Water pump multi-working-condition performance testing device and testing method based on digital twinning

Abstract

The invention relates to the technical field of pumps, in particular to a water pump multi-working-condition performance testing device and method based on digital twinning, and the water pump multi-working-condition performance testing device based on digital twinning comprises a water pump control system, a water pump service system and a water pump virtual system. The water pump control system comprises a water pump module, a control system module and a collection module, the water pump module and the collection module are connected with the control system module, the control system module is connected with the water pump service system and the water pump virtual system, and the water pump virtual system is connected with the water pump service system. A water pump multi-working-condition performance test method based on digital twinning comprises the following steps of regulating and controlling operation of a water pump, collecting and processing data during operation of the water pump, conducting numerical simulation on the multi-working-condition operation process of the water pump, constructing a neural network, obtaining an optimal operation strategy and feeding back the optimal operation strategy till a target working condition is achieved. The system is ingenious in design, high in response speed, high in safety and good in stability.

Description

Technical Field

[0001] The present invention relates to the field of pump technology, and specifically to a water pump multi-operating condition performance test device and test method based on digital twin. Background Art

[0002] With the improvement of residents' living standards and the development of modern industry and agriculture, water pumps are used more and more widely. During operation, water pumps mainly respond to user needs quickly and provide energy for fluids. In the fields of water supply and drainage, power systems, petrochemicals, etc., water pumps will not always operate at the designed operating point, and need to switch operating points according to user needs. In order to improve energy utilization efficiency, reduce energy waste, and extend equipment life, the water pump system needs to quickly judge and respond to demand conditions. In this process, the water pump needs to transition from one stable operating condition to another, that is, the water pump will be in an unstable transition process with drastic performance changes. The existing technology usually uses empiricism to judge and adjust the transition process of the pump, which has a slow response speed and great safety hazards. Summary of the invention

[0003] The purpose of the present invention is to solve the shortcomings of the prior art and to provide a water pump multi-condition performance test device and test method based on digital twins with ingenious design, fast response speed, high safety and good stability.

[0004] To achieve the above purpose, the technical solution adopted by the present invention is: A water pump multi-operating condition performance test device based on digital twin, characterized in that it includes a water pump control system, a water pump service system and a water pump virtual system, the water pump control system includes a water pump module, a control system module and an acquisition module, the water pump module and the acquisition module are respectively connected to the control system module, the control system module is respectively connected to the water pump service system and the water pump virtual system, the water pump virtual system is connected to the water pump service system, so as to facilitate the control of the water pump operation by the control system module, the data of the water pump operation is collected by the acquisition module and transmitted to the control system module, the control system module transmits the data to the water pump service system and the water pump virtual system respectively, the water pump virtual system performs numerical simulation on the water pump multi-operating conditions based on the operation database, obtains the value obtained by the numerical simulation of the water pump virtual system, the water pump service system receives the data collected by the control system module, constructs a database and a neural network, processes the value obtained by the numerical simulation of the water pump virtual system, obtains the optimal control strategy and feeds it back to the control system module, and updates the operation database at the same time until the target operating condition is reached.

[0005] The water pump module of the present invention includes a water pump, a water tank, a first solenoid valve, a second solenoid valve and a variable frequency motor, the control system module includes a control cabinet, an electrical cabinet and a frequency converter, the acquisition module includes a first static pressure sensor, a first dynamic pressure sensor, a second static pressure sensor, a second dynamic pressure sensor, a flow meter, a torque speed sensor, a host computer and an acquisition card, The water inlet of the water pump is connected to the water outlet of the water tank via a water inlet pipeline, and the water outlet of the water pump is connected to the water inlet of the water tank via a water outlet pipeline. The first solenoid valve, the first static pressure sensor, the first dynamic pressure sensor, and the flow meter are installed on the outlet pipe of the water pump, and the second solenoid valve, the second static pressure sensor, and the second dynamic pressure sensor are installed on the inlet pipe of the water pump. The water pump is driven by a variable frequency motor, the frequency converter is connected to the variable frequency motor, the torque speed sensor is connected to the driving shaft of the water pump, and the water pump, the first solenoid valve, the first static pressure sensor, the first dynamic pressure sensor, the second solenoid valve, the second static pressure sensor, the second dynamic pressure sensor, the flow meter, the host computer, the acquisition card, the electrical cabinet, and the frequency converter are respectively connected to the control cabinet to facilitate controlling the valve opening of the water pump outlet through the first solenoid valve, detecting the static pressure value and the dynamic pressure value at the water pump outlet through the first static pressure sensor and the first dynamic pressure sensor, measuring the flow rate at the pump outlet through the flow meter, controlling the valve opening of the water pump inlet through the second solenoid valve, detecting the static pressure value and the dynamic pressure value at the water pump inlet through the second static pressure sensor and the second dynamic pressure sensor, detecting the torque and speed of the water pump through the torque speed sensor, adjusting the speed of the water pump through the frequency converter, and then adjusting the output power of the water pump.

