Disclosure of Invention
The invention provides an automatic positioning and aligning method and system for guardrail plate installation, which can effectively solve the problems in the background technology.
In order to achieve the above purpose, the technical scheme adopted by the invention is as follows:
a method of fence panel installation automatic positioning alignment, the method comprising:
acquiring parameter information of the guardrail plate, and regulating and controlling installation equipment according to the parameter information;
acquiring the motion state of the mounting equipment when carrying the guardrail plate, and predicting the motion positioning influence through the motion state;
obtaining road topography information, and establishing a guardrail plate mounting guideline according to the road topography information;
establishing a visual recognition system, detecting and recognizing the positions of the guardrail plates according to the guardrail plate installation guideline, and obtaining guardrail plate positioning information;
establishing an automatic alignment model, inputting the guardrail plate positioning information, and primarily judging the matching degree of the guardrail plate positioning information and the motion state of the installation equipment;
performing primary alignment on the installation track of the installation equipment according to the matching degree;
inputting the motion positioning influence into the automatic alignment model, and realigning the installation track of the installation equipment.
Further, establishing a guardrail plate mounting guideline according to the road topography information, including:
acquiring road side gradient information, wherein the gradient direction is the forward direction of a road side installation guardrail plate;
determining the mounting positions of the two ends of the guardrail plate according to the road side gradient information;
and connecting the mounting positions of the two ends of the guardrail plate, and establishing the guardrail plate mounting guideline.
Further, detecting and identifying the position of the guardrail plate according to the guardrail plate mounting alignment line to obtain guardrail plate positioning information, including:
collecting an image of the guardrail plate mounting area;
performing feature recognition on the installation area image;
and determining the installation position of the guardrail plate based on the guardrail plate installation guideline according to the characteristic recognition result to obtain guardrail plate positioning information.
Further, performing feature recognition on the installation area image includes:
converting the installation area image into a feature matrix capable of performing convolution operation;
setting a convolution layer according to a preset contour of the guardrail plate;
and performing convolution operation on the feature matrix through the convolution layer to finish feature identification of the installation region image.
Further, establishing an automatic alignment model includes:
collecting historical installation information, and cleaning data in the historical installation information;
deep learning is carried out on the cleaned data based on a machine learning algorithm;
extracting a certain amount of cleaned data to construct a training set and a verification set;
checking the adjustment result after deep learning through the training set and the verification set;
the automatic alignment model is established based on the machine learning algorithm.
Further, performing a preliminary alignment of the mounting trace of the mounting apparatus according to the degree of matching, including:
setting a matching threshold value for the guardrail plate positioning information and the installation equipment through the automatic alignment model;
judging whether the matching degree is within the matching threshold value or not;
if not, the installation position of the installation equipment is adjusted again until the matching degree is within the matching threshold value, and initial alignment of the installation track of the installation equipment is completed.
Further, the movement state of the mounting device in carrying the guardrail plate is acquired, wherein the movement state comprises movement deviation and mounting shake of the mounting device.
Further, inputting the kinematic positioning effect into the auto-alignment model and realigning the mounting trajectory of the mounting device, comprising:
correcting the matching threshold according to the movement deviation and the installation jitter;
and aligning the installation track of the installation equipment again according to the corrected matching threshold value.
An guardrail plate mounting automatic positioning alignment system, the system comprising:
the parameter information acquisition module is used for acquiring parameter information of the guardrail plate and regulating and controlling the installation equipment according to the parameter information;
the motion state acquisition module acquires the motion state of the installation equipment when carrying the guardrail plate, and predicts the motion positioning influence through the motion state;
the land form information acquisition module is used for acquiring road land form information and establishing a guardrail plate installation guideline according to the road land form information;
the visual identification module is used for establishing a visual identification system, detecting and identifying the positions of the guardrail plates according to the guardrail plate installation guideline and obtaining guardrail plate positioning information;
an alignment model building module for building an automatic alignment model, inputting the guardrail plate positioning information, and primarily judging the matching degree of the guardrail plate positioning information and the motion state of the installation equipment,
the primary alignment module is used for carrying out primary alignment on the installation track of the installation equipment according to the matching degree;
and the realignment module inputs the motion positioning influence into the automatic alignment model and realigns the installation track of the installation equipment.
