CN117824487A - High-precision intelligent detection method for differential mechanism tool of pipeline robot - Google Patents

Abstract

The invention discloses a high-precision intelligent detection method for a differential mechanism tool of a pipeline robot, which comprises the following steps: cylindricity error evaluation, error optimization calculation technology and differential detection evaluation. Firstly, determining the distance from a measuring point of a high-precision inductive sensor to the axis of a working machine of a differential case, and determining cylindricity error assessment indexes of a piece to be measured and a standard piece; then, optimizing the cylindricity error by using the provided neighborhood particle intelligent optimization method, so as to determine the section radius of the piece to be measured; finally, differential detection assessment is performed. The method provided by the invention overcomes the defects of complicated operation, low efficiency, low measurement precision, incapability of being applied to the field and the like in the prior detection technology, can realize online rapid detection, reduces errors introduced by human factors, and can finish rapid high-precision detection of the differential mechanism tool.

Description

A high-precision intelligent detection method for differential tooling of pipeline robots

Technical Field

The present invention relates to the field of robot differential tooling, and in particular to a high-precision intelligent detection method for a pipeline robot differential tooling.

Background technique

The differential is an important component of the pipeline robot drive system. When the pipeline robot passes through a bend, the differential can enable the left and right (or front and rear) drive wheels to rotate at different speeds, adjust the speed difference between the left and right wheels, and make the left and right wheels roll at different speeds, thereby ensuring that the drive wheels on both sides perform pure rolling motion. As an important basic component of the differential, the differential housing directly affects the assembly of the internal gears of the differential and the accuracy of the assembly of the differential and the drive axle. For special pipeline robots, quality must be controlled in both differential manufacturing and testing, which can effectively improve the stability of the quality of the differential housing.

The inspection of the differential housing is a very important process, with many measurement items and high precision requirements. At present, the commonly used form and position tolerance measurement methods are mainly manual measurement and three-coordinate measurement. The manual approximate measurement method mainly measures the roundness by approximating the micrometer. This measurement method has the following problems: 1) The operation is cumbersome and the labor intensity of the workers is high; 2) The approximate judgment of the qualified status of the parts may result in misjudgment of qualified status; 3) The measurement accuracy is low, which is closely related to the workers' operating proficiency and working status; 4) The differential housing needs to be manually transported from the machine tool to the measuring turntable, and then sorted and stored after the measurement, which is time-consuming and laborious. The three-coordinate measurement method solves the above problems to a certain extent. It can quantitatively detect the shape and position size of the parts, and can also make intuitive graphical descriptions of the part morphology, converting abstract numbers into intuitive images, which brings greater convenience to quality control. However, the measurement efficiency of the three-coordinate measuring machine is also low, and it has strict requirements on the measurement environment and poor measurement repeatability, requiring a special indoor constant temperature measurement place.

In the prior art, there have been many studies on the form and position tolerance measurement methods, mainly focusing on the analysis of measurement methods, evaluation principles and measurement uncertainty, which has greatly promoted the development of form and position tolerance measurement technology. The geometric feature evaluation of cylindrical geometric bodies provides mathematical models and evaluation methods for both cylindricity and axis straightness. Cylindricity can be evaluated by the least squares cylinder method, the maximum circumscribed cylinder method, the minimum inscribed cylinder method and the minimum area cylinder method. However, this manual measurement operation is relatively complicated, and the measurement error is uncertain. Based on the neural network intelligent optimization algorithm, the measurement process of the low-speed roundness instrument is improved, and the ability to quickly obtain results and judge abnormal patterns is achieved, but the actual application of this method is difficult.

Summary of the invention

The technical problem to be solved by the present invention is: to propose a high-precision intelligent detection method for the differential tooling of a pipeline robot, to improve the efficiency and accuracy of the high-precision intelligent detection of the differential tooling of the pipeline robot on the basis of determining the error evaluation index, aiming at the problems of cumbersome operation, low efficiency and low measurement accuracy existing in the manual measurement, three-coordinate measurement and intelligent optimization algorithm detection of the differential housing.

The present invention adopts the following technical solution: a high-precision intelligent detection method for differential tooling of a pipeline robot, comprising the following steps:

S1. Determine the cylindrical error evaluation index: According to the relative measurement principle, first place multiple high-precision inductive sensors on the cylindrical surface of the differential housing assembly machine, then place the standard parts, adjust the readings of each sensor to zero, and then place the test piece, and use the distance from the measuring point to the axis to determine the cylindrical error evaluation index of the test piece and the standard part;

S2, Error optimization calculation: Propose a neighborhood particle intelligent optimization method to determine parameters;

S3. Differential inspection and evaluation: Calculate the error between the cross-sectional radius of the differential test piece and the cross-sectional radius of the differential standard part. The smaller the value, the closer the test piece is to the standard part.

Furthermore, the determination of the cylindricity error evaluation index in step S1 is performed in the following manner:

S1.1. Use Q sensors to measure and sample. Each sensor corresponds to x, y, and z coordinates in three-dimensional space. The coordinates of the intersection of the axis of the cylinder and the xoy plane are ( a , b , 0), and the direction vector of the axis is ( p , q , 1).

Among them, a , b , p , q are parameters to be determined;

S1.2. The z coordinate of the i -th sensor is determined by the sensor installation height. The relationship between the x, y coordinates and the sensor reading δ _i is:

;

Where, i =1,2,..., Q _, R0 is the radius of the standard profile (unit: mm), θ is the angle of the test piece relative to the initial position (unit: degree);

S1.3. Calculate any sampling point on the cylindrical surface Distance to the axis d _i :

;

in, ;

S1.4. According to the least squares principle, the sum of the squares of the distances from each measuring point to the cylindrical surface should be the smallest, so the error evaluation index function of cylindricity is:

;

Among them, a , b , p , q are parameter variables to be determined, and R is the cross-sectional radius of the differential test piece.

Furthermore, in step S2, the error optimization calculation is performed as follows:

S2.1. Initialization: Determine the particle population size M , the iteration parameter k = 1, the maximum number of iterations is K _max , and the initial position vector of the particle is:

;

The initial velocity vector of the particle is:

;

The maximum velocity is V _max , and the number of particles in each neighborhood is N ;

They represent the position parameter variables corresponding to the rth particle in the first iteration, Respectively represent and/> The corresponding speed parameter variable;

S2.2. Calculation error evaluation index: Substitute into the formula in step S1.4 to obtain the error assessment index function value;

S2.3. Determine the neighborhood solution: If k < K _max , use:

Construct N _m neighborhoods, ;

Instead, use:

Construct Nm neighborhoods. The first particle in _each neighborhood can receive global information, and other particles only receive neighborhood information.

S2.4. Update particles: In the k +1th iteration, the particle updates its velocity and position according to the following formula:

;

Among them, w is the inertia weight; c ₁ , c ₂ , c ₃ are learning factors, r ₁ , r ₂ are random numbers uniformly distributed between [0,1];

is the best position (individual extreme value) experienced by the ith particle,/> is the best position experienced by all particle populations (global extremum),/> is the best position experienced by all particles in the neighborhood corresponding to the i -th particle (neighborhood extreme value);

Y _r is the neighborhood learning ability function, which takes a constant or random number between [0,1] to distinguish the ability of particles to acquire global or neighborhood knowledge;

S2.5. Determine the termination conditions:

If the maximum number of iterations is reached ( k = K _max ), the optimization ends and the global optimal value is obtained. That is, the optimal parameter values of a , b , p , q , and R ;

otherwise, , go to step S2.2.

Compared with the prior art, the present invention adopts the above technical solution and has the following technical effects:

The high-precision intelligent detection method for the differential tooling of the pipeline robot of the present invention collects information in real time through a built-in high-precision inductive sensor, thereby overcoming the shortcomings of the three-coordinate measurement method that requires point-by-point coordinate collection, has a slow measurement speed, and cannot be applied on site; the parameter optimization algorithm integrates the idea of neighborhood information sharing into the optimization algorithm, fully utilizes the evolution information between particles in the search process, promotes the information exchange between particles in the search process, increases the diversity of particles, avoids premature convergence of the algorithm, and can realize collaborative and intelligent exploration of spatial areas, thereby greatly improving the optimization search efficiency and performance, and enhancing the global search capability of the algorithm.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG1 is a flowchart of the steps of a high-precision intelligent detection method for differential tooling of a pipeline robot according to the present invention;

FIG2 is a schematic diagram of the distribution of high-precision inductive sensors of the present invention;

FIG3 is a flow chart of calculation of cylindricity error evaluation index of the present invention;

FIG4 is a diagram showing the relationship between the sensor reading and the coordinates of the measured point according to an embodiment of the present invention;

FIG5 is a flowchart of error optimization calculation of the present invention.

Detailed ways

The technical solution of the present invention is further described in detail below in conjunction with the accompanying drawings:

It will be understood by those skilled in the art that, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as generally understood by those skilled in the art in the field to which the present invention belongs. It should also be understood that terms such as those defined in common dictionaries should be understood to have meanings consistent with the meanings in the context of the prior art, and will not be interpreted with idealized or overly formal meanings unless defined as herein.

The high-precision intelligent detection method for differential tooling of a pipeline robot of the present invention, as shown in FIG1 , comprises the following steps:

S1. Cylindricity error assessment: Based on the relative measurement principle, several high-precision inductive sensors are placed on the cylindrical surface of the differential housing assembly machine, and the differential standard parts are placed. The readings of each sensor are adjusted to zero, and then the differential to be tested is placed. The distance from the measuring point to the axis is used to determine the cylindricity error assessment index of the differential to be tested and the standard part.

S2. Error optimization calculation: Use the neighborhood particle intelligent optimization method to determine the parameters in the cylindrical error evaluation;

S3. Differential test and evaluation: Calculate the error between the cross-sectional radius R of the differential to be tested and the cross-sectional radius R0 _of the differential standard part .

In one embodiment of the present invention, taking measurement sampling with 10 sensors as an example, in order to determine the cylindricity error evaluation index, firstly, a plurality of high-precision inductive sensors are placed on the cylindrical surface of the differential housing assembly machine, as shown in FIG. 2 .

The calculation process of the cylindricity error evaluation index is shown in Figure 3. The specific steps are as follows:

Step 1: Each sensor corresponds to x, y, z coordinates in three-dimensional space, the coordinates of the intersection of the axis of the cylinder and the xoy plane are marked as ( a , b , 0), and the direction vector of the axis is ( p , q , 1);

Among them, a , b , p , q are parameters to be determined;

Step 2: The z coordinate of the i -th sensor is determined by the sensor installation height. The relationship between the sensor reading and the coordinates of the measured point is shown in Figure 4. The dotted line in Figure 4 is the standard contour of the differential standard part, and the solid line is the measured contour of the differential part to be tested. The relationship between the x, y coordinates and the sensor reading δ _i is determined as follows:

;

Where, i = 1, 2, ..., 10, R ₀ is the radius of the standard profile (unit: mm), θ is the angle of the test piece relative to the initial position (unit: degree);

Step 3: Calculate any sampling points on the cylindrical surface Distance to the axis d _i :

;

in, ;

Step 4: According to the least squares principle, the sum of the squares of the distances from each measuring point to the cylindrical surface should be minimized, so the error evaluation index function of cylindricity is:

;

Among them, a , b , p , q , and R are parameter variables to be determined.

Then, the error optimization calculation is performed, and the neighborhood particle intelligent optimization method is used to determine the parameter variables in the calculation of the cylindrical error evaluation index. The specific implementation process is shown in Figure 5. The steps of the neighborhood particle intelligent optimization method are as follows:

(1) Initialization:

In this embodiment, the particle population size M = 40, the iteration parameter k = 1, and the maximum number of iterations K _max = 500 are determined;

Use a pseudo-random number generator to generate the initial position vector of the particle:

;

and the particle's initial velocity vector:

;

The learning factor c ₁ = c ₂ = c ₃ = 1.5, the maximum velocity V _max = 10, and the number of particles in each neighborhood N = 5;

(2) Calculation error evaluation index:

Will Substitute into the formula in step 4 to obtain the error assessment index function value;

(3) Determine the neighborhood solution:

If k < K _max , use:

Construct N _m neighborhoods, ;

Instead, use:

Construct N _m neighborhoods;

The first particle in each neighborhood can receive global information, and other particles only receive neighborhood information;

(4) Update particles:

In the k +1th iteration, the particle updates its velocity and position according to the following formula:

;

Wherein, w is the inertia weight, and in particular, a linear decreasing weight is adopted in this embodiment; r ₁ , r ₂ are random numbers uniformly distributed between [0,1];

is the best position (individual extreme value) experienced by the ith particle,/> is the best position experienced by all particle populations (global extremum),/> is the best position experienced by all particles in the neighborhood corresponding to the i -th particle (neighborhood extreme value);

Y _r is the neighborhood learning ability function, which takes a constant or random number between [0,1] to distinguish the ability of particles to acquire global or neighborhood knowledge;

(5) Determine the termination conditions:

If the maximum number of iterations is reached ( k = K _max ), the optimization ends and the global optimal value is obtained. That is, the optimal parameter value of a , b , p , q , and R ; otherwise, /> , go to step (2).

Finally, the differential detection and evaluation in the high-precision intelligent detection method of the pipeline robot differential tooling is carried out, that is, the error _e between the cross-sectional radius R of the differential to be tested and the cross-sectional radius R0 of the differential standard part is calculated:

;

The smaller the value of error e is, the closer the part to be tested is to the standard part.

Through experiments, the high-precision intelligent detection method for differential tooling of the pipeline robot of the present invention can reduce the error e by 10% compared with the intelligent detection method based on conventional particle swarm.

In an embodiment of the present invention, an electronic device is also provided, including: one or more processors; a storage device on which one or more programs are stored; when the one or more programs are executed by the one or more processors, the one or more processors implement the high-precision intelligent detection method for differential tooling of a pipeline robot described in any of the above embodiments.

In an embodiment of the present invention, a computer-readable storage medium is further provided, on which a computer program is stored. When the program is executed by a processor, the steps of the high-precision intelligent detection method for differential tooling of a pipeline robot in any of the above embodiments are implemented.

This embodiment is only for illustrating the technical idea of the present invention, and cannot be used to limit the protection scope of the present invention. Any changes made on the basis of the technical solution in accordance with the technical idea proposed by the present invention shall fall within the protection scope of the present invention.

Claims (7)

1. A high-precision intelligent detection method for a differential mechanism tool of a pipeline robot is characterized by comprising the following steps:

s1, evaluating cylindricity errors: according to a relative measurement principle, a plurality of high-precision inductive sensors are placed on a cylindrical surface of a differential housing tooling machine, differential standard components are placed, readings of the sensors are zeroed, then the differential to-be-measured components are placed, and cylindricity error assessment indexes of the differential to-be-measured components and the standard components are determined by utilizing the distance from a measurement point to an axis;

s2, error optimization calculation: determining parameters in cylindricity error assessment and the section radius of a differential to-be-measured part by using a neighborhood particle intelligent optimization method;

s3, differential detection and assessment: and calculating the error between the section radius of the differential to-be-measured piece and the section radius of the differential standard piece.

2. The intelligent high-precision detection method for the differential tool of the pipeline robot according to claim 1, wherein the cylindricity error assessment in the step S1 is carried out by adopting the following steps:

s1.1, useQThe inductive sensors perform measurement sampling, and each sensor corresponds to x, y, the z coordinate marks the intersection point coordinate of the axis of the cylindrical surface of the differential housing tooling machine and the xoy plane as%a,b0), the direction vector of the axis is%p,q1), wherein, the mixture is prepared from the components of the mixture,a、b、p、qis a parameter to be determined;

s1.2, theiZ-coordinates of individual sensorsz _i Determined by the sensor mounting height, x, y coordinatesx _i、 y _i And sensor readingsδ _i The relation of (2) is:

；

wherein,i=1,2,...,Q，R ₀ is the radius of the cross section of the standard part of the differential mechanism,θthe angle of the differential mechanism to-be-measured piece relative to the initial position is changed;

s1.3, calculating any measuring point on the cylindrical surfaceDistance to axisd _i ：

；

Wherein,；

s1.4, according to the least square principle, the sum of squares of the distances from each measuring point to the cylindrical surface is minimum, and the error evaluation index function of the cylindricity is as follows:

；

wherein,Rthe radius of the section of the differential to be measured is the radius of the section of the differential to be measured.

3. The high-precision intelligent detection method for the differential tool of the pipeline robot according to claim 2, wherein in the error optimization calculation of the step S2, the neighborhood particle intelligent optimization method comprises the following steps:

s2.1, initializing: determining particle population sizeMIteration parametersk=1, the maximum number of iterations isK _max Maximum speed ofV _max Each neighborhood particle number isN；

The initial position vector of the particle is:

；

the initial velocity vector of the particle is:

；

wherein the parameter variablea、b、p、q、RThe upper 1 of (1) represents the iteration number, the lowerrIndicating the number of particles;

respectively represent the 1 st iterationrThe position parameter variables corresponding to the individual particles,respectively indicate and->Corresponding speed parameter variables;

s2.2, calculating an error evaluation index: first, thekIn the second iterationrPosition of individual particlesThe expression is as follows:

；

will beSubstituting the values into the formula in the step S1.4 to obtain error evaluation index function values;

s2.3, determining a neighborhood scheme: for the firstkConstructing particles by multiple iterationsN _m The first particle of each neighborhood receives global information, and other particles only receive neighborhood information;

s2.4, updating particles: first, thek +1 iterations, the particles update speed and position according to the following formula:

；

wherein,was the weight of the inertia is given,c ₁ , c ₂ , c ₃ in order for the learning factor to be a function of,r ₁ , r ₂ take [0,1 ]]Random numbers uniformly distributed among the two;

represents the extreme value of the individual, namelyiThe best location for individual particles to experience; />Representing a global extremum, which is the best location experienced by all particle populations; />Representing the neighborhood extremum as the firstiThe individual particles correspond to the best locations experienced by all particles in the neighborhood;

Y _r for neighborhood learning ability function, take [0,1]Constant or random number between them to distinguish the ability of the particle to acquire global or neighborhood knowledge;indicating particle numberkThe velocity of the secondary iteration particles;

s2.5, judging termination conditions: if the maximum number of iterations is reachedk=K _max Then the optimization is finished, and the obtained global optimal valueNamely, isa、b、p、q、RIs determined by the parameter values of the parameter values;

otherwise the first set of parameters is selected,the step S2.2 is followed by an iterative search for an optimal value.

4. The high-precision intelligent detection method for the differential tool of the pipeline robot according to claim 3, wherein the neighborhood scheme is determined in the step S2.3, and the particles are constructed by adopting the following formulaN _m The following neighbors:

if it isk<K _max ：

；

Conversely:

；

wherein,。

5. the high-precision intelligent detection method for the differential tool of the pipeline robot according to claim 1, wherein in step S3, the radius of the section of the differential part to be detected is calculatedRRadius of section of standard part of differential mechanismR ₀ Error of (2)e：

；

Error ofeThe smaller the value, the closer the part to be measured is to the standard.

6. An electronic device, comprising:

one or more processors;

a storage device having one or more programs stored thereon;

when executed by the one or more processors, causes the one or more processors to implement the method of any of claims 1 to 5.

7. A computer readable storage medium, characterized in that it has stored thereon a computer program which, when executed by a processor, implements the steps of the method for high-precision intelligent detection of a pipe robot differential tooling according to any one of claims 1 to 5.