Control method for reverse pumping of unmanned operation disk pump

Abstract

The present invention provides a control method for unmanned disk pump reverse pumping, which obtains real-time information of multiple data and equipment status, and controls the whole process of disk pump reverse pumping time, control variables, and controlled variables according to the obtained information, so that the whole process of unmanned operation can stabilize the requirements of various process control indicators and the pump body is not damaged; according to the abnormal variables that may be generated in the process, a corresponding logical sequence control program is creatively designed, and when a certain condition or multiple conditions are triggered, the corresponding program control strategy will automatically respond. The patent of the present invention adopts the above control method, which can effectively solve the problems of process instability, manual operation errors and pump damage in the reverse pumping process, and it is unmanned automatic control, and the process stability rate can reach a higher level.

Description

Technical Field

[0001] The invention relates to the field of disc pump reverse pumping, and in particular to a control method for an unmanned disc pump reverse pumping. Background Art

[0002] Centrifugal pumps are widely used for transporting liquid materials in chemical, electric power, food and other manufacturing industries, meeting the needs of most industrial applications. During use, most people manually turn the pumps regularly according to the material characteristics, and manually empty the pumps regularly according to system requirements. The process of turning the pumps and emptying the pumps has many manual operation points, unstable process control, high labor intensity, and a large number of machine pumps in the device area, which consumes a lot of manpower and has extremely low work efficiency.

[0003] In order to solve the above problems and utilize the conditions of remote start and stop of centrifugal pumps and control frequency, combined with the requirements of equipment, process, system, etc., a control method for unmanned disc pump reverse pumping was creatively invented to efficiently complete the disc pump reverse pumping task. Summary of the invention

[0004] The main purpose of the present invention is to provide a control method for unmanned disc pump reversal, so as to solve the problems of manual periodic disc pumping according to material characteristics and regular manual reversal according to system requirements, multiple manual operation points in the disc pump reversal process, unstable process control and high labor intensity.

[0005] In order to solve the above technical problems, the technical solution adopted by the present invention is: a control method for unmanned disc pump reverse pump, the disc pump control interface sets the running time, frequency, frequency increase rate and date and time, the DCS system collects the real-time status of the AB pump, screens and detects the machine seal water flow, remote / local status and electrical status of the standby pump; According to the real-time status of AB pumps collected by the system, it is confirmed that one AB pump is in use and the other is in standby. The disc pump control center controls the mutual standby interlocking cut-off and automatic recovery to perform the disc pump reverse operation.

[0006] In a preferred embodiment, the disc pump control method includes: S1. Set the pump operation time, frequency, frequency increase rate and date and time on the pump control interface; S2, DCS system collects the real-time status of AB pumps, screens and detects the machine seal water flow, remote / local status and electrical status of the standby pump; S3. Confirm that the standby pump is ready, cut off the mutual standby interlock, start the standby pump, and gradually increase its frequency to the preset value; After reaching the set running time, stop the standby pump; Automatic recovery of mutual standby interlocking.

[0007] In the preferred solution, when it is confirmed that the seal water flow of the standby pump is normal, the standby pump is in the remote state and there is no electrical fault, the pump rotation program will cut off the interlock switch between the two pumps and automatically issue a command to start the standby pump; Once the operation signal of the standby pump is obtained, the frequency of the standby pump will be increased incrementally according to the set requirements until the rotation frequency of the pump is reached; When the pump rotation timer reaches the set operation duration, a command to stop the standby pump is issued, ending the pump rotation operation; After that, the system will automatically reactivate the interlock.

[0008] In the preferred solution, the pump switching control method includes: A1. Set the pump switching operation time, frequency, frequency increase rate and date and time on the pump switching control interface; A2. The DCS system collects the real-time status of pumps A and B, screens and detects the seal water flow, remote / local status and electrical status of the standby pump; A3. Confirm that the standby pump is ready, automatically start the pump switching program at the specified date and time point, and remember the original main pump frequency value; A4. Gradually increase the frequency of the standby pump while decreasing the frequency of the original main pump to keep the total flow stable; Gradually increase the frequency of the standby pump while decreasing the frequency of the original main pump to keep the total flow stable.

[0009] Perform dynamic adjustment according to the set value (SP) and actual value (PV) of the flow automatic control valve: SP > PV + 1: Increase the frequency of the standby pump; SP < PV - 1: Decrease the frequency of the original main pump.

[0010] When the frequency of the original main pump drops to 0, stop the original main pump.

[0011] Automatically restore the interlock.

[0012] In the preferred solution, under the conditions that the seal water flow of the standby pump is normal, the standby pump is in the remote state and there is no electrical fault, the pump switching program will automatically start at the date and time set by the engineer; At this time, the program remembers the frequency value of the original main pump, starts to gradually increase the frequency of the standby pump, and at the same time the original main pump gradually decreases the frequency. The increase and decrease rates during this process are linked and controlled with the flow regulating valve of the pump output.

[0013] When the set value (SP) of the flow automatic control valve is greater than the actual value (PV) + 1, increase the frequency of the standby pump and keep the original operating pump stationary; When SP is less than PV - 1, decrease the frequency of the original main pump; The control is cyclically carried out in this way until the original main pump frequency drops to 0, and the standby pump frequency is increased to the memory value of the original main pump frequency. After confirming that the original main pump frequency is 0, the original main pump is stopped and the two pumps are re-engaged in mutual standby interlocking.

[0014] In the preferred solution, a predictive maintenance algorithm is introduced. Based on historical data and real-time monitoring data, a machine learning model is used to predict potential pump failures, schedule maintenance in advance, and reduce unplanned downtime. Specifically, it includes: Collect and store historical operation data of the pump, including key parameters such as shaft temperature, vibration, flow rate, and pressure; Use machine learning models to train historical data and establish a fault prediction model, in which a long short-term memory network model is selected as the machine learning model; Monitor the operating status of the pump in real time, input current data into the model, and predict possible future failures; Real-time monitoring and fault prediction stage: based on time window Collect pump operation data in real time, normalize it and input it into the trained LSTM model; the model outputs the fault prediction probability ,when It is determined that a fault may occur when the current time is , predicting future failure times The formula is: ,in is the time step, is the fault development rate coefficient; this formula estimates the time of fault occurrence based on the fault probability growth trend and fault development characteristics, which helps to arrange maintenance plans reasonably.