[0006] The water pump virtual system of the present invention comprises a numerical simulation module and a dynamic display module. The numerical simulation module is used to simulate the operation data obtained by the control system module through a numerical simulation method and a result preprocessing method, and the dynamic display module is used to visualize the system operation status.

[0007] The water pump service system of the present invention includes a machine learning module, an optimal strategy selection module and a judgment module. The machine learning module is used to determine various parameters of the water pump operation through machine learning technology: the relationship between the water pump's flow rate, head, outlet static pressure, dynamic pressure, inlet static pressure, dynamic pressure, and valve opening, build a database, and construct a neural network. The optimal strategy selection module is used to find the corresponding strategy through the neural network to obtain the optimal operating strategy. The judgment module is used to calculate the response time of the optimal strategy and judge the rationality of the strategy selection. If the time is too long, the number of strategy groups will be reselected.

[0008] The optimal strategy selection module of the present invention uses a determined neural network to find (P'-P) / P'≤0.01 according to P' required by the user, where P is the current output power of the water pump and P' is the expected output power of the water pump.

[0009] The expression of the strategy optimization mode of the optimal strategy selection module of the present invention is: Y=(Valve 1 opening'-Valve 1 opening) 2 + (valve 2 opening'-valve 2 opening) 2 , Among them, valve 1 opening' is the valve target opening of the first solenoid valve, and valve 1 opening is the valve current opening of the first solenoid valve; valve 2 opening' is the valve target opening of the second solenoid valve, and valve 2 opening is the valve current opening of the second solenoid valve. Several groups of strategies are selected, and the group with the smallest Y value is the optimal strategy.

[0010] The control system module of the present invention transmits the data of water pump operation to the water pump virtual system, so that the boundary conditions during numerical simulation in the virtual system and the real-time status of the water pump remain highly consistent, and performs numerical simulation on the strategy scheme given by the water pump service system, and outputs torque value, power value and pressure pulsation to the water pump service system. After the judgment module of the water pump service system makes a judgment, the strategy is passed to the water pump control system.

[0011] A multi-operating-condition performance test method for a water pump based on digital twins, characterized in that it comprises the following steps: Step 1: Build a database. The water pump control system transmits the water pump multi-condition operation data to the water pump service system. The water pump service system receives the data of the water pump control system to build a database and apply machine learning methods to fit the neural network. Step 2: In the water pump control system, the water flow direction is: water tank-second static pressure sensor, second dynamic pressure sensor-second solenoid valve-water pump-first solenoid valve-first static pressure sensor, first dynamic pressure sensor-flow meter-water tank. The control module opens the second solenoid valve, and the water in the water tank flows into the water pump through the water inlet pipe. When the water fills the entire pump cavity of the water pump, the control module controls the variable frequency motor, and the variable frequency motor starts the water pump. The control module opens the first solenoid valve and adjusts the valve opening of the first solenoid valve. The flow meter monitors the flow at the outlet of the water pump at the same time until the outlet flow of the water pump reaches the preset target flow value. The control module controls the first solenoid valve and the second solenoid valve, and monitors the speed of the water pump through the torque speed sensor. After the speed of the water pump reaches 90% of the rated speed, the control module adjusts the valve opening of the first solenoid valve and the valve opening of the second solenoid valve so that the output power of the water pump reaches the current test power. Step 3: The real-time operating data in the water pump control system is transmitted to the water pump virtual system, and the water pump virtual system performs numerical simulation on the current working condition of the water pump; Step 4: Pass the simulated user demand target to the water pump service system, use the database established in the first step, and find the control strategy through the neural network according to the simulated target operating conditions, with a limit of 50 groups of strategies. If more than 50 groups are found, stop looking for strategies.

