1. The utility model provides a finished product oil pipeline contaminated product section laser on-line measuring device which characterized in that: the system comprises a main pipeline (12), a detection device connecting pipe (13), a mixed oil output connecting pipe (9), a flange ball valve I (1), an ultraviolet laser transmitter (2), a detachable optical detection window (3), an ultraviolet laser receiver (4), an optical signal processor (5), an electromagnetic valve controller (6), a flange ball valve II (7), a lifting check valve (8), an electromagnetic valve I (10) and an electromagnetic valve II (11); installing a detection device connecting pipe (13) on a main pipeline (12) of product oil, installing a flange ball valve I (1) on the detection device connecting pipe (13), installing a detachable optical detection window (3) on the detection device connecting pipe (13), installing a lifting check valve (8) on the detection device connecting pipe (13), installing a flange ball valve II (7) on the detection device connecting pipe (13), installing an ultraviolet laser emitter (2) on one side of the detachable optical detection window (3), installing an ultraviolet laser receiver (4) on the other side of the detachable optical detection window (3), connecting the ultraviolet laser receiver (4) with an optical signal processor (5), connecting the other end of the optical signal processor (5) with an electromagnetic valve controller (6), connecting the electromagnetic valve controller (6) with an electromagnetic valve I (10) and an electromagnetic valve II (11), and connecting one end of the electromagnetic valve I (10) with a mixed oil output connecting pipe (9), the other end of the mixed oil output connecting pipe (9) is connected with the main pipeline (12); and a second electromagnetic valve (11) is arranged on the main pipeline (12).

2. The laser online detection device for the oil mixing section of the product oil conveying pipeline according to claim 1, characterized in that: the detection device is characterized in that a detachable optical window (3) is installed on the detection device connecting pipe (13), and an ultraviolet laser transmitter (2) and an ultraviolet laser receiver (4) are installed on two sides of the detachable optical window (3) respectively.

3. The laser online detection device for the oil mixing section of the product oil conveying pipeline according to claim 1, characterized in that: the flange ball valve I (1) and the flange ball valve II (7) of detection device connecting pipe (13) facial make-up are convenient for install and dismantle optical window, reduce simultaneously to have the oil to leak out in the pipeline, the over-and-under type check valve (8) of detection device connecting pipe (12) facial make-up are arranged in guaranteeing the directional flow of oil in the detection tube.

4. The laser online detection device for the oil mixing section of the product oil conveying pipeline according to claim 1, characterized in that: the ultraviolet wavelength adopted by the ultraviolet laser transmitter (2) is 480 nm.

5. The laser online detection device for the oil mixing section of the product oil conveying pipeline according to claim 1, characterized in that: one end of the electromagnetic valve controller (6) is connected with the optical signal processor (5), the other end of the electromagnetic valve controller is connected with the first electromagnetic valve (10) and the second electromagnetic valve (11), and the opening or closing of the first electromagnetic valve (10) and the second electromagnetic valve (11) is controlled through signals output by the optical signal processor (5).

6. The laser online detection device for the oil mixing section of the product oil conveying pipeline according to claim 1, characterized in that: the mixed oil output connecting pipe (9) is connected with the mixed oil storage tank, and can convey the mixed oil section of the finished oil to the mixed oil storage tank for centralized treatment.

Background

The transportation mode of the finished oil mainly comprises four transportation modes of a waterway, a railway, a highway and a pipeline. Compared with other transportation modes, the pipeline transportation has the advantages of strong terrain and climate adaptability, low oil product loss, reduced possibility of accidents, easy realization of automatic management and acceleration of turnover speed, and low comprehensive transportation cost, and the prior finished oil transportation mostly adopts a sequential transportation mode. The oil product is conveyed in sequence, namely, different types of oil products are conveyed through one pipeline according to a certain sequence, and a plurality of input points and output points can be arranged on the pipeline, so that the utilization rate of the pipeline is improved. In the sequential conveying pipeline, when two oil products are alternated, an oil mixing section is generated, and the oil mixing forming mode mainly comprises accidental oil mixing, local oil mixing and oil mixing along the line. Because of the uneven flow velocity caused by the pipe pressure, the two oil products are inevitably mixed with each other, and a longer oil mixing section is formed at the joint of the two oil products. Among them, the oil mixing along the line is the most important reason, and is mainly based on convection transfer and diffusion transfer.