[0015] Based on the prediction results, maintenance reminders are issued in advance and appropriate maintenance measures are recommended to avoid sudden failures.

[0016] In the preferred solution, an optimization scheduling algorithm is introduced to automatically adjust the schedule of pumping and reversing by analyzing the operating status and load of the pump, thereby optimizing energy consumption and equipment life. Specifically, it includes: Analyze the operating load and energy consumption of the pump and identify high-load and low-load periods; Automatically adjust the schedule of disk pumping and reverse pumping according to production plan and equipment maintenance needs to avoid peak load periods and reduce energy consumption; Combine weather forecasts and electricity price fluctuations to further optimize scheduling plans and achieve energy conservation; Optimizing Scheduling Steps: Gaining the Future Weather forecast temperature ,humidity And electricity price fluctuation function ; Considering the influence of temperature and humidity on pump performance and energy consumption, the energy consumption calculation formula is modified: , , is the average temperature and humidity, , is the influence coefficient; construct a new objective function , and then use the simulated annealing algorithm to optimize the scheduling; modify the energy consumption formula to reflect the impact of meteorology, integrate the electricity price factor into the new objective function, and use the simulated annealing algorithm to jump out of the local optimum and obtain the optimal energy-saving scheduling plan; Regularly update model parameters based on newly collected data and actual operation feedback; Based on historical temperature, humidity and energy consumption data, the least squares method was used to recalibrate the influence coefficient. , : , maintain the accuracy of the model and the effectiveness of the scheduling plan; this formula optimizes parameters based on historical data, improves the accuracy of energy consumption calculations, and ensures that scheduling adapts to changes in the environment and equipment operating conditions.

[0017] In the preferred solution, a fault diagnosis expert system is introduced, which integrates the failure mode and effect analysis FMEA database, combines real-time data, quickly diagnoses and recommends solutions, and shortens the fault handling time. Specifically, it includes: Integrated FMEA database covering common failure types and their impacts; Collect pump operation data in real time and automatically diagnose fault causes in combination with FMEA database; Provide detailed fault reports and solution suggestions to help operators quickly handle problems and shorten downtime; Real-time data collection and fault diagnosis steps: frequency Collect pump operating parameters to form vector , where the shaft temperature , vibration amplitude ,flow ,pressure ; For each failure mode , set diagnostic rules based on historical data and expert experience: If ( ), a malfunction may occur , is the diagnostic function, is the threshold; the fault calculation is: ; is the weight, etc. are reference values; the diagnostic function integrates multi-parameter changes to evaluate faults, and the threshold determines the possibility of fault occurrence to achieve accurate diagnosis; Steps to generate fault reports and solutions: Diagnose faults After that, a report is generated containing the failure mode, cause ( ),Influence( )、 and solution suggestions; the operator handles the fault according to the report and the processing time , calculate downtime loss costs , To calculate the cost coefficient, evaluate the treatment effect, provide feedback to improve the diagnostic rules and solution library, reduce downtime losses, and improve system reliability; Regularly update the FMEA database, modify parameters based on new failure cases and technological developments; use Bayesian reasoning to update failure frequency : , is the number of old samples, For failure The number of new samples, is the total number of samples; updates ensure the timeliness and accuracy of the database and improve diagnostic reliability and foresight.

[0018] In the preferred solution, a multivariable collaborative control algorithm is introduced during the disc pumping and reverse pumping process, and the mutual influence between multiple controlled pressures, flows, and temperatures is considered at the same time to achieve more precise control and improve the stability and response speed of the system, specifically including: Simultaneously monitor multiple controlled pressures, flows, and temperatures, and analyze the relationships between them; Establish a multivariable control model to ensure the coordination and consistency between variables and avoid the impact of fluctuations in a single variable on other variables; Through advanced PID control strategy, more precise control can be achieved and the stability and response speed of the system can be improved.

[0019] In the preferred solution, the variable relationship analysis step is: based on the sampling period Synchronous pressure acquisition ,flow ,temperature Data, duration , get the data set , ; Calculate the correlation coefficient matrix between variables ,element ( ), is the mean value; Determine the strength of variable association and lag relationship and construct a directed graph Describe the causal flow, is a node set, The mutual correlation matrix quantifies the linear association of variables, and the directed graph intuitively presents the causal structure, which helps to understand the dynamic characteristics of the system and lays the foundation for the design of the control model. Steps for establishing a multivariable control model: Based on the variable relationship, select the state space method to build the model; set the state vector According to the laws of physics and experience, the equation of state is , output equation ; , , , is the coefficient matrix, is the control input, , The noise is estimated by using system identification technology data to make the model fit the data. This model describes the dynamics of variables from the global perspective of the system, provides a theoretical framework for collaborative control, and improves control accuracy and stability. PID control strategy implementation steps: Set up a multivariable PID controller, control law , is the error vector, is the reference value, , , is the PID gain matrix; Use particle swarm optimization algorithm PSO to adjust parameters and fitness function , , The positive definite matrix weighs the control error and energy consumption; the particle position in PSO is the gain value, and the optimization is sought based on the speed update to achieve the optimal parameter combination; multivariable PID and PSO work together to take into account multivariable correlation and control performance, enhance system response and stability, and accurately track the set value; Real-time monitoring of system operation, online updating of model parameters using recursive least squares method based on new data; assuming new data , update the estimate , is the parameter estimation vector, , ; Online updates ensure that the model tracks system changes, optimizes control performance, and adapts to operating condition fluctuations.