[0012] Step 5: Select the optimal strategy. The strategy with the smallest Y value among the 50 strategies is the optimal strategy. Step 6: The optimal strategy data value is passed to the water pump virtual system. In the water pump virtual system, the change value of the valve opening is split, and the splitting index is that the relative change rate θ does not exceed 5%. Op in the formula d Op represents the valve opening at a moment after the split, Op represents the valve opening at a moment before the split, and the valve opening of the first solenoid valve and the valve opening of the second solenoid valve are split respectively; Step 7: The split valve opening is transmitted to the ANSYS Workbench platform, and the previous working condition is used as the initial working condition to form a numerical simulation calculation of the multi-working condition transition process, and the calculation results are transmitted to the water pump service system; Step 8: In the water pump service system, the pressure pulsation value obtained by numerical simulation is evaluated to ensure that the amplitude of the pressure pulsation is not greater than the maximum pressure pulsation value that the unit can withstand, and the main frequency is different from the natural frequency, that is, to ensure that the amplitude of the pressure pulsation is within a safe range. When the frequency of the pressure pulsation is inconsistent with the resonant frequency of the unit, the new valve opening is transmitted to the water pump control system; Step 9: Transmit the real-time data of the water pump control system to the water pump service system, update the database, and repeat steps 1 to 8 until the output power reaches the target value.

[0013] The present invention has the advantages of clever design, fast response speed, high safety, good stability, etc. due to the adoption of the above structure. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 It is a schematic diagram of the interconnection relationship among the water pump control system, water pump service system and water pump virtual system of the present invention.

[0015] Figure 2 It is a schematic diagram of the water pump control system in the present invention.

[0016] Figure 3 Schematic diagram of the water pump virtual system in the present invention. DETAILED DESCRIPTION

[0017] The specific implementation modes of the present invention are further described in detail below in conjunction with the accompanying drawings.

[0018] A water pump multi-operating condition performance test device based on digital twin, characterized in that it includes a water pump control system, a water pump service system and a water pump virtual system, the water pump control system includes a water pump module, a control system module and an acquisition module, the water pump module and the acquisition module are respectively connected to the control system module, the control system module is respectively connected to the water pump service system and the water pump virtual system, the water pump virtual system is connected to the water pump service system, so as to facilitate the control of the water pump operation by the control system module, the data of the water pump operation is collected by the acquisition module and transmitted to the control system module, the control system module transmits the data to the water pump service system and the water pump virtual system respectively, the water pump virtual system performs numerical simulation on the water pump multi-operating conditions based on the operation database, obtains the value obtained by the numerical simulation of the water pump virtual system, the water pump service system receives the data collected by the control system module, constructs a database and a neural network, processes the value obtained by the numerical simulation of the water pump virtual system, obtains the optimal control strategy and feeds it back to the control system module, and updates the operation database at the same time until the target operating condition is reached.

