CN116822999B - Method and system for predicting monitoring density of oil product of oil mixing interface of finished oil pipeline - Google Patents
Method and system for predicting monitoring density of oil product of oil mixing interface of finished oil pipeline Download PDFInfo
- Publication number
- CN116822999B CN116822999B CN202311107413.5A CN202311107413A CN116822999B CN 116822999 B CN116822999 B CN 116822999B CN 202311107413 A CN202311107413 A CN 202311107413A CN 116822999 B CN116822999 B CN 116822999B
- Authority
- CN
- China
- Prior art keywords
- oil
- downstream
- monitoring
- upstream
- density
- 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
Links
- 238000012544 monitoring process Methods 0.000 title claims abstract description 181
- 238000002156 mixing Methods 0.000 title claims abstract description 52
- 238000000034 method Methods 0.000 title claims abstract description 33
- 238000011144 upstream manufacturing Methods 0.000 claims abstract description 111
- 238000009826 distribution Methods 0.000 claims description 25
- 238000012549 training Methods 0.000 claims description 24
- 238000012937 correction Methods 0.000 claims description 22
- 239000000203 mixture Substances 0.000 claims description 16
- 238000004364 calculation method Methods 0.000 claims description 13
- 238000012545 processing Methods 0.000 abstract description 3
- 239000003921 oil Substances 0.000 description 238
- 238000005520 cutting process Methods 0.000 description 8
- 230000000875 corresponding effect Effects 0.000 description 7
- 230000008569 process Effects 0.000 description 6
- 238000012512 characterization method Methods 0.000 description 5
- 230000014509 gene expression Effects 0.000 description 5
- 239000011159 matrix material Substances 0.000 description 5
- 238000010586 diagram Methods 0.000 description 4
- 230000006870 function Effects 0.000 description 4
- 238000005259 measurement Methods 0.000 description 4
- 239000003209 petroleum derivative Substances 0.000 description 4
- 239000002283 diesel fuel Substances 0.000 description 3
- 238000004519 manufacturing process Methods 0.000 description 3
- 238000012935 Averaging Methods 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 2
- 230000008901 benefit Effects 0.000 description 2
- 238000003066 decision tree Methods 0.000 description 2
- 238000013508 migration Methods 0.000 description 2
- 230000005012 migration Effects 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 239000013598 vector Substances 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000001276 controlling effect Effects 0.000 description 1
- 230000002596 correlated effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000017105 transposition Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N9/00—Investigating density or specific gravity of materials; Analysing materials by determining density or specific gravity
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
- G06F18/254—Fusion techniques of classification results, e.g. of results related to same input data
- G06F18/256—Fusion techniques of classification results, e.g. of results related to same input data of results relating to different input data, e.g. multimodal recognition
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/27—Regression, e.g. linear or logistic regression
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0637—Strategic management or analysis, e.g. setting a goal or target of an organisation; Planning actions based on goals; Analysis or evaluation of effectiveness of goals
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/02—Agriculture; Fishing; Forestry; Mining
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Business, Economics & Management (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Human Resources & Organizations (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Strategic Management (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Evolutionary Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Computational Biology (AREA)
- Economics (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Health & Medical Sciences (AREA)
- General Business, Economics & Management (AREA)
- Marketing (AREA)
- Tourism & Hospitality (AREA)
- General Health & Medical Sciences (AREA)
- Educational Administration (AREA)
- Entrepreneurship & Innovation (AREA)
- Mining & Mineral Resources (AREA)
- Analytical Chemistry (AREA)
- Animal Husbandry (AREA)
- Marine Sciences & Fisheries (AREA)
- Game Theory and Decision Science (AREA)
- Primary Health Care (AREA)
- Biochemistry (AREA)
- Agronomy & Crop Science (AREA)
- Pathology (AREA)
- Development Economics (AREA)
- Quality & Reliability (AREA)
- Immunology (AREA)
- Operations Research (AREA)
- Chemical & Material Sciences (AREA)
- Production Of Liquid Hydrocarbon Mixture For Refining Petroleum (AREA)
- Pipeline Systems (AREA)
Abstract
The invention belongs to the technical field of data processing, and provides a method and a system for predicting the monitoring density of a backward oil product in a mixed oil interface of a finished oil pipeline, which are used for solving the problem that the monitoring density of the backward oil product cannot be reliably predicted; and combining the obtained monitoring density value of the upstream station yard of the oil product of the finished oil pipeline to obtain a prediction result. The monitoring density of the oil product after the oil mixing interface can be accurately predicted.
Description
Technical Field
The invention belongs to the technical field of data processing, and particularly relates to a method and a system for predicting the monitoring density of a subsequent oil product of an oil mixing interface of a finished oil pipeline.
Background
The statements in this section merely provide background information related to the present disclosure and may not necessarily constitute prior art.
