Safe and economical oil quality prediction method for mechanical equipment

Abstract

The invention relates to the technical field of oil quality prediction, in particular to a safe and economical oil quality prediction method for mechanical equipment, which comprises the following steps: establishing a database, confirming a prediction model by using information of the database, and calculating an early warning value according to the prediction model. According to the method, the data automatically measured through sensor integration is collected and concluded, the oil product database is established through the collected data, the prediction model is obtained through modeling, the early warning value is calculated, and the method is not limited by factors such as time and space and is high in flexibility and high in reliability. Meanwhile, the problem that professional and scientific quality prediction and evaluation technology is needed in the industrial production process is solved, and the method is used for guiding work such as the oil change period and the oil change frequency so as to reduce oil waste and promote safer and more economical operation of mechanical equipment.

Description

Technical Field

[0001] The invention relates to the technical field of oil quality prediction, and in particular to a safe and economical method for predicting the quality of oil used in mechanical equipment. Background Art

[0002] Relevant standards include GB / T 7596-2017 "Quality of Mineral Turbine Oil in Power Plant Operation", GB 11118.1-2011 "Hydraulic Oil", DL / T 571-2014 "Guidelines for Operation and Maintenance of Phosphate Ester Fire-Resistant Oil for Power Plants" and other oil quality standards.

[0003] The oil used in mechanical equipment includes lubricating oil, hydraulic oil, fire-resistant oil, etc., which can realize the functions of lubrication, sealing, cooling, cleaning, and corrosion protection of equipment or components. The daily monitoring indicators of the oil quality of mechanical equipment oil mainly include moisture, kinematic viscosity, acid value, demulsification degree, flash point, cleanliness, etc. When the above indicators exceed the standard or are abnormal, it is easy to cause the relevant performance of the oil system equipment to decline, affecting production and economic operation. In the actual production process, the oil index exceeds the standard and requires comprehensive, complex and comprehensive cause analysis work, but due to various reasons, the management of mechanical equipment oil only accounts for a small part of the production and operation management of the enterprise, and the use, maintenance and management of mechanical equipment oil are more likely to be weakened by the enterprise. Therefore, many enterprises have a wide variety of mechanical equipment oil, and the oil consumption is large. Some oil products are replaced before the oil change period. It can be seen that the quality monitoring of mechanical equipment oil involved in the industrial production process urgently needs professional and scientific quality prediction and evaluation technology. Summary of the invention

[0004] The purpose of the present invention is to provide a safe and economical method for predicting the quality of oil used in mechanical equipment.

[0005] To achieve the above object, the present invention provides the following technical solution: a safe and economical method for predicting the quality of oil used in mechanical equipment, comprising the following steps:

[0006] S1, establish database: including user elements U i 、Oil for mechanical equipment i,j 、Performance Index I i,j,k,l and the number of tests TC i,j,k,l , and complete data information collection and data statistics;

[0007] The i is the user type, i≥1, i=1,2,3.......n, n is a positive integer;

[0008] j is the type of oil used in mechanical equipment, j≥1, j=1,2,3...m, m is a positive integer;

[0009] The k is the type of performance indicator, k≥1, k=1,2,3...o, o is a positive integer;

[0010] The l represents the number of tests, l≥1, l=1,2,3...p, p is a positive integer;

[0011] S2, confirm the prediction model: extract database information, carry out oil quality prediction modeling for the performance indicators of the type of oil used in the mechanical equipment of the predicted user, use the error analysis method to analyze the accuracy of different models, and select the model with the smallest error as the prediction model for the corresponding performance indicators of the corresponding mechanical equipment oil used in the predicted user;

[0012] S3, calculate the warning value: substitute the test times of the performance index of the type of oil used in the user's mechanical equipment into the prediction model. When the value of a certain performance index is equal to or does not meet the limit value of the test index, the calculation is stopped, and the test times exceeding the limit value of the performance index are recorded. The time node corresponding to the test times is the quality deterioration warning time of the oil used in the user's mechanical equipment.

[0013] Furthermore, the S2 is further specifically as follows: for the user element U i 、Oil for mechanical equipment i,j Specific performance indicators I i,j,k,l , taking the test number value l as the horizontal axis, extract l≥20 performance index data I i,j,k,l Draw a data curve, use one or more functions in the function to perform mathematical operations to establish a fitting curve model. The fitting curve should be infinitely close to the data curve or even overlap with the data curve. Use the error analysis method to calculate the error value RMSE between different fitting curve models and the data curve i,j,k , the smallest min(RMSE i,j,k ) is the prediction model y corresponding to the user, the oil used in the mechanical equipment, and the performance index. i,j,k =f(TC i,j,k,l ), where y i,j,k is the performance indicator value.