[0019] The water pump module of the present invention includes a water pump, a water tank, a first solenoid valve, a second solenoid valve and a variable frequency motor, the control system module includes a control cabinet, an electrical cabinet and a frequency converter, the acquisition module includes a first static pressure sensor, a first dynamic pressure sensor, a second static pressure sensor, a second dynamic pressure sensor, a flow meter, a torque speed sensor, a host computer and an acquisition card, The water inlet of the water pump is connected to the water outlet of the water tank via a water inlet pipeline, and the water outlet of the water pump is connected to the water inlet of the water tank via a water outlet pipeline. The first solenoid valve, the first static pressure sensor, the first dynamic pressure sensor, and the flow meter are installed on the outlet pipe of the water pump, and the second solenoid valve, the second static pressure sensor, and the second dynamic pressure sensor are installed on the inlet pipe of the water pump. The water pump is driven by a variable frequency motor, the frequency converter is connected to the variable frequency motor, the torque speed sensor is connected to the driving shaft of the water pump, and the water pump, the first solenoid valve, the first static pressure sensor, the first dynamic pressure sensor, the second solenoid valve, the second static pressure sensor, the second dynamic pressure sensor, the flow meter, the host computer, the acquisition card, the electrical cabinet, and the frequency converter are respectively connected to the control cabinet to facilitate controlling the valve opening of the water pump outlet through the first solenoid valve, detecting the static pressure value and the dynamic pressure value at the water pump outlet through the first static pressure sensor and the first dynamic pressure sensor, measuring the flow rate at the pump outlet through the flow meter, controlling the valve opening of the water pump inlet through the second solenoid valve, detecting the static pressure value and the dynamic pressure value at the water pump inlet through the second static pressure sensor and the second dynamic pressure sensor, detecting the torque and speed of the water pump through the torque speed sensor, adjusting the speed of the water pump through the frequency converter, and then adjusting the output power of the water pump.

[0020] The water pump virtual system of the present invention comprises a numerical simulation module and a dynamic display module. The numerical simulation module is used to simulate the operation data obtained by the control system module through a numerical simulation method and a result preprocessing method, and the dynamic display module is used to visualize the system operation status.

[0021] The water pump service system of the present invention includes a machine learning module, an optimal strategy selection module and a judgment module. The machine learning module is used to determine various parameters of the water pump operation through machine learning technology: the relationship between the water pump's flow rate, head, outlet static pressure, dynamic pressure, inlet static pressure, dynamic pressure, and valve opening, and construct a neural network. The optimal strategy selection module is used to find the corresponding strategy through the neural network to obtain the optimal operating strategy. The judgment module is used to calculate the response time of the optimal strategy and judge the rationality of the strategy selection.

[0022] The optimal strategy selection module of the present invention uses a determined neural network to find (P'-P) / P'≤0.01 according to P' required by the user, where P is the current output power of the water pump and P' is the expected output power of the water pump.

[0023] The expression of the strategy optimization mode of the optimal strategy selection module of the present invention is: Y=(Valve 1 opening'-Valve 1 opening) 2 + (valve 2 opening'-valve 2 opening) 2 , Among them, valve 1 opening' is the valve target opening of the first solenoid valve, and valve 1 opening is the valve current opening of the first solenoid valve; valve 2 opening' is the valve target opening of the second solenoid valve, and valve 2 opening is the valve current opening of the second solenoid valve. Several groups of strategies are selected, and the group with the smallest Y value is the optimal strategy.

[0024] The control system module of the present invention transmits the data of water pump operation to the water pump virtual system, so that the boundary conditions during numerical simulation in the virtual system and the real-time status of the water pump remain highly consistent, and performs numerical simulation on the strategy scheme given by the water pump service system, and outputs torque value, power value and pressure pulsation to the water pump service system. After the judgment module of the water pump service system makes a judgment, the strategy is passed to the water pump control system.

[0025] This embodiment simulates the situation where the current water pump control system is in an output power of 2Kw and needs to switch to an output power of 3Kw.