Further, the alignment model building module includes:
the historical data acquisition unit acquires historical installation information and cleans data in the historical installation information;
the deep learning unit is used for deep learning the cleaned data based on a machine learning algorithm;
a training verification set unit is constructed, and a training set and a verification set are constructed by extracting a certain amount of cleaned data;
and training a verification model unit, checking the adjustment result after deep learning through the training set and the verification set, and establishing the automatic alignment model based on the machine learning algorithm.
By the technical scheme of the invention, the following technical effects can be realized:
the problem that the automatic installation robot is poor in positioning and aligning degree of the guardrail plates is effectively solved, the manual flow of installing the guardrail plates of units is shortened, the labor cost is saved, and the installation efficiency is further improved.
The foregoing description is only an overview of the technical solutions of the present application, and may be implemented according to the content of the specification in order to make the technical means of the present application more clearly understood, and in order to make the above-mentioned and other objects, features and advantages of the present application more clearly understood, the following detailed description of the present application will be given.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments.
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. The terminology used in the description of the invention herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and/or" as used herein includes any and all combinations of one or more of the associated listed items.
Example 1
As shown in fig. 1, a method for automatically positioning and aligning a guardrail plate, the method comprises:
s100: acquiring parameter information of the guardrail plate, and regulating and controlling the installation equipment according to the parameter information;
specifically, it is necessary to acquire parameter information of the size, shape, etc. of the guardrail plate and apply the parameter information to a regulation and control process of the installation apparatus according to the information, by regulating and controlling the installation apparatus, stability and consistency of influence on the operation parameters such as speed and holding pressure can be maintained, if parameter settings of one installation apparatus are used for guardrail plates of different models and materials, it is difficult to predict and determine the operation state when installing different guardrail plates by the installation apparatus, and it is difficult to control the positioning alignment of the guardrail plates.
S200: the method comprises the steps of collecting the motion state of the mounting equipment in carrying the guardrail plate, and predicting the motion positioning influence through the motion state;
in this step, where the motion state refers to a series of movements of the mounting apparatus after the preparation work and movements of the robot arm during the mounting, when the mounting apparatus performs the actions to be performed by the mounting apparatus, the movements are regarded as actions that may affect the alignment deviation, and then considering the possible positioning influence of the action prediction generated by the work may consider as many factors affecting the alignment of the fence.
S300: obtaining road topography information, and establishing a guardrail plate mounting guideline according to the road topography information;
s400: establishing a visual recognition system, detecting and recognizing the positions of the guardrail plates according to guardrail plate installation guidelines, and obtaining guardrail plate positioning information;
in the above steps, the alignment line is a reference line established for assisting the visual recognition system to detect the position of the guardrail plate, and because the actual construction scene is not a very standard alignment line, the alignment line which is convenient for positioning the guardrail plate is established after the road topography information is considered, so that the more accurate detection and recognition of the position of the guardrail plate can be realized.
S500: establishing an automatic alignment model, inputting guardrail plate positioning information, and primarily judging the matching degree of the guardrail plate positioning information and the motion state of the installation equipment;
s600: performing primary alignment on the installation track of the installation equipment according to the matching degree;
selecting a proper algorithm, such as a correlation analysis algorithm, a Pearson correlation coefficient or a Spearman rank correlation coefficient, a linear regression or support vector regression, and a deep learning algorithm, establishing an automatic alignment model, wherein the automatic alignment model can make corresponding judgment on acquired data, and adjust the specific position of the installation equipment according to the judgment, so as to adjust the positioning alignment of the guardrail plate; and (3) inputting the positioning information of the visual guardrail plate into a model, and carrying out preliminary judgment on the matching degree of the positioning information and the movement state data of the installation equipment by combining the positioning information and the movement state data of the visual guardrail plate, and correspondingly adjusting the position of the installation equipment according to a judgment result so as to avoid larger deviation of the installation equipment in the accuracy.
S700: and inputting the motion positioning influence into an automatic alignment model, and realigning the installation track of the installation equipment.
The motion positioning influence is a factor which possibly causes accuracy deviation and appears in the working process, the model is set to be aligned for the second time so as to improve the alignment, manual secondary adjustment is reduced, the motion positioning influence can possibly fluctuate according to specific construction environments and construction projects, the installation equipment can not influence the installation and positioning of the guardrail plate sometimes, and then the automatic alignment model can feed back signals which do not need secondary alignment.
According to the technical scheme, the problem that the automatic installation robot is poor in positioning and aligning degree of the guardrail plates is effectively solved, the manual flow for installing the guardrail plates of units is shortened, the labor cost is saved, and the installation efficiency is further improved.
Further, as shown in fig. 2, the building of the guardrail plate mounting guideline according to the road topography information includes:
s310: acquiring road side gradient information, wherein the gradient direction is the forward direction of a road side installation guardrail plate;
s320: determining the mounting positions of two ends of the guardrail plate according to the road side gradient information;
s330: and connecting the mounting positions of the two ends of the guardrail plate, and establishing a guardrail plate mounting standard line.
On the basis of the embodiment, the road side gradient information is data with the greatest influence on the establishment of alignment lines in all road topography information, and the gradient information of the corresponding installation road section can be obtained through a measuring instrument or navigation of the road in advance, wherein the forward direction is the road direction, and the guardrail plate after installation and the road direction can be kept parallel and are not influenced by special topography.
Further, detecting and identifying the position of the guardrail plate according to the guardrail plate mounting alignment line to obtain guardrail plate positioning information, including:
collecting an image of a guardrail plate mounting area;
performing feature recognition on the installation area image;
and determining the installation position of the guardrail plate based on the guardrail plate installation guideline according to the characteristic identification result to obtain guardrail plate positioning information.
In this embodiment, the capturing of the image of the mounting area of the guardrail plate is to pay attention to the position representing the alignment line, the mounting alignment line is critical to the positioning of the guardrail plate, and is to determine the mark and the reference object of the contour position of the guardrail plate, wherein the capturing of the image of the mounting area of the guardrail plate can capture a plurality of angles to form a three-dimensional space, preferably two angles are adopted in this embodiment, one is the top view angle of the perpendicular road to the road surface, and the other is the top view angle from the road center to the road side, i.e. the mounting view angle, and the two angles can accurately position the pre-mounting position of the guardrail plate in combination with the position of the mounting alignment line, and in addition, a plurality of different definitions can be made for the mark of the features, for example, besides the alignment line feature, the connecting support rod for mounting the guardrail plate can also be one of the features for capturing the image recognition, and the extracted feature principle is to help the positioning of the guardrail plate more accurately, and can be set according to the specific construction project and scene.
Further, performing feature recognition on the installation area image includes:
converting the installation area image into a feature matrix capable of performing convolution operation;
setting a convolution layer according to a preset contour of the guardrail plate;
and carrying out convolution operation on the feature matrix through the convolution layer to finish feature identification on the image of the installation area.
On the basis of the embodiment, the convolution check image is used for filtering operation, and local features in the image are effectively captured through relatively small local perception, so that the network can better process the spatial information of the image; in the convolution layer, the same convolution kernel weight is used for input data at different positions, so that the number of parameters to be learned is reduced, and the parameter sharing mechanism enables the network to have stronger generalization capability and more efficiently extract image features.
Further, as shown in fig. 3, establishing the auto-alignment model includes:
s510: collecting historical installation information, and cleaning data in the historical installation information;
s520: deep learning is carried out on the cleaned data based on a machine learning algorithm;
s530: extracting a certain amount of cleaned data to construct a training set and a verification set;
s540: checking the adjustment result after deep learning through the training set and the verification set;
s550: an auto-alignment model is established based on a machine learning algorithm.
In this embodiment, experience and law can be obtained from a large amount of actual data by collecting history installation information, so that accuracy and robustness of an automatic alignment model are improved, an execution position of installation equipment is conveniently calibrated, the cleaned history data are learned through a machine learning algorithm, and deep learning content can include characteristics of position, posture, speed, acceleration and the like of a robot for installing a guardrail plate, wherein the machine learning method can select to use a deep learning model, such as a convolutional neural network, a cyclic neural network, a Transformer or the like, so as to capture complex relations in the data, so that the model can better understand an adjustment process of the installation equipment and characteristics of the guardrail plate, and the alignment accuracy is improved; the training set and the verification set are utilized to test the model, wherein the training set and the verification set can extract targeted data, namely historical data information of similar projects, from the historical installation information, and the performance of the model can be evaluated and optimized through the constructed training set and verification set, so that the reliability and stability of the automatic alignment model are improved.
Further, the primary alignment of the mounting track of the mounting device according to the matching degree includes:
setting a matching threshold value for the positioning information of the guardrail plate and the installation equipment through an automatic alignment model;
judging whether the matching degree is within a matching threshold value or not;
if not, the installation position of the installation equipment is adjusted again until the matching degree is within the matching threshold value, and initial adjustment of the installation track of the installation equipment is completed.
Specifically, the determination of the matching degree is given by an automatic alignment model, and the establishment of the automatic alignment model is based on historical installation information, so the matching degree is obtained according to the determination of a machine deep learning model according to historical similar projects and road topography location factors, the matching threshold is according to deviation which can be allowed to be positioned and aligned in the current construction project, the alignment operation can be realized only within the threshold, the matching threshold can be temporarily set in specific construction, the deviation value can be used as a reference for setting the matching threshold for the manual correction generated by machine installation of the guardrail plate at the beginning of the construction project, the approximate matching threshold can be set according to experience, and the matching threshold can be temporarily corrected according to the construction.
Further, the movement state of the mounting device in carrying the guardrail plate is collected, wherein the movement state comprises movement deviation and mounting shake of the mounting device.
Inputting the motion positioning influence into an automatic alignment model and realigning the installation track of the installation equipment, wherein the method comprises the following steps of:
correcting the matching threshold according to the movement deviation and the installation jitter;
and aligning the mounting track of the mounting equipment again according to the corrected matching threshold value.
Specifically, due to the different plate materials of the facing guard rail plates in the working advancing process of the installation equipment, unexpected shaking and installation equipment displacement deviation generated due to topography reasons can be generated during load-bearing working, temporary adjustment can be made according to actual working conditions based on prediction of influence on the movement deviation and the installation shaking, positioning and alignment accuracy can be improved by considering the factors, and the probability of manual alignment is reduced.
Example two
Based on the same inventive concept as the method for automatically positioning and aligning the installation of the guardrail plate in the foregoing embodiment, the present invention further provides an automatic positioning and aligning system for installing the guardrail plate, as shown in fig. 4, the system includes:
the parameter information acquisition module is used for acquiring parameter information of the guardrail plate and regulating and controlling the installation equipment according to the parameter information;
the motion state acquisition module acquires the motion state of the installation equipment in carrying the guardrail plate, and predicts the motion positioning influence through the motion state;
the land form information acquisition module is used for acquiring road land form information and establishing a guardrail plate installation guideline according to the road land form information;
the visual identification module is used for establishing a visual identification system, detecting and identifying the positions of the guardrail plates according to the guardrail plate installation guideline and obtaining guardrail plate positioning information;
an alignment model building module for building an automatic alignment model, inputting the positioning information of the guardrail plate, and primarily judging the matching degree of the positioning information of the guardrail plate and the motion state of the installation equipment,
the primary alignment module is used for carrying out primary alignment on the installation track of the installation equipment according to the matching degree;
and the realignment module inputs the motion positioning influence into the automatic alignment model and realigns the installation track of the installation equipment.
The adjusting system of the invention can effectively realize the automatic positioning and aligning method for installing the guardrail plate, and has the technical effects as described in the embodiment, and the description is omitted here.
Further, the alignment model building module includes:
the historical data acquisition unit acquires historical installation information and cleans data in the historical installation information;
deep learning unit, based on machine learning algorithm, to deep learn the data after washing;
a training verification set unit is constructed, and a training set and a verification set are constructed by extracting a certain amount of cleaned data;
and training a verification model unit, checking the adjustment result after deep learning through a training set and a verification set, and establishing an automatic alignment model