Correction method for measurement error of linear array sensor
阅读说明：本技术 一种线列阵传感器测量误差的修正方法 (Correction method for measurement error of linear array sensor ) 是由 顾汉洋 肖瑶 张亨伟 刘莉 于 2019-10-29 设计创作，主要内容包括：一种线列阵传感器测量误差的修正方法,包括步骤：S1：将线列阵传感器测量两相流得到的横截面导电相份额分布矩阵B赋值给中间矩阵C；S2：将矩阵C作为边界条件,利用线列阵传感器电场模拟的方式模拟线列阵传感器得到模拟的测量矩阵C′；S3：判断矩阵C′和矩阵B相应元素的最大差值是否小于规定的阈值,如果小于阈值,则认为中间矩阵C等于实际的导电相份额分布矩阵A；否则对中间矩阵C进行修正,获得修正后的矩阵C″；S4：对矩阵C″中每一个元素进行修正,将修正之后的矩阵C″赋值给矩阵C,并返回步骤S2,直至矩阵C′和矩阵B相应元素的最大差值小于规定的阈值。可以将线列阵传感器由于均匀敏感体的假设带来的误差降低至3％以下。(A method for correcting measurement errors of a linear array sensor comprises the following steps: s1: assigning a cross-section conductive phase fraction distribution matrix B obtained by measuring two-phase flow by a linear array sensor to an intermediate matrix C; s2: simulating the linear array sensor by using the matrix C as a boundary condition in a mode of simulating an electric field of the linear array sensor to obtain a simulated measurement matrix C'; s3: judging whether the maximum difference value of corresponding elements of the matrix C' and the matrix B is smaller than a specified threshold value, and if so, considering that the intermediate matrix C is equal to the actual conductive phase share distribution matrix A; otherwise, correcting the middle matrix C to obtain a corrected matrix C'; s4: and modifying each element in the matrix C ', assigning the modified matrix C ' to the matrix C, and returning to the step S2 until the maximum difference value of the corresponding elements of the matrix C ' and the matrix B is smaller than a specified threshold value. The error of the linear array sensor caused by the assumption of a uniform sensitive body can be reduced to below 3 percent.)
1. A method for correcting measurement errors of a linear array sensor is characterized by comprising the following steps:
s1: assigning a cross-section conductive phase fraction distribution matrix B obtained by measuring two-phase flow by a linear array sensor to an intermediate matrix C;
s2: simulating the linear array sensor by using the matrix C as a boundary condition in a mode of simulating an electric field of the linear array sensor to obtain a simulated measurement matrix C';
s3: judging whether the maximum difference value of corresponding elements of the matrix C' and the matrix B is smaller than a specified threshold value, and if so, considering that the intermediate matrix C is equal to the actual conductive phase share distribution matrix A; otherwise, correcting the middle matrix C to obtain a corrected matrix C';
s4: and modifying each element in the matrix C ', assigning the modified matrix C ' to the matrix C, and returning to the step S2 until the maximum difference value of the corresponding elements of the matrix C ' and the matrix B is smaller than a specified threshold value.
2. The correction method according to claim 1, wherein the correction formula of the matrix C "obtained after the correction in step S3 is: c ″ ═ C + k × (B-C'), where k is the modified relaxation factor.
3. The correction method according to claim 1, wherein in step S4, each element in the matrix C "is corrected by:
judging whether the element value in the matrix C' is larger than 1, if so, setting the element value to 1;
and judging whether the element value in the matrix C' is less than 0, if so, setting the element value to 0.
4. The correction method as claimed in claim 1, wherein the step S2 is performed by simulating the electric field of the linear array sensor in a manner as follows:
establishing a simulation area, wherein the simulation area comprises the transmitting electrodes and the receiving electrodes which are the same as the transmitting electrodes and the receiving electrodes in number of the actual linear array sensors;
dividing the simulation area into a plurality of square grids, wherein each transmitting electrode and each receiving electrode are represented by a string of grids;
setting the boundary of the simulation area as an insulation boundary;
setting the potential of one transmitting electrode to be 1, setting the potentials of the other transmitting electrodes and all receiving electrodes to be 0, calculating the current value received by each receiving electrode through a Laplace equation, and repeating the steps until all the transmitting electrodes are excited, finishing the electric field simulation of the linear array sensor, and obtaining the current value corresponding to each intersection point in the grid;
and dividing the current value corresponding to each intersection point by the current value acquired by the linear array sensor in the pure conductive phase to obtain the distribution of the conductive phase portion in the square sensitive body measured by the linear array sensor.
5. The correction method according to claim 1, wherein the threshold value is a positive number smaller than 2%.
6. The correction method according to claim 1, wherein k is a positive number smaller than 4.
Technical Field
The invention relates to the technical field of two-phase flow measurement, in particular to a method for correcting measurement errors of a linear array sensor.
Background
The two-phase flow is widely applied to the fields of nuclear power generation, thermal power generation, oil exploitation, chemical engineering, pharmacy and the like. In two-phase flow, the gas phase is generally a non-conductive phase and the liquid phase is a conductive phase. The linear array sensor measures the conductivity distribution of the whole flow section to calculate the distribution of the conductive phase portion. The linear array sensor is composed of two groups of metal wire meshes which are perpendicular to each other but not in contact with each other. The wires in one set of meshes are referred to as the transmit electrodes and the wires in the other set of meshes are referred to as the receive electrodes. The emitting electrodes are sequentially excited, and pulse square waves are introduced. When one emitter electrode is excited, the other emitter electrodes and all the receiver electrodes are kept at 0 potential; and simultaneously recording the current received by all the receiving electrodes. When all the transmitting electrodes are excited, the measurement of one frame data of the linear array sensor is completed. The acquired data represents the conductivity of a sensitive volume near the intersection of the corresponding transmit and receive electrodes.
In the prior art process, the sensing body is considered to be a square centered at the intersection of the emitter and receiver electrodes, the side length of the square being equal to the lateral distance between the wires. All the sensitive bodies together constitute a measuring cross section. However, from simulations of the electric field, this assumption is not accurate. The sensitive body is an irregular geometric body, and the assumption that the sensitive body is uniformly distributed brings systematic errors. For a linear array sensor with a wire mesh spacing La/wire spacing L of 0.5, the systematic error in measuring the conductive phase fraction distribution due to the assumption of a uniform sensitive body can reach 30% at the maximum. Therefore, correction of the linear array sensor data is necessary.
Disclosure of Invention
The application provides a correction method of measurement errors of a linear array sensor, which comprises the following steps:
s1: assigning a cross-section conductive phase fraction distribution matrix B obtained by measuring two-phase flow by a linear array sensor to an intermediate matrix C;
s2: simulating the linear array sensor by using the matrix C as a boundary condition in a mode of simulating an electric field of the linear array sensor to obtain a simulated measurement matrix C';
s3: judging whether the maximum difference value of corresponding elements of the matrix C' and the matrix B is smaller than a specified threshold value, and if so, considering that the intermediate matrix C is equal to the actual conductive phase share distribution matrix A; otherwise, correcting the middle matrix C to obtain a corrected matrix C';
s4: and modifying each element in the matrix C ', assigning the modified matrix C ' to the matrix C, and returning to the step S2 until the maximum difference value of the corresponding elements of the matrix C ' and the matrix B is smaller than a specified threshold value.
In one embodiment, the modified matrix C ″ obtained in step S3 has a modification formula as follows: c ″ ═ C + k × (B-C'), where k is the modified relaxation factor.
In one embodiment, the step S4 of correcting each element in the matrix C ″ specifically includes:
judging whether the element value in the matrix C' is larger than 1, if so, setting the element value to 1;
and judging whether the element value in the matrix C' is less than 0, if so, setting the element value to 0.
In one embodiment, the electric field simulation method of the linear array sensor in step S2 includes:
establishing a simulation area, wherein the simulation area comprises the transmitting electrodes and the receiving electrodes which are the same as the transmitting electrodes and the receiving electrodes in number of the actual linear array sensors;
dividing the simulation area into a plurality of square grids, wherein each transmitting electrode and each receiving electrode are represented by a string of grids;
setting the boundary of the simulation area as an insulation boundary;
setting the potential of one transmitting electrode to be 1, setting the potentials of the other transmitting electrodes and all receiving electrodes to be 0, calculating the current value received by each receiving electrode through a Laplace equation, and repeating the steps until all the transmitting electrodes are excited, finishing the electric field simulation of the linear array sensor, and obtaining the current value corresponding to each intersection point in the grid;
and dividing the current value corresponding to each intersection point by the current value acquired by the linear array sensor in the pure conductive phase to obtain the distribution of the conductive phase portion in the square sensitive body measured by the linear array sensor.
In one embodiment, the threshold is a positive number less than 2%.
In one embodiment, k is a positive number less than 4.
According to the correction method of the embodiment, the error of the linear array sensor caused by the assumption of the uniform sensitive body can be reduced to be less than 3%.
Drawings
FIG. 1 is a flow chart of a correction method;
FIG. 2 is a distribution diagram of true conductive phase fractions;
FIG. 3 is a distribution diagram of conductive phase fraction measured directly with a linear array sensor;
FIG. 4 is a distribution diagram of conductive phase fraction after correction;
FIG. 5 is a grid schematic diagram of electric field simulation of a linear array sensor.
Detailed Description
The present invention will be described in further detail with reference to the following detailed description and accompanying drawings.
The current measured at each intersection of the linear-array sensor (wire-mesh sensor) can characterize the conductive phase fraction of a sensitive body (sensitive volume) near the intersection. In the conventional processing method, the sensitive body is assumed to be a rectangular solid with uniform distribution. This assumption introduces systematic errors. In the use process of an actual linear array sensor, the systematic error enables the conductive phase share measured in a sensitive body completely wrapped by the non-conductive phase to be larger than 0; and the non-conductive phase portion in the sensitive body completely wrapped by the conductive phase is more than 0.
The invention provides a correction method aiming at errors generated by the assumption of a linear array sensor. The error caused by the assumption of the linear array sensor can be reduced to below 3 percent by using the correction method. The correction method of the present invention includes the following steps, and its flow chart is shown in fig. 1.
S1: and assigning the cross-section conductive phase fraction distribution matrix B obtained by measuring two-phase flow by the linear array sensor to the intermediate matrix C.
The distribution of true conductive phase fractions measured by the linear array sensor is assumed to be matrix a (unknown), see fig. 2. The numbers in each sensitive volume in fig. 2 represent the fraction of conductive phase in that sensitive volume. The unmodified conductive phase fraction distribution matrix measured by the linear array sensor is B (known), and the number in each sensitive body in fig. 3 represents the conductive phase fraction measured by the linear array sensor in the sensitive body, see fig. 3. The systematic error due to the homogeneous sensitive volume assumption results in a large difference between the values in matrix B and a.
By constructing the intermediate matrix C, the cross-sectional conductive phase fraction distribution matrix B obtained by measuring the two-phase flow by the linear array sensor is assigned to the intermediate matrix C, and is corrected by the following steps S2 and S3.
S2: and simulating the linear array sensor by using the matrix C as a boundary condition in a mode of simulating the electric field of the linear array sensor to obtain a simulated measurement matrix C'.
Specifically, a cuboid simulation area is established, and the simulation area comprises the emitting electrodes and the receiving electrodes which are the same as the emitting electrodes and the receiving electrodes in number of the actual linear array sensors; the simulation area of this example contains 9 transmitting electrodes and 9 receiving electrodes;
dividing the simulation area into a plurality of square grids, for example, dividing the simulation area into 270 × 270 × 46 square grids, each of the transmitting electrodes and the receiving electrodes being represented by a string of grids; as shown in fig. 5;
setting the boundary of the simulation area as an insulation boundary;
setting the potential of one transmitting electrode to be 1, setting the potentials of the other transmitting electrodes and all receiving electrodes to be 0, calculating the current value received by each receiving electrode through a Laplace equation, and repeating the steps until all the transmitting electrodes are excited, finishing the electric field simulation of the linear array sensor, and obtaining the current value corresponding to each intersection point in the grid; specifically, the laplace equation for the electric field is:
and dividing the obtained current value corresponding to each intersection point by the current value acquired by the linear array sensor in the pure conductive phase to obtain a measurement matrix C' of conductive phase share distribution in the square sensitive body measured by the linear array sensor.
S3: judging whether the maximum difference value of corresponding elements of the matrix C' and the matrix B is smaller than a specified threshold value, and if so, considering that the intermediate matrix C is equal to the actual conductive phase share distribution matrix A; otherwise, the intermediate matrix C is corrected to obtain a corrected matrix C'.
And assigning the matrix B to the middle matrixes C and C as boundary conditions, and simulating the measurement process of the linear array sensor by utilizing a potential field simulation mode. The simulation yields a measurement matrix C'. If the maximum difference max | C '-B | between the matrix C' and the matrix B is less than the threshold ε, ε is less than 2%, then the matrix C and the matrix A are considered equal; if the maximum difference between matrix C' and matrix B is greater than the threshold ε, then matrix C and matrix A are considered to be different and need to be corrected.
The basic idea of correction is that the conductive phase share of a certain intersection point obtained by direct measurement of the linear array sensor and the conductive phase share of the uniformly distributed sensitive bodies corresponding to the intersection point have positive first correlation; if the fraction of conductive phase in a particular sensitive volume in matrix C' is less than (greater than) the fraction of conductive phase in the corresponding sensitive volume in matrix B, it is also assumed that the fraction of conductive phase in that sensitive volume in intermediate matrix C is also less than (greater than) the fraction of conductive phase in actual matrix a. Therefore, the intermediate matrix C is corrected by using the difference value between the matrix B and the matrix C' to obtain a corrected conductive phase share distribution matrix C ", wherein the correction formula is as follows: c ″ ═ C + k × (B-C'), where k is the modified relaxation factor and k is a positive number less than 4, and in this example k is 1.
S4: and modifying each element in the conductive phase share distribution matrix C ', assigning the modified matrix C ' to the matrix C, and returning to the step S2 until the maximum difference value of the corresponding elements of the matrix C ' and the matrix B is smaller than a specified threshold value.
In the correction matrix C ", it may happen that the fraction of the conductive phase in a certain sensitive volume is greater than 1 or less than 0. Since the actual conductive phase share distribution matrix a does not have a condition greater than 1 or less than 0, each element in the matrix C ″ is corrected, specifically, all the element values greater than 1 in the matrix C ″ are set to 1, and the element values less than 0 are set to 0, that is, whether the element values in the matrix C ″ are greater than 1 is determined, and if yes, the element values are set to 1; and judging whether the element value in the matrix C 'is less than 0, if so, setting the element value to be 0, and further correcting the elements in the matrix C' to reduce errors and be beneficial to ensuring the stability of the simulation calculation.
And the corrected matrix C 'is re-introduced into the matrix C, and the step S2 is carried out to circulate until the error between the matrix C' and the matrix B is less than the threshold value of 2%.
To obtain the intermediate matrix C after the correction is completed, i.e. after the error between matrix C' and matrix B is less than 2%, see fig. 4, the error is already less than 3% with respect to matrix a.
The present invention has been described in terms of specific examples, which are provided to aid understanding of the invention and are not intended to be limiting. For a person skilled in the art to which the invention pertains, several simple deductions, modifications or substitutions may be made according to the idea of the invention.
- 上一篇：一种医用注射器针头装配设备
- 下一篇：