A lubricating oil flow detection method and system for a filling production line

Abstract

The present application relates to the field of flow detection, more particularly, the present application relates to a lubricating oil flow detection method and system for filling production line.The method comprises: collecting the flow value of lubricating oil and the rotation time and blade reset time after pretreatment of turbine blade; constructing a round window, calculating the first adhesiveness of turbine blade in any round; calculating the level value and trend value of reset time sequence by using Holt linear trend method, to calculate the second adhesiveness of any round; taking the average value of first adhesiveness and second adhesiveness as the total adhesiveness of any round, traversing to obtain the total adhesiveness of each blade of the same turbine in the same round, for correcting the flow value, obtaining the real flow value, and completing flow detection. Through the technical scheme of the present application, the accuracy of lubricating oil flow detection result can be improved, and sufficient lubrication can be obtained during the operation of the production line.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of flow detection. More particularly, the present application relates to a lubricating oil flow detection method and system for a filling production line. BACKGROUND

[0002] In modern industrial production processes, filling production lines are a crucial part of manufacturing, especially in industries such as petroleum, chemical, food, and beverage, where the degree of automation is extremely high. Filling equipment needs to accurately, quickly, and stably fill products into packaging containers, while lubricating oil, as a key fluid for the operation of mechanical equipment, must maintain appropriate flow and stable supply to reduce friction between mechanical parts and improve equipment efficiency and service life.

[0003] However, as production scales expand and production line speeds increase, traditional lubricating oil flow detection methods face significant challenges. Mechanical or electrical flowmeters often cannot meet the needs of high precision and high stability, and have problems such as being difficult to maintain, easily damaged, and easily affected by the environment. To solve this problem, using a turbine flowmeter to measure the flow of lubricating oil has become an effective solution. When fluid flows through the turbine, the speed of the turbine rotating is proportional to the flow of the fluid. By monitoring the flow of lubricating oil in real time, it can ensure that the equipment obtains sufficient lubrication during operation, preventing wear, failure, and downtime caused by insufficient lubrication, thereby improving production efficiency, extending equipment life, and reducing maintenance costs.

[0004] However, even if the turbine flowmeter has been calibrated and calibrated at the factory, changes in the viscosity of the lubricating oil will still affect the measurement accuracy. The viscosity of lubricating oil changes at different temperatures, and when the viscosity increases, the turbine blades encounter more resistance when rotating in the fluid, causing the blade speed to slow down. The reduction in speed means that the measurement signal of the flowmeter will also decrease accordingly, so the reading of the flowmeter will be lower than the actual flow, resulting in inaccurate lubricating oil flow detection results. SUMMARY

[0005] To solve the problem of inaccurate lubricating oil flow detection results, the present application provides solutions in the following aspects.

[0006] In a first aspect, the present application discloses a lubricating oil flow detection method for a filling production line, comprising: collecting the flow value of lubricating oil and the rotation time and blade reset time after pretreatment of turbine blades; constructing a round window of any round, and calculating the first adhesion of turbine blades in any round according to the rotation time in the round window; constructing the blade reset time of each round in the round window as a reset time sequence, and calculating the level value and trend value of the reset time sequence by using the Holt linear trend method to calculate the second adhesion of any round; taking the average value of the first adhesion and the second adhesion as the total adhesion of any round, and obtaining the total adhesion of each blade of the same turbine in the same round for correcting the flow value to obtain the real flow value and complete flow detection.

[0007] By detecting the adhesion degree of the blades in real time, the influence of lubricating oil flow on the movement performance of the blades can be reflected, thereby providing more accurate data support for flow correction. Through the real flow value after correction, the accuracy of flow detection can be effectively improved, errors caused by adhesion problems can be avoided, and the running efficiency and stability of the turbine system under different working conditions can be ensured. In addition, this method can capture the changes of blade state in different time periods and provide long-term reliable operation data.

[0008] Preferably, the pretreatment comprises: removing the noise of the rotation time and the blade reset time by using a filtering algorithm, and filling the rotation time and the blade reset time by using a polynomial interpolation.

[0009] Preferably, the round window comprises: taking any round as a target round, and constructing a continuous preset number of rounds as a round window in the order of round size, wherein the target round is the largest round in the round window.

[0010] Preferably, the first adhesion satisfies the relationship:

[0011] , represents the first adhesion of round , represents the rotation time of round , represents the average value of the rotation time in the round window, represents the rotation time of round , represents the length of the round window, represents a normalization function.