[0026] A multi-operating-condition performance test method for a water pump based on digital twins, characterized in that it comprises the following steps: Step 1: Build a database, adjust the variable frequency motor through the control cabinet, the variable frequency motor drives the water pump, the control module opens the second solenoid valve, the water in the water tank flows into the water pump through the water inlet pipe, when the water fills the entire pump cavity of the water pump, the control module controls the variable frequency motor, the variable frequency motor starts the water pump, the control module opens the first solenoid valve, adjusts the valve opening of the first solenoid valve, and the flow meter monitors the flow at the outlet of the water pump until the outlet flow of the water pump reaches a preset target flow value. The control module controls the first solenoid valve and the second solenoid valve, and monitors the speed of the water pump through the torque speed sensor. After the speed of the water pump reaches 90% of the rated speed, the control module adjusts the valve opening of the first solenoid valve and the valve opening of the second solenoid valve so that the output power of the water pump reaches the current test power, and obtains multiple sets of data values ​​of the valve opening value of the first solenoid valve, the valve opening value of the second solenoid valve, the outlet static pressure value, the outlet dynamic pressure value, the inlet static pressure value, the inlet dynamic pressure value, the torque value, the flow rate, the head, and the power value, and transmits the above data to the water pump service system; Step 2: The water pump service system establishes a database for the data obtained in step 1, applies machine learning methods to establish a neural network, and passes the data X and power value extracted from the database into the neural network. The neural network automatically starts to iterate until the iteration error is less than 0.01, and the update is completed. The details are as follows: 1. Grid initialization, extract data X and power P from the database, determine the number of neurons in the grid input layer, hidden layer and output layer, set the connection weight v at each node between the input layer and hidden layer, hidden layer and output layer ij ,w jk , initialize the hidden layer threshold a, output layer threshold b, 2. Hidden layer output calculation: In the formula, v i0 =-1,x0=a j ; Transfer function for the hidden layer.

[0027] x i represents the kth group of input values ​​(including data values ​​collected such as flow rate, head, outlet static pressure, etc., k is the number of data groups in the database), y j According to the above formula, x i The corresponding hidden layer output, j represents the number of neurons in the hidden layer; 3. Output layer output calculation: o k Indicates the valve opening corresponding to each set of inputs obtained by the neural network; 4. Error calculation: d k Indicates the actual valve opening value corresponding to each set of input values ​​in the database; 5. Weight update: represents a value that characterizes a characteristic of gradient propagation, Where η is the learning rate, which is generally 0.1 by default; 6. When the model error is less than 0.01, the update ends; Step 3: Adjust the valve opening of the first solenoid valve and the valve opening of the second solenoid valve to make the water pump work at an output power of 2Kw, that is, the current test power is 2Kw and the target flow value is 73.6m 3 / h, the head is 10 m, and the calculation method of the flow rate, head and power of the water pump is the same as that of the prior art and will not be repeated; Step 4: The data in the water pump control system at this time is transferred to the water pump virtual system, and the working condition with an output power of 2Kw is numerically simulated using the ANSYS Workbench platform; Step 5: The user demand target, i.e., transition to the operating condition with an output power of 3Kw, is transmitted to the water pump service system. In the neural network established in step 2, the corresponding strategy is searched according to the target operating condition, with a limit of 50 sets of corresponding strategies. When the requirement is reached, the strategy search is stopped; Step 6: Select the optimal strategy and the optimal evaluation index: Y=(valve 1 opening'-valve 1 opening) 2 + (valve 2 opening'-valve 2 opening) 2； Among them, valve 1 opening' is the valve target opening of the first solenoid valve, and valve 1 opening is the valve current opening of the first solenoid valve; valve 2 opening' is the valve target opening of the second solenoid valve, and valve 2 opening is the valve current opening of the second solenoid valve. 50 groups of strategies are selected, and the group with the smallest Y value is the optimal strategy; Step 7: The optimal strategy data value is passed to the water pump virtual system. In the water pump virtual system, the change value of the valve opening is split, and the splitting index is that the change rate θ does not exceed 5%. Op in the formula d Op represents the valve opening at the moment after the split, and Op represents the valve opening at the moment before the split.

[0028] Step 8: Import the split valve opening into the ANSYS Workbench platform, take the previous working condition as the initial working condition, form a numerical simulation calculation of the transition process, and pass the calculation results to the water pump service system; Step 9: In the water pump service system, the pressure pulsation value obtained by numerical simulation is evaluated to ensure that the amplitude of the pressure pulsation does not exceed the maximum pressure pulsation value that the unit can withstand, and the main frequency is different from the natural frequency, that is, to ensure that the amplitude of the pressure pulsation is within a safe range. When the frequency of the pressure pulsation is inconsistent with the resonant frequency of the unit, the new valve opening is transmitted to the water pump control system; Step 10: Transfer the operating data of the water pump control system to the water pump service system, update the database, and repeat steps 1 to 9 until the output power reaches 3Kw and then stops.