[0014] Furthermore, the S3 is further specifically as follows: the test times l=1, 2, 3...p, where p is a positive integer, are substituted into the prediction model y i,j,k Calculate, when a certain test number TC i,j,k,l The corresponding y i,j,k Equal to or not meeting the corresponding performance index I i,j,k,l The standard limit value I i,j,k,0 When TC is reached, the calculation stops; the test times TC of all performance indicators corresponding to the corresponding user and the corresponding mechanical equipment oil i,j,k,l Select the smallest min(TC i,j,k,l) is the early warning value EW of the oil quality deterioration of the user and the mechanical equipment t That is EW t =min(TC i,j,k,l ).

[0015] Furthermore, the performance index is an index reflecting the quality status of the oil product.

[0016] Further, based on the data collection and statistical analysis, performance index I i,j,k,l Equal to or not meeting the performance index limit value specified in the relevant standards I i,j,k,0 The requirements are unified as not meeting the requirements, and the rest are meeting the requirements; the cumulative performance index I i,j,k,l In all test times TC i,j,k,p The number of times that the requirements are not met is NT i,j,k,q , when NT i,j,k,q The percentage of all test times of this performance indicator TC i,j,k,p percentage When , the performance index is considered to be a sensitivity performance index, where q is the number of tests corresponding to the case where the performance index does not meet the requirements. The larger it is, the more sensitive the sensitivity factor is to the change in the quality of the oil used in the mechanical equipment;

[0017] Among the sensitivity performance indicators, select the performance indicator limit value I that meets the requirements of the relevant standards. i,j,k,0 The data is used to perform model prediction in step S2.

[0018] Compared with the prior art, the present invention collects and summarizes the data automatically measured by sensor integration, establishes an oil product database through the collected data, models a prediction model, and calculates the warning value. The present invention is not limited by time, space and other factors and has strong flexibility. At the same time, the present invention solves the problem that the industrial production process requires professional and scientific quality prediction and evaluation technology, and is used to guide the oil change cycle, oil change frequency and other work to reduce oil waste and promote safer and more economical operation of mechanical equipment. DETAILED DESCRIPTION

[0019] The technical scheme of the present invention is further described below in conjunction with specific examples. It should be understood that the following examples are only exemplary descriptions and explanations of the present invention, and should not be construed as limiting the scope of protection of the present invention. All technologies implemented based on the above content of the present invention are included in the scope that the present invention is intended to protect.

[0020] Among them, the relevant standards can refer to the oil quality standards described in the background technology.

[0021] Example 1

[0022] First, for a certain company's U1 turbine lubricating oil O 1,1 Moisture I 1,1,1 , Kinematic viscosity I 1,1,2 、Cleanliness I 1,1,3 Performance test, I 1,1,1 ,I 1,1,2 ,I 1,1,3 Test times TC 1,1,1,100 TC 1,1,2,100 TC 1,1,3,100 100 times respectively, and the moisture I 1,1,1,1 ~I 1,1,1,100 , Kinematic viscosity I 1,1,2,1 ~I 1,1,2,100 、Cleanliness I 1,1,3,1 ~I 1,1,3,100 100 test data each to form a database;

[0023] Among them, the limit value of moisture in this case is I 1,1,1,0 ≤30mg / L, kinematic viscosity I 1,1,2,0 39.1~52.9mm 2 / s, cleanliness limit value I 1,1,3,0 ≤8, according to the database data, does not meet the requirements NT 1,1,1,20 20 times, NT 1,1,2,0 0 times, NT 1,1,3,10 is 10 times, then in, and If the moisture content is greater than 5%, 1,1,1 、Cleanliness I 1,1,3 For the enterprise U1 turbine lubricating oil O 1,1 Sensitive performance indicators;

[0024] Next, extract the sensitive performance index moisture I that meets the requirements from the database 1,1,1 、Cleanliness I 1,1,3 50 test data of each, and use one or more functions such as polynomial function, power function, exponential function, etc. to perform mathematical operations such as addition, subtraction, multiplication and division to respectively analyze the sensitive performance index moisture I 1,1,1 、Cleanliness I 1,1,3 Extract data for fitting and obtain a series of fitting models y 1,1,1 =f(TC 1,1,1,50 ),y 1,1,3 =f(TC 1,1,3,50 ), and calculate the RMSE corresponding to the fitted model respectively 1,1,1 、RMSE 1,1,3 value, select the smallest min(RMSE 1,1,1 )、min(RMSE 1,1,3 ) is the quality prediction model y1,1,1,b =f b (TC 1,1,1,50 ),y 1,1,3,b =f b (TC 1,1,3,50 );

[0025] Finally, substitute l=1,2,3...p into the quality prediction model y 1,1,1,b =f b (TC 1,1,1,50 ),y 1,1,3,b =f b (TC 1,1,3,50 ) for moisture I 1,1,1 、Cleanliness I 1,1,3 The quality prediction calculation of TC 1,1,1,55 That is, the 55th moisture I 1,1,1,55 TC 1,1,3,60 That is, the 60th cleanliness level I 1,1,3,60 If the limit value is not met, then the enterprise U1 turbine lubricating oil O 1,1 The quality deterioration warning value EW t It is 55, which means that the oil quality exceeded the standard during the 55th test.

[0026] Although the specific implementation modes of the present invention are described above, those skilled in the art should understand that the specific implementation modes described are only illustrative and are not intended to limit the scope of the present invention. Effective modifications and changes made by those skilled in the art in accordance with the spirit of the present invention should be included in the scope of protection of the claims of the present invention.

Claims

1. A safe and economical method for predicting the quality of oil used in mechanical equipment, characterized by: The following steps are involved: S1, establish database: including user elements U i 、Oil for mechanical equipment i,j 、Performance Index I i,j,k,l and the number of tests TC i,j,k,l , and complete data information collection and data statistics; The i is the user type, i≥1, i=1,2,3.......n, n is a positive integer; j is the type of oil used in mechanical equipment, j≥1, j=1,2,3...m, m is a positive integer; The k is the type of performance indicator, k≥1, k=1,2,3...o, o is a positive integer; The l represents the number of tests, l≥1, l=1,2,3...p, p is a positive integer; S2, confirm the prediction model: extract database information, carry out oil quality prediction modeling for the performance indicators of the type of oil used in the mechanical equipment of the predicted user, use the error analysis method to analyze the accuracy of different models, and select the model with the smallest error as the prediction model for the corresponding performance indicators of the corresponding mechanical equipment oil used in the predicted user; S3, calculate the warning value: substitute the test times of the performance index of the type of oil used in the user's mechanical equipment into the prediction model. When the value of a certain performance index is equal to or does not meet the limit value of the test index, the calculation is stopped, and the test times exceeding the limit value of the performance index are recorded. The time node corresponding to the test times is the quality deterioration warning time of the oil used in the user's mechanical equipment.

2. A safe and economical mechanical equipment oil quality prediction method according to claim 1, characterized in that: The S2 is further specifically as follows: for the user element U i 、Oil for mechanical equipment i,j Specific performance indicators I i,j,k,l , taking the test number value l as the horizontal axis, extract l≥20 performance index data I i,j,k,l Draw a data curve, use one or more functions in the function to perform mathematical operations to establish a fitting curve model. The fitting curve should be infinitely close to the data curve or even overlap with the data curve. Use the error analysis method to calculate the error value RMSE between different fitting curve models and the data curve i,j,k , the smallest min(RMSE i,j,k ) is the prediction model y corresponding to the user, the oil used in the mechanical equipment, and the performance index. i,j,k =f(TC i,j,k,l ), where y i,j,k is the performance indicator value.

3. A safe and economical mechanical equipment oil quality prediction method according to claim 2, characterized in that: S3 is further specifically as follows: Substitute the test times l=1, 2, 3...p, where p is a positive integer, into the prediction model y i,j,k Calculate, when a certain test number TC i,j,k,l The corresponding y i,j,k Equal to or not meeting the corresponding performance index I i,j,k,l The standard limit value I i,j,k,0 When TC is reached, the calculation stops; the test times TC of all performance indicators corresponding to the corresponding user and the corresponding mechanical equipment oil i,j,k,l Select the smallest min(TC i,j,k,l ) is the early warning value EW of the oil quality deterioration of the user and the mechanical equipment t That is EW t =min(TC i,j,k,l ).

4. A safe and economical mechanical equipment oil quality prediction method according to claim 1, characterized in that: The performance index is an index reflecting the quality status of oil products.

5. A safe and economical mechanical equipment oil quality prediction method according to claim 1, characterized in that: Complete data collection and statistical analysis based on which performance indicators I i,j,k,l Equal to or not meeting the performance index limit value specified in the relevant standards I i,j,k,0 The requirements are unified as not meeting the requirements, and the rest are meeting the requirements; the cumulative performance index I i,j,k,l In all test times TC i,j,k,p The number of times that the requirements are not met is NT i,j,k,q , when NT i,j,k,q The percentage of all test times of this performance indicator TC i,j,k,p Percent φ i,j,k ≥5%, the performance index is considered to be a sensitive performance index, where q is the number of tests corresponding to the case where the performance index does not meet the requirements, φ i,j,k The larger it is, the more sensitive the sensitivity factor is to the change in the quality of the oil used in the mechanical equipment; Among the sensitivity performance indicators, select the performance indicator limit value I that meets the requirements of the relevant standards. i,j,k,0 The data is used to perform model prediction in step S2.