[0020] The present invention provides a control method for unmanned disc pump reverse pumping, with high automatic control rate and unmanned operation: highly automated control of the device is achieved, and there is no need for continuous manual intervention in the disc pump reverse pumping process, which greatly reduces manpower input and improves production efficiency.

[0021] Precise control and process stability: Through precise control strategies, such as dynamically adjusting the pump frequency according to the set value and actual value of the flow automatic control valve, the process control indicators can be effectively stabilized to ensure the stability of the production process and the consistency of product quality.

[0022] Self-diagnosis and safety assurance: The system can obtain various status information of the pump such as shaft temperature, vibration, and electrical status in real time, and respond quickly when a fault occurs. For example, when the main pump or standby pump fails, it will switch and lock the fault output in time, and at the same time, it will sound and light alarms, provide complete self-diagnosis information, ensure the safety of the control process, effectively prevent the fault from further expanding, extend the service life of the machine pump, and reduce the equipment maintenance cost. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] The present invention will be further described below in conjunction with the accompanying drawings and embodiments: Figure 1 A process flow chart of a production device provided by an embodiment of the present invention; Figure 2 A flow chart of an unmanned disc pump control method provided by an embodiment of the present invention; Figure 3 A flow chart of an unmanned reverse pump control method provided in an embodiment of the present invention.

[0024] In the figure: liquid material tank 1; pump A 2; pump B 3; sealing water flow sensor for pump A 4; sealing water flow sensor for pump B 5; remote pressure gauge at pump outlet 6; flow meter 7; regulating valve 8; remote detection sensor for pump A 9; local detection sensor for pump A 10; remote detection signal sensor for pump B 11; local detection sensor for pump B 12; shaft temperature sensor for pump A 13; shaft vibration sensor for pump A 14; shaft temperature sensor for pump B 15; shaft vibration sensor for pump B 16; mutual standby interlock 17. DETAILED DESCRIPTION

[0025] Example 1 like Figures 1 to 3 As shown, a control method for unmanned disc pump reverse pumping is provided. The disc pump control interface sets the operation time, frequency, frequency increase rate and date and time. The DCS system collects the real-time status of the AB pumps, screens and detects the machine seal water flow, remote / local status and electrical status of the standby pump. According to the real-time status of AB pumps collected by the system, it is confirmed that one AB pump is in use and the other is in standby. The disc pump control center controls the mutual standby interlock 17 to be cut off and automatically restored to perform the disc pump reverse operation.

[0026] Acquire real-time information on multiple data and equipment status, and control the entire process of the pump reversal time, control variables, and controlled variables based on the acquired information, so that the unmanned entire process can stabilize the requirements of various process control indicators and the pump body will not be damaged; according to the abnormal variables that may occur in the process, creatively design a corresponding logical sequence control program, and when one or more conditions are triggered, the corresponding program control strategy will automatically respond. The patent of this invention adopts the above control method, which can effectively solve the routine operation, emergency operation and highly automated control and processing in different scenes of the production site, while ensuring the stability of the process and solving the difficulties in equipment management.

[0027] In a preferred embodiment, the disc pump control method includes: S1. Set the pump operation time, frequency, frequency increase rate and date and time on the pump control interface; S2, DCS system collects the real-time status of AB pumps, screens and detects the machine seal water flow, remote / local status and electrical status of the standby pump; S3, confirm that the standby pump is ready, cut off the mutual standby interlock 17, start the standby pump, and gradually increase its frequency to the preset value; After reaching the set running time, stop the standby pump; Automatic recovery of mutual standby interlock 17.

[0028] In the preferred solution, when it is confirmed that the machine seal water flow of the standby pump is normal, the standby pump is in remote state and there is no electrical fault, the disk pump program will cut off the two pumps mutual standby interlock 17 switch and automatically issue a command to start the standby pump; Once the standby pump operation signal is obtained, the standby pump frequency will be increased incrementally according to the set requirements until it reaches the disk pump operation frequency; When the disc pump timer reaches the set running time, a command to stop the standby pump is issued to end the disc pump operation; After that, the system will automatically re-enter the mutual standby interlock 17.

[0029] In the preferred embodiment, the reverse pump control method includes: A1. The reverse pump control interface sets the reverse pump operation time, frequency, frequency increase rate and date and time; A2. The DCS system collects the real-time status of the AB pumps, screens and detects the machine seal water flow, remote / local status and electrical status of the standby pumps; A3. Confirm that the standby pump is ready, automatically start the reverse pump program at the specified date and time, and remember the original main pump frequency value; A4. Gradually increase the frequency of the standby pump and reduce the frequency of the original main pump to keep the total flow stable; Gradually increase the frequency of the standby pump while reducing the frequency of the original main pump to keep the total flow stable.

[0030] Dynamically adjust according to the set value (SP) and actual value (PV) of the flow control regulating valve: SP > PV + 1: Increase the frequency of the standby pump; SP < PV - 1: Decrease the frequency of the original main pump.

[0031] When the frequency of the original main pump drops to 0, stop the original main pump.

[0032] Automatically resume the mutual backup interlock 17.

[0033] In the preferred solution, under the conditions of normal seal water flow of the standby pump, remote status of the standby pump, and no electrical faults, the pump switching program will automatically start at the date and time set by the engineer; At this time, the program remembers the frequency value of the original main pump, starts to gradually increase the frequency of the standby pump, and at the same time the original main pump gradually decreases the frequency. The increase and decrease rates during this process are linked and controlled with the flow control valve of the pump output.

[0034] When the set value (SP) of the flow control regulating valve is greater than the actual value (PV) + 1, increase the frequency of the standby pump, and the originally operating pump remains stationary; When SP is less than PV - 1, decrease the frequency of the original main pump; Control in this way in a loop until the frequency of the original main pump drops to 0, the frequency of the standby pump increases to the remembered value of the original main pump frequency. After confirming that the frequency of the original main pump is 0, stop the original main pump, and re - engage the mutual backup interlock (17) of the two pumps.