[0029] As attached Figure 1-3 The present invention is cleverly designed and simple in structure. It only has three systems: a water pump control system for regulating the operation of the water pump, collecting data when the water pump is running, and processing the collected data when the water pump is running to obtain an operation database; a water pump virtual system for numerically simulating the multi-operating condition operation process of the water pump based on the operation database to obtain the value obtained by the numerical simulation of the water pump virtual system; a water pump service system is responsible for building a database for the data from the water pump control module, constructing a neural network, processing the values ​​obtained by the numerical simulation of the water pump virtual system, obtaining the optimal operation strategy and feeding it back to the water pump control system, and updating the operation database at the same time until the target operating condition is reached. The three systems interact with each other, and the use of neural networks makes the optimal strategy selection fast, the response speed fast, the amplitude of the pressure pulsation does not exceed the maximum pressure pulsation value that the unit can withstand, the safety is high, and the water pump operation stability is good.

[0030] The present invention has the advantages of clever design, fast response speed, high safety, good stability, etc. due to the adoption of the above structure.

Claims

1. A water pump multi-operating condition performance test device based on digital twin, characterized by: It includes a water pump control system, a water pump service system and a water pump virtual system. The water pump control system includes a water pump module, a control system module and a collection module. The water pump module and the collection module are respectively connected to the control system module. The control system module is respectively connected to the water pump service system and the water pump virtual system. The water pump virtual system is connected to the water pump service system.

2. According to claim 1, a water pump multi-operating condition performance test device based on digital twin is characterized in that: The water pump module includes a water pump, a water tank, a first solenoid valve, a second solenoid valve and a variable frequency motor, the control system module includes a control cabinet, an electrical cabinet and a frequency converter, the acquisition module includes a first static pressure sensor, a first dynamic pressure sensor, a second static pressure sensor, a second dynamic pressure sensor, a flow meter, a torque speed sensor, a host computer and an acquisition card, the water inlet of the water pump is connected to the water outlet of the water tank via a water inlet pipeline, the water outlet of the water pump is connected to the water inlet of the water tank via a water outlet pipeline, the first solenoid valve, the first static pressure sensor, the first dynamic pressure sensor, the second static pressure sensor, the second dynamic pressure sensor, the flow meter, the torque speed sensor, the host computer and the acquisition card, the water inlet of the water pump is connected to the water outlet of the water tank via a water outlet pipeline, the first solenoid valve, the first static pressure sensor, the first dynamic pressure sensor, the second static pressure sensor, the second dynamic pressure sensor, the flow meter, the torque speed sensor, the host computer and the acquisition card, The pressure sensor and the flow meter are installed on the water outlet pipe of the water pump, the second solenoid valve, the second static pressure sensor, and the second dynamic pressure sensor are installed on the inlet pipe of the water pump, the water pump is driven by a variable frequency motor, the frequency converter is connected to the variable frequency motor, the torque speed sensor is connected to the driving shaft of the water pump, the water pump, the first solenoid valve, the first static pressure sensor, the first dynamic pressure sensor, the second solenoid valve, the second static pressure sensor, the second dynamic pressure sensor, the flow meter, the host computer, the acquisition card, the electrical cabinet, and the frequency converter are respectively connected to the control cabinet.

3. A water pump multi-operating condition performance test device based on digital twin according to claim 1 or 2, characterized in that: The water pump virtual system includes a numerical simulation module and a dynamic display module. The numerical simulation module is used to simulate the operation data obtained by the control system module through a numerical simulation method and a result preprocessing method, and the dynamic display module is used to visualize the system operation status.

4. According to claim 2, a water pump multi-operating condition performance test device based on digital twin is characterized in that: The water pump service system includes a machine learning module, an optimal strategy selection module and a judgment module. The machine learning module is used to determine the relationship between various parameters of the water pump operation through machine learning technology and construct a neural network. The optimal strategy selection module is used to find the corresponding strategy through the neural network and obtain the optimal operation strategy. The judgment module is used to calculate the response time of the optimal strategy.

5. According to claim 4, a water pump multi-operating condition performance test device based on digital twin is characterized in that: The optimal strategy selection module uses a determined neural network to find (P'-P) / P'≤0.01 according to P' required by the user, where P is the current output power of the water pump and P' is the expected output power of the water pump.

6. A water pump multi-operating condition performance test device based on digital twin according to claim 4 or 5, characterized in that: The expression of the strategy optimization method of the optimal strategy selection module is: Y= (valve 1 opening'-valve 1 opening) 2 + (valve 2 opening'-valve 2 opening) 2 , where valve 1 opening' is the valve target opening of the first solenoid valve, and valve 1 opening is the valve current opening of the first solenoid valve; valve 2 opening' is the valve target opening of the second solenoid valve, and valve 2 opening is the valve current opening of the second solenoid valve.

7. A water pump multi-operating condition performance test device based on digital twin according to claim 1, 2, 4 or 5, characterized in that: The control system module transmits the water pump operation data to the water pump virtual system, so that the boundary conditions during numerical simulation in the virtual system and the real-time status of the water pump remain highly consistent, and performs numerical simulation on the strategy scheme given by the water pump service system, and outputs the torque value, power value and pressure pulsation to the water pump service system. After the judgment module of the water pump service system makes a judgment, it passes the strategy to the water pump control system.

8. A method for testing the multi-operating performance of a water pump based on digital twins according to any one of claims 1 to 7, characterized in that: The following steps are involved: Step 1: Build a database. The water pump control system transmits the water pump multi-condition operation data to the water pump service system. The water pump service system receives the data of the water pump control system to build a database and apply machine learning methods to fit the neural network. Step 2: In the water pump control system, the control module opens the second solenoid valve, and the water in the water tank flows into the water pump through the water inlet pipe. When the water fills the entire pump cavity of the water pump, the control module controls the variable frequency motor, and the variable frequency motor starts the water pump. The control module opens the first solenoid valve, adjusts the valve opening of the first solenoid valve, and monitors the flow rate at the water pump outlet at the same time by the flow meter until the water pump outlet flow rate reaches a preset value. The control module controls the first solenoid valve and the second solenoid valve, and the torque speed sensor monitors the speed of the water pump so that the speed of the water pump reaches a preset speed value. The control module adjusts the valve opening of the first solenoid valve and the valve opening of the second solenoid valve so that the output power of the water pump reaches the current test power. Step 3: The real-time operating data in the water pump control system is transmitted to the water pump virtual system, and the water pump virtual system performs numerical simulation on the current working condition of the water pump; Step 4: pass the simulated user demand target to the water pump service system, use the database established in the first step, and find the control strategy through the neural network according to the simulated target working conditions. If the strategy exceeds the preset number of groups, stop looking for the strategy; Step 5: Select the optimal strategy: Among the preset number of strategies, the one with the smallest Y value is the optimal strategy, Y= (valve 1 opening'-valve 1 opening) 2 + (valve 2 opening'-valve 2 opening) 2 , where valve 1 opening' is the valve target opening of the first solenoid valve, and valve 1 opening is the valve current opening of the first solenoid valve; valve 2 opening' is the valve target opening of the second solenoid valve, and valve 2 opening is the valve current opening of the second solenoid valve; Step 6: The optimal strategy data value is passed to the water pump virtual system. In the water pump virtual system, the change value of the valve opening is split, and the splitting index is that the relative change rate θ does not exceed 5%. Op in the formula d Op represents the valve opening at the moment after the split, Op represents the valve opening at the moment before the split; Step 7: The split valve opening is transmitted to the ANSYS Workbench platform, and the previous working condition is used as the initial working condition to form a numerical simulation calculation of the multi-working condition transition process, and the calculation results are transmitted to the water pump service system; Step 8: In the water pump service system, the pressure pulsation value obtained by numerical simulation is evaluated to ensure that the amplitude of the pressure pulsation is not greater than the maximum pressure pulsation value that the unit can withstand. When the frequency of the pressure pulsation is inconsistent with the resonant frequency of the unit, the new valve opening is transmitted to the water pump control system; Step 9: Transmit the real-time data of the water pump control system to the water pump service system, update the database, and repeat steps 1 to 8 until the output power reaches the target value.

Images