The currently common oil-mixed interface detection technologies include a fluorescence labeling method, a tracer atom labeling method, a density type detection method, a capacitance type detection method and a sound wave type detection method. The fluorescence labeling method is safe and convenient, and the tracer generally does not lag behind, but the detection range is limited by considering that some oil products have fluorescence (such as heated diesel). The tracer atom marking method has high sensitivity and is convenient for accurately cutting oil products, but the method can generate radioactive pollution. The density type detection method detects interfaces of different oil products by using density, but has poor detection effect on two oil products with slightly different densities. The capacitance type detection method is convenient to operate and simple in equipment, but the difference of the dielectric constants of the oil products can bring great influence on the measurement precision. The sound wave type detection method can detect the interface of two oil products with small density difference, but if water and gas are contained in the conveyed oil product, the detection result can be influenced.

Disclosure of Invention

The invention aims to overcome the defects of the prior art, and provides the laser online detection device for the oil mixing section of the product oil conveying pipeline, which solves the problems that the existing oil mixing interface detection technology is difficult to carry out real-time detection, the detection precision is insufficient, the monitoring condition is limited and the like.

In order to solve the problems existing in the background technology, the invention adopts the following technical scheme: the system comprises a main pipeline, a detection device connecting pipe, a mixed oil output connecting pipe, a flange ball valve I, an ultraviolet laser emitter, a detachable optical detection window, an ultraviolet laser receiver, an optical signal processor, an electromagnetic valve controller, a flange ball valve II, a lifting check valve, an electromagnetic valve I and an electromagnetic valve II; installing a detection device connecting pipe on a main pipeline of the product oil, installing a flange ball valve I on the detection device connecting pipe, installing a detachable optical detection window on the detection device connecting pipe, installing a lifting check valve on the detection device connecting pipe, installing a flange ball valve II on the detection device connecting pipe, installing an ultraviolet laser emitter on one side of the detachable optical detection window, installing an ultraviolet laser receiver on the other side of the detachable optical detection window, connecting the ultraviolet laser receiver with an optical signal processor, connecting the other end of the optical signal processor with a solenoid valve controller, connecting the solenoid valve controller with the solenoid valve I and the solenoid valve II, connecting one end of the solenoid valve I with a mixed oil output connecting pipe, and connecting the other end of the mixed oil output connecting pipe with the main pipeline; and a second electromagnetic valve is installed on the main pipeline.

The detection device connecting pipe on install and to dismantle the optical window, can dismantle optical window both sides and install ultraviolet laser emitter and ultraviolet laser receiver respectively.

The flange ball valve I and the flange ball valve II which are arranged on the detection device connecting pipe are convenient for mounting and dismounting the optical window, oil loss in the pipeline is reduced, and the lifting check valve arranged on the detection device connecting pipe is used for ensuring the directional flow of oil in the detection pipe.

The ultraviolet wavelength adopted by the ultraviolet laser transmitter is 480 nm.

One end of the electromagnetic valve controller is connected with the optical signal processor, the other end of the electromagnetic valve controller is connected with the first electromagnetic valve and the second electromagnetic valve, and the opening or closing of the first electromagnetic valve and the second electromagnetic valve is controlled through signals output by the optical signal processor.

The mixed oil output connecting pipe is connected with the mixed oil storage tank, and the mixed oil section of the finished oil can be conveyed to the mixed oil storage tank for centralized treatment.

The invention has the beneficial effects that: the invention has the advantages of real-time detection, high detection accuracy and easy maintenance and management, the oil product is irradiated by the ultraviolet laser, the data of the irradiated oil product is transmitted to the signal processing device by the ultraviolet laser receiver, and the ultraviolet laser signal is processed by the data processing principle. When the signal processing device outputs the oil mixing signal, the valve control device controls the electromagnetic valve on the main pipeline to be closed, and the electromagnetic valve on the oil mixing output pipeline is opened; when the signal processing device outputs the single-product oil signal, the electromagnetic valve on the valve control device control main pipeline is opened, the electromagnetic valve on the mixed oil output pipeline is closed, and the signal output by the valve control device can be adjusted according to the flow velocity of the oil product, so that the mixed oil in the mixed oil section is prevented from being discharged completely, and meanwhile, the single-product oil can be prevented from entering the mixed oil pipeline to cause the waste of the oil product.