For economic reasons, multiple finished oils are typically continuously transported in the same finished oil pipeline in a batch sequence. In the process of pipe transportation, an oil mixing interface is generated between adjacent batches, and although a densimeter installed in a station is difficult to obtain the actual density value of the forward and backward oil products of the oil mixing interface, the condition that the actual feedback monitoring density of the oil mixing interface changes along with time is core data of an operator for processing the oil mixing interface. Specifically, a field generally converts a time-varying curve of a density monitoring value of a mixed oil interface into a time-varying curve of a mixed oil concentration distribution, and when the mixed oil concentration value is reduced to a certain threshold value, a batch cutting method is adopted to treat the mixed oil interface. However, the densimeter can only sense the current monitoring value of the density of the oil mixing interface of the station, and cannot predict the monitoring value of the density of the pure post-running oil which does not arrive at the station. If the monitoring value of the density of the oil product after the control can be accurately controlled, key data support can be provided for accurately guiding the on-site batch cutting work of the oil product. However, the hydraulic thermodynamic conditions in the pipe conveying process are complex and changeable, and the monitoring density of the afterflow oil product cannot be accurately obtained based on the existing oil product density calculation formula; hardware equipment generally faces zero drift phenomenon, so that the monitored density values of the same oil products in upstream and downstream stations are obviously different, and batch cutting work of the oil products is plagued. In addition, the frequent change of the pipe transportation working condition causes the data to have multi-mode characteristics, if the data-driven modeling method is directly adopted to predict the density of the backward oil product, the model is easy to fall into a fitting error area, and a reliable prediction result of the monitoring density of the backward oil product cannot be provided.
Disclosure of Invention
In order to overcome the defects of the prior art, the invention provides a method and a system for predicting the monitoring density of the backward oil product of the oil mixing interface of a finished oil pipeline, which are used for carrying out multi-mode identification on the data of the finished oil pipeline through a Gaussian mixture regression algorithm, and a maximum expected algorithm is combined with a multi-mode identification result to carry out training to obtain a correction model for the difference value of the monitoring density of the upstream and downstream stations of the predicted backward oil product, so that the monitoring density of the backward oil product of the oil mixing interface can be accurately predicted.
In order to achieve the above object, a first aspect of the present invention provides a method for predicting a monitoring density of a petroleum product at a petroleum product mixing interface of a petroleum product pipeline, comprising:
acquiring historical data of an oil mixing interface of a finished oil pipeline;
the hydraulic power thermal information of the upstream and downstream yards of the finished oil pipeline in the historical data is selected, the upstream yard monitoring density of the forward and backward oil products and the upstream and downstream yard monitoring density difference of the forward oil products are used as key input characteristic variables, the upstream and downstream yard monitoring density difference of the backward oil products is selected as output variables, and the key input characteristic variables and the output variables are subjected to multi-mode identification by means of Gaussian mixture regression algorithm, so that mode identification results under all modes are obtained;
training by using a maximum expected algorithm according to the modal identification result to obtain a model for correcting the monitoring density difference of the upstream and downstream yards of the backward oil product corresponding to the mode;
and predicting the oil mixing interface of the finished oil pipeline to be predicted by using a downstream station monitoring density difference correction model of the downstream oil product, and combining the obtained upstream station monitoring density value of the downstream oil product of the finished oil pipeline to obtain the downstream oil monitoring density value of the downstream station of the finished oil pipeline to be predicted.
The second aspect of the invention provides a system for predicting the monitoring density of a subsequent oil product of an oil mixing interface of a finished oil pipeline, comprising:
the acquisition module is used for: acquiring historical data of an oil mixing interface of a finished oil pipeline;
and a Gaussian module: the hydraulic power thermal information of the upstream and downstream yards of the finished oil pipeline in the historical data is selected, the upstream yard monitoring density of the forward and backward oil products and the upstream and downstream yard monitoring density difference of the forward oil products are used as key input characteristic variables, the upstream and downstream yard monitoring density difference of the backward oil products is selected as output variables, and the key input characteristic variables and the output variables are subjected to multi-mode identification by means of Gaussian mixture regression algorithm, so that mode identification results under all modes are obtained;
training module: training by using a maximum expected algorithm according to the modal identification result to obtain a model for correcting the monitoring density difference of the upstream and downstream yards of the backward oil product corresponding to the mode;
and a prediction module: and predicting the oil mixing interface of the finished oil pipeline to be predicted by using a downstream station monitoring density difference correction model of the downstream oil product, and combining the obtained upstream station monitoring density value of the downstream oil product of the finished oil pipeline to obtain the downstream oil monitoring density value of the downstream station of the finished oil pipeline to be predicted.
The one or more of the above technical solutions have the following beneficial effects:
in the invention, historical data of a finished oil pipeline is obtained, hydraulic thermal information of an upstream station and a downstream station is selected from the historical data, upstream station monitoring density of a forward oil product and downstream station monitoring density difference of the forward oil product are used as key input characteristic variables, the upstream station monitoring density difference and the downstream station monitoring density difference of the forward oil product are used as output variables, a Gaussian regression algorithm is utilized for modal identification, a model identification result is trained based on a maximum expected algorithm, a downstream station monitoring density difference correction model of the forward oil product is obtained, and the upstream station monitoring density difference and downstream station monitoring density difference of the backward oil product to be predicted are predicted by the obtained downstream station monitoring density difference correction model; the multi-mode characteristic is identified through the Gaussian regression algorithm, so that the accuracy of oil monitoring density prediction after the oil mixing interface is improved, and the method has important significance in guiding the on-site development of oil batch management.
Additional aspects of the invention will be set forth in part in the description which follows and, in part, will be obvious from the description, or may be learned by practice of the invention.
Drawings
The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description serve to explain the invention.
FIG. 1 is a schematic diagram of oil mixing interface measurement migration in accordance with an embodiment of the present invention;
FIG. 2 is a graph showing a typical density distribution of a mixed oil according to a first embodiment of the present invention;
FIG. 3 is a graph showing distortion of a mixed oil concentration distribution curve caused by a prediction deviation of a monitoring density of a backward oil product according to the first embodiment of the present invention;
FIG. 4 is a flowchart of a model for predicting the density monitoring value of a subsequent petroleum product according to an embodiment of the present invention;
FIG. 5 is a schematic diagram of a correction model for the density monitoring difference between upstream and downstream yards of a backward oil product according to an embodiment of the present invention.
Detailed Description
It should be noted that the following detailed description is exemplary and is intended to provide further explanation of the invention. Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
It is noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of exemplary embodiments according to the present invention.
Embodiments of the invention and features of the embodiments may be combined with each other without conflict.
Example 1
The embodiment discloses a method for predicting the monitoring density of a subsequent oil product of an oil mixing interface of a finished oil pipeline, which comprises the following steps:
acquiring historical data of an oil mixing interface of a finished oil pipeline;
the hydraulic power thermal information of the upstream and downstream yards of the finished oil pipeline in the historical data is selected, the upstream yard monitoring density of the forward and backward oil products and the upstream and downstream yard monitoring density difference of the forward oil products are used as key input characteristic variables, the upstream and downstream yard monitoring density difference of the backward oil products is selected as output variables, and the key input characteristic variables and the output variables are subjected to multi-mode identification by means of Gaussian mixture regression algorithm, so that mode identification results under all modes are obtained;
training by using a maximum expected algorithm according to the modal identification result to obtain a model for correcting the monitoring density difference of the upstream and downstream yards of the backward oil product corresponding to the mode;
and predicting the oil mixing interface of the finished oil pipeline to be predicted by using a downstream station monitoring density difference correction model of the downstream oil product, and combining the obtained upstream station monitoring density value of the downstream oil product of the finished oil pipeline to obtain the downstream oil monitoring density value of the downstream station of the finished oil pipeline to be predicted.
The method for predicting the monitoring density of the oil product of the oil mixing interface of the finished oil pipeline mainly comprises the following three parts: and (3) establishing an oil mixing database of an upstream and a downstream stations of the finished oil pipeline, and establishing a correction model of the monitoring density difference value of the upstream and the downstream stations of the post-production oil, and predicting the monitoring density of the post-production oil at the oil mixing interface.
(1) Establishing an oil mixing database of an upstream and a downstream station of a finished oil pipeline
The method is characterized in that an oil mixing database of an upstream station and a downstream station of a finished oil pipeline is established, based on upstream and downstream station measuring equipment, the hydraulic and thermal information of the pipeline and the monitoring density of the front and rear oil products are obtained, and the database is established, so that data support is provided for the subsequent establishment of a correction model of the difference value of the monitoring density of the rear oil products. The method comprises the steps of determining the outlet temperature and pressure of an upstream station yard, the inlet temperature and pressure of a downstream station yard, the monitoring density of oil products in the front and back of the upstream station yard and the monitoring density of oil products in the front and back of the downstream station yard in a certain period.
The method for determining the outlet temperature and the pressure of the upstream station is specifically as follows: collecting the temperature and pressure of the outlet of an upstream station of the finished oil pipeline in a certain period, wherein the temperature monitoring sample size is thatThe pressure monitoring sample size is->Averaging to obtain the time average temperature of the outlet of the upstream station>C, controlling the temperature; pressure equalizing force at time->Units: MPa:
（1）
（2）
wherein,the>Temperature data points, DEG C; />The>Pressure data points, MPa.
The inlet temperature and pressure of the downstream station site are determined specifically as follows: collecting the inlet temperature and pressure of a station downstream of a finished product pipeline within a certain period of time, wherein the temperature monitoring sample size is thatThe pressure monitoring sample size is->Averaging to obtain the time average temperature of the inlet of the downstream station yard>Units: the temperature is lower than the temperature; pressure equalizing force at time->Unit MPa:
（3）
（4）
wherein,the>Temperature data points, DEG C, ">The>Pressure data points, MPa.
Determining the difference value of the monitoring density of the forward oil obtained from the upstream and downstream stationsMonitoring density differences for a post-run oilThe method comprises the following steps: based on an upstream and downstream station densimeter, acquiring a forward oil density monitoring value +.>Units: kg/m3, upstream station downstream oil density monitoring value +.>Units: kg/m3; downstream station forward oil density monitoring value +.>Units: kg/m3; downstream station downstream oil density monitoring value +.>Units: kg/m3; calculating the upstream and downstream forward oil product monitoring density difference value +.>Difference between the density of the oil product and the upstream and downstream oil products>：
（5）
（6）
(2) Establishing a post-production oil upstream and downstream station yard monitoring density difference correction model
The method for establishing the downstream and upstream station yard monitoring density difference correction model of the afterflow oil product is based on the historical data of the existing pipe-conveying oil-mixing interface and combines a Gaussian mixture regression algorithm (Gaussian Mixture Regression Model, GMR) to form the afterflow oil product monitoring density difference correction model, and specifically comprises the following steps:
selecting a modal identification key feature variable: order theAnd->Respectively representing an input variable matrix and an output variable matrix of the oil mixing data, < >>For training set sample size, +.>For the matrix transposition operation,characterization of->The variable vectors are input by the training set samples.
To accurately explore multi-mode information of oil mixing data hiding of pipe transportation, temperatures acquired by upstream and downstream stations are selectedAnd (3) withPressure->And->Upstream station acquired forward oil monitoring density +.>Oil product for post-treatmentMonitoring Density->And the difference value of the monitoring density of the forward oil products acquired by the upstream and downstream stations>As a key input feature variable for identifying data multi-mode by the GMR algorithm.
The input variable is assumed to be defined by GMRA gaussian distribution. Input variable in->Distribution of (i.e.)>Edge probability Density function expression under individual modalities +.>The following are provided:
（7）
wherein,is Gaussian distribution->,/>Respectively refer to->Mean vector and covariance matrix in gaussian distribution under each mode.
Determining regression key input variables: monitoring density difference value of upstream and downstream stations of forward oil obtained from upstream and downstream stationsKey input variable of regression process of GMR algorithm>Because of the difference between the upstream and downstream stations of the backward oil productsFor a variable to be predicted, the relationship between the two can be expressed as:
+/>（8）
wherein,representation of->Gaussian white noise in each mode, subject to a mean of 0, variance +.>T represents the transpose operation. In GMR (GMR)>The individual modes are defined by the correspondence->Individual local model characterization,/->Is->Regression coefficients in the individual local models.
Training a post-oil product monitoring density difference correction model: based on existing tubing-to-oil interface history data, model parameters can be estimated in combination with a maximum expectation algorithm (Expectation Maximization, EM). Using hidden variablesCharacterization of the modality of each sample, when +.>Time indicate->The training set samples belong to->A modality. Calculate->Belonging to->Posterior probability of individual modalities->And define statistics->：
（9）
（10）
Wherein,for giving +.>Input/output characteristic information of each sample +.>The samples belong to->Strip of individual modesProbability of a piece; />Characterization of->The samples belong to->In the individual modes, the conditional probability of the output variable with respect to the input variable; />Is->The samples belong to->And (3) inputting the conditional probability distribution of the variable in each mode.
Based on the acquired posterior probabilityThe Gaussian distribution weight of each mode can be calculated>Mean->Covariance->The expression is as follows:
（11）
（12）
（13）
（14）
（15）
wherein,；/>is a diagonal matrix，/>。
The prediction of the monitoring density of the backward oil product of the oil mixing interface of the finished oil pipeline is to predict the monitoring density of the backward oil product, and the difference value of the monitoring density of the backward oil product is predicted based on the hydraulic thermal information and the monitoring density information of the oil product obtained by the measuring equipment of the upstream and downstream stationsThe monitoring density value of the backward oil product obtained by combining an upstream station yard is +.>Predicting downstream station monitoring density value of downstream oil product +.>The method specifically comprises the following steps:
determining key variable information of a sample to be predicted: in the first placeTaking samples to be predicted as examples, combining upstream and downstream measuring equipment to obtain upstream and downstream station site temperatures +.>And->Pressure->And->Upstream station acquired forward oil monitoring density +.>Monitoring Density of the afterrun oil>And the difference value of the monitoring density of the upstream and downstream stations of the obtained forward oil products>Modal recognition variable +.>Upstream and downstream station monitoring density difference of forward oil product +.>Regression variables in the composition GMR algorithm。
Predicting a post-run oil monitoring density difference: for->Samples to be predicted->，/>In order to be a known quantity,for the quantity to be predicted, its assignment can be calculated based on formula (16)>Posterior probability of individual modalities->：
（16）
Wherein,for giving +.>Input/output characteristic information of each sample +.>The samples belong to->Conditional probability of individual modality->Is->The samples belong to->Probability of individual modality>Is->The samples belong to->And (3) inputting the conditional probability distribution of the variable in each mode.
Accordingly, the predicted amount can be obtainedConditional probability expression +.>：
（17）
Wherein,given input variables, the quantity to be predicted +.>Conditional probability expressions of (2);for the output variable +.>The compliance mean is +.>Variance is->Is a gaussian distribution of (c).
Consider to beIs expected as a predictive result->Therefore, the following oil product monitoring density difference value +.>Predictive value->The expression is:
（18）
wherein,is->At a given +.>The following conditions are expected.
Predicting a post-run oil monitoring density value: downstream oil monitoring density value obtained in combination with upstream station yard +.>Predicting downstream station downstream oil product monitoring density value +.>：
（19）
FIG. 1 is a schematic diagram of oil blending interface measurement migration with multiple finished oil products pumped from an upstream manufacturer to a downstream customer on a finished oil pipeline by batch basis. In the process of passing the station of the oil mixing interface, the monitoring density of the subsequent oil product is accurately estimated, and the distribution information of the oil mixing concentration is determined, so that the method is a key premise for realizing the accurate cutting of oil product batches and improving the running economic benefit of a finished oil pipe network.
FIG. 2 shows a typical blended oil density profile where a yard-mounted density sensor first monitors the previous oil monitoring density timing information of a batch sequence followed by a gradual sensing of the subsequent oil density monitoring information as the blended oil interface passes through an upstream yard. The oil mixing concentration distribution information needs to be calculated based on the information of the monitoring density of the oil products in the front and the back, however, when the pipeline between two adjacent stations is moved, the downstream station can only sense the monitoring density of the oil products in the front. At present, a method for directly monitoring the density of the afterflow oil is not available, so that the on-site data are necessary to be combined, a high-precision prediction method for the monitoring density of the afterflow oil is established, and data support is provided for batch cutting work of the oil.
Fig. 3 shows the error of the information of the oil mixing concentration distribution caused by the prediction error of the monitoring density of the subsequent oil product, wherein δ represents the prediction error of the monitoring density of the subsequent oil product, and δ=0 represents the actual oil mixing concentration distribution curve. Along with the gradual increase of the prediction error of the monitoring density of the backward oil product, the deviation degree of the corresponding mixed oil concentration distribution curve and the true curve is more and more obvious: the oil mixing concentration distribution should be in the interval of [0,1], however, when the error occurs in the monitoring density prediction result of the subsequent oil product, the oil mixing concentration value may have negative number, and the calculation result violates the physical rule. In addition, if the value of the mixed oil concentration is defined as 1% as the oil batch cutting threshold point, when delta is increased from 0 to 1,3 and 5, the corresponding time deviation of the oil batch cutting threshold point for representing the mixed oil interface is 216s,450s and 624s respectively. The above results again illustrate the significance of accurate prediction of the monitoring density of the subsequent oil products to accurately perform the batch cutting operation of the oil products.
Fig. 4 shows a flow of a model for predicting the monitoring value of the density of the oil product after the backward flow, and the temperature, pressure and the density information of the oil product are respectively obtained based on measuring equipment installed in an upstream station and a downstream station. It should be noted that for accurate modeling, the upstream station densimeter can know the batch sequence of the oil products after monitoring the forward and backward oil product density, such as pushing diesel oil by gasoline, and abbreviated as gasoline-diesel interface; and pushing gasoline by diesel oil, namely a diesel oil interface for short, and the like, so as to establish a corresponding backward oil density monitoring difference correction model.
FIG. 5 shows a schematic diagram of a model for correcting the difference value of the density monitoring of the oil product after the oil product is subjected to the backward motion; reasons for the accurate prediction of the monitoring density of the aftermarket oil product include: the density of diesel and gasoline varies with temperature and pressure, and complex operating conditions result in a highly nonlinear relationship between density and hydrothermal, i.e., pressure temperature, data; hardware devices are commonly subject to zero drift, which means that even under similar temperature and pressure conditions, the upstream and downstream densitometers may get different measurements for the same oil. Therefore, if the monitoring density of the backward oil product is directly predicted during modeling, unreliable prediction results can be caused by inherent measurement errors of the hydraulic thermodynamic conditions and different hardware devices.
However, the preceding and following oils undergo similar hydrodynamic thermal processes during tubing and both use the same equipment upstream and downstream to measure their density information. Therefore, the difference value of the density of the upstream and downstream stations of the previous oil product is monitoredDifference between monitoring density and downstream station of backward oil product +.>And necessarily highly correlated. The functional relation between the capture of the data driving model is adopted, namely, the density difference value is monitored for the upstream and downstream stations of the forward oil product>Difference between monitoring density and downstream station of backward oil product +.>Modeling is carried out according to the dependency relationship between the two stations, and the difference value of the monitoring density of the upstream and downstream stations of the oil product after the correction is +.>Is a more reasonable modeling method.
In the process, a mode identification function is introduced by combining with a GMR algorithm, so that the model can monitor density difference values of upstream and downstream stations of the backward oil product under different mode conditionsDifference between monitoring density and upstream and downstream stations of forward oil product +.>And the functional relation between the two models is accurately modeled, so that the model prediction accuracy is ensured.
Equation (20) shows the theoretical equation of oil density as a function of temperature, whereinCharacterizing the standard density of the oil product at 20 DEG CDegree, in kg/m3. In order to demonstrate the effectiveness and superiority of the modeling method provided by the embodiment, a theoretical formula and two common machine learning algorithms are adopted, namely a gradient lifting decision tree (Gradient Boosted Decision Tree) and an artificial neural network algorithm (Artificial Neural network, ANN) and a prediction result obtained by directly predicting the monitoring density of the subsequent oil product by adopting a GMR algorithm are adopted. Furthermore, to evaluate model performance as comprehensively as possible, 121 samples were divided into training and prediction sets, with test set ratios stepped up from 0.3 to 0.6.
Tables 1 and 2 respectively show the accuracy of predicting the monitoring density of the backward oil products of each model based on the representation of Root Mean Square Error (RMSE) under the conditions of a gasoline-diesel interface and a diesel-gasoline interface, and the calculation formula of the error index is as follows, whereinAnd->Characterization of the +.>True and fit values for individual samples, +.>To test the sample size. The smaller the RMSE, the lower the overall error.
（20）
（21）
Table 2 prediction results for each model under the gasoline-diesel interface:
table 3 prediction results for each model under the diesel-gasoline interface:
as can be obtained from tables 2 and 3, the backward oil product monitoring density prediction method provided by the embodiment has obvious advantages, and proves that the method provided by the invention has important significance for improving the positioning of the oil mixing interface of the finished oil pipeline.
Example two
The purpose of this embodiment is to provide the oil product monitoring density prediction system of the oil mixing interface of finished oil pipeline afterward, include:
the system for predicting the monitoring density of the oil product of the oil mixing interface of the finished oil pipeline is characterized by comprising the following components:
the acquisition module is used for: acquiring historical data of an oil mixing interface of a finished oil pipeline;
and a Gaussian module: the hydraulic power thermal information of the upstream and downstream yards of the finished oil pipeline in the historical data is selected, the upstream yard monitoring density of the forward and backward oil products and the upstream and downstream yard monitoring density difference of the forward oil products are used as key input characteristic variables, the upstream and downstream yard monitoring density difference of the backward oil products is selected as output variables, and the key input characteristic variables and the output variables are subjected to multi-mode identification by means of Gaussian mixture regression algorithm, so that mode identification results under all modes are obtained;
training module: training by using a maximum expected algorithm according to the modal identification result to obtain a model for correcting the monitoring density difference of the upstream and downstream yards of the backward oil product corresponding to the mode;
and a prediction module: and predicting the oil mixing interface of the finished oil pipeline to be predicted by using a downstream station monitoring density difference correction model of the downstream oil product, and combining the obtained upstream station monitoring density value of the downstream oil product of the finished oil pipeline to obtain the downstream oil monitoring density value of the downstream station of the finished oil pipeline to be predicted.
In this embodiment, the training module includes:
a first calculation module: calculating posterior probability of modal classification of key input characteristic variables in each training sample;
a second calculation module: calculating the variance, mean and covariance of the Gaussian distribution of each mode based on the posterior probability of the mode sub-genus of the key feature variable in each acquired training sample;
a third calculation module: establishing a relation between key input characteristic variables and output variables based on the calculated variance of the Gaussian distribution;
a fourth calculation module: and establishing an edge probability density function of the key input characteristic variable based on the mean value and the covariance of the calculated Gaussian distribution.
In this embodiment, the prediction module includes:
and a determination module: taking the temperature and pressure of an upstream station and a downstream station of a finished product pipeline to be predicted, the monitoring density of the forward and backward oil products obtained by the upstream station as a modal identification variable in a Gaussian mixture regression algorithm, and taking the difference value of the monitoring density of the forward oil products obtained by the upstream station and the downstream station as a regression variable in the Gaussian mixture regression algorithm;
probability calculation module: calculating posterior probability of modal identification variable modal classification;
and a difference prediction module: and obtaining a post-oil product monitoring density difference predictive value based on the post-oil product monitoring density difference correction model, the posterior probability of modal identification variable modal classification and the regression variable.
In this embodiment, the prediction module further includes:
and an addition module: and adding the predicted value of the difference value of the monitoring density of the backward oil product to the obtained monitoring density value of the upstream station of the backward oil product of the finished oil pipeline to obtain the monitoring density value of the downstream station of the backward oil product of the finished oil pipeline to be predicted.
While the foregoing description of the embodiments of the present invention has been presented in conjunction with the drawings, it should be understood that it is not intended to limit the scope of the invention, but rather, it is intended to cover all modifications or variations within the scope of the invention as defined by the claims of the present invention.
Claims (3)
1. The method for predicting the monitoring density of the oil product of the oil mixing interface of the finished oil pipeline is characterized by comprising the following steps:
acquiring historical data of an oil mixing interface of a finished oil pipeline;
the hydraulic power thermal information of the upstream and downstream yards of the finished oil pipeline in the historical data is selected, the upstream yard monitoring density of the forward and backward oil products and the upstream and downstream yard monitoring density difference of the forward oil products are used as key input characteristic variables, the upstream and downstream yard monitoring density difference of the backward oil products is selected as output variables, and the key input characteristic variables and the output variables are subjected to multi-mode identification by means of Gaussian mixture regression algorithm, so that mode identification results under all modes are obtained;
training by using a maximum expected algorithm according to the modal identification result to obtain a model for correcting the monitoring density difference of the upstream and downstream yards of the backward oil product corresponding to the mode;
predicting the oil mixing interface of the finished oil pipeline to be predicted by using a downstream station monitoring density difference correction model of the downstream oil product, and combining the obtained upstream station monitoring density value of the downstream oil product of the finished oil pipeline to obtain a downstream oil monitoring density value of the downstream station of the finished oil pipeline to be predicted;
the hydraulic and thermal information of the upstream and downstream stations comprises the outlet temperature and pressure of the upstream stations and the inlet temperature and pressure of the downstream stations; determining a difference value of upstream and downstream station monitoring densities of the forward oil products based on the upstream station density monitoring value and the downstream station density monitoring value of the forward oil products of the finished oil pipeline; determining a downstream and upstream station monitoring density difference value of the downstream oil product based on the downstream station density monitoring value of the downstream oil product and the upstream station density monitoring value of the downstream oil product of the finished oil pipeline;
training by using a maximum expected algorithm according to a modal identification result to obtain a model for correcting the monitoring density difference of the upstream and downstream yards of the backward oil product corresponding to the modal, wherein the model specifically comprises the following steps:
calculating posterior probability of modal classification of key input characteristic variables in each training sample;
calculating the variance, mean and covariance of the Gaussian distribution of each mode based on the posterior probability of the mode sub-genus of the key feature variable in each acquired training sample;
based on the calculated variance of the Gaussian distribution, establishing a relation between a key input characteristic variable and an output variable;
establishing an edge probability density function of the key input characteristic variable based on the mean value and covariance of the calculated Gaussian distribution;
predicting an oil mixing interface of a finished oil pipeline to be predicted by using a backward oil monitoring density difference correction model, wherein the method specifically comprises the following steps:
taking the temperature and pressure of upstream and downstream stations of a finished oil pipeline to be predicted, the obtained upstream and downstream station monitoring density of the forward and backward oil product as a modal identification variable in a Gaussian mixture regression algorithm, and taking the obtained upstream and downstream station monitoring density difference value of the forward oil product as a regression variable in the Gaussian mixture regression algorithm;
calculating posterior probability of modal identification variable modal classification;
obtaining a predicted value of the monitoring density difference of the downstream station of the backward oil based on the backward oil monitoring density difference correction model, the posterior probability of modal identification variable modal classification and the regression variable;
and adding the predicted value of the difference value of the monitoring density of the backward oil product to the obtained monitoring density value of the upstream station of the backward oil product of the finished oil pipeline to obtain the monitoring density value of the downstream station of the backward oil product of the finished oil pipeline to be predicted.
2. The method for predicting the monitoring density of the downstream oil product of the oil mixing interface of the finished oil pipeline according to claim 1, wherein the obtained expectation of the monitoring density difference value of the downstream oil product station is taken as a predicted value of the monitoring density difference value of the downstream oil product station according to posterior probability and regression variables of modal identification variable modal classification.
3. The system for predicting the monitoring density of the oil product of the oil mixing interface of the finished oil pipeline is characterized by comprising the following components:
the acquisition module is used for: acquiring historical data of an oil mixing interface of a finished oil pipeline;
and a Gaussian module: the hydraulic power thermal information of the upstream and downstream yards of the finished oil pipeline in the historical data is selected, the upstream yard monitoring density of the forward and backward oil products and the upstream and downstream yard monitoring density difference of the forward oil products are used as key input characteristic variables, the upstream and downstream yard monitoring density difference of the backward oil products is selected as output variables, and the key input characteristic variables and the output variables are subjected to multi-mode identification by means of Gaussian mixture regression algorithm, so that mode identification results under all modes are obtained;
training module: training by using a maximum expected algorithm according to the modal identification result to obtain a model for correcting the monitoring density difference of the upstream and downstream yards of the backward oil product corresponding to the mode;
and a prediction module: predicting the oil mixing interface of the finished oil pipeline to be predicted by using a downstream station monitoring density difference correction model of the downstream oil product, and combining the obtained upstream station monitoring density value of the downstream oil product of the finished oil pipeline to obtain a downstream oil monitoring density value of the downstream station of the finished oil pipeline to be predicted;
the hydraulic and thermal information of the upstream and downstream stations comprises the outlet temperature and pressure of the upstream stations and the inlet temperature and pressure of the downstream stations; determining a difference value of upstream and downstream station monitoring densities of the forward oil products based on the upstream station density monitoring value and the downstream station density monitoring value of the forward oil products of the finished oil pipeline; determining a downstream and upstream station monitoring density difference value of the downstream oil product based on the downstream station density monitoring value of the downstream oil product and the upstream station density monitoring value of the downstream oil product of the finished oil pipeline;
the training module comprises:
a first calculation module: calculating posterior probability of modal classification of key input characteristic variables in each training sample;
a second calculation module: calculating the variance, mean and covariance of the Gaussian distribution of each mode based on the posterior probability of the mode sub-genus of the key feature variable in each acquired training sample;
a third calculation module: based on the calculated variance of the Gaussian distribution, establishing a relation between a key input characteristic variable and an output variable;
a fourth calculation module: establishing an edge probability density function of the key input characteristic variable based on the mean value and covariance of the calculated Gaussian distribution;
the prediction module comprises:
and a determination module: taking the temperature and pressure of an upstream station and a downstream station of a finished product pipeline to be predicted, the monitoring density of the forward and backward oil products obtained by the upstream station as a modal identification variable in a Gaussian mixture regression algorithm, and taking the difference value of the monitoring density of the forward oil products obtained by the upstream station and the downstream station as a regression variable in the Gaussian mixture regression algorithm;
probability calculation module: calculating posterior probability of modal identification variable modal classification;
and a difference prediction module: obtaining a predicted value of the monitoring density difference value of the backward oil based on the backward oil monitoring density difference value correction model, the posterior probability of modal identification variable modal classification and the regression variable;
the prediction module further includes: and an addition module: and adding the predicted value of the difference value of the monitoring density of the backward oil product to the obtained monitoring density value of the upstream station of the backward oil product of the finished oil pipeline to obtain the monitoring density value of the downstream station of the backward oil product of the finished oil pipeline to be predicted.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311107413.5A CN116822999B (en) | 2023-08-31 | 2023-08-31 | Method and system for predicting monitoring density of oil product of oil mixing interface of finished oil pipeline |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311107413.5A CN116822999B (en) | 2023-08-31 | 2023-08-31 | Method and system for predicting monitoring density of oil product of oil mixing interface of finished oil pipeline |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116822999A CN116822999A (en) | 2023-09-29 |
| CN116822999B true CN116822999B (en) | 2023-12-05 |
Family
ID=88127906
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311107413.5A Active CN116822999B (en) | 2023-08-31 | 2023-08-31 | Method and system for predicting monitoring density of oil product of oil mixing interface of finished oil pipeline |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116822999B (en) |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP2416153A1 (en) * | 2010-08-02 | 2012-02-08 | ENI S.p.A. | Method for predicting the quality and yields of a crude oil |
| RU2613385C1 (en) * | 2016-03-10 | 2017-03-16 | Федеральное государственное бюджетное образовательное учреждение высшего образования "Тихоокеанский государственный университет" | Automated control system of oil quality |
| WO2020261940A1 (en) * | 2019-06-27 | 2020-12-30 | 積水ポリマテック株式会社 | Battery module |
| CN112182885A (en) * | 2020-09-29 | 2021-01-05 | 中国民用航空飞行学院 | Fuel consumption deviation prediction method and system based on Gaussian mixture model |
| CN113196550A (en) * | 2018-07-30 | 2021-07-30 | 凯尊创新有限公司 | Housing for rechargeable battery |
| CN114861759A (en) * | 2022-04-06 | 2022-08-05 | 中国石油大学(华东) | Distributed training method of linear dynamic system model |
| CN116307303A (en) * | 2023-05-24 | 2023-06-23 | 中国石油大学(华东) | Mechanism-data dual-drive oil mixing length prediction method and system for finished oil pipeline |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| AU2014395111B2 (en) * | 2014-05-23 | 2017-06-22 | Landmark Graphics Corporation | Robust viscosity estimation methods and systems |
| US10628750B2 (en) * | 2014-11-17 | 2020-04-21 | Massachusetts Institute Of Technology | Systems and methods for improving petroleum fuels production |
-
2023
- 2023-08-31 CN CN202311107413.5A patent/CN116822999B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP2416153A1 (en) * | 2010-08-02 | 2012-02-08 | ENI S.p.A. | Method for predicting the quality and yields of a crude oil |
| RU2613385C1 (en) * | 2016-03-10 | 2017-03-16 | Федеральное государственное бюджетное образовательное учреждение высшего образования "Тихоокеанский государственный университет" | Automated control system of oil quality |
| CN113196550A (en) * | 2018-07-30 | 2021-07-30 | 凯尊创新有限公司 | Housing for rechargeable battery |
| WO2020261940A1 (en) * | 2019-06-27 | 2020-12-30 | 積水ポリマテック株式会社 | Battery module |
| CN113906611A (en) * | 2019-06-27 | 2022-01-07 | 积水保力马科技株式会社 | Battery module |
| CN112182885A (en) * | 2020-09-29 | 2021-01-05 | 中国民用航空飞行学院 | Fuel consumption deviation prediction method and system based on Gaussian mixture model |
| CN114861759A (en) * | 2022-04-06 | 2022-08-05 | 中国石油大学(华东) | Distributed training method of linear dynamic system model |
| CN116307303A (en) * | 2023-05-24 | 2023-06-23 | 中国石油大学(华东) | Mechanism-data dual-drive oil mixing length prediction method and system for finished oil pipeline |
Non-Patent Citations (3)
| Title |
|---|
| Automatic method for estimation of in situ effective contact angle from X-ray micro tomography images of two-phase flow in porous media;Alessio Scanziani等人;《Journal of Colloid and Interface Science》;全文 * |
| 成品油混输监测系统的关键技术问题及解决办法;黄懿雪;李国栋;郭长滨;;自动化博览(07);全文 * |
| 融合机制与高斯混合回归算法的成品油管道顺序输送混油长度预测模型;袁子云等;《中国石油大学学报（自然科学版）》;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116822999A (en) | 2023-09-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107153874B (en) | Water quality prediction method and system | |
| TWI521360B (en) | Metrology sampling method and computer program product thereof | |
| Fang et al. | Inverse Gaussian processes with correlated random effects for multivariate degradation modeling | |
| CN105929340B (en) | A method of battery SOC is estimated based on ARIMA | |
| CN107085750A (en) | A kind of mixing dynamic fault Forecasting Methodology based on ARMA and ANN | |
| Zhang et al. | Extended design cost: a performance criterion for real-time optimization systems | |
| Jiang et al. | On estimating quantile sensitivities via infinitesimal perturbation analysis | |
| Akpinar et al. | Naive forecasting of household natural gas consumption with sliding window approach | |
| CN116307303B (en) | Mechanism-data dual-drive oil mixing length prediction method and system for finished oil pipeline | |
| CN111143768A (en) | Air quality prediction algorithm based on ARIMA-SVM combined model | |
| Lin et al. | A novel efficient model for gas compressibility factor based on GMDH network | |
| Duan et al. | Bayesian analysis for the transformed exponential dispersion process with random effects | |
| CN114782099A (en) | Budget estimation and analysis method for construction cost | |
| CN116822999B (en) | Method and system for predicting monitoring density of oil product of oil mixing interface of finished oil pipeline | |
| Ioshchikhes et al. | Assessing energy efficiency measures for hydraulic systems using a digital twin | |
| Li et al. | An improved multiplicative acceptability consistency-driven group decision making with triangular fuzzy reciprocal preference relations | |
| Zhang et al. | An integrated framework of preventive maintenance and task scheduling for repairable multi-unit systems | |
| Hou et al. | Observing a three-tank system | |
| CN111241629B (en) | Intelligent prediction method for performance change trend of hydraulic pump of airplane based on data driving | |
| Pang et al. | RUL prediction for bivariate degradation process considering individual differences | |
| Gao et al. | Using transient measurements in iterative steady-state optimizing control | |
| Piplani et al. | Simplification strategies for simulation models of semiconductor facilities | |
| CN106651462A (en) | Evaluation method for second-hand device | |
| Salazar-Alvarez et al. | Regressing sample quantiles to perform nonparametric capability analysis | |
| Mauro et al. | Extreme values calculation of multi-modal peak distributions |
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 |