[0012] The first viscosity measure reflects the amplitude of the time fluctuation of turbine blade rotation. Specifically, it measures the time fluctuation caused by changes in lubricating oil flow and friction during blade rotation. Through normalization, this metric can eliminate the interference of external factors such as temperature fluctuations or system load changes, ensuring more accurate viscosity calculations. Therefore, the first viscosity measure not only reveals the impact of lubricating oil on blade rotation efficiency but also reflects the frictional conditions of the blades under different operating conditions.

[0013] Preferably, the first adhesion satisfies the following relationship:

[0014] , Indicates round First adhesion, Indicates round Rotation time, This represents the average rotation time within the round window. Indicates round Rotation time, Indicates the length of the round window. This represents the normalization function.

[0015] Normalization effectively reduces the impact of external interference factors, ensuring more accurate viscosity calculations. Specifically, the first viscosity reflects the relationship between lubricating oil flow and blade rotation efficiency, revealing the influence of frictional changes during blade operation. This method not only provides a more accurate indicator for evaluating blade performance but also allows for early identification of changes in lubricating oil flow rate by monitoring viscosity variations.

[0016] Preferably, the second adhesion satisfies the following relationship:

[0017] , Indicates round The second adhesion, Indicates round The level value, This represents the mean of the horizontal values ​​within the round window. Indicates round The trend value, This represents the mean of the trend values ​​within the round window. This represents the normalization function.

[0018] Preferably, the second adhesion also satisfies the following relationship:

[0019] , Indicates round The second adhesion, Indicates round a level value, denotes the rotation of the blade reset time, denotes the average of the level value in the rotation window, denotes the rotation of the trend value, denotes the average of the trend value in the rotation window, denotes the normalized function.

[0020] The abnormal fluctuation of the blade reset time can be captured sharply, whether the blade is affected by external load changes, accumulated resistance or impurities, or the overall performance of the turbine system is declining. In this way, it can be identified whether the blade has adhesion and the degree of adhesion, providing strong data support for preventive maintenance of turbine equipment, and helping to improve the stability of turbine operation and improve the accuracy of lubricating oil flow detection results.

[0021] Preferably, the total adhesion degree further comprises: obtaining the influence weight of the first adhesion degree and the second adhesion degree respectively by a particle swarm optimization algorithm, and taking the sum of the product of the first adhesion degree and the second adhesion degree and the influence weight as the total adhesion degree.

[0022] Preferably, the obtaining of the real flow value comprises: calculating the average cumulative value of the total adhesion degree of each blade of the same turbine in the same rotation, and calculating the product of the normalized average cumulative value and the preset weight, taking the sum of 1 and the product as a correction factor of the flow value, and using the correction factor to correct the flow value to obtain the real flow value.

[0023] By introducing the correction factor and applying it to the correction of the flow value, not only the accuracy of flow measurement can be improved, but also the flow state of the turbine in actual operation can be more truly reflected, and this process ensures that the performance evaluation of the turbine system under different working conditions is more accurate.

[0024] In a second aspect, the present application discloses a lubricating oil flow detection system for a filling production line, comprising: a processor and a memory, the memory stores computer program instructions, when the computer program instructions are executed by the processor, any one of the lubricating oil flow detection methods for the filling production line is realized.

[0025] The beneficial effects of the present application are:

[0026] 1、The present application combines the dynamic changes of turbine blade rotation time and reset time, and uses the concept of adhesion degree to accurately detect the lubricating oil flow, which can effectively correct the deviation caused by flow measurement error, so as to obtain more accurate real flow value.

[0027] 2. The application predicts the trend change of the reset time sequence by the Holt linear trend method, so that the application can still stably and accurately detect the flow under the working environment with fast dynamic change. The particle swarm optimization algorithm is used to dynamically adjust the influence weight of the first adhesion and the second adhesion, and further ensure the high-precision detection under different working conditions.

[0028] 3. The application has good robustness and adaptability in actual production, can greatly improve the precision of flow measurement, reduce the influence of artificial error on the production process, and ensure the smooth operation of the filling production line under different load conditions. BRIEF DESCRIPTION OF DRAWINGS

[0029] Figure 1 is a flow chart of a lubricating oil flow detection method for a filling production line according to an embodiment of the application. DETAILED DESCRIPTION

[0030] The technical solutions in the embodiments of the application will be described clearly and completely below with reference to the drawings in the embodiments of the application. Obviously, the described embodiments are part of the embodiments of the application, rather than all the embodiments of the application.

[0031] The specific embodiments of the application will be described in detail below with reference to the drawings.

[0032] Referring to Figure 1 A lubricating oil flow detection method for a filling production line includes steps S1-S4, which will be described in detail below.

[0033] S1: Collect the flow value of the lubricating oil and the rotation time and blade reset time after pretreatment of the turbine blade.

[0034] In an embodiment, selecting a suitable installation position is the key to ensuring that the turbine flowmeter accurately measures the flow. The turbine flowmeter should be installed on a straight pipe section to avoid being installed in a bend, valve or other place that may cause fluid disturbance, because the fluid flow in these areas is unstable and can easily affect the accuracy of the flowmeter. After installation, the turbine flowmeter needs to be calibrated using a standard device with a known flow to ensure that its measurement results are accurate and reliable.

[0035] In a turbine flowmeter, in order to accurately measure the flow value of lubricating oil, a fixed-position magnetic sensor can be installed near the turbine blade, and magnetic material can be installed on each blade of the turbine to facilitate identification and tracking of each blade. When the turbine blade rotates, the magnetic material passes through the sensor, causing the magnetic field to change, thereby triggering the sensor to generate a voltage pulse signal. The Hall effect sensor is installed on the turbine housing to ensure that the signal of the magnetic field change on the blade can be detected. By measuring the time interval between these pulse signals, the rotation time and blade reset time of each turbine blade can be accurately obtained. The rotation time and blade reset time are collected in real time to ensure that the flowmeter can accurately reflect the change of fluid flow.

[0036] The rotation time of each turbine blade represents the time taken by the blade to pass through the magnetic sensor, that is, the pulse interval time recorded by the magnetic sensor each time the blade passes through the magnetic sensor.

[0037] The blade reset time represents the time from when a certain blade passes through the sensor to when the blade returns to its original position. Since the position of the turbine blade is fixed and the relative position between the blades does not change, the reset time data of any blade can be used as a representative of the overall blade reset time in the current rotation process.

[0038] In order to improve the accuracy and reliability of the rotation time and blade reset time, it is necessary to first use a filtering algorithm to remove noise from the rotation time and blade reset time data. Since the data collected by the magnetic sensor may be affected by environmental interference or system errors, using appropriate filtering algorithms can effectively remove high-frequency noise, retain useful signals, and ensure that the data is smoother and more stable. Filtering algorithms include low-pass filtering, Kalman filtering, etc.

[0039] Polynomial interpolation method is used to fill in the missing values or incomplete data in the rotation time and blade reset time. Polynomial interpolation can calculate the value of unknown points using mathematical models based on known sample data points, and then smooth the data, fill in the gaps, and ensure the continuity and integrity of the data.

[0040] Preprocessing can effectively eliminate noise and fill in data gaps, ultimately obtaining more accurate and reliable rotation time and reset time data, and improving the accuracy of flow measurement and the stability of the system.

[0041] S2: Construct a round window for any round, and calculate the first adhesion degree of the turbine blade in any round according to the rotation time in the round window.

[0042] It should be noted that the rotation time data of each turbine blade represents the time consumed by the blade when passing through the magnetic sensor. If all the blades are of the same size and no adhesion occurs during the lubricating oil flow, the rotation time of each blade when passing through the sensor should be the same, because the flow resistance of the lubricating oil they receive is consistent, and the rotation speed is also the same. However, if some blades are adhered, the blades adhered will receive greater flow resistance due to the increased friction between the blades and the fluid caused by adhesion, or due to the uneven distribution of flow resistance. Therefore, these adhered blades will take more time when passing through the sensor, which is manifested as longer rotation time.

[0043] In an embodiment, a round window of any round is constructed, and the method for constructing the round window is as follows: taking any round as a target round, and constructing a continuous preset number of rounds in the order of round size as a round window, wherein the target round is the largest round in the round window.

[0044] The first adhesion degree is calculated, and the first adhesion degree satisfies the relationship:

[0045] , The first adhesion degree of round , The rotation time of round , The average rotation time in the round window, The rotation time of round , The length of the round window, The normalization function.

[0046] Wherein, The deviation of the rotation time of round , that is, the difference between the rotation time of the round and the average rotation time. If the deviation is large, it indicates that the blade of the round is abnormal, that is, there is adhesion; and The volatility of the rotation time of round , if the volatility is large, it indicates that there is large volatility in the rotation time in the overall system, which may indicate that some blades are affected by uneven resistance or adhesion.

[0047] In an embodiment, the first adhesion degree satisfies the relationship:

[0048] , The first adhesion degree of round , The rotation time of round , The average rotation time in the round window, Indicates round Rotation time, Indicates the length of the round window. This represents the normalization function.

[0049] use To measure the round The difference between the rotation time and the mean reveals the deviation in rotation time. The denominator balances the volatility between different rounds by weighted summation of the fluctuations in rotation time across all rounds, ensuring that the scheme has higher stability and accuracy in handling fluctuations in rotation time.

[0050] S3: Construct a reset time series for the blade reset time of each round in the round window, and use the Holt linear trend method to calculate the level and trend values ​​of the reset time series in order to calculate the second adhesion of any round.

[0051] It should be noted that the reset time data for each turbine blade reflects the time required for the blade to return to its original position after passing the sensor. Under normal circumstances, since all blades are of uniform size and affected by the same lubricating oil flow resistance, the reset time is relatively constant. However, when the turbine begins to experience resistance, impurities or dirt gradually accumulate on the blades, increasing the frictional resistance and making blade rotation more difficult. This increase in resistance causes the turbine blade reset time to gradually lengthen, exhibiting an approximately linear upward trend. Simultaneously, because fluid flow is affected by resistance, the flow rate typically decreases with increasing resistance, also exhibiting an approximately linear downward trend. This phenomenon indicates that as the turbine blades experience more and more resistance, the blade reset time increases and the flow rate decreases, reflecting a decline in turbine system efficiency.

[0052] In one embodiment, the blade reset time for each round within a round window is constructed as a reset time series, and the level and trend values ​​of the reset time series are calculated using the Holt linear trend method. The Holt linear trend method is an existing technique for time series forecasting, suitable for data with a linear trend. It calculates the level and trend values ​​using two smoothing equations. The level value represents the current level or average of the data; the trend value represents the trend of data change, i.e., the rise or fall of the data.

[0053] Calculate the second adhesion degree, which satisfies the following relationship:

[0054] , Indicates round The second adhesion, Indicates round The level value, a mean value of the level values in the round window, a trend value of the round , a mean value of the trend values in the round window, a normalization function.

[0055] The Holt linear trend method obtains the level value and the trend value of the reset time through smoothing processing, the level value reflects the average state of the reset time, and the trend value reveals the direction and amplitude of the change of the reset time over time. When calculating the second adhesion, the deviation of the level value and the trend value is combined, which can further quantify the degree of abnormal fluctuation of the reset time. When the turbine blade is affected by gradually increasing resistance or impurities, the level value and the trend value of the reset time will be significantly deviated, and the calculation of the second adhesion can accurately reflect this change, helping to discover potential blade adhesion problems in time. Through this method, the system can accurately monitor the state of the blade during the operation of the turbine, effectively improve the accuracy of fault prediction, provide reliable data support for the maintenance and optimization of the turbine, reduce the occurrence of faults, and improve the overall efficiency and stability of the system.

[0056] In one embodiment, the second adhesion also satisfies the relationship:

[0057] , a second adhesion of the round , a level value of the round , a blade reset time of the round , a mean value of the level values in the round window, a trend value of the round , a mean value of the trend values in the round window, a normalization function.

[0058] The second adhesion reflects the deviation of the reset time from its smoothed expected level and trend, especially through the cubic root operation to smooth the deviation, effectively avoiding the influence of a single large deviation on the result.

[0059] S4: Take the mean value of the first adhesion and the second adhesion as the total adhesion of any round, traverse to obtain the total adhesion of each blade of the same turbine in the same round, and use it to correct the flow value to obtain the true flow value, and complete the flow detection.

[0060] It should be noted that the blade is affected by the flow of lubricating oil during rotation, which may produce different first adhesions, i.e. the friction between the blade and the lubricating oil changes during the movement of the blade, which directly affects the movement performance of the blade. The reset time data reflects the overall adhesion of the blade when it returns to the starting position after completing a cycle, i.e. the second adhesion, which includes the adhesion change of the blade under the action of external factors (such as airflow, resistance, etc.). When the first adhesion and the second adhesion are combined, the adhesion of the blade during rotation can be comprehensively evaluated, and by quantifying the influence of the two, the overall working state of the blade can be more accurately understood.

[0061] In one embodiment, the average of the first adhesion and the second adhesion is taken as the total adhesion of any round, and the total adhesion of each blade of the same turbine in the same round is obtained by traversal. The average cumulative value of the total adhesion of each blade of the same turbine in the same round is calculated, and the product of the normalized average cumulative value and the preset weight is calculated. The sum of 1 and the product is taken as a correction factor of the flow value, which is used to correct the flow value to obtain a true flow value, and the flow detection is completed. For example, the preset weight is set to 0.5.

[0062] In one embodiment, the total adhesion further comprises: obtaining the influence weight of the first adhesion and the second adhesion respectively by a particle swarm optimization algorithm, and taking the sum of the product of the first adhesion and the second adhesion and the influence weight respectively as the total adhesion.

[0063] The system comprises a processor and a memory, and the memory stores computer program instructions which, when executed by the processor, implement a steam leakage detection method for a waste heat recovery system according to the first aspect of the application.

[0064] The system also includes a communication bus and a communication interface and other components well known to those skilled in the art, the settings and functions of which are known in the art, and therefore will not be described here.

[0065] It should be noted that for those skilled in the art, without departing from the concept of the present application, several modifications and improvements can be made, which all belong to the protection scope of the present application. Therefore, the protection scope of the present application patent should be subject to the appended claims.

Claims

1. A lubricating oil flow rate detection method for a filling production line, characterized by, The method comprises: collecting the flow value of lubricating oil and the rotation time and blade reset time after pretreatment of turbine blades; constructing a round window of any round, calculating the first adhesion of turbine blades in any round according to the rotation time in the round window; constructing the blade reset time of each round in the round window as a reset time sequence, calculating the level value and trend value of the reset time sequence by using the Holt linear trend method, to calculate the second adhesion of any round; taking the average value of the first adhesion and the second adhesion as the total adhesion of any round, traversing to obtain the total adhesion of each blade of the same turbine in the same round, for correcting the flow value to obtain the real flow value and complete flow detection.

2. The lubricating oil flow detection method for a filling production line according to claim 1, characterized by, The pretreatment comprises: using a filtering algorithm to remove noise of the rotation time and the blade reset time, and using a polynomial interpolation to fill the rotation time and the blade reset time.

3. The lubricating oil flow detection method for a filling production line according to claim 1, characterized by, The round window comprises: taking any round as a target round, and constructing a continuous preset number of rounds in the order of round size as a round window, wherein the target round is the largest round in the round window.

4. The lubricating oil flow detection method for a filling production line according to claim 1, characterized by, The first adhesion satisfies the relationship: , denotes the round of the first stickiness, denotes the round time, denotes the mean of the round times in the round window, denotes the round time, denotes the round window length, denotes the normalization function.

5. The lubricating oil flow detection method for a filling production line according to claim 1, characterized by, The first adhesion satisfies the relationship: , denotes the round of the first stickiness, denotes the rotation time of the round , denotes the mean of the rotation times in the round window, denotes the rotation time of the round , denotes the length of the round window, denotes a normalization function.

6. The lubricating oil flow detection method for a filling production line according to claim 1, characterized by, The second adhesion satisfies the relationship: , Indicates round The second adhesion, Indicates round The level value, This represents the mean of the horizontal values ​​in the round window. Indicates round The trend value, This represents the mean of the trend values ​​within the round window. This represents the normalization function.

7. A method for detecting the flow rate of lubricating oil for a filling production line according to claim 6, characterized in that, The second adhesion also satisfies the relationship: , represents the round of the second stickiness, represents the level value of the round , represents the blade reset time of the round , represents the mean of the level values in the round window, represents the trend value of the round , represents the mean of the trend values in the round window, represents a normalization function.

8. The lubricating oil flow detection method for a filling production line according to claim 1, characterized by, The total adhesion also comprises: obtaining the influence weight of the first adhesion and the second adhesion by a particle swarm optimization algorithm, taking the product of the first adhesion and the second adhesion and the influence weight respectively, and taking the sum of the products as the total adhesion.

9. The lubricating oil flow detection method for a filling production line according to claim 1, characterized by, The method for obtaining the real flow value comprises: calculating the average cumulative value of the total adhesion of each blade of the same turbine in the same round, calculating the product of the normalized average cumulative value and a preset weight, taking the sum of 1 and the product as a correction factor of the flow value, and correcting the flow value to obtain the real flow value.

10. A lubricating oil flow detection system for a filling line, characterized in that, The method comprises: a processor and a memory, the memory storing computer program instructions, when the computer program instructions are executed by the processor, realizing a lubricating oil flow detection method for a filling production line according to any one of claims 1-9.

Images