Description of the drawings:

FIG. 1 is a block diagram of the present invention.

The specific implementation mode is as follows:

referring to fig. 1, the present invention specifically adopts the following embodiments: the system comprises a main pipeline 12, a detection device connecting pipe 13, a mixed oil output connecting pipe 9, a flange ball valve I1, an ultraviolet laser emitter 2, a detachable optical detection window 3, an ultraviolet laser receiver 4, an optical signal processor 5, an electromagnetic valve controller 6, a flange ball valve II 7, a lifting check valve 8, an electromagnetic valve I10 and an electromagnetic valve II 11; installing a detection device connecting pipe 13 on a main pipeline 12 of product oil, installing a flange ball valve I1 on the detection device connecting pipe 13, installing a detachable optical detection window 3 on the detection device connecting pipe 13, installing a lifting check valve 8 on the detection device connecting pipe 13, installing a flange ball valve II 7 on the detection device connecting pipe 13, installing an ultraviolet laser emitter 2 on one side of the detachable optical detection window 3, installing an ultraviolet laser receiver 4 on the other side of the detachable optical detection window 3, connecting the ultraviolet laser receiver 4 with an optical signal processor 5, connecting the other end of the optical signal processor 5 with an electromagnetic valve controller 6, connecting the electromagnetic valve controller 6 with an electromagnetic valve I10 and an electromagnetic valve II 11, connecting one end of the electromagnetic valve I10 with a mixed oil output connecting pipe 9, and connecting the other end of the mixed oil output connecting pipe 9 with the main pipeline 12; and a second electromagnetic valve 11 is installed on the main pipeline 12. The detection device connecting pipe 13 is provided with a detachable optical window 3, and two sides of the detachable optical window 3 are respectively provided with an ultraviolet laser transmitter 2 and an ultraviolet laser receiver 4. The flange ball valve I1 and the flange ball valve II 7 which are arranged on the detection device connecting pipe 13 are convenient for mounting and dismounting the optical window, oil loss in a pipeline is reduced, and the lifting check valve 8 arranged on the detection device connecting pipe 12 is used for ensuring the directional flow of oil in the detection pipe. The ultraviolet wavelength adopted by the ultraviolet laser transmitter 2 is 480 nm. One end of the electromagnetic valve controller 6 is connected with the optical signal processor 5, and the other end of the electromagnetic valve controller is connected with the first electromagnetic valve 10 and the second electromagnetic valve 11, and the opening or closing of the first electromagnetic valve 10 and the second electromagnetic valve 11 is controlled through signals output by the optical signal processor 5. The mixed oil output connecting pipe 9 is connected with a mixed oil storage tank, and can convey the mixed oil section of the finished oil to the mixed oil storage tank for centralized treatment.

Detection device connecting pipe 13 on install and to dismantle optical window 3, can dismantle optical window 3 both sides and install ultraviolet laser emitter 2 and ultraviolet laser receiver 4 respectively, can dismantle optical window 3 and can solve because the optical window light transmissivity that the live time overlength leads to worsens the problem that influences the detection precision, can solve through dismantling the window and wash or change new window.

The detection device is characterized in that the detection device is provided with two flange ball valves I1, two flange ring ball valves II 7 and a lifting check valve 8, wherein the flange ball valves I are used for cutting off pipelines and facilitating installation and disassembly of an optical window, and the lifting check valve II is used for ensuring the directional flow of oil in the detection pipe. The types of the flange ball valve I1 and the flange ring ball valve II are Q41F; the lift check valve is model number H41T-16.

One end of the electronic valve controller 6 is connected with the optical signal processor 5, and the other end is connected with the ZCT-₅₀₀Solenoid valve A10 and ZCT-₅₀₀The type of the second electromagnetic valve 11, the first electromagnetic valve 10 and the second electromagnetic valve 11 is ZCT-₅₀₀And A, controlling the opening and closing of a first electromagnetic valve 10 and a second electromagnetic valve 12 through an oil mixing signal output by the optical signal processor 5, wherein the first electromagnetic valve 10 is opened and the second electromagnetic valve 11 is closed when the optical signal outputs oil mixing data, and the first electromagnetic valve 10 is closed and the second electromagnetic valve 11 is opened when the optical signal outputs single oil data.

The main product oil conveying pipeline 12 is a steel pipe D500mm, the detection device connecting pipe 13 is a stainless steel pipe D200mm, and the two parts are connected through flanges.

The working principle is as follows, the relationship between the absorbance and the concentration of the liquid to be detected is established based on the Lambert beer law, namely when a beam of parallel light vertically irradiates the liquid to be detected, a water sample absorbs light with specific wavelength, the absorbance A of the water sample is in direct proportion to the concentration c and the light absorption thickness l of the component to be detected, the transmittance T is in inverse proportion to the concentration c and the light absorption thickness l, and the expression is as follows:

（1）

in the formula: a (lambda) -absorbance of the liquid to be measured;

i0 — incident light intensity;

it-transmitted light intensity;

tt-the transmittance, i.e., the ratio of It to I0;

c, the concentration of the component to be detected;

l-the thickness of the absorption;

ε -molar absorptivity.

When the liquid to be detected contains a plurality of components which do not influence each other, the total absorbance of the liquid is equal to the sum of the absorbance of each component in the solution at the same wavelength. The total absorbance calculation formula is as follows:

（2）

in the formula: a1, a2, A3, An — the absorbance of each component;

epsilon i-the extinction coefficient of each component;

ci-concentration of each component;

A-Total Absorbance of Mixed Components.

In the process of carrying out spectrum measurement on a sample, the measured spectrum information contains the characteristic information of the sample and the interference of external environment influence factors, and data preprocessing is required before quantitative and qualitative analysis is carried out by utilizing spectrum data, so that the aims of reducing noise and improving modeling precision are fulfilled.

The noise reduction methods generally used are: Savitzky-Golay polynomial smoothing, standard normal transformation, vector normalization, min-max normalization, multivariate scatter correction, spectral data derivation, etc. The design adopts vector normalization to preprocess the spectral data, because the method can reduce the influence of the optical path on the spectral information and is suitable for spectral noise reduction treatment of the liquid to be detected with large concentration change. The formula is as follows:

（3）

in the formula:-average value of the jth column element of the data matrix;

-the sum of squares root of the jth column element of the data matrix.

The design adopts an ultraviolet laser transmitter, pure oil and oil-mixed oil products are detected through experiments, and an optimal spectral characteristic interval is selected through a combined interval PLS algorithm, so that the wavelength of the ultraviolet laser transmitter adopted by the design is determined to be 480 nm.

The method comprises the steps of calculating a combined model of two or more subintervals based on iPLS (induced nearest neighbor PLS), selecting a combined Interval with the minimum root mean square error and the maximum correlation coefficient, and performing PLS regression analysis on the combined Interval to obtain the combined Interval partial least square model with the strongest prediction capability.

The relationship between the spectral information and the concentration or the content of the liquid to be measured is established by a Principal Component regression Analysis (PCA), so that the prediction of the specific Component or the content of the liquid to be measured is realized. The principle is that the spectral information is subjected to dimensionality reduction, spectral information with linear correlation difference is removed, and then the principal components are utilized to perform multivariate linear regression. The method can greatly reduce the complexity of the operation.

First, through singular value decomposition, the argument matrix X can be decomposed into:

（4）

in the formula: s is the square root of the eigenvalues of the covariance matrix XTX;

u-an orthogonal matrix of standard columns;

PT-orthogonal matrix of standard rows.

At that time, T is the score matrix in the principal component analysis, and P is the load matrix. The principal component regression model can be expressed as:

（5）

in the formula: b-correlation coefficient matrix

The least squares solution of can be expressed as:

（6）

principal component regression (PCA) can effectively reduce the interference of noise to a prediction model, and is convenient for further analysis. However, this method has a disadvantage that, in the modeling process, the accuracy of the prediction model is affected because part of the spectral data is discarded due to the reduction of the data dimension.

The performance of the established model can be evaluated by calculating the evaluation indexes of the regression model: and the correlation coefficient R, the decision coefficient R2, the correction set cross validation root mean square error RMSECV, the validation set external validation root mean square error RMSEP and the system deviation Bias reflect the accuracy reliability of the model. When the value of the correlation coefficient R or the determination coefficient R2 is larger, the values of the correction set cross validation root mean square error RMSECV, the validation set external validation root mean square error RMSEP and the system deviation Bias are smaller, the higher the prediction accuracy of the established regression model is, and the better the prediction effect is. If the RMSECV is much smaller than the RMSEP, this means that the representative of the correction set samples is poor. If RMSEP is much smaller than RMSECV, it indicates poor representation of the prediction set samples. Wherein:

（7）

（8）

（9）

（10）

（11）

in the formula:-a spectral modeling prediction value;

-an arithmetic mean of the spectral modeling predictors;

-arithmetic mean of the values determined by the standard method.

The mahalanobis distance can effectively calculate the similarity of the two sample sets, namely the closest distance between the "barycenters" of the two sample sets. Mahalanobis distance can take into account the link between various characteristics and can eliminate correlation interference between variables. The calculation formula is as follows:

（12）

in the formula: xi-data of ith sample

xj-data of j-th sample

Cov-sample covariance matrix

For qualitative analysis and identification of liquid multi-component mixtures, the design combines principal component analysis and Mahalanobis distance to establish a qualitative analysis model, and the method specifically comprises the following steps:

(1) firstly, the spectrum matrix of the correction setC _n×m Performing principal component dimensionality reduction to obtain a principal component score matrixS _n×f And load matrixL _m×f ；

(2) Carrying out centralization processing on the scoring matrix;

(3) computing covariance matrices for the samples of the correction setCov；

(4) Calculating the mahalanobis distance of the correction set samples;

(5) calculating the standard deviation of the Mahalanobis distance of the correction set samples, and setting the weight coefficient of the correction set samples;

(6) repeating the steps (1) to (4) to calculate the mahalanobis distance of the prediction set samples;

(7) and judging the prediction set.

The correction set spectrum dimensionality reduction formula is as follows:

（13）

in the formula: n-number of samples in the calibration set;

m is the number of variables;

f-number of principal Components

Score matrixS _n×f The centering treatment of (2) is performed according to the following formula:

（14）

in the formula: i =1,2,3, …, n, j =1,2,3, …, n

（15）

（16）

（17）

In the formula:-an average score matrix of n correction set samples;

a mean centering matrix of Sc-S;

s _i -correcting the spectral score vector of the sample i in the set;

M _cal,i -mahalanobis distance of the correction set sample i;

Cov _cal -correcting the covariance matrix of the set samples.

The mahalanobis distance standard deviation calculation formula of the correction set samples is as follows:

（18）

in the formula:-average value of mahalanobis distance of correction set samples

Note the bookM _pre, iTo predict a set sampleiThe distance between the two adjacent channels of the channel,G _pre for the spectral vector of the unknown sample,G _cal in order to correct the population of set samples,eand the weight coefficient is the Markov distance standard deviation and is used for setting a threshold range. The qualitative discrimination rule based on mahalanobis distance is as follows:

at that timeThen, then；

At that timeThen, then；

At that timeThen reselect the thresholde。

To sum up, this product oil pipeline oil mixing section laser on-line measuring device detects in real time, and it is high to detect the accuracy, easy to maintain management, shines the oil through ultraviolet laser, and the data conduction after the oil will be shone to ultraviolet laser receiver is for signal processing device, handles ultraviolet laser signal through above data processing principle. When the signal processing device outputs the oil mixing signal, the valve control device controls the electromagnetic valve on the main pipeline to be closed, and the electromagnetic valve on the oil mixing output pipeline is opened; when the signal processing device outputs the single-product oil signal, the electromagnetic valve on the valve control device control main pipeline is opened, the electromagnetic valve on the mixed oil output pipeline is closed, and the signal output by the valve control device can be adjusted according to the flow velocity of the oil product, so that the mixed oil in the mixed oil section is prevented from being discharged completely, and meanwhile, the single-product oil can be prevented from entering the mixed oil pipeline to cause the waste of the oil product.