[0035] I. Overview of the overall control logic Control objectives: Achieve unmanned operation of pump switching, stabilize process control indicators, ensure no damage to the pump body, improve the automatic control rate of the device, ensure process stability, achieve complete self - diagnosis information and safe control process, and extend the service life of the pump and motor.

[0036] Basic control conditions Hardware equipment and connections: Involve equipment such as liquid material tank 1, Pump A 2, Pump B 3, etc. Data is collected through the seal water flow sensor 4 of Pump A, the seal water flow sensor 5 of Pump B, the remote pressure gauge 6 at the pump outlet, the flowmeter 7, the regulating valve 8, the remote detection sensor 9 of Pump A, the local detection sensor 10 of Pump A, the remote detection signal sensor 11 of Pump B, the local detection sensor 12 of Pump B, the shaft temperature sensor 13 of Pump A, the shaft vibration sensor 14 of Pump A, the shaft temperature sensor 15 of Pump B, and the shaft vibration sensor 16 of Pump B. Each device works in coordination, and the mutual backup interlock 17 ensures the safety of pump switching and operation.

[0037] Data acquisition and transmission: The above - mentioned sensors collect data in real - time and transmit it to the distributed control system DCS to provide a basis for subsequent control decisions.

[0038] 2. Operation steps and principles of disc pump Parameter setting (S1) Operation: Set the pump running time, frequency, frequency increase rate and date and time on the pump control soft panel.

[0039] Principle: These parameters provide basic operating instructions for the disc pumping process, determining the duration of the disc pumping, operating speed changes and start timing.

[0040] Data collection and status detection (S2) Operation: The DCS system collects the real-time online data (shaft temperature, vibration, flow, pressure, etc.) and status (remote / local, electrical, etc.) of the AB pump, selects the spare pump and detects its machine seal water flow, remote / local status, and electrical status.

[0041] Principle: Fully grasp the real-time status of the pump to ensure that the standby pump is ready for starting. Normal sealing water can prevent leakage and other problems. Remote status and electrical fault-free status ensure remote control and normal operation of the motor.

[0042] Startup and frequency increase (S3) Operation: When the standby pump has normal sealing water flow, normal remote status and no electrical faults, cut off the interlock switch between the two pumps, start the standby pump, and increase the frequency to the disc pump operating frequency at the set rate.

[0043] Principle: Mutual interlocking is cut off to avoid false operation during switching. Increasing the frequency at the set rate can start the standby pump smoothly to prevent impact.

[0044] Operation monitoring and main pump fault handling Operation: During the pumping process, the main pump and the standby pump are controlled to back up each other. When the main pump fails and stops, the standby pump is quickly converted to the main pump, copying the original main pump operation frequency. The regulating valve coordinates to control the discharge flow rate within a certain range, and locks the main pump fault output at the same time.

[0045] Principle: The mutual backup function ensures continuous operation, the standby pump quickly takes over to ensure production continuity, the regulating valve maintains a stable flow to avoid process fluctuations, and the fault lock is easy to troubleshoot.

[0046] Disc pump end operation Operation: When the pump timer reaches the set time, the standby pump stops and the mutual standby interlock is automatically activated after the pump reaches the stop state.

[0047] Principle: End the disk pump operation as planned, restore the mutual standby interlock to prepare for the next switch, and ensure that the system is always in a safe state for switchability.

[0048] 3. Steps and principles of reverse pump operation Parameter setting (A1): Set the pump change - over operation time, frequency, frequency increase rate, and date and time on the soft panel for pump change - over control. It is similar to the pump inspection setting but applies to the pump change - over process.

[0049] Data acquisition and status detection (A2): Similar to step S2 of pump inspection, prepare data for pump change - over operation to ensure that the standby pump is ready.

[0050] Start - up and frequency adjustment (A3 - A4) Operation: When the standby pump conditions are met and the set date and time point is reached, remove the mutual standby interlock, start the standby pump and remember the original main pump frequency value. Then, dynamically adjust the frequencies of the main and standby pumps according to the set value (SP) and actual value (PV) of the flow control valve. If SP > PV + 1, increase the frequency of the standby pump; if SP < PV - 1, decrease the frequency of the original main pump. Cycle the control until the frequency of the original main pump drops to 0 and the standby pump increases to the remembered value of the original main pump frequency.

[0051] Principle: Precisely control the frequency switching between the main and standby pumps, and adjust according to the flow deviation to ensure the stability of the total flow and meet the production requirements.

[0052] Fault handling during pump change - over Operation: During pump change - over, the original main pump and the standby pump are in mutual standby control. If the original main pump fails and trips, the standby pump quickly increases its frequency to the original main pump frequency; if the standby pump fails and trips, the original main pump automatically resumes to its original frequency. Obtain the status of the main pump and the standby pump during the process. In case of a fault - stopped pump, give an audible and visual alarm and lock the fault information.

[0053] Principle: Ensure the stable operation of the system during pump change - over, respond to faults in a timely manner and provide accurate information for maintenance.

[0054] Operations after pump change - over Operation: Stop the original main pump when its frequency is 0 and engage the mutual standby interlock of the two pumps.

[0055] Principle: Complete the pump change - over in an orderly manner, restore the mutual standby interlock, and make the system return to the normal standby state.

[0056] IV. Association with Embodiments 2 and 3 and Overall Significance Association Shared data basis: The data acquisition part in Embodiment 1 provides a real - time data source for Embodiments 2 and 3. For example, the predictive maintenance algorithm and fault diagnosis expert system in Embodiment 3 rely on these data for analysis.

[0057] Control process coordination: The disc pump reverse pump operation process of Example 1 is the core of the overall system operation. Example 2 further describes its application in the actual production process. Example 3 adds an intelligent algorithm based on this to optimize the entire process. For example, the optimization scheduling algorithm can adjust the disc pump reverse pump schedule, which is associated with parameters such as the operation time of Example 1.

[0058] Overall significance: Example 1 constructs the basic framework and operation process of unmanned disc pump reverse pumping, Example 2 places it in an actual production scenario to illustrate its feasibility and practicality, and Example 3 improves system performance from an intelligent perspective. The combination of the three realizes efficient, stable and intelligent disc pump reverse pumping control, solves the pain points of manual operation, and improves production efficiency and equipment management level.

[0059] Example 2 Further illustrate with reference to Example 1, Figure 1-3 The structure shown, see Figure 1 Described: Taking the A pump 2 process operation line as description, when the detection signal of the A pump machine sealing water flow sensor 4 is normal, the A pump inlet manual valve is fully open, the A pump outlet manual valve is fully open, and the A pump remote detection sensor 9 detects a normal signal, which are all within the allowable conditions, the material coming out of the liquid material tank 11 enters the A pump 2, and the A pump 2 starts to operate, and the material is transported to the next production unit after being controlled by the flow meter 7 and the regulating valve 8. During the production operation of the A pump 2, the regulating valve 8 and the flow meter 7 form a PID control loop to automatically control the material conveying amount. The pump outlet remote pressure gauge 6, the A pump shaft temperature sensor 13 and the A pump shaft vibration sensor 14 simultaneously monitor the A pump shaft temperature and the A pump shaft vibration in real time online. The three items of the pump outlet remote pressure are all within the normal control range, and the pump mutual standby interlock 17 switching switch is in the input state 17. This is the operation status of the conveying process.

[0060] Disc pump: Step 1: Set the pump running time, pump running frequency, pump frequency increase rate, pump date and time on the pump control panel.

[0061] Step 2: The distributed control system (DCS) obtains the pumping time, pumping frequency, and pumping frequency increase rate set by the engineer on the pump panel. It obtains the real-time online data and status of the AB pumps, selects the standby pumps according to the pump operation status, and detects the standby pump sealing water flow, standby pump remote / local status, and electrical status.

[0062] Step 3. Under the conditions that the sealing water flow of the standby pump is normal, the standby pump is in remote status, and there is no electrical fault, the pump program cuts off the mutual standby interlock switch of the two pumps and automatically issues a command to start the standby pump. After obtaining the standby pump running signal, the standby pump frequency is increased incrementally according to the pump frequency increase rate setting requirements. When the frequency reaches the pump running frequency and the pump timer reaches the set running time, a command to stop the standby pump is issued, the standby pump pumping ends, the standby pump stop status is obtained after the standby pump is pumped, and the mutual standby interlock is automatically activated.

[0063] In the steps described, it should be noted that in the process of adjusting the standby pump, the control of the main pump and the standby pump is still mutual backup. When the main pump stops due to a fault, the standby pump will immediately switch to the main pump, and the standby pump quickly copies the operating frequency of the original main pump. Under the coordination of the regulating valve, the discharge flow rate is controlled within a certain range. At this time, the main pump fault signal is obtained, the fault output of the main pump is locked, and the cause of the fault is indicated. In the process of adjusting the standby pump, the vibration, shaft temperature, electrical and operating status of the standby pump shaft are obtained. When the standby pump fails, the pump is immediately interlocked and stopped, and the fault output of the standby pump is locked to indicate the cause of the fault.

[0064] In this embodiment, comprehensive control can be performed based on the operating status of the main equipment directly associated with the pump. Since routine or non-routine inspection and maintenance operations may occur during device production, when the stop status of the main equipment associated with the pump is obtained, the pump program is automatically blocked and the pump operation task is stopped. When the operating status of the main equipment is obtained, the pump program is automatically reset and unlocked, and comprehensive control is performed according to the parameters set by the engineer.

[0065] Back Pump: Step 1: Set the reverse pump running time, reverse pump running frequency, pump frequency increase rate, reverse pump date and time on the reverse pump control soft panel.

[0066] Step 2: The DCS (Distributed Control System) obtains the pump reversal date and time set by the engineer on the pump reversal panel. It obtains the real-time online data and operating status of the AB pumps, selects the standby pumps based on the operating status of the pumps, and detects the standby pump sealing water flow, remote / local status, and electrical status.

[0067] Step 3, obtain the date and time set by the engineer to automatically start the program. Under the conditions that the sealing water flow of the standby pump is normal, the standby pump is in remote state, and there is no electrical fault, the reverse pump program cuts off the mutual standby interlock switch of the two pumps, and then starts the standby pump. At this time, the program remembers the frequency value of the original main pump, and the standby pump begins to gradually increase the frequency, and the original main pump gradually reduces the frequency. The rate of increase and decrease is linked with the flow control valve output by the pump. When the set value (SP) of the flow automatic control valve is greater than the actual value (PV) + 1, the frequency of the standby pump is increased, and the original running pump remains unchanged. When the SP of the flow automatic control valve is less than the PV value -1, the frequency of the original main pump is reduced. The original main pump frequency is reduced to 0 in turn through the reincarnation control, and the standby pump is increased to the memory value of the original main pump frequency. If the frequency of the original main pump is determined to be 0, the original main pump is stopped, and the two pumps are put into mutual standby interlock.

[0068] In the steps described, it should be noted that, during the pump reversal process, the control of the original main pump and the standby pump is still mutual backup. When the original main pump stops due to a fault, the standby pump will immediately be converted to the main pump, and the pump frequency will be increased to the original main pump frequency value; when the standby pump stops due to a fault during the pump reversal process, the original main pump automatically returns to the original frequency; during the pump reversal process, the shaft vibration, shaft temperature, electrical and operating status of the main pump and the standby pump are obtained, and if a fault occurs and the pump stops, an audible and visual alarm will be issued, and the fault information will be locked.

[0069] In this embodiment, comprehensive control can be performed based on the operating status of the main equipment directly associated with the pump. Since routine or non-routine inspection and maintenance operations may occur during device production, when the stop status of the main equipment associated with the pump is obtained, the pump reverse program is automatically blocked and the pump reverse operation task is stopped. When the operating status of the main equipment is obtained, the pump reverse program is automatically reset and unlocked, and comprehensive control is performed according to the parameters set by the engineer.

[0070] Compared with the current technical methods, the unmanned disc pump reverse pump control method of the present invention has the following beneficial effects: (1) the device has a high automatic control rate and is unmanned; (2) the control method is accurate and the process stability is high; (3) the self-diagnosis information is complete, the control process is safe, and the service life of the machine pump is extended; Example 3 Further illustrate with reference to Example 1, Figure 1-3 The structure shown introduces a predictive maintenance algorithm. Based on historical data and real-time monitoring data, a machine learning model is used to predict potential pump failures, schedule maintenance in advance, and reduce unplanned downtime. Specifically, it includes: Collect and store historical operation data of the pump, including key parameters such as shaft temperature, vibration, flow rate, and pressure; Use machine learning models to train historical data and establish a fault prediction model, in which a long short-term memory network model is selected as the machine learning model; Monitor the operating status of the pump in real time, input current data into the model, and predict possible future failures; Real-time monitoring and fault prediction stage: based on time window Collect pump operation data in real time, normalize it and input it into the trained LSTM model; the model outputs the fault prediction probability ,when It is determined that a fault may occur when the current time is , predicting future failure times The formula is: ,in is the time step, is the fault development rate coefficient; this formula estimates the time of fault occurrence based on the fault probability growth trend and fault development characteristics, which helps to arrange maintenance plans reasonably.

[0071] Based on the prediction results, maintenance reminders are issued in advance and appropriate maintenance measures are recommended to avoid sudden failures.

[0072] In the preferred solution, an optimization scheduling algorithm is introduced to automatically adjust the schedule of pumping and reversing by analyzing the operating status and load of the pump, thereby optimizing energy consumption and equipment life. Specifically, it includes: Analyze the operating load and energy consumption of the pump and identify high-load and low-load periods; Automatically adjust the schedule of disk pumping and reverse pumping according to production plan and equipment maintenance needs to avoid peak load periods and reduce energy consumption; Combine weather forecasts and electricity price fluctuations to further optimize scheduling plans and achieve energy conservation; Optimizing Scheduling Steps: Gaining the Future Weather forecast temperature ,humidity And electricity price fluctuation function ; Considering the influence of temperature and humidity on pump performance and energy consumption, the energy consumption calculation formula is modified: , , is the average temperature and humidity, , is the influence coefficient; construct a new objective function , and then use the simulated annealing algorithm to optimize the scheduling; modify the energy consumption formula to reflect the impact of meteorology, integrate the electricity price factor into the new objective function, and use the simulated annealing algorithm to jump out of the local optimum and obtain the optimal energy-saving scheduling plan; Regularly update model parameters based on newly collected data and actual operation feedback; Based on historical temperature, humidity and energy consumption data, the least squares method was used to recalibrate the influence coefficient. , : , maintain the accuracy of the model and the effectiveness of the scheduling plan; this formula optimizes parameters based on historical data, improves the accuracy of energy consumption calculations, and ensures that scheduling adapts to changes in the environment and equipment operating conditions.

[0073] In the preferred solution, a fault diagnosis expert system is introduced, which integrates the failure mode and effect analysis FMEA database, combines real-time data, quickly diagnoses and recommends solutions, and shortens the fault handling time. Specifically, it includes: Integrated FMEA database covering common failure types and their impacts; Collect pump operation data in real time and automatically diagnose fault causes in combination with FMEA database; Provide detailed fault reports and solution suggestions to help operators quickly handle problems and shorten downtime; Real-time data collection and fault diagnosis steps: frequency Collect pump operating parameters to form vector , where the shaft temperature , vibration amplitude ,flow ,pressure ; For each failure mode , set diagnostic rules based on historical data and expert experience: If ( ), a malfunction may occur , is the diagnostic function, is the threshold; the fault calculation is: ; is the weight, etc. are reference values; the diagnostic function integrates multi-parameter changes to evaluate faults, and the threshold determines the possibility of fault occurrence to achieve accurate diagnosis; Steps to generate fault reports and solutions: Diagnose faults After that, a report is generated containing the failure mode, cause ( ),Influence( )、 and solution suggestions; the operator handles the fault according to the report and the processing time , calculate downtime loss cost , To calculate the cost coefficient, evaluate the treatment effect, provide feedback to improve the diagnostic rules and solution library, reduce downtime losses, and improve system reliability; Regularly update the FMEA database, modify parameters based on new failure cases and technological developments; use Bayesian reasoning to update failure frequency : , is the number of old samples, For failure The number of new samples, is the total number of samples; updates ensure the timeliness and accuracy of the database and improve diagnostic reliability and foresight.

[0074] In the preferred solution, a multivariable collaborative control algorithm is introduced during the disc pumping and reverse pumping process, and the mutual influence between multiple controlled pressures, flows, and temperatures is considered at the same time to achieve more precise control and improve the stability and response speed of the system, specifically including: Simultaneously monitor multiple controlled pressures, flows, and temperatures, and analyze the relationships between them; Establish a multivariable control model to ensure the coordination and consistency between variables and avoid the impact of fluctuations in a single variable on other variables; Through advanced PID control strategy, more precise control can be achieved and the stability and response speed of the system can be improved.

[0075] In the preferred solution, the variable relationship analysis step is: based on the sampling period Synchronous pressure acquisition ,flow ,temperature Data, duration , get the data set , ; Calculate the correlation coefficient matrix between variables ,element ( ), is the mean; Determine the strength of variable association and lag relationship and construct a directed graph Describe the causal flow, is a node set, The mutual correlation matrix quantifies the linear association of variables, and the directed graph intuitively presents the causal structure, which helps to understand the dynamic characteristics of the system and lays the foundation for the design of the control model. Steps for establishing a multivariable control model: Based on the variable relationship, select the state space method to build the model; set the state vector According to the laws of physics and experience, the equation of state is , output equation ; , , , is the coefficient matrix, is the control input, , The noise is estimated by using system identification technology data to make the model fit the data. This model describes the dynamics of variables from the global perspective of the system, provides a theoretical framework for collaborative control, and improves control accuracy and stability. PID control strategy implementation steps: Set up a multivariable PID controller, control law , is the error vector, is the reference value, , , is the PID gain matrix; Use particle swarm optimization algorithm PSO to adjust parameters and fitness function , , The positive definite matrix weighs the control error and energy consumption; the particle position in PSO is the gain value, and the optimization is sought based on the speed update to achieve the optimal parameter combination; multivariable PID and PSO work together to take into account multivariable correlation and control performance, enhance system response and stability, and accurately track the set value; Real-time monitoring of system operation, online updating of model parameters using recursive least squares method based on new data; assuming new data , update the estimate , is the parameter estimation vector, , ; Online updates ensure that the model tracks system changes, optimizes control performance, and adapts to operating condition fluctuations.

[0076] The above embodiments are only preferred technical solutions of the present invention and should not be regarded as limiting the present invention. The protection scope of the present invention shall be the technical solutions recorded in the claims, including equivalent replacement solutions of the technical features in the technical solutions recorded in the claims. That is, equivalent replacement improvements within this scope are also within the protection scope of the present invention.

Claims

1. A control method for an unmanned disc pump reverse pump, characterized in that: The pump control interface sets the operating time, frequency, frequency increase rate and date and time. The DCS system collects the real-time status of the AB pumps, screens and detects the machine seal water flow, remote / local status and electrical status of the standby pumps; According to the real-time status of AB pumps collected by the system, it is confirmed that one AB pump is in use and the other is in standby mode. The disc pump control center controls the mutual standby interlock (17) to be cut off and automatically restored to perform the disc pump reverse pump operation.

2. According to claim 1, a method for controlling an unmanned disc pump reverse pump is characterized by: Disc pump control methods include: S1. Set the pump operation time, frequency, frequency increase rate and date and time on the pump control interface; S2, DCS system collects the real-time status of AB pumps, screens and detects the machine seal water flow, remote / local status and electrical status of the standby pump; S3. Confirm that the standby pump is ready, remove the mutual standby interlock (17), start the standby pump, and gradually increase its frequency to the preset value; After reaching the set running time, stop the standby pump; Automatic recovery of mutual standby interlock (17).

3. According to claim 2, a method for controlling an unmanned disk pump reverse pump is characterized by: When it is confirmed that the machine seal water flow of the standby pump is normal, the standby pump is in remote state and there is no electrical fault, the pump program will cut off the mutual standby interlock (17) switch of the two pumps and automatically issue a command to start the standby pump; Once the standby pump operation signal is obtained, the standby pump frequency will be increased incrementally according to the set requirements until it reaches the disk pump operation frequency; When the disc pump timer reaches the set running time, a command to stop the standby pump is issued to end the disc pump operation; Afterwards, the system will automatically re-enter the mutual standby interlock (17).

4. According to claim 1, a method for controlling an unmanned disc pump reverse pump is characterized by: Back pump control methods include: A1. The reverse pump control interface sets the reverse pump operation time, frequency, frequency increase rate and date and time; A2. The DCS system collects the real-time status of the AB pumps, screens and detects the machine seal water flow, remote / local status and electrical status of the standby pumps; A3. Confirm that the standby pump is ready, automatically start the reverse pump program at the specified date and time, and remember the original main pump frequency value; A4. Gradually increase the frequency of the standby pump and reduce the frequency of the original main pump to keep the total flow stable; Gradually increase the frequency of the standby pump while reducing the frequency of the original main pump to keep the total flow stable. Dynamic adjustment is performed based on the set value (SP) and actual value (PV) of the flow automatic control valve: SP > PV + 1: Increase the standby pump frequency; SP < PV - 1: Reduce the original main pump frequency. When the original main pump frequency drops to 0, stop the original main pump. Automatic recovery of mutual standby interlock (17).

5. The control method of unmanned operation of the disc pump according to claim 4 is characterized in that: If the standby pump sealing water flow is normal, the standby pump is in remote status, and there is no electrical fault, the pump reversal program will automatically start at the date and time set by the engineer; At this time, the program remembers the frequency value of the original main pump and begins to gradually increase the frequency of the standby pump. At the same time, the original main pump gradually reduces the frequency. The increase and decrease rates in this process are linked to the flow control valve of the pump output. When the set value (SP) of the flow automatic control valve is greater than the actual value (PV) + 1, the frequency of the standby pump is increased and the original running pump remains unchanged; When SP is less than PV-1, the original main pump frequency is reduced; The control is cyclically carried out in this way until the original main pump frequency drops to 0, and the standby pump frequency is increased to the memory value of the original main pump frequency. After confirming that the original main pump frequency is 0, the original main pump is stopped and the mutual standby interlocking of the two pumps is re-activated (17).

6. According to the unmanned operation disc pump reverse pump control method of claim 1, characterized in that: The predictive maintenance algorithm uses machine learning models based on historical data and real-time monitoring data to predict potential pump failures, schedule maintenance in advance, and reduce unplanned downtime. Specifically, it includes: Collect and store historical operation data of the pump, including key parameters such as shaft temperature, vibration, flow rate, and pressure; Use machine learning models to train historical data and establish a fault prediction model, in which a long short-term memory network model is selected as the machine learning model; Monitor the operating status of the pump in real time, input current data into the model, and predict possible future failures; Real-time monitoring and fault prediction stage: based on time window Collect pump operation data in real time, normalize it and input it into the trained LSTM model; the model outputs the fault prediction probability ,when It is determined that a fault may occur when the current time is , predicting future failure times The formula is: ,in is the time step, is the fault development rate coefficient; this formula estimates the time of fault occurrence based on the fault probability growth trend and fault development characteristics, which helps to arrange maintenance plans reasonably. Based on the prediction results, maintenance reminders are issued in advance and appropriate maintenance measures are recommended to avoid sudden failures.

7. According to claim 1, a method for controlling the reverse pump of an unmanned disc pump is characterized by: introducing The optimized scheduling algorithm automatically adjusts the schedule of pumping and reversing by analyzing the operating status and load of the pump to optimize energy consumption and equipment life, including: Analyze the operating load and energy consumption of the pump and identify high-load and low-load periods; Automatically adjust the schedule of disk pumping and reverse pumping according to production plan and equipment maintenance needs to avoid peak load periods and reduce energy consumption; Combine weather forecasts and electricity price fluctuations to further optimize scheduling plans and achieve energy conservation; Optimizing Scheduling Steps: Gaining the Future Weather forecast temperature ,humidity And electricity price fluctuation function ; Considering the influence of temperature and humidity on pump performance and energy consumption, the energy consumption calculation formula is modified: , , is the average temperature and humidity, , is the influence coefficient; construct a new objective function , and then use the simulated annealing algorithm to optimize the scheduling; modify the energy consumption formula to reflect the impact of meteorology, integrate the electricity price factor into the new objective function, and use the simulated annealing algorithm to jump out of the local optimum and obtain the optimal energy-saving scheduling plan; Regularly update model parameters based on newly collected data and actual operation feedback; Based on historical temperature, humidity and energy consumption data, the least squares method was used to recalibrate the influence coefficient. , : , maintain the accuracy of the model and the effectiveness of the scheduling plan; this formula optimizes parameters based on historical data, improves the accuracy of energy consumption calculations, and ensures that scheduling adapts to changes in the environment and equipment operating conditions.

8. According to claim 1, a method for controlling an unmanned disc pump reverse pump is characterized by: introducing The fault diagnosis expert system integrates the failure mode and effect analysis FMEA database, combines real-time data, quickly diagnoses and recommends solutions, and shortens the fault handling time. Specifically, it includes: Integrated FMEA database covering common failure types and their impacts; Collect pump operation data in real time and automatically diagnose fault causes in combination with FMEA database; Provide detailed fault reports and solution suggestions to help operators quickly handle problems and shorten downtime; Real-time data collection and fault diagnosis steps: frequency Collect pump operating parameters to form vector , where the shaft temperature , vibration amplitude ,flow ,pressure ; For each failure mode , set diagnostic rules based on historical data and expert experience: If ( ), a malfunction may occur , is the diagnostic function, is the threshold; the fault calculation is: ； is the weight, etc. are reference values; the diagnostic function integrates multi-parameter changes to evaluate faults, and the threshold determines the possibility of fault occurrence to achieve accurate diagnosis; Steps to generate fault reports and solutions: Diagnose faults After that, a report is generated containing the failure mode, cause ( ),Influence( )、 and solution suggestions; the operator handles the fault according to the report and the processing time , calculate downtime loss costs , To calculate the cost coefficient, evaluate the treatment effect, provide feedback to improve the diagnostic rules and solution library, reduce downtime losses, and improve system reliability; Regularly update the FMEA database, modify parameters based on new failure cases and technological developments; use Bayesian reasoning to update failure frequency : , is the number of old samples, For failure The number of new samples, is the total number of samples; updates ensure the timeliness and accuracy of the database and improve diagnostic reliability and foresight.

9. The control method for unmanned disc pump reverse pump according to claim 1, characterized in that: In the process of disc pumping and reverse pumping, a multivariable collaborative control algorithm is introduced to simultaneously consider the mutual influence between multiple controlled pressures, flows, and temperatures to achieve more precise control and improve the stability and response speed of the system. Specifically, it includes: Simultaneously monitor multiple controlled pressures, flows, and temperatures, and analyze the relationships between them; Establish a multivariable control model to ensure the coordination and consistency between variables and avoid the impact of fluctuations in a single variable on other variables; Through advanced PID control strategy, more precise control can be achieved and the stability and response speed of the system can be improved.

10. The control method of unmanned disc pump reverse pump according to claim 9, characterized in that: the variable Relationship analysis steps: Sampling period Synchronous pressure acquisition ,flow ,temperature Data, duration , get the data set , ; Calculate the correlation coefficient matrix between variables ,element ( ), is the mean; Determine the strength of variable association and lag relationship and construct a directed graph Describe the causal flow, is a node set, The mutual correlation matrix quantifies the linear association of variables, and the directed graph intuitively presents the causal structure, which helps to understand the dynamic characteristics of the system and lays the foundation for the design of the control model. Steps for establishing a multivariable control model: Based on the variable relationship, select the state space method to build the model; set the state vector According to the laws of physics and experience, the equation of state is , output equation ; , , , is the coefficient matrix, is the control input, , The noise is estimated by using system identification technology data to make the model fit the data. This model describes the dynamics of variables from the global perspective of the system, provides a theoretical framework for collaborative control, and improves control accuracy and stability. PID control strategy implementation steps: Set up a multivariable PID controller, control law , is the error vector, is the reference value, , , is the PID gain matrix; Use particle swarm optimization algorithm (PSO) to adjust parameters and fitness function , , The positive definite matrix weighs the control error and energy consumption; the particle position in PSO is the gain value, and the optimization is sought based on the speed update to achieve the optimal parameter combination; multivariable PID and PSO work together to take into account multivariable correlation and control performance, enhance system response and stability, and accurately track the set value; Real-time monitoring of system operation, online updating of model parameters using recursive least squares method based on new data; assuming new data , update the estimate , is the parameter estimation vector, , ; Online updates ensure that the model tracks system changes, optimizes control performance, and adapts to operating condition fluctuations.

Images