CN111898088B - Method and system for determining full viscosity temperature curve of crude oil - Google Patents

Method and system for determining full viscosity temperature curve of crude oil Download PDF

Info

Publication number
CN111898088B
CN111898088B CN202010766855.0A CN202010766855A CN111898088B CN 111898088 B CN111898088 B CN 111898088B CN 202010766855 A CN202010766855 A CN 202010766855A CN 111898088 B CN111898088 B CN 111898088B
Authority
CN
China
Prior art keywords
crude oil
temperature
viscosity
newtonian fluid
determining
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN202010766855.0A
Other languages
Chinese (zh)
Other versions
CN111898088A (en
Inventor
姚博
张健铭
李传宪
杨飞
孙广宇
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
China University of Petroleum East China
Original Assignee
China University of Petroleum East China
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by China University of Petroleum East China filed Critical China University of Petroleum East China
Priority to CN202010766855.0A priority Critical patent/CN111898088B/en
Publication of CN111898088A publication Critical patent/CN111898088A/en
Application granted granted Critical
Publication of CN111898088B publication Critical patent/CN111898088B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10—Complex mathematical operations
    • G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N11/00—Investigating flow properties of materials, e.g. viscosity, plasticity; Analysing materials by determining flow properties
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10—Complex mathematical operations
    • G06F17/17—Function evaluation by approximation methods, e.g. inter- or extrapolation, smoothing, least mean square method
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/004—Artificial life, i.e. computing arrangements simulating life
    • G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N11/00—Investigating flow properties of materials, e.g. viscosity, plasticity; Analysing materials by determining flow properties
    • G01N2011/0026—Investigating specific flow properties of non-Newtonian fluids
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N11/00—Investigating flow properties of materials, e.g. viscosity, plasticity; Analysing materials by determining flow properties
    • G01N2011/006—Determining flow properties indirectly by measuring other parameters of the system
    • G01N2011/0093—Determining flow properties indirectly by measuring other parameters of the system thermal properties

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Mathematical Physics (AREA)
  • General Engineering & Computer Science (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Pure & Applied Mathematics (AREA)
  • Software Systems (AREA)
  • Mathematical Optimization (AREA)
  • Mathematical Analysis (AREA)
  • Computational Mathematics (AREA)
  • Health & Medical Sciences (AREA)
  • General Health & Medical Sciences (AREA)
  • Evolutionary Computation (AREA)
  • Computational Linguistics (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • Artificial Intelligence (AREA)
  • Biomedical Technology (AREA)
  • Biophysics (AREA)
  • Algebra (AREA)
  • Databases & Information Systems (AREA)
  • Analytical Chemistry (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Chemical & Material Sciences (AREA)
  • Geometry (AREA)
  • Biochemistry (AREA)
  • Immunology (AREA)
  • Pathology (AREA)
  • Computer Hardware Design (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Evolutionary Biology (AREA)
  • Operations Research (AREA)
  • Probability & Statistics with Applications (AREA)
  • Feedback Control In General (AREA)
  • Control Of Temperature (AREA)

Abstract

The invention relates to a method and a system for determining a full viscosity temperature curve of crude oil, wherein the method comprises the following steps: acquiring a density value and a condensation point value of crude oil; according to the values, a crude oil Newtonian fluid viscosity-temperature coefficient prediction model is adopted to determine the state viscosity-temperature coefficient of the Newtonian fluid; according to the viscosity-temperature coefficient of the Newtonian fluid state and the wax content, a crude oil non-Newtonian fluid viscosity-temperature coefficient prediction model is adopted to determine the viscosity-temperature coefficient of the non-Newtonian fluid state; selecting a temperature value which is larger than the abnormal point temperature value of the crude oil from the crude oil temperature set as a crude oil Newtonian fluid temperature set; the remainder being a set of crude oil non-newtonian fluid temperatures; determining a viscosity-temperature curve of a temperature interval of the crude oil Newtonian fluid according to the temperature set of the crude oil Newtonian fluid and the viscosity-temperature coefficient of the state of the Newtonian fluid; and determining a viscosity-temperature curve of the temperature interval of the crude oil non-Newtonian fluid according to the temperature set of the crude oil non-Newtonian fluid and the viscosity-temperature coefficient of the state of the non-Newtonian fluid. The full viscosity temperature curve of the crude oil is obtained by the method disclosed by the invention, so that the safety of crude oil pipeline transportation and storage is ensured.

Description

Method and system for determining full viscosity temperature curve of crude oil
Technical Field
The invention relates to the technical field of crude oil exploration, in particular to a method and a system for determining a full viscosity temperature curve of crude oil.
Background
The viscosity of crude oil is a key parameter in the safe transportation and storage links of crude oil. On the one hand, for long-distance oil delivery pipelines, the increase of the viscosity of the oil products can increase the friction resistance of the pipelines along the flow direction, the energy consumption for delivery is increased, and when the viscosity of the oil products fluctuates, the load of a motor and an oil delivery pump can be caused to fluctuate, so that the motor and the oil delivery pump deviate from the range of a high-efficiency working area to cause waste. On the other hand, when the offshore tanker discharges oil at the dock, if the viscosity of the crude oil is too high, great potential safety hazards are brought to the receiving, discharging, storing and subsequent pipeline conveying of the crude oil dock. Especially under the abominable operating mode of lower temperature or temperature abrupt change in winter, crude oil viscosity is showing and is increasing, mobility worsens, and the pier is difficult to effectively develop and connects and unloads oil operation, and the oil drum must adjust and connect and unload the place, has not only increased oil drum lag time, also brought pressure for refining enterprise crude oil resource guaranty. Therefore, accurately obtaining the viscosity-temperature curve of crude oil is a basic condition for normal operation of the storage and transportation link.
China crude oil is characterized by higher wax content and higher viscosity at normal temperature. The crude oil is in Newtonian fluid state when the temperature is higher than the abnormal point, and the viscosity of the crude oil only changes with the change of the temperature for one oil product; when the temperature is lower than the abnormal point, the wax originally dissolved in the crude oil is crystallized and separated out to form a certain three-dimensional network structure, and the crude oil is in a non-Newtonian fluid state at the moment, and the viscosity of the oil changes along with the change of the temperature and the shear rate. The measurement of the viscosity-temperature curve of crude oil is usually carried out by a high-precision viscometer or rheometer, the viscosity of crude oil is reflected by the viscosity of crude oil along with the change condition of temperature and shear rate, and when the viscosity of crude oil is obtained, the viscosity of crude oil at any temperature and shear rate is obtained. Because of limited field experimental conditions, measuring crude oil full viscosity temperature profiles using viscometers or rheometers under engineering field conditions is difficult to achieve.
Since the viscosity-temperature properties of crude oil are closely related to the physical properties of density, congealing point, wax content, etc., some scholars have sought to estimate the viscosity of crude oil by means of data fitting. Zhang Chunming et al derive a model for predicting the viscosity of crude oil at any temperature based on a viscosity value at 50 c by fitting the viscosity-temperature data of 20 crude oil samples, see equation 2-1.
μ T ＝(0.01481n T +0.9421)μ (3.1613-0.5525lnT) 50 (2-1)
Saeed M.AL-Zahrani et al demonstrate by controlled variable methods that temperature and wax content directly affect the viscosity of crude oil, and model non-Newtonian fluids at a single temperature and wax content, and propose equations 2-2 for predicting the viscosity of waxy crude oil at different temperatures and different shear rates.
Figure BDA0002615012570000021
Wherein mu is the predicted viscosity of the oil, A, B, C and D are fitted parameters, T is the temperature, and W is the mass content of wax in crude oil.
Shivanjali et al propose a model for calculating the viscosity of Indian waxy crude oil by least squares, see equations 2-3.
Figure BDA0002615012570000022
Wherein mu is the predicted viscosity of the oil, W is the mass content of wax in crude oil, gamma is the shear rate, and tau is the shear stress.
The formula fitted by the method simply fits the relation between the viscosity of the crude oil and the temperature and shear rate from the data angle through the change trend of the viscosity, and the wax content in the waxy crude oil is considered, but other physical properties of the waxy crude oil are not applied to the prediction of the viscosity of the waxy crude oil, and the reason for influencing the viscosity of the crude oil is difficult to search from the physical property angle of the crude oil. Secondly, the existing research is only aimed at specific waxy crude oil, and the obtained formula form once fixed limits the application range of the crude oil, so that the crude oil sample with the viscosity to be predicted cannot be adapted later. Again, the existing crude oil viscosity prediction does not consider the viscosity change condition of crude oil in the temperature range of non-newtonian fluid and newtonian fluid, and is difficult to meet the requirement of full viscosity temperature curve required by actual pipeline transportation.
Disclosure of Invention
The invention aims to provide a method and a system for determining a full viscosity temperature curve of crude oil, which fully consider the basic physical properties of the crude oil, respectively treat the viscosity of the crude oil in a non-Newtonian fluid temperature range and a Newtonian fluid temperature range to obtain the full viscosity temperature curve of the crude oil, and ensure the safety of the pipeline transportation and storage of the crude oil.
In order to achieve the above object, the present invention provides the following solutions:
a method for determining a full viscosity temperature curve of crude oil, comprising:
acquiring a density value and a condensation point value of crude oil;
according to the density value and the condensation point value, a crude oil Newtonian fluid viscosity-temperature coefficient prediction model is adopted to determine a Newtonian fluid state viscosity-temperature coefficient; the crude oil Newtonian fluid viscosity-temperature coefficient prediction model is established according to a quantum particle swarm optimization algorithm and a generalized regression neural network;
obtaining the wax content of crude oil;
according to the Newtonian fluid state viscosity-temperature coefficient and the wax content, a crude oil non-Newtonian fluid viscosity-temperature coefficient prediction model is adopted to determine a non-Newtonian fluid state viscosity-temperature coefficient; the crude oil non-Newtonian fluid viscosity-temperature coefficient prediction model is established according to a quantum particle swarm optimization algorithm and a generalized regression neural network;
Acquiring a crude oil temperature set; the crude oil temperature set comprises different temperature values;
selecting a temperature value which is larger than the abnormal point temperature value of the crude oil from the crude oil temperature set as a crude oil Newtonian fluid temperature set;
selecting a temperature value less than or equal to the temperature value of the abnormal point of the crude oil from the crude oil temperature set as a crude oil non-Newtonian fluid temperature set;
determining a viscosity-temperature curve of a temperature interval of the crude oil Newtonian fluid according to the crude oil Newtonian fluid temperature set and the viscosity-temperature coefficient of the state of the Newtonian fluid;
and determining a viscosity-temperature curve of the temperature interval of the crude oil non-Newtonian fluid according to the temperature set of the crude oil non-Newtonian fluid and the viscosity-temperature coefficient of the state of the non-Newtonian fluid.
Optionally, the crude oil newton fluid viscosity temperature coefficient prediction model obtaining process specifically includes:
acquiring a crude oil sample density value set and a crude oil sample condensation point value set; the density value set of the crude oil samples comprises a density value of each crude oil sample; the set of the condensation point values of the crude oil samples comprises the condensation point value of each crude oil sample;
inputting the crude oil sample density value set and the crude oil sample condensation point value set into a first generalized regression neural network to obtain a first root mean square error;
Judging whether the first iteration times are smaller than the first preset iteration times or not to obtain a first judgment result;
if the first judgment result indicates that the first iteration number is smaller than the first preset iteration number, adopting a quantum particle swarm optimization algorithm to adjust the super parameters in the first generalized regression neural network, updating the adjusted generalized regression neural network into the first generalized regression neural network, and returning to the step of inputting the crude oil sample density value set and the crude oil sample condensation point value set into the first generalized regression neural network to obtain a first root mean square error;
and if the first judgment result indicates that the first iteration times are larger than or equal to the first preset iteration times, selecting a first generalized regression neural network corresponding to the first root mean square error value with the minimum value as a crude oil Newtonian fluid viscosity-temperature coefficient prediction model.
Optionally, the crude oil non-newtonian fluid viscosity temperature coefficient prediction model obtaining process specifically includes:
inputting the crude oil sample density value set and the crude oil sample condensation point value set into the crude oil Newtonian fluid viscosity-temperature coefficient prediction model to obtain a crude oil sample Newtonian fluid state viscosity-temperature coefficient set;
Acquiring a crude oil sample wax content set; the set of crude oil sample wax content includes a wax content value for each crude oil sample;
inputting the crude oil sample Newtonian fluid state viscosity-temperature coefficient set and the crude oil sample wax content set into a second generalized regression neural network to obtain a second root mean square error;
judging whether the second iteration times are smaller than second preset iteration times or not to obtain a second judgment result;
if the second judgment result indicates that the second iteration number is smaller than the second preset iteration number, adopting a quantum particle swarm optimization algorithm to adjust the super-parameters in the second generalized regression neural network, updating the adjusted generalized regression neural network into the second generalized regression neural network, and returning to the step of inputting the crude oil sample Newtonian fluid state viscosity-temperature coefficient set and the crude oil sample wax content set into the second generalized regression neural network to obtain a second root mean square error;
and if the second judgment result indicates that the second iteration times are larger than or equal to the second preset iteration times, selecting a second generalized regression neural network corresponding to the minimum second root mean square error value as a crude oil non-Newtonian fluid viscosity-temperature coefficient prediction model.
Optionally, the determining the viscosity-temperature curve of the temperature interval of the crude oil newtonian fluid according to the temperature set of the crude oil newtonian fluid and the viscosity-temperature coefficient of the newtonian fluid state specifically includes:
according to formula mu 1 ＝aT 1 2 +bT 1 +c determining a viscous temperature curve of the crude oil Newtonian fluid temperature interval;
wherein mu 1 At a temperature T 1 Viscosity, a, b and c are viscosity-temperature coefficients, T, of Newtonian fluid states 1 Is the temperature value in the crude oil Newtonian fluid temperature set.
Optionally, the determining the viscosity-temperature curve of the temperature interval of the crude oil non-newtonian fluid according to the temperature set of the crude oil non-newtonian fluid and the viscosity-temperature coefficient of the non-newtonian fluid state specifically includes:
according to the formula
Figure BDA0002615012570000041
Determining a viscosity-temperature curve of a non-Newtonian fluid temperature interval of crude oil;
wherein mu 2 At a temperature T 2 The viscosity at the time of the formation of the gel,
Figure BDA0002615012570000042
n′＝fT 2 +g, d, e, f, g are viscosity-temperature coefficients of non-Newtonian fluid states, K is a consistency coefficient, n' is a rheological behavior index, and gamma is shearCutting rate, T 2 Is the temperature value in the temperature set of the crude oil non-Newtonian fluid.
A system for determining a full viscosity temperature curve of crude oil, comprising:
the crude oil density value and condensation point value acquisition module is used for acquiring the density value and condensation point value of crude oil;
the Newtonian fluid state viscosity-temperature coefficient determining module is used for determining the Newtonian fluid state viscosity-temperature coefficient by adopting a crude oil Newtonian fluid viscosity-temperature coefficient prediction model according to the density value and the condensation point value; the crude oil Newtonian fluid viscosity-temperature coefficient prediction model is established according to a quantum particle swarm optimization algorithm and a generalized regression neural network;
The crude oil wax content acquisition module is used for acquiring the wax content of crude oil;
the non-Newtonian fluid state viscosity-temperature coefficient determination module is used for determining the non-Newtonian fluid state viscosity-temperature coefficient by adopting a crude oil non-Newtonian fluid viscosity-temperature coefficient prediction model according to the Newtonian fluid state viscosity-temperature coefficient and the wax content; the crude oil non-Newtonian fluid viscosity-temperature coefficient prediction model is established according to a quantum particle swarm optimization algorithm and a generalized regression neural network;
the crude oil temperature set obtaining module is used for obtaining a crude oil temperature set; the crude oil temperature set comprises different temperature values;
the crude oil Newtonian fluid temperature set selection module is used for selecting a temperature value which is larger than the temperature value of the abnormal point of the crude oil from the crude oil temperature set as a crude oil Newtonian fluid temperature set;
the crude oil non-Newtonian fluid temperature set selection module is used for selecting a temperature value smaller than or equal to the crude oil abnormal point temperature value from the crude oil temperature set as a crude oil non-Newtonian fluid temperature set;
the crude oil Newtonian fluid temperature interval viscosity-temperature curve determining module is used for determining a crude oil Newtonian fluid temperature interval viscosity-temperature curve according to the crude oil Newtonian fluid temperature set and the Newtonian fluid state viscosity-temperature coefficient;
And the crude oil non-Newtonian fluid temperature interval viscosity-temperature curve determining module is used for determining a crude oil non-Newtonian fluid temperature interval viscosity-temperature curve according to the crude oil non-Newtonian fluid temperature set and the non-Newtonian fluid state viscosity-temperature coefficient.
Optionally, the crude oil newton fluid viscosity temperature coefficient prediction model obtaining process specifically includes:
the crude oil sample density value set and crude oil sample condensation point value set acquisition unit is used for acquiring the crude oil sample density value set and the crude oil sample condensation point value set; the density value set of the crude oil samples comprises a density value of each crude oil sample; the set of the condensation point values of the crude oil samples comprises the condensation point value of each crude oil sample;
the first root mean square error obtaining unit is used for inputting the crude oil sample density value set and the crude oil sample condensation point value set into a first generalized regression neural network to obtain a first root mean square error;
the first judging unit is used for judging whether the first iteration times are smaller than the first preset iteration times or not to obtain a first judging result;
the first parameter adjusting unit is configured to adjust the super parameter in the first generalized regression neural network by using a quantum particle swarm optimization algorithm if the first judgment result indicates that the first iteration number is smaller than a first preset iteration number, update the adjusted generalized regression neural network to be the first generalized regression neural network, and return to the first root mean square error obtaining unit;
And the crude oil Newton fluid viscosity-temperature coefficient prediction model determining unit is used for selecting a first generalized regression neural network corresponding to the minimum first root mean square error value as a crude oil Newton fluid viscosity-temperature coefficient prediction model if the first judgment result indicates that the first iteration number is larger than or equal to the first preset iteration number.
Optionally, the crude oil non-newtonian fluid viscosity temperature coefficient prediction model obtaining process specifically includes:
the crude oil sample Newtonian fluid state viscosity-temperature coefficient set obtaining unit is used for inputting the crude oil sample density value set and the crude oil sample condensation point value set into the crude oil Newtonian fluid viscosity-temperature coefficient prediction model to obtain a crude oil sample Newtonian fluid state viscosity-temperature coefficient set;
the crude oil sample wax content collection acquisition unit is used for acquiring a crude oil sample wax content collection; the set of crude oil sample wax content includes a wax content value for each crude oil sample;
the second root mean square error obtaining unit is used for inputting the crude oil sample Newtonian fluid state viscosity-temperature coefficient set and the crude oil sample wax content set into a second generalized regression neural network to obtain a second root mean square error;
the second judging unit is used for judging whether the second iteration times are smaller than second preset iteration times or not to obtain a second judging result;
The second parameter adjusting unit is configured to adjust the super parameter in the second generalized regression neural network by using a quantum particle swarm optimization algorithm if the second judgment result indicates that the second iteration number is smaller than the second preset iteration number, update the adjusted generalized regression neural network to be the second generalized regression neural network, and return to the second root mean square error obtaining unit;
and the crude oil non-Newtonian fluid viscosity-temperature coefficient prediction model determining unit is used for selecting a second generalized regression neural network corresponding to the minimum second root mean square error value as the crude oil non-Newtonian fluid viscosity-temperature coefficient prediction model if the second judgment result indicates that the second iteration number is larger than or equal to the second preset iteration number.
Optionally, the module for determining the viscosity-temperature curve of the temperature interval of the crude oil newtonian fluid specifically includes:
a viscosity-temperature curve determining unit for temperature interval of crude oil Newtonian fluid, which is used for determining the viscosity-temperature curve according to the formula mu 1 ＝aT 1 2 +bT 1 +c determining a viscous temperature curve of the crude oil Newtonian fluid temperature interval;
wherein mu 1 At a temperature T 1 Viscosity, a, b and c are viscosity-temperature coefficients, T, of Newtonian fluid states 1 Is the temperature value in the crude oil Newtonian fluid temperature set.
Optionally, the module for determining the viscosity-temperature curve of the temperature interval of the crude oil non-newtonian fluid specifically includes:
The viscosity-temperature curve determining unit for the temperature interval of the crude oil non-Newtonian fluid is used for determining the viscosity-temperature curve according to the formula
Figure BDA0002615012570000071
Determining a viscosity-temperature curve of a non-Newtonian fluid temperature interval of crude oil;
wherein mu 2 At a temperature T 2 The viscosity at the time of the formation of the gel,
Figure BDA0002615012570000072
n′＝fT 2 +g, d, e, f, g are viscosity-temperature coefficients of non-Newtonian fluid states, K is a consistency coefficient, n' is a rheological behavior index, gamma is a shear rate, T 2 Is the temperature value in the temperature set of the crude oil non-Newtonian fluid.
According to the specific embodiment provided by the invention, the invention discloses the following technical effects:
the invention provides a method and a system for determining a full viscosity temperature curve of crude oil, which relate the basic physical density, the condensation point and the wax content of the crude oil to a viscosity-temperature curve of the crude oil, establish a viscosity-temperature coefficient prediction model of a Newtonian fluid of the crude oil and a viscosity-temperature coefficient prediction model of a non-Newtonian fluid of the crude oil according to a quantum particle swarm optimization algorithm and a generalized regression neural network, determine the viscosity-temperature coefficient according to the basic physical property of the crude oil by utilizing the models, and finally determine the full viscosity-temperature curve according to the viscosity-temperature coefficient and the temperature of the crude oil, thereby solving the technical problem of difficult field acquisition of viscosity in the transportation and storage process of a crude oil pipeline and ensuring the transportation and storage safety of the crude oil pipeline.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions of the prior art, the drawings that are needed in the embodiments will be briefly described below, it being obvious that the drawings in the following description are only some embodiments of the present invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of a method for determining a full viscosity temperature curve of crude oil according to an embodiment of the present invention;
FIG. 2 is a graph showing the importance of the basic physical properties of crude oil according to the second embodiment of the present invention;
FIG. 3 is a schematic diagram of a generalized recurrent neural network according to a second embodiment of the present invention;
FIG. 4 is a flowchart of a procedure for obtaining a hyper-parameter in a generalized recurrent neural network according to a second embodiment of the present invention;
fig. 5 is a schematic structural diagram of a system for determining a full viscosity temperature curve of crude oil according to a third embodiment of the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
The invention aims to provide a method and a system for determining a full viscosity temperature curve of crude oil, which fully consider the basic physical properties of the crude oil, respectively treat the viscosity of the crude oil in a non-Newtonian fluid temperature range and a Newtonian fluid temperature range to obtain the full viscosity temperature curve of the crude oil, and ensure the safety of the pipeline transportation and storage of the crude oil.
In order that the above-recited objects, features and advantages of the present invention will become more readily apparent, a more particular description of the invention will be rendered by reference to the appended drawings and appended detailed description.
Example 1
Fig. 1 is a flowchart of a method for determining a full viscosity temperature curve of crude oil according to an embodiment of the present invention, as shown in fig. 1, the method for determining a full viscosity temperature curve of crude oil according to the present invention includes:
s101, obtaining a density value and a condensation point value of crude oil.
S102, determining a Newtonian fluid state viscosity-temperature coefficient by adopting a crude oil Newtonian fluid viscosity-temperature coefficient prediction model according to the density value and the condensation point value; the crude oil Newtonian fluid viscosity-temperature coefficient prediction model is established according to a quantum particle swarm optimization algorithm and a generalized regression neural network.
S103, obtaining the wax content of the crude oil.
S104, determining the viscosity-temperature coefficient of the non-Newtonian fluid state by adopting a crude oil non-Newtonian fluid viscosity-temperature coefficient prediction model according to the viscosity-temperature coefficient of the Newtonian fluid state and the wax content; the crude oil non-Newtonian fluid viscosity-temperature coefficient prediction model is established according to a quantum particle swarm optimization algorithm and a generalized regression neural network.
S105, acquiring a crude oil temperature set; the crude oil temperature set includes different temperature values. Specific: for a crude oil, its viscosity changes with temperature when it is in newtonian fluid state, that is, a temperature corresponds to a viscosity; when in a non-newtonian fluid state, the viscosity varies not only with temperature, but also with shear rate, that is, one temperature, one shear rate corresponds to one viscosity.
S106, selecting a temperature value larger than the temperature value of the abnormal point of the crude oil from the crude oil temperature set as the crude oil Newtonian fluid temperature set.
S107, selecting a temperature value smaller than or equal to the temperature value of the abnormal point of the crude oil from the crude oil temperature set as a crude oil non-Newtonian fluid temperature set.
S108, determining a viscosity-temperature curve of the temperature interval of the crude oil Newtonian fluid according to the temperature set of the crude oil Newtonian fluid and the viscosity-temperature coefficient of the state of the Newtonian fluid. Specifically, according to formula μ 1 ＝aT 1 2 +bT 1 +c determining a viscous temperature curve of the crude oil Newtonian fluid temperature interval; wherein mu 1 At a temperature T 1 Viscosity, a, b and c are viscosity-temperature coefficients, T, of Newtonian fluid states 1 Is the temperature value in the crude oil Newtonian fluid temperature set.
S109, determining a viscosity-temperature curve of the temperature interval of the crude oil non-Newtonian fluid according to the temperature set of the crude oil non-Newtonian fluid and the viscosity-temperature coefficient of the state of the non-Newtonian fluid. Specifically, according to the formula
Figure BDA0002615012570000091
Determining a viscosity-temperature curve of a non-Newtonian fluid temperature interval of crude oil; wherein mu 2 At a temperature T 2 Viscosity at time>
Figure BDA0002615012570000092
n′＝fT 2 +g, d, e, f, g are viscosity-temperature coefficients of non-Newtonian fluid states, K is a consistency coefficient, n' is a rheological behavior index, gamma is a shear rate, T 2 Is the temperature value in the temperature set of the crude oil non-Newtonian fluid. />
The obtaining process of the viscosity-temperature coefficient prediction model of the crude oil Newtonian fluid in S102 specifically comprises the following steps:
step 201, acquiring a crude oil sample density value set and a crude oil sample condensation point value set; the density value set of the crude oil samples comprises a density value of each crude oil sample; the set of crude oil sample condensation point values includes the condensation point value of each crude oil sample.
Step 202, inputting the set of crude oil sample density values and the set of crude oil sample condensation point values into a first generalized regression neural network to obtain a first root mean square error.
Step 203, determining whether the first iteration number is smaller than a first preset iteration number, so as to obtain a first determination result.
Step 204, if the first judgment result indicates that the first iteration number is smaller than the first preset iteration number, adjusting the super-parameters in the first generalized regression neural network by adopting a quantum particle swarm optimization algorithm, updating the adjusted generalized regression neural network to be the first generalized regression neural network, and returning to step 202
Step 205, if the first determination result indicates that the first iteration number is greater than or equal to the first preset iteration number, selecting a first generalized regression neural network corresponding to the minimum first root mean square error value as a crude oil newton fluid viscosity-temperature coefficient prediction model.
The obtaining process of the viscosity-temperature coefficient prediction model of the crude oil non-Newtonian fluid in the S104 specifically comprises the following steps:
and step 401, inputting the set of crude oil sample density values and the set of crude oil sample condensation point values into the crude oil Newtonian fluid viscosity-temperature coefficient prediction model to obtain a set of crude oil sample Newtonian fluid state viscosity-temperature coefficients.
Step 402, obtaining a crude oil sample wax content set; the set of crude sample wax content includes a wax content value for each crude sample.
And step 403, inputting the set of viscosity-temperature coefficients of the Newtonian fluid state of the crude oil sample and the set of wax content of the crude oil sample into a second generalized regression neural network to obtain a second root mean square error.
Step 404, determining whether the second iteration number is smaller than the second preset iteration number, to obtain a second determination result.
And step 405, if the second judgment result indicates that the second iteration number is smaller than the second preset iteration number, adjusting the super-parameters in the second generalized regression neural network by adopting a quantum particle swarm optimization algorithm, updating the adjusted generalized regression neural network into the second generalized regression neural network, and returning to step 403.
Step 406, if the second determination result indicates that the second iteration number is greater than or equal to the second preset iteration number, selecting a second generalized regression neural network corresponding to the minimum second root mean square error value as the crude oil non-newton fluid viscosity-temperature coefficient prediction model.
Example two
In order to achieve the above object of the present invention, the present invention provides a second embodiment.
1. Temperature coefficient calculation
For waxy crude oil, a heating and conveying process is generally adopted to carry out long-distance pipeline conveying on the waxy crude oil. According to national standard GB50253-2014, when heating and conveying are required, the conveying temperature of crude oil at each point along a pipeline is preferably 3-5 ℃ higher than the condensation point of the crude oil. Thus, the oil delivery temperature may be below the normal point of the crude oil, i.e., the crude oil is delivered in a non-newtonian fluid regime. Therefore, for crude oil transportation, the transportation temperature t can be divided into two sections, see formula 4-1:
Figure BDA0002615012570000101
wherein t is R Is the temperature of the crude oil when going out, t F Is the abnormal point temperature of crude oil, t L Is the temperature at which the crude oil enters the station.
1) In the newtonian fluid temperature regime:
when a waxy crude oil is in a Newtonian fluid state, the viscosity of the waxy crude oil is only related to the temperature, and a second order polynomial fitting can be used to obtain the relationship between the viscosity of the crude oil and the temperature, as shown in the formula 4-2:
μ 1 ＝aT 1 2 +bT 1 +c (4-2)
Thus, three parameters a, b, c can be defined as the viscosity temperature coefficient of waxy crude oil in newtonian fluid state as a measure of the viscosity of crude oil in the newtonian fluid regime. For selected crude oils, after determining the three parameters a, b, c, the viscosity at any temperature (temperature above the abnormal point) can be calculated. The viscosity-temperature relation curve of the temperature interval of the Newtonian fluid of the selected crude oil can be obtained.
2) Within the non-newtonian fluid temperature regime:
considering that the lowest oil transportation temperature is still 3-5 ℃ above the condensation point of crude oil and the shear rate of the oil transportation is not high, and the convenience of a power law equation in calculation, engineering commonly applies the power law equation to describe the rheological equation of the crude oil in the temperature region of the non-Newtonian fluid, and the formula is shown in the formula 4-3:
Figure BDA0002615012570000111
wherein, the consistency coefficient K and the rheological behavior index n' can be used to represent the change of the apparent viscosity of the non-Newtonian fluid at a certain temperature along with the shear rate, and the formulas 4-4 and 4-5 are as follows:
Figure BDA0002615012570000112
n′＝fT 2 +g (4-5)
according to the change trend of K, n ' along with the temperature, a change formula of K, n ' along with the temperature is fitted, and K, n ' values at different temperatures can be obtained only by obtaining four parameters, namely d, e, f and g, so that apparent viscosity of the non-Newtonian fluid at different temperatures and different shear rates can be calculated. The relation 4-6 can be obtained:
Figure BDA0002615012570000113
Thus, four parameters d, e, f, g can be defined as viscosity temperature coefficients for waxy crude oil in a non-newtonian fluid state as a measure of viscosity for waxy crude oil in a non-newtonian fluid state.
In summary, the full viscosity-temperature curve suitable for the transportation of waxy crude oil can be obtained by completely predicting a, b, c, d, e, f and g7 parameter values.
The prediction of the viscosity of the waxy crude oil is converted into the prediction of the viscosity-temperature coefficient, so that the output value of a prediction model is reduced, the connection between the characteristic vector and the target value is facilitated to be established, and feasibility is provided for the prediction of the full viscosity-temperature curve.
2. Feature selection
The importance degree of the characteristics such as moisture, condensation point, mechanical impurities, density, sulfur content, crude oil wax content and net heat to viscosity-temperature coefficient prediction is calculated through an XGboost model, the importance ordering of the characteristics is shown in fig. 2, it can be seen from fig. 2 that the influence of the density and the condensation point on the viscosity of the crude oil is maximum, and in order to reduce the dimension of an input characteristic vector as much as possible, therefore, the density and the viscosity are selected as the input characteristic vector.
When a waxy crude oil is in a non-newtonian fluid state, the main factor affecting its viscosity is the wax content, and therefore, the viscosity temperature coefficients a, b, c and wax content characterizing its newtonian fluid state are taken as input eigenvectors.
3. Crude oil viscosity-temperature coefficient prediction
The generalized regression neural network (Generalized Regression Neural Network, GRNN) is proposed by Spicht, is a branch of a radial basis network, has a structure similar to that of the radial basis network, is only slightly different in the output linear layer, is very suitable for function approximation, has better performance in the prediction of small sample data, can approximate any nonlinear function, and overcomes the problem of local minima. The GRNN is structurally composed of four layers, an input layer, a mode layer, a summing layer, and an output layer, respectively, as shown in fig. 3. The number of neurons of the input layer is equal to the dimension of the input vector in the learning sample, and each neuron is a simple distribution unit and directly transmits the input variable to the mode layer. The number of neurons of the pattern layer is equal to the number of learning samples, and each neuron corresponds to a different sample. Two types of neurons are used in the summation layer for summation. The number of neurons in the output layer is equal to the dimension k of the output vector in the learning sample, each neuron divides the output of the summing layer, and the output of element j corresponds to the j-th element of the estimation result Y (X).
And constructing a crude oil Newtonian fluid viscosity temperature coefficient prediction model and a crude oil non-Newtonian fluid viscosity temperature coefficient prediction model. The input characteristic of the crude oil Newtonian fluid viscosity-temperature coefficient prediction model is the density and the condensation point of waxy crude oil, and the target value is Newtonian fluid viscosity-temperature coefficients a, b and c; the input of the crude oil non-Newtonian fluid viscosity temperature coefficient prediction model is characterized in that the output values a, b, c and the wax content of the crude oil non-Newtonian fluid viscosity temperature coefficient prediction model are the viscosity temperature coefficients d, e, f and g of the non-Newtonian fluid.
And calling a crude oil physical property data set in the crude oil physical property database, and sorting the crude oil physical property data set to be a sample data set of a model, wherein 80% of the sample data is used as a training set, and 20% is used as a test set.
The prediction model acquisition process is as follows:
(1) Crude oil Newtonian fluid viscosity-temperature coefficient prediction model:
1) Input layer: the crude oil density and the condensation point are used as characteristic vectors to be input into an input layer, and the input characteristic vector data structure is as follows:
Figure BDA0002615012570000121
wherein->
Figure BDA0002615012570000122
For the density value of the ith crude oil sample, +.>
Figure BDA0002615012570000123
For the congeal point value of the ith crude sample, L is the total number of crude samples, i.e., L different crude oils.
2) Mode layer: and calculating the value of a Gaussian (Gauss) function of each sample in the test sample and the training sample at the mode layer, wherein the number of nodes is equal to the number of the training samples. Ith test sample tex i And the jth training sample trx i The Gauss function between them takes the value:
Figure BDA0002615012570000131
where δ is the hyper-parameter of the generalized recurrent neural network.
3) Summation layer: the number of the summing layer nodes is equal to the dimension of the output sample plus 1, and the dimension of the output sample of the model is 3, so that the number of the summing layer nodes is 4. The summation layer is divided into two parts, wherein the first node is the arithmetic sum of the mode layer outputs, and the outputs of the other k nodes are the weighted sum of the mode layer outputs. Assume that for a test sample tex, the output of the pattern layer is
{g 1 ,g 2 ,g 3 ,…,g m }
The output of the first node of the summing layer is:
Figure BDA0002615012570000132
the outputs of the remaining k nodes are:
Figure BDA0002615012570000133
wherein the weighting coefficient y ij And j elements of the label of the training sample corresponding to the j-th mode layer node.
4) Output layer: the number of output layer nodes is equal to the dimension of the tag vector, and the output of each node is equal to the output of the corresponding summation layer divided by the output of the first node of the summation layer. The output value data structure of the model is as follows:
Figure BDA0002615012570000134
wherein a is i ，b i ，c i The viscosity-temperature coefficient of the Newtonian fluid region corresponding to the ith test sample crude oil is given, and n is the capacity of the test sample.
The super parameter delta of the GRNN plays a key role in the prediction accuracy of the model, and the super parameter tuning of the GRNN model is performed by using a Quantum Particle Swarm (QPSO) algorithm, as shown in fig. 4, and the specific steps are as follows:
step 1, defining an initial position vector of a vector particle swarm optimization algorithm (Quantumparticle swarm optimization, QPSO) as a super-parameter delta in GRNN.
And 2, taking the root mean square error obtained by the GRNN as an adaptability function of the QPSO, and taking a position vector which enables the adaptability function value to be minimum as a global position vector.
And step 3, if the number of QPSO iterations meets the set maximum value, outputting the global position vector as a final super-parameter delta.
And 4, if the number of QPSO iterations does not meet the set maximum value, redefining an initial position vector of the population, and returning to the step 2.
(2) Crude oil non-Newtonian fluid viscosity-temperature coefficient prediction model:
1) Input layer: the output a, b, c and wax content of the crude oil Newtonian fluid viscosity-temperature coefficient prediction model are used as characteristic vectors to be input into an input layer, and the input characteristic vector data structure is as follows:
Figure BDA0002615012570000141
wherein->
Figure BDA0002615012570000142
The values of the viscosity-temperature coefficients a, b and c of the ith sample are respectively +.>
Figure BDA0002615012570000143
The wax content of the ith sample is given, and L is the total sample amount.
2) Mode layer: and calculating the Gauss function value of each sample in the test sample and the training sample at the mode layer, wherein the number of nodes is equal to the number of the training samples. Ith test sample tex i With the j-th training sample tex i The Gauss function between them takes the value:
Figure BDA0002615012570000144
wherein delta is a hyper-parameter of the generalized recurrent neural network.
3) Summation layer: the number of the summing layer nodes is equal to the dimension of the output sample plus 1, and the dimension of the output sample of the model is 4, so that the number of the summing layer nodes is 5. The summation layer is divided into two parts, wherein the first node is the arithmetic sum of the mode layer outputs, and the outputs of the other k nodes are the weighted sum of the mode layer outputs. Assume that for a test sample tex, the output of the pattern layer is
{g 1 ,g 2 ,g 3 ,…,g m }
The output of the first node of the summing layer is:
Figure BDA0002615012570000145
the outputs of the remaining k nodes are:
Figure BDA0002615012570000146
wherein the weighting coefficient y ij And j elements of the label of the training sample corresponding to the j-th mode layer node.
4) Output layer: the number of output layer nodes is equal to the dimension of the tag vector, and the output of each node is equal to the output of the corresponding summation layer divided by the output of the first node of the summation layer. The output value data structure of the model is as follows:
Figure BDA0002615012570000151
wherein d i ，e i ，f i ，g i The viscosity-temperature coefficient of the non-Newtonian fluid region corresponding to the ith test sample crude oil is given, and n is the capacity of the test sample.
The super parameter delta of the GRNN plays a key role in the prediction accuracy of the model, and the super parameter tuning of the GRNN model is performed by using a Quantum Particle Swarm (QPSO) algorithm, as shown in fig. 4, and the specific steps are as follows:
step 1, defining an initial position vector of a vector particle swarm optimization algorithm (Quantumparticle swarm optimization, QPSO) as a super-parameter delta in GRNN.
And 2, taking the root mean square error obtained by the GRNN as an adaptability function of the QPSO, and taking a position vector which enables the adaptability function value to be minimum as a global position vector.
And step 3, if the number of QPSO iterations meets the set maximum value, outputting the global position vector as a final super-parameter delta.
And 4, if the number of QPSO iterations does not meet the set maximum value, redefining an initial position vector of the population, and returning to the step 2.
Inputting the density and the condensation point of the crude oil to be detected into a crude oil Newtonian fluid viscosity temperature coefficient prediction model to obtain viscosity temperature coefficients a, b and c of the crude oil to be detected in a Newtonian fluid region; and inputting the viscosity-temperature coefficients a, b, c and the wax content of the oil sample to be measured in the Newtonian fluid region into a crude oil non-Newtonian fluid viscosity-temperature coefficient prediction model to obtain the viscosity-temperature coefficients d, e, f and g of the crude oil to be measured in the non-Newtonian fluid region. Thus, the complete viscosity-temperature coefficient of the oil sample to be measured can be obtained, and the complete viscosity-temperature curve of the oil sample to be measured can be obtained according to formulas 4-2, 4-3, 4-4 and 4-5.
Example III
The invention also provides a system for determining the full viscosity temperature curve of crude oil, as shown in fig. 5, the system for determining the full viscosity temperature curve comprises:
the crude oil density value and condensation point value acquisition module 1 is used for acquiring the density value and condensation point value of crude oil.
The Newtonian fluid state viscosity-temperature coefficient determining module 2 is used for determining the Newtonian fluid state viscosity-temperature coefficient by adopting a crude oil Newtonian fluid viscosity-temperature coefficient prediction model according to the density value and the condensation point value; the crude oil Newtonian fluid viscosity-temperature coefficient prediction model is established according to a quantum particle swarm optimization algorithm and a generalized regression neural network.
And the crude oil wax content acquisition module 3 is used for acquiring the wax content of the crude oil.
The non-Newtonian fluid state viscosity-temperature coefficient determining module 4 is used for determining the non-Newtonian fluid state viscosity-temperature coefficient by adopting a crude oil non-Newtonian fluid viscosity-temperature coefficient prediction model according to the Newtonian fluid state viscosity-temperature coefficient and the wax content; the crude oil non-Newtonian fluid viscosity-temperature coefficient prediction model is established according to a quantum particle swarm optimization algorithm and a generalized regression neural network.
The crude oil temperature set obtaining module 5 is used for obtaining a crude oil temperature set; the crude oil temperature set includes different temperature values.
And the crude oil Newtonian fluid temperature set selection module 6 is used for selecting a temperature value which is larger than the abnormal point temperature value of the crude oil from the crude oil temperature set as the crude oil Newtonian fluid temperature set.
And the crude oil non-Newtonian fluid temperature set selection module 7 is used for selecting a temperature value smaller than or equal to the crude oil abnormal point temperature value from the crude oil temperature set as the crude oil non-Newtonian fluid temperature set.
And the crude oil Newtonian fluid temperature interval viscosity-temperature curve determining module 8 is used for determining a crude oil Newtonian fluid temperature interval viscosity-temperature curve according to the crude oil Newtonian fluid temperature set and the Newtonian fluid state viscosity-temperature coefficient.
The viscosity-temperature curve determining module 9 of the temperature interval of the crude oil non-Newtonian fluid is used for determining the viscosity-temperature curve of the temperature interval of the crude oil non-Newtonian fluid according to the temperature set of the crude oil non-Newtonian fluid and the viscosity-temperature coefficient of the state of the non-Newtonian fluid.
Preferably, the crude oil newtonian fluid viscosity temperature coefficient prediction model obtaining process specifically includes:
the crude oil sample density value set and crude oil sample condensation point value set acquisition unit is used for acquiring the crude oil sample density value set and the crude oil sample condensation point value set; the density value set of the crude oil samples comprises a density value of each crude oil sample; the set of crude oil sample condensation point values includes the condensation point value of each crude oil sample.
The first root mean square error obtaining unit is used for inputting the crude oil sample density value set and the crude oil sample condensation point value set into a first generalized regression neural network to obtain a first root mean square error.
The first judging unit is used for judging whether the first iteration times are smaller than the first preset iteration times or not to obtain a first judging result.
And the first parameter adjusting unit is used for adjusting the super parameters in the first generalized regression neural network by adopting a quantum particle swarm optimization algorithm if the first judgment result indicates that the first iteration number is smaller than the first preset iteration number, updating the adjusted generalized regression neural network into the first generalized regression neural network, and returning to the first root mean square error obtaining unit.
And the crude oil Newton fluid viscosity-temperature coefficient prediction model determining unit is used for selecting a first generalized regression neural network corresponding to the minimum first root mean square error value as a crude oil Newton fluid viscosity-temperature coefficient prediction model if the first judgment result indicates that the first iteration number is larger than or equal to the first preset iteration number.
Preferably, the crude oil non-newtonian fluid viscosity temperature coefficient prediction model obtaining process specifically includes:
the crude oil sample Newtonian fluid state viscosity-temperature coefficient set obtaining unit is used for inputting the crude oil sample density value set and the crude oil sample condensation point value set into the crude oil Newtonian fluid viscosity-temperature coefficient prediction model to obtain a crude oil sample Newtonian fluid state viscosity-temperature coefficient set.
The crude oil sample wax content collection acquisition unit is used for acquiring a crude oil sample wax content collection; the set of crude sample wax content includes a wax content value for each crude sample.
And the second root mean square error obtaining unit is used for inputting the crude oil sample Newtonian fluid state viscosity-temperature coefficient set and the crude oil sample wax content set into a second generalized regression neural network to obtain a second root mean square error.
The second judging unit is used for judging whether the second iteration number is smaller than the second preset iteration number or not, and obtaining a second judging result.
And the second parameter adjusting unit is used for adjusting the super parameters in the second generalized regression neural network by adopting a quantum particle swarm optimization algorithm if the second judgment result indicates that the second iteration number is smaller than the second preset iteration number, updating the adjusted generalized regression neural network into the second generalized regression neural network, and returning to the second root mean square error obtaining unit.
And the crude oil non-Newtonian fluid viscosity-temperature coefficient prediction model determining unit is used for selecting a second generalized regression neural network corresponding to the minimum second root mean square error value as the crude oil non-Newtonian fluid viscosity-temperature coefficient prediction model if the second judgment result indicates that the second iteration number is larger than or equal to the second preset iteration number.
Preferably, the module for determining the viscosity-temperature curve of the temperature interval of the crude oil newtonian fluid specifically comprises:
a viscosity-temperature curve determining unit for temperature interval of crude oil Newtonian fluid, which is used for determining the viscosity-temperature curve according to the formula mu 1 ＝aT 1 2 +bT 1 +c determining a viscous temperature curve of the crude oil Newtonian fluid temperature interval; wherein mu 1 At a temperature T 1 Viscosity, a, b and c are viscosity-temperature coefficients, T, of Newtonian fluid states 1 Is the temperature value in the crude oil Newtonian fluid temperature set.
Preferably, the module for determining the viscosity-temperature curve of the temperature interval of the crude oil non-newtonian fluid specifically comprises:
Determination unit for viscosity-temperature curve of crude oil non-Newtonian fluid temperature intervalA cell for according to the formula
Figure BDA0002615012570000171
Determining a viscosity-temperature curve of a non-Newtonian fluid temperature interval of crude oil; wherein mu 2 At a temperature T 2 Viscosity at time>
Figure BDA0002615012570000172
n′＝fT 2 +g, d, e, f, g are viscosity-temperature coefficients of non-Newtonian fluid states, K is a consistency coefficient, n' is a rheological behavior index, gamma is a shear rate, T 2 Is the temperature value in the temperature set of the crude oil non-Newtonian fluid.
The invention provides a method and a system for determining a full viscosity temperature curve of crude oil, which relate the basic physical density, the condensation point and the wax content of the crude oil to a viscosity-temperature curve of the crude oil, establish a viscosity temperature coefficient prediction model of a Newtonian fluid of the crude oil and a viscosity temperature coefficient prediction model of a non-Newtonian fluid of the crude oil according to a quantum particle swarm optimization algorithm and a generalized regression neural network, determine the viscosity temperature coefficient according to the basic physical property of the crude oil by utilizing the model, and finally determine the full viscosity temperature curve according to the viscosity temperature coefficient and the temperature of the crude oil, thereby solving the technical problem of difficult field acquisition of viscosity in the transportation and storage process of a crude oil pipeline, ensuring the transportation and storage safety of the crude oil pipeline, and belongs to the field of petroleum and natural gas engineering-oil gas storage and transportation engineering. And by the method and the system, the viscosity-temperature curve of unknown crude oil within-20 ℃ to 100 ℃ can be predicted.
In the present specification, each embodiment is described in a progressive manner, and each embodiment is mainly described in a different point from other embodiments, and identical and similar parts between the embodiments are all enough to refer to each other.
The principles and embodiments of the present invention have been described herein with reference to specific examples, the description of which is intended only to assist in understanding the methods of the present invention and the core ideas thereof; also, it is within the scope of the present invention to be modified by those of ordinary skill in the art in light of the present teachings. In view of the foregoing, this description should not be construed as limiting the invention.

Claims (10)

1. A method for determining a full viscosity temperature curve of crude oil, comprising:
acquiring a density value and a condensation point value of crude oil;
according to the density value and the condensation point value, a crude oil Newtonian fluid viscosity-temperature coefficient prediction model is adopted to determine a Newtonian fluid state viscosity-temperature coefficient; the crude oil Newtonian fluid viscosity-temperature coefficient prediction model is established according to a quantum particle swarm optimization algorithm and a generalized regression neural network;
obtaining the wax content of crude oil;
according to the Newtonian fluid state viscosity-temperature coefficient and the wax content, a crude oil non-Newtonian fluid viscosity-temperature coefficient prediction model is adopted to determine a non-Newtonian fluid state viscosity-temperature coefficient; the crude oil non-Newtonian fluid viscosity-temperature coefficient prediction model is established according to a quantum particle swarm optimization algorithm and a generalized regression neural network;
Acquiring a crude oil temperature set; the crude oil temperature set comprises different temperature values;
selecting a temperature value which is larger than the abnormal point temperature value of the crude oil from the crude oil temperature set as a crude oil Newtonian fluid temperature set;
selecting a temperature value less than or equal to the temperature value of the abnormal point of the crude oil from the crude oil temperature set as a crude oil non-Newtonian fluid temperature set;
determining a viscosity-temperature curve of a temperature interval of the crude oil Newtonian fluid according to the crude oil Newtonian fluid temperature set and the viscosity-temperature coefficient of the state of the Newtonian fluid;
and determining a viscosity-temperature curve of the temperature interval of the crude oil non-Newtonian fluid according to the temperature set of the crude oil non-Newtonian fluid and the viscosity-temperature coefficient of the state of the non-Newtonian fluid.
2. The method for determining the full viscosity temperature curve of crude oil according to claim 1, wherein the crude oil newton fluid viscosity temperature coefficient prediction model obtaining process specifically comprises:
acquiring a crude oil sample density value set and a crude oil sample condensation point value set; the density value set of the crude oil samples comprises a density value of each crude oil sample; the set of the condensation point values of the crude oil samples comprises the condensation point value of each crude oil sample;
inputting the crude oil sample density value set and the crude oil sample condensation point value set into a first generalized regression neural network to obtain a first root mean square error;
Judging whether the first iteration times are smaller than the first preset iteration times or not to obtain a first judgment result;
if the first judgment result indicates that the first iteration number is smaller than the first preset iteration number, adopting a quantum particle swarm optimization algorithm to adjust the super parameters in the first generalized regression neural network, updating the adjusted generalized regression neural network into the first generalized regression neural network, and returning to the step of inputting the crude oil sample density value set and the crude oil sample condensation point value set into the first generalized regression neural network to obtain a first root mean square error;
and if the first judgment result indicates that the first iteration times are larger than or equal to the first preset iteration times, selecting a first generalized regression neural network corresponding to the first root mean square error value with the minimum value as a crude oil Newtonian fluid viscosity-temperature coefficient prediction model.
3. The method for determining the full viscosity temperature curve of crude oil according to claim 2, wherein the crude oil non-newtonian fluid viscosity temperature coefficient prediction model obtaining process specifically comprises:
inputting the crude oil sample density value set and the crude oil sample condensation point value set into the crude oil Newtonian fluid viscosity-temperature coefficient prediction model to obtain a crude oil sample Newtonian fluid state viscosity-temperature coefficient set;
Acquiring a crude oil sample wax content set; the set of crude oil sample wax content includes a wax content value for each crude oil sample;
inputting the crude oil sample Newtonian fluid state viscosity-temperature coefficient set and the crude oil sample wax content set into a second generalized regression neural network to obtain a second root mean square error;
judging whether the second iteration times are smaller than second preset iteration times or not to obtain a second judgment result;
if the second judgment result indicates that the second iteration number is smaller than the second preset iteration number, adopting a quantum particle swarm optimization algorithm to adjust the super-parameters in the second generalized regression neural network, updating the adjusted generalized regression neural network into the second generalized regression neural network, and returning to the step of inputting the crude oil sample Newtonian fluid state viscosity-temperature coefficient set and the crude oil sample wax content set into the second generalized regression neural network to obtain a second root mean square error;
and if the second judgment result indicates that the second iteration times are larger than or equal to the second preset iteration times, selecting a second generalized regression neural network corresponding to the minimum second root mean square error value as a crude oil non-Newtonian fluid viscosity-temperature coefficient prediction model.
4. The method for determining a full viscosity temperature curve of crude oil according to claim 1, wherein the determining a viscosity temperature curve of a crude oil newtonian fluid temperature interval according to the set of crude oil newtonian fluid temperatures and the newtonian fluid state viscosity temperature coefficient specifically comprises:
according to formula mu 1 ＝aT 1 2 +bT 1 +c determining a viscous temperature curve of the crude oil Newtonian fluid temperature interval;
wherein mu 1 At a temperature T 1 Viscosity, a, b and c are viscosity-temperature coefficients, T, of Newtonian fluid states 1 Is the temperature value in the crude oil Newtonian fluid temperature set.
5. The method for determining a full viscosity temperature curve of crude oil according to claim 1, wherein the determining a viscosity temperature curve of a non-newtonian fluid temperature interval of crude oil according to the set of non-newtonian fluid temperatures of crude oil and the viscosity temperature coefficient of non-newtonian fluid state specifically comprises:
according to the formula
Figure FDA0002615012560000031
Determination ofA viscosity-temperature curve of the crude oil non-Newtonian fluid in a temperature interval;
wherein mu 2 At a temperature T 2 The viscosity at the time of the formation of the gel,
Figure FDA0002615012560000032
n′＝fT 2 +g, d, e, f, g are viscosity-temperature coefficients of non-Newtonian fluid states, K is a consistency coefficient, n' is a rheological behavior index, gamma is a shear rate, T 2 Is the temperature value in the temperature set of the crude oil non-Newtonian fluid.
6. A system for determining a full viscosity temperature profile of crude oil, comprising:
The crude oil density value and condensation point value acquisition module is used for acquiring the density value and condensation point value of crude oil;
the Newtonian fluid state viscosity-temperature coefficient determining module is used for determining the Newtonian fluid state viscosity-temperature coefficient by adopting a crude oil Newtonian fluid viscosity-temperature coefficient prediction model according to the density value and the condensation point value; the crude oil Newtonian fluid viscosity-temperature coefficient prediction model is established according to a quantum particle swarm optimization algorithm and a generalized regression neural network;
the crude oil wax content acquisition module is used for acquiring the wax content of crude oil;
the non-Newtonian fluid state viscosity-temperature coefficient determination module is used for determining the non-Newtonian fluid state viscosity-temperature coefficient by adopting a crude oil non-Newtonian fluid viscosity-temperature coefficient prediction model according to the Newtonian fluid state viscosity-temperature coefficient and the wax content; the crude oil non-Newtonian fluid viscosity-temperature coefficient prediction model is established according to a quantum particle swarm optimization algorithm and a generalized regression neural network;
the crude oil temperature set obtaining module is used for obtaining a crude oil temperature set; the crude oil temperature set comprises different temperature values;
the crude oil Newtonian fluid temperature set selection module is used for selecting a temperature value which is larger than the temperature value of the abnormal point of the crude oil from the crude oil temperature set as a crude oil Newtonian fluid temperature set;
The crude oil non-Newtonian fluid temperature set selection module is used for selecting a temperature value smaller than or equal to the crude oil abnormal point temperature value from the crude oil temperature set as a crude oil non-Newtonian fluid temperature set;
the crude oil Newtonian fluid temperature interval viscosity-temperature curve determining module is used for determining a crude oil Newtonian fluid temperature interval viscosity-temperature curve according to the crude oil Newtonian fluid temperature set and the Newtonian fluid state viscosity-temperature coefficient;
and the crude oil non-Newtonian fluid temperature interval viscosity-temperature curve determining module is used for determining a crude oil non-Newtonian fluid temperature interval viscosity-temperature curve according to the crude oil non-Newtonian fluid temperature set and the non-Newtonian fluid state viscosity-temperature coefficient.
7. The system for determining the full viscosity temperature curve of crude oil according to claim 6, wherein the crude oil newtonian fluid viscosity temperature coefficient prediction model obtaining process specifically comprises:
the crude oil sample density value set and crude oil sample condensation point value set acquisition unit is used for acquiring the crude oil sample density value set and the crude oil sample condensation point value set; the density value set of the crude oil samples comprises a density value of each crude oil sample; the set of the condensation point values of the crude oil samples comprises the condensation point value of each crude oil sample;
The first root mean square error obtaining unit is used for inputting the crude oil sample density value set and the crude oil sample condensation point value set into a first generalized regression neural network to obtain a first root mean square error;
the first judging unit is used for judging whether the first iteration times are smaller than the first preset iteration times or not to obtain a first judging result;
the first parameter adjusting unit is configured to adjust the super parameter in the first generalized regression neural network by using a quantum particle swarm optimization algorithm if the first judgment result indicates that the first iteration number is smaller than a first preset iteration number, update the adjusted generalized regression neural network to be the first generalized regression neural network, and return to the first root mean square error obtaining unit;
and the crude oil Newton fluid viscosity-temperature coefficient prediction model determining unit is used for selecting a first generalized regression neural network corresponding to the minimum first root mean square error value as a crude oil Newton fluid viscosity-temperature coefficient prediction model if the first judgment result indicates that the first iteration number is larger than or equal to the first preset iteration number.
8. The system for determining the full viscosity temperature curve of crude oil according to claim 7, wherein the process for obtaining the crude oil non-newtonian fluid viscosity temperature coefficient prediction model specifically comprises:
The crude oil sample Newtonian fluid state viscosity-temperature coefficient set obtaining unit is used for inputting the crude oil sample density value set and the crude oil sample condensation point value set into the crude oil Newtonian fluid viscosity-temperature coefficient prediction model to obtain a crude oil sample Newtonian fluid state viscosity-temperature coefficient set;
the crude oil sample wax content collection acquisition unit is used for acquiring a crude oil sample wax content collection; the set of crude oil sample wax content includes a wax content value for each crude oil sample;
the second root mean square error obtaining unit is used for inputting the crude oil sample Newtonian fluid state viscosity-temperature coefficient set and the crude oil sample wax content set into a second generalized regression neural network to obtain a second root mean square error;
the second judging unit is used for judging whether the second iteration times are smaller than second preset iteration times or not to obtain a second judging result;
the second parameter adjusting unit is configured to adjust the super parameter in the second generalized regression neural network by using a quantum particle swarm optimization algorithm if the second judgment result indicates that the second iteration number is smaller than the second preset iteration number, update the adjusted generalized regression neural network to be the second generalized regression neural network, and return to the second root mean square error obtaining unit;
And the crude oil non-Newtonian fluid viscosity-temperature coefficient prediction model determining unit is used for selecting a second generalized regression neural network corresponding to the minimum second root mean square error value as the crude oil non-Newtonian fluid viscosity-temperature coefficient prediction model if the second judgment result indicates that the second iteration number is larger than or equal to the second preset iteration number.
9. The system for determining the full viscosity temperature curve of crude oil according to claim 6, wherein the module for determining the viscosity temperature curve of the newtonian fluid temperature interval of crude oil specifically comprises:
a viscosity-temperature curve determining unit for temperature interval of crude oil Newtonian fluid, which is used for determining the viscosity-temperature curve according to the formula mu 1 ＝aT 1 2 +bT 1 +c determining a viscous temperature curve of the crude oil Newtonian fluid temperature interval;
wherein mu 1 At a temperature T 1 Viscosity, a, b and c are viscosity-temperature coefficients, T, of Newtonian fluid states 1 Is the temperature value in the crude oil Newtonian fluid temperature set.
10. The system for determining a full viscosity temperature curve of crude oil according to claim 6, wherein the module for determining a viscosity temperature curve of a non-newtonian fluid of crude oil specifically comprises:
the viscosity-temperature curve determining unit for the temperature interval of the crude oil non-Newtonian fluid is used for determining the viscosity-temperature curve according to the formula
Figure FDA0002615012560000051
Determining a viscosity-temperature curve of a non-Newtonian fluid temperature interval of crude oil;
Wherein mu 2 At a temperature T 2 The viscosity at the time of the formation of the gel,
Figure FDA0002615012560000052
n′＝fT 2 +g, d, e, f, g are viscosity-temperature coefficients of non-Newtonian fluid states, K is a consistency coefficient, n' is a rheological behavior index, gamma is a shear rate, T 2 Is the temperature value in the temperature set of the crude oil non-Newtonian fluid. />
CN202010766855.0A 2020-08-03 2020-08-03 Method and system for determining full viscosity temperature curve of crude oil Active CN111898088B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010766855.0A CN111898088B (en) 2020-08-03 2020-08-03 Method and system for determining full viscosity temperature curve of crude oil

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010766855.0A CN111898088B (en) 2020-08-03 2020-08-03 Method and system for determining full viscosity temperature curve of crude oil

Publications (2)

Publication Number Publication Date
CN111898088A CN111898088A (en) 2020-11-06
CN111898088B true CN111898088B (en) 2023-05-26

Family

ID=73183117

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010766855.0A Active CN111898088B (en) 2020-08-03 2020-08-03 Method and system for determining full viscosity temperature curve of crude oil

Country Status (1)

Country Link
CN (1) CN111898088B (en)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114898818B (en) * 2022-04-06 2024-06-18 中国石油大学(北京) Mixed crude oil condensation point prediction model training method, device and application method
CN115270651B (en) * 2022-06-20 2024-03-15 北京科技大学 Monocular video-oriented non-Newtonian fluid simulation reconstruction method

Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110377879A (en) * 2019-07-17 2019-10-25 中国石油化工股份有限公司 A kind of calculation method of the defeated radius of mixed transportation of oil and gas pipeline room temperature collection

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2018031021A1 (en) * 2016-08-11 2018-02-15 Halliburton Energy Services, Inc. Drilling fluid contamination determination for downhole fluid sampling tool
AU2017424316A1 (en) * 2017-07-21 2020-01-02 Landmark Graphics Corporation Deep learning based reservoir modeling

Patent Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110377879A (en) * 2019-07-17 2019-10-25 中国石油化工股份有限公司 A kind of calculation method of the defeated radius of mixed transportation of oil and gas pipeline room temperature collection

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
基于混沌量子粒子群BP算法的研究;何晓云;许江淳;陈文绪;;自动化仪表(第12期);全文 *

Also Published As

Publication number Publication date
CN111898088A (en) 2020-11-06

Similar Documents

Publication Publication Date Title
CN111898088B (en) Method and system for determining full viscosity temperature curve of crude oil
CN109472110B (en) Method for predicting residual service life of aeroengine based on LSTM network and ARIMA model
Chelgani et al. Prediction of coal grindability based on petrography, proximate and ultimate analysis using multiple regression and artificial neural network models
Laory et al. Methodologies for predicting natural frequency variation of a suspension bridge
CN103389472B (en) A kind of Forecasting Methodology of the cycle life of lithium ion battery based on ND-AR model
CN104820781B (en) Consider that temperature follows the BGA thermal fatigue life of solder joint Forecasting Methodologies of load sequence loading effect
Ghasemi et al. Principal component analysis-based predictive modeling and optimization of permanent deformation in asphalt pavement: elimination of correlated inputs and extrapolation in modeling
Behbahani et al. Investigation of wax precipitation in crude oil: Experimental and modeling
Korres et al. A neural network approach to the prediction of diesel fuel lubricity
Farzaneh-Gord et al. Accurate determination of natural gas compressibility factor by measuring temperature, pressure and Joule-Thomson coefficient: Artificial neural network approach
Mo et al. Few-shot RUL estimation based on model-agnostic meta-learning
Li et al. An integrated method for evaluating and predicting long-term operation safety of concrete dams considering lag effect
Lin et al. A novel efficient model for gas compressibility factor based on GMDH network
Zhang et al. Application of ANN to predict the apparent viscosity of waxy crude oil
Fallahi et al. Leakage detection in water distribution networks using hybrid feedforward artificial neural networks
Sulaimon et al. New two-phase and three-phase thermodynamic models for predicting wax precipitation in hydrocarbon mixtures
CN114912364A (en) Natural gas well flow prediction method, device, equipment and computer readable medium
CN113469409A (en) Gaussian process-based state prediction method and device for electric energy metering device
CN112580254A (en) Multi-dimensional analysis and early warning method for converter transformer based on multi-level decision fusion
Yao et al. Prediction of wax deposits for crude pipelines using time-dependent data mining
Voitov Structural identification of the mathematical model of the functioning of tribosystems under conditions of boundary lubrication
Baldo et al. Artificial neural network prediction of airport pavement moduli using interpolated surface deflection data
Kozłowski et al. Conveyor Belts Joints Remaining Life Time Forecasting with the Use of Monitoring Data and Mathematical Modelling
Moçayd et al. Rheology data-driven machine learning models for phosphate slurry pipeline in Morocco
Song et al. A Hybrid Supervised Deep Learning and Nonlinear Finite Element Framework for Efficient Fatigue Life Predictions of Rotary Shouldered Threaded Connections

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant