An ultrasonic self-cleaning filter and an intelligent control method thereof

Abstract

The application discloses an ultrasonic self-cleaning filter and an intelligent control method thereof, and belongs to the technical field of irrigation water treatment equipment. The filter comprises a shell assembly, which is a sealed cylinder structure, and is provided with a blowdown port at the top, a water inlet at the bottom and a water outlet on the side wall. The filter assembly comprises an inner filter screen in a cylindrical structure, which is arranged inside the shell assembly and is sealingly connected with the top end and the bottom end of the side wall of the shell assembly through a sealing ring. The ultrasonic cleaning assembly comprises an ultrasonic generator and a plurality of vibration units connected with the ultrasonic generator. The control system comprises a controller connected with the ultrasonic generator, a differential pressure sensor and a flow sensor electrically connected with the controller and arranged at the water outlet, and the controller is connected with a cloud platform through a wireless communication module. The controller is used for determining the optimal cleaning time and cleaning period of the filter under the current irrigation cycle according to the historical operation data and cleaning data of the filter.

Description

Technical Field

[0001] This invention relates to the field of irrigation water treatment equipment technology, specifically to an ultrasonic self-cleaning filter and its intelligent control method. Background Technology

[0002] Traditional self-cleaning filters consist of a cylinder, an internal filter screen housed within the cylinder, and a backwashing system or scraper cleaning system. The filter operates for a certain period of time, after which the backwashing system or scraper cleaning system is activated to clean the internal filter screen. While this method can effectively remove sediment adhering to the internal filter screen, it still has the following drawbacks:

[0003] 1. Incomplete cleaning: Sediment easily adheres to the filter screen pores, requiring a large amount of water for backwashing (statistically 10-20% of the total flow rate), and it is difficult to remove fine particles. 2. Filter screen wear: Cleaning and scraping the filter screen leads to wear and tear over long-term use. 3. Low operating efficiency: During backwashing, the filter outlet needs to be closed, preventing filtration operations, and backwashing consumes relatively high amounts of water and energy. 4. Insufficient intelligence: Relying on fixed time or pressure differential threshold control, it cannot adapt to dynamic changes in water quality. Summary of the Invention

[0004] In view of the above-mentioned shortcomings in the prior art, the ultrasonic self-cleaning filter and its intelligent control method provided by the present invention solve the problems of incomplete cleaning and insufficient intelligence in the existing methods.

[0005] To achieve the above-mentioned objectives, the technical solution adopted by this invention is as follows:

[0006] In one aspect, an ultrasonic self-cleaning filter is provided, including a housing assembly, which is a sealed cylindrical structure, with a drain port at the top, a water inlet at the bottom, and a water outlet on the side wall.

[0007] A filter assembly includes an inner filter screen with a cylindrical structure, the inner filter screen being disposed inside a housing assembly and being sealed to the top and bottom ends of the side wall of the housing assembly via a sealing ring;

[0008] An ultrasonic cleaning assembly includes an ultrasonic generator disposed on the outer surface of a housing assembly and multiple vibration units evenly distributed on an inner filter screen and connected to the ultrasonic generator, wherein the ultrasonic generator is electrically connected to a control system.

[0009] The control system includes a controller and a differential pressure sensor and a flow sensor electrically connected to the controller and installed at the outlet. The controller is connected to a cloud platform via a wireless communication module.

[0010] The controller is used to determine the optimal cleaning duration and cleaning cycle of the filter under the current irrigation cycle based on the filter's historical operating data and cleaning data.

[0011] Furthermore, the ultrasonic self-cleaning filter also includes a turbidity sensor installed inside the water inlet and electrically connected to the controller.

[0012] Furthermore, the method for determining the operating frequency of the vibration unit includes:

[0013] Calculate the ultrasonic wave attenuation coefficient within the filter medium based on the rated operating frequency of the vibration unit:

[0014]

[0015] in, The sound wave attenuation coefficient; is the dielectric constant; f is the rated operating frequency; b is the frequency index;

[0016] Determine the penetration distance and attenuation of the ultrasonic waves within the housing assembly, and calculate the maximum frequency of the vibrating unit:

[0017]

[0018] Where m is the sound wave attenuation value; D is the penetration distance;

[0019] The maximum frequency is set as a percentage of the operating frequency of the vibration unit. The set percentage in this scheme is neither equal to nor the preset percentage.

[0020] The beneficial effects of the above technical solution are: the calculation formula accurately matches the characteristics of the medium, improves energy utilization, and utilizes the sound wave attenuation model. Calculate the attenuation coefficient of the medium containing sediment, and scientifically calculate the maximum frequency by combining the penetration distance and attenuation threshold. To avoid excessive sound energy attenuation due to excessively high frequency, and to ensure that ultrasonic waves effectively penetrate the filter and that cavitation effect covers the entire area, the operating frequency is set to [value missing]. The set ratio, with a safety margin, can prevent insufficient acoustic energy under extreme water quality (such as high turbidity), while avoiding transducer overload damage and reducing equipment failure rate.

[0021] Furthermore, the vibration units are evenly arranged at the upper end of the inner filter screen, with the maximum spacing between adjacent vibration units being [missing information]. , Where is the operating frequency, and c is the speed at which ultrasound propagates in the medium.

[0022] The beneficial effects of the above technical solution are as follows: Designed according to the requirements of vibration unit spacing, it can form a uniform sound field by the interference of sound waves from adjacent transducers, so that there are no blind spots on the filter surface and the removal rate of fine particles is improved.

[0023] Secondly, a smart control method for an ultrasonic self-cleaning filter is provided, comprising the following steps:

[0024] S1. Obtain historical flow data, differential pressure data, and cleaning cycle of the filter in multiple irrigation cycles prior to the current irrigation cycle;

[0025] S2. Input the acquired data into the trained LSTM time series prediction network to predict the flow rate data, differential pressure data and cleaning cycle in the current irrigation cycle.

[0026] S3. Determine the occurrence time of the inefficient zone based on the differential pressure data during the first cleaning cycle of the current irrigation cycle;

[0027] S4. Read the time corresponding to each pressure difference between the occurrence time and the end of the first cleaning cycle as the new cleaning cycle, and determine the total number of new cleaning cycles for the current irrigation cycle.

[0028] S5. Set the flow rate data and differential pressure data for each new cleaning cycle corresponding to each differential pressure to be equal to the corresponding data for the first new cleaning cycle.

[0029] S6. Based on the cleaning time and flow rate data and pressure difference data corresponding to each pressure difference, calculate the total cost of water used by the filter and electricity used by the filter and water pump in the current irrigation cycle.

[0030] S7. Select the cleaning time and new cleaning cycle under the pressure difference corresponding to the minimum total cost, and take them as the optimal cleaning time and cleaning cycle for the filter under the current irrigation cycle.

[0031] The beneficial effects of the above technical solution are as follows: This solution can reduce ineffective cleaning and lower the false trigger rate by dynamically predicting the cleaning timing through the above method. By optimizing the cleaning time and cycle, the overall energy consumption is reduced and water consumption is reduced.

[0032] Furthermore, the expression for calculating the total cost of water used for the filter and electricity used for the filter and pump during the current irrigation cycle is as follows:

[0033]

[0034]

[0035]

[0036]

[0037] Where F is the pressure difference The corresponding total cost; The j-th pressure difference between the occurrence time and the end of the first cleaning cycle, 1≤j≤J, where J is the total number of pressure differences selected between the occurrence time and the end of the first cleaning cycle; , and All are weighting coefficients; , and These are the energy costs of the ultrasonic cleaning components, the water pump, and the wastewater discharge costs. M is the depreciation factor; M is the cost of internal filter wear. Ultrasonic power; and They are respectively The corresponding duration and wastewater discharge volume for the i-th cleaning cycle, where I is the pressure difference. Total number of cleaning cycles; The inverter efficiency coefficient of the ultrasonic cleaning component; For industrial electricity prices; and Both represent the pressure difference and flow rate at time t; T is the total duration of the current irrigation cycle. The density of the medium; For water pump efficiency; This is the water quality impact coefficient; For water prices.

[0038] The beneficial effects of the above technical solution: The above formula quantifies the energy consumption of ultrasonic cleaning. Energy consumption of water pump operation Wastewater discharge costs Including filter wear cost M, an irrigation system was constructed under a specific pressure difference. The comprehensive cost model is as follows. The energy consumption cost of high-frequency vibration cleaning is reflected in the following: Precisely quantify the real-time energy consumption of water pumps due to changes in pipeline resistance; It reflects the impact of water quality on water consumption costs; the calculation formula can prioritize energy consumption, water consumption, and equipment wear and tear, supporting multi-objective cost optimization; the model provides a dynamic cost evaluation framework for intelligent irrigation systems, helping to synergistically optimize water conservation, energy saving, and equipment lifespan.

[0039] Furthermore, when there is a pressure difference in the pressure difference data that is greater than the preset pressure difference, it indicates that the filter has entered the inefficient zone, and the earliest time when the pressure difference that is greater than the preset pressure difference is taken as the occurrence time.

[0040] Furthermore, methods for obtaining the dataset for training the LSTM time series prediction network include:

[0041] A1. Differential pressure sensor and flow sensor are used to collect differential pressure and flow rate in real time;

[0042] A2. Calculate the efficiency coefficient based on the real-time flow rate, and start the ultrasonic cleaning component when the efficiency coefficient is less than or equal to the preset threshold and the real-time pressure difference is greater than or equal to the preset pressure difference.

[0043] A3. When the real-time differential pressure equals the sewage discharge termination differential pressure, turn off the ultrasonic cleaning component, and then return to step A1 until one irrigation cycle is completed.

[0044] A4. After the irrigation cycle is completed, record the differential pressure data, flow rate data, and irrigation cycle for the entire cycle.

[0045] A5. Use steps A1 to A4 to obtain differential pressure data, flow rate data, and irrigation cycle data for multiple different irrigation cycles, and use the differential pressure data, flow rate data, and irrigation cycle data for multiple irrigation cycles as a dataset.

[0046] The beneficial effects of the above technical solution are as follows: By collecting differential pressure and flow rate data in real time, and dynamically triggering cleaning decisions based on efficiency coefficients, the ultrasonic component is precisely shut down when the termination differential pressure is reached, forming a complete closed-loop control process; subsequently, full-time data within the irrigation cycle is recorded, and finally, multi-cycle data is integrated to construct an LSTM training dataset. This solution achieves three core benefits:

[0047] ① Precise data acquisition: Multi-sensor fusion ensures the real-time and completeness of differential pressure and flow data, providing high-quality input for LSTM;

[0048] ② Intelligent cleaning decision-making: The efficiency coefficient and dual threshold mechanism (efficiency ≤ preset value & pressure difference ≥ threshold) reduce false triggering and optimize resource consumption;

[0049] ③ Efficient Dataset Construction: Closed-loop recording of the entire irrigation cycle data directly generates time-series datasets covering complex working conditions, accelerating model training and optimization.

[0050] The overall technical solution significantly improves the response accuracy, water and energy saving, and model generalization ability of the self-cleaning system.

[0051] Furthermore, the discharge termination pressure difference The expression is ; Preset differential pressure; This refers to the pressure difference of clean water.

[0052] Furthermore, the method for determining the preset pressure difference includes:

[0053] B1. Read the turbidity data collected by the turbidity sensor and calculate the difference between two adjacent turbidity data points;

[0054] B2. Determine whether the absolute value of the difference is greater than the turbidity threshold. If yes, proceed to step B3; otherwise, return to step B1 if the filter is not turned off.

[0055] B3. When the difference indicates an increase in turbidity, the preset differential pressure is increased by a preset ratio; when the difference indicates a decrease in turbidity, the preset differential pressure is decreased by a preset ratio, and then the process returns to step B1 if the filter is not closed.

[0056] The beneficial effects of this invention are as follows: This solution cleans the inner filter screen using ultrasonic vibration of the ultrasonic cleaning component. During cleaning, the detached mud and sand particles remain inside the inner filter screen, eliminating the need to close the water outlet. The filtered irrigation water can flow out normally through the outlet. During cleaning, only the drain outlet needs to be opened, and the cleaned mud and sand water can be discharged from the top drain outlet. Throughout the cleaning process, the filter can normally filter and output irrigation water, with only a small amount of water lost to discharge mud and sand. This solution ensures the normal delivery of irrigation water, guaranteeing uninterrupted irrigation and improving irrigation efficiency.

[0057] Because the vibration units of this solution are evenly distributed on the inner filter screen, the entire inner filter screen can be vibrated during the cleaning process, which can avoid the inability to remove some fine particles during existing scraping and backwashing; and since there are no mechanical parts in contact with the surface of the inner filter screen during cleaning, the possibility of mechanical damage to the filter screen can be avoided.

[0058] At the beginning of each irrigation cycle, this solution can predict the pressure difference and flow rate at all time points within the current irrigation cycle by using historical pressure difference and flow rate information of the filter. Based on this, the solution further combines the pressure difference data of entering the inefficient zone to select the optimal irrigation cycle and cleaning duration. This solution cleans when the pressure difference is in the inefficient zone, which firstly ensures the filtration efficiency of the filter. By starting the cleaning under the appropriate pressure difference, it can also minimize the operating costs of the water pump and filter during the irrigation period, thereby reducing irrigation costs. Attached Figure Description

[0059] Figure 1 This is a 3D view of an ultrasonic self-cleaning filter.

[0060] Figure 2 This is a top view of an ultrasonic self-cleaning filter.

[0061] Figure 3 This is a flowchart of an intelligent control method for an ultrasonic self-cleaning filter.

[0062] The components include: 1. Shell assembly; 11. Cylinder; 111. Outlet; 12. Base; 121. Inlet; 13. Cover; 131. Drain; 2. Inner filter; 3. Vibration unit. Detailed Implementation

[0063] The specific embodiments of the present invention are described below to enable those skilled in the art to understand the present invention. However, it should be understood that the present invention is not limited to the scope of the specific embodiments. For those skilled in the art, various changes are obvious as long as they are within the spirit and scope of the present invention as defined and determined by the appended claims. All inventions utilizing the concept of the present invention are protected.

[0064] like Figure 1 and Figure 2 As shown, the ultrasonic self-cleaning filter provided by this solution includes a housing assembly 1, a filter assembly, an ultrasonic cleaning assembly, and a control system. The housing assembly 1 is a sealed cylindrical structure 11, specifically including a cylindrical body 11, a base 12, and a pressure cap 13. The upper and lower openings of the cylindrical body 11 are sealed by the pressure cap 13 and the base 12. The top of the pressure cap 13 is provided with a drain port 131, the side wall of the cylindrical body 11 is provided with a water outlet 111, and the base 12 is provided with a water inlet 121.

[0065] The filter assembly includes a cylindrical inner filter screen 2, which is located inside the housing assembly 1 and is sealed to the top and bottom of the side wall of the housing assembly 1 by a sealing ring; wherein the outlet 111 is located on the side wall of the cylindrical body 11 where the inner filter screen 2 is located, so as to ensure that the irrigation water enters the outlet 111 after passing through the inner filter screen 2.

[0066] The ultrasonic cleaning assembly includes an ultrasonic generator mounted on the outer surface of the housing assembly and multiple vibration units 3 evenly distributed on the inner filter screen and connected to the ultrasonic generator. The even distribution ensures coverage of the entire inner filter screen area. In this design, the vibration unit 3 is an ultrasonic transducer. The ultrasonic generator is electrically connected to the control system. To ensure the service life of the vibration unit 3, a corrosion-resistant rectangular cylindrical sealing shell is provided on the outer surface of each vibration unit. The power module of the vibration unit 3 is located on the outer wall of the cylinder 11 to achieve water-electricity separation.

[0067] The control system includes a controller and a differential pressure sensor and a flow sensor electrically connected to the controller and installed at the outlet 111. To facilitate information retrieval by personnel, it may also include a human-machine interface electrically connected to the controller. The controller is connected to a cloud platform via a wireless communication module.

[0068] The controller is used to determine the optimal cleaning time and cleaning cycle of the filter under the current irrigation cycle based on the filter's historical operating data and cleaning data. The detailed implementation process of this process can be found in the intelligent control method of steps S1 to S7.

[0069] In implementation, this solution preferably includes an ultrasonic self-cleaning filter that also includes a turbidity sensor installed inside the inlet 121 and electrically connected to the controller. The collected turbidity data can be used to accurately determine the time when the filter enters the inefficient zone.

[0070] In one embodiment of the present invention, the method for determining the operating frequency of the vibration unit 3 includes:

[0071] Based on the rated operating frequency of vibration unit 3, calculate the sound wave attenuation coefficient of the ultrasonic wave in the filter medium:

[0072]

[0073] in, The sound wave attenuation coefficient; is the dielectric constant; f is the rated operating frequency; b is the frequency index;

[0074] Determine the penetration distance and sound wave attenuation value of the ultrasonic wave within the housing assembly 1, and calculate the maximum frequency of the vibration unit 3:

[0075]

[0076] Where m is the sound wave attenuation value; D is the penetration distance;

[0077] The maximum frequency is set as the operating frequency of vibration unit 3.

[0078] To ensure that the ultrasonic waves from the vibration unit 3 can completely cover the entire inner filter screen 2 area, this design preferably arranges the vibration units 3 evenly at the upper end of the inner filter screen 2, with the maximum spacing between adjacent vibration units 3 being... , Where is the operating frequency, and c is the speed at which ultrasound propagates in the medium.

[0079] like Figure 3 As shown, this solution also provides an intelligent control method for an ultrasonic self-cleaning filter, which includes steps S1 to S7.

[0080] In step S1, the historical flow rate data, differential pressure data, and cleaning cycle of the filter in multiple irrigation cycles before the current irrigation cycle are obtained; the cleaning cycle in this scheme refers to the time between two adjacent starts of the vibration unit 3.

[0081] In step S2, the acquired data is input into the trained LSTM time series prediction network to predict the flow rate data, differential pressure data, and cleaning cycle within the current irrigation cycle.

[0082] In one embodiment of the present invention, the method for obtaining the dataset for training the LSTM time series prediction network includes:

[0083] A1. Differential pressure sensor and flow sensor are used to collect differential pressure and flow rate in real time;

[0084] A2. Calculate the efficiency coefficient η1 = Q / Q based on the real-time traffic flow. design Q design The ultrasonic cleaning component is activated when the efficiency coefficient is less than or equal to a preset threshold (preferably 85%) and the real-time differential pressure is greater than or equal to a preset differential pressure.

[0085] In practice, this scheme is preferably implemented in a way that the method for determining the preset differential pressure includes:

[0086] B1. Read the turbidity data collected by the turbidity sensor and calculate the difference between two adjacent turbidity data points;

[0087] B2. Determine whether the absolute value of the difference is greater than the turbidity threshold. If yes, proceed to step B3; otherwise, return to step B1 if the filter is not turned off.

[0088] B3. When the difference indicates an increase in turbidity, the preset differential pressure is increased by a preset ratio; when the difference indicates a decrease in turbidity, the preset differential pressure is decreased by a preset ratio, and then the process returns to step B1 if the filter is not closed.

[0089] A3. When the real-time differential pressure equals the sewage discharge termination differential pressure, turn off the ultrasonic cleaning component, and then return to step A1 until one irrigation cycle is completed.

[0090] Among them, the sewage discharge termination pressure difference The expression is ; Preset differential pressure; This refers to the pressure difference of clean water.

[0091] A4. After the irrigation cycle is completed, record the differential pressure data, flow rate data, and irrigation cycle for the entire cycle.

[0092] A5. Use steps A1 to A4 to obtain differential pressure data, flow rate data, and irrigation cycle data for multiple different irrigation cycles, and use the differential pressure data, flow rate data, and irrigation cycle data for multiple irrigation cycles as a dataset.

[0093] In step S3, the occurrence time of the inefficient zone is determined based on the differential pressure data in the first cleaning cycle of the current irrigation cycle. In this scheme, when there is a differential pressure data that is greater than the preset differential pressure, it indicates that the filter has entered the inefficient zone, and the earliest time when the differential pressure that is greater than the preset differential pressure is taken as the occurrence time.

[0094] In step S4, the time corresponding to each pressure difference between the occurrence time and the end of the first cleaning cycle is read as the new cleaning cycle, and the total number of new cleaning cycles for the current irrigation cycle is determined; to facilitate understanding of the new cleaning cycle, an example is given below:

[0095] Assuming the irrigation filtration cycle lasts 48 hours, after processing with an LSTM time-series prediction network, 12 cleaning cycles are generated, meaning cleaning occurs every 4 hours (including filtration and cleaning time). In the first cycle, assuming the filter enters an inefficient zone at 2.5 hours, a pressure differential can be selected at intervals between this point and the end of the first cleaning cycle, for example, 10 minutes. At 2 hours and 40 minutes, a pressure differential is selected, requiring 10 minutes of cleaning. The new cleaning cycle is then 2 hours and 50 minutes. Dividing 48 hours by 2 hours and 50 minutes and rounding down gives the total number of new cleaning cycles, i.e., the number of cleaning cycles. At 2 hours and 50 minutes, another pressure differential is selected, requiring 12 minutes of cleaning. The new cleaning cycle is then 3 hours and 2 minutes. This process continues until a new cleaning cycle is defined as 3 hours and 50 minutes.

[0096] In step S5, the flow rate and pressure difference data for each new cleaning cycle corresponding to each pressure difference are set to be equal to the corresponding data for the first new cleaning cycle. At this time, a new cycle is defined as 2 hours and 50 minutes. After cleaning is completed, the flow rate will return to the optimal state. Theoretically, the flow rate and pressure difference for each subsequent cycle are the same as those for the first cycle. Therefore, this scheme sets the pressure difference and flow rate for subsequent cycles to be the same as those for the first new cleaning cycle.

[0097] In step S6, based on the cleaning time and flow rate data and pressure difference data corresponding to each pressure difference, the total cost of water used by the filter and electricity used by the filter and water pump in the current irrigation cycle is calculated.

[0098] In implementation, this scheme preferably uses the following expression to calculate the total cost of water used for the filter and electricity used for the filter and pump during the current irrigation cycle:

[0099]

[0100]

[0101]

[0102]

[0103] Where F is the pressure difference The corresponding total cost; The j-th pressure difference between the occurrence time and the end of the first cleaning cycle, 1≤j≤J, where J is the total number of pressure differences selected between the occurrence time and the end of the first cleaning cycle; , and All are weighting coefficients; , and These are the energy costs of the ultrasonic cleaning components, the water pump, and the wastewater discharge costs. M is the depreciation factor; M is the cost of internal filter 2 wear and tear. Ultrasonic power; and They are respectively The corresponding duration and wastewater discharge volume for the i-th cleaning cycle, where I is the pressure difference. Total number of cleaning cycles; The inverter efficiency coefficient of the ultrasonic cleaning component; For industrial electricity prices; and Both represent the pressure difference and flow rate at time t; T is the total duration of the current irrigation cycle. The density of the medium; For water pump efficiency; This is the water quality impact coefficient; For water prices.

[0104] This solution calculates costs using the methods described above, enabling a comprehensive value-based calculation of vibration energy consumption, differential pressure energy consumption, and wastewater discharge water consumption, thus facilitating optimal cost control throughout the entire lifecycle.

[0105] In step S7, the cleaning time and new cleaning cycle corresponding to the pressure difference with the minimum total cost are selected and used as the optimal cleaning time and cleaning cycle for the filter under the current irrigation cycle.

[0106] In summary, this solution uses ultrasonic cleaning, which can improve cleaning efficiency while reducing wear on the inner filter 2. Combined with the intelligent control method of this solution, by selecting the pressure difference in the inefficient zone and considering cost, the cleaning time and cleaning cycle with the lowest cost and reasonable filtration effect can be selected, thus balancing cost and filtration effect.

Claims

1. An ultrasonic self-cleaning filter, characterized by, The application relates to a filter system and a method for determining the optimal cleaning time and cleaning cycle of a filter. The filter system comprises a shell assembly, an ultrasonic cleaning assembly, a filter assembly and a control system. The shell assembly is a sealed cylinder structure, the top of the shell assembly is provided with a sewage outlet, the bottom of the shell assembly is provided with a water inlet, and the side wall of the shell assembly is provided with a water outlet. The filter assembly comprises an inner filter screen in a cylindrical structure, the inner filter screen is arranged in the shell assembly, and the top end and the bottom end of the side wall of the shell assembly are sealed and connected through a sealing ring. The ultrasonic cleaning assembly comprises an ultrasonic generator arranged on the outer surface of the shell assembly and a plurality of vibration units uniformly distributed on the inner filter screen and connected with the ultrasonic generator, and the ultrasonic generator is electrically connected with the control system. The control system comprises a controller, a differential pressure sensor and a flow sensor arranged at the water outlet and electrically connected with the controller, and the controller is connected with a cloud platform through a wireless communication module. The controller is used for determining the optimal cleaning time and cleaning cycle of the filter in the current irrigation cycle according to historical operation data and cleaning data of the filter. The method for determining the working frequency of the vibration unit comprises the following steps: wherein is the sound wave attenuation coefficient; is the medium characteristic constant; f is the nominal operating frequency; b is the frequency exponent; According to the rated working frequency of the vibration unit, the sound wave attenuation coefficient of the ultrasonic wave in the filter medium is calculated. The penetration distance and the sound wave attenuation value of the ultrasonic wave in the shell assembly are determined, and the maximum frequency of the vibration unit is calculated. Wherein, m is the sound wave attenuation value; D is the penetration distance. The vibration units are uniformly arranged at the upper end of the inner filter screen, and the maximum distance between adjacent vibration units is less than 1 / 4 of the wavelength of the vibration frequency , is the working frequency, and c is the propagation speed of the ultrasonic wave in the medium.

2. The ultrasonic self-cleaning filter according to claim 1, wherein The set proportion of the maximum frequency is used as the working frequency of the vibration unit.

3. The intelligent control method of the ultrasonic self-cleaning filter according to claim 1 or 2, characterized in that, The turbidity sensor arranged in the water inlet and electrically connected with the controller is further included. The method comprises the following steps: S1, historical flow data, differential pressure data and cleaning cycles of the filter in a plurality of irrigation cycles before the current irrigation cycle are obtained. S2, the obtained data is input into a trained LSTM time series prediction network to obtain the flow data, differential pressure data and cleaning cycles in the current irrigation cycle. S3, the occurrence time of the low-efficiency area is determined according to the differential pressure data in the first cleaning cycle of the current irrigation cycle. S4, the time corresponding to each differential pressure between the occurrence time and the end of the first cleaning cycle is read as a new cleaning cycle, and the total number of new cleaning cycles in the current irrigation cycle is determined. S5, the flow data and differential pressure data in each new cleaning cycle corresponding to each differential pressure are set to be equal to the corresponding data in the first new cleaning cycle. S6, the total cost of water used by the filter and electricity used by the filter and the water pump in the current irrigation cycle is calculated according to the cleaning time corresponding to each differential pressure and the flow data and differential pressure data.

4. The intelligent control method of claim 3, wherein, S7, the cleaning time and the new cleaning cycle corresponding to the differential pressure with the minimum total cost are selected as the optimal cleaning time and cleaning cycle of the filter in the current irrigation cycle. wherein F is the pressure difference the corresponding total cost; is the jth pressure difference between the occurrence time and the end of the first cleaning cycle, 1≤j≤J, J is the total number of pressure differences selected between the occurrence time and the end of the first cleaning cycle; and are weight coefficients; and are the ultrasonic cleaning assembly, the water pump energy consumption cost and the sewage water consumption cost, respectively; is the depreciation factor; M is the inner filter screen loss cost; is the ultrasonic power; are the corresponding ith cleaning duration and sewage water volume, I is the total number of cleanings under the pressure difference ; is the frequency converter efficiency coefficient of the ultrasonic cleaning assembly; is the industrial electricity price; are the pressure difference and flow at time t; T is the total duration of the current irrigation cycle; is the density of the medium; is the water pump efficiency; is the water quality influence coefficient; is the water price.​​​​ 5. The intelligent control method of claim 3, wherein, The expression for calculating the total cost of water used by the filter and electricity used by the filter and the water pump in the current irrigation cycle is as follows:

6. The intelligent control method of claim 3, wherein, When there is a differential pressure greater than the preset differential pressure in the differential pressure data, it indicates that the filter runs into the low-efficiency area, and the time when the differential pressure greater than the preset differential pressure is met for the first time is taken as the occurrence time. The data set for training the LSTM time series prediction network comprises the following steps: A1, the differential pressure and the flow are collected in real time by using the differential pressure sensor and the flow sensor. A2, when the efficiency coefficient is less than or equal to the preset threshold value and the real-time differential pressure is greater than or equal to the preset differential pressure, the ultrasonic cleaning assembly is started according to the real-time flow. A3、When the real-time pressure difference is equal to the blowdown termination pressure difference, the ultrasonic cleaning assembly is closed, and then the step A1 is returned until a watering cycle is completed; A4、After the watering cycle is completed, the pressure difference data, the flow data and the watering cycle in the entire cycle are recorded; A5、The pressure difference data, the flow data and the watering cycle in multiple different watering cycles are obtained by using the steps A1-A4, and the pressure difference data, the flow data and the watering cycle of the multiple watering cycles are taken as a data set.

7. The intelligent control method of claim 6, wherein, The expression of the blowdown termination pressure differential is ; is a preset pressure differential; is a clean water pressure differential.

8. The intelligent control method of claim 6, wherein, The method for determining the preset pressure difference comprises: B1、Read the turbidity data collected by the turbidity sensor, and calculate the difference between the adjacent two turbidity data; B2、Determine whether the absolute value of the difference is greater than the turbidity threshold value, if yes, go to step B3, otherwise return to step B1 when the filter is not closed; B3、When the difference indicates that the turbidity increases, the preset pressure difference is increased by a preset proportion; when the difference indicates that the turbidity decreases, the preset pressure difference is decreased by a preset proportion, and then return to step B1 when the filter is not closed.

Images