both sides of the bearing to monitor the bearing temperature in real time; The bearing oil monitoring module is installed on the bearing oil supply side and is used to monitor the lubricating oil content and the impurity content of the lubricating oil; The bearing fault diagnosis part includes an original data acquisition module, a vibration data preprocessing module and a diagnosis network module; Wherein, the raw data acquisition module is installed on the bearing housing and is used to collect the raw vibration signal when the bearing is in operation; The vibration data preprocessing module is installed on the main control board and is used to perform filtering and cepstrum preprocessing operations on the original vibration data; The diagnostic network module is installed on the main control board and is used to diagnose bearing faults.

10. A bearing fault diagnosis system for natural energy water pumping equipment, characterized in that: include: Photovoltaic power generation modules are used to directly convert solar radiation into electrical energy to power equipment; The motor module is used to drive the water pump; A water pump module, used for pumping groundwater into a water storage tank under the drive of a motor; The filter module comprises a coarse filtration unit and a fine filtration unit; wherein the coarse filtration unit is used to preliminarily treat the groundwater through an absorption pipe network; and the fine filtration unit is used to perform multi-layer purification and disinfection operations on the groundwater after the coarse filtration; The sensor and detector module has a built-in vibration sensor, a temperature sensor, an oil detector and a turbidity detector; wherein the vibration sensor is used to collect the original vibration data of the water pump bearing in real time; the temperature sensor and the oil detector are used to monitor the temperature and oil state of the water pump bearing in real time; the turbidity detector is used to detect the water impurity content during groundwater filtration; The pulse pump and liquid-gas device module comprises a pulse pump and a liquid-gas device; wherein the pulse pump is used to generate water flow by utilizing the drop in height, periodically stopping the water flow instantaneously, generating impact force to press the water to a high place; the liquid-gas device is used to generate high-pressure air by utilizing the drop in water flow, and pressurizing the water to a high place by means of the high-pressure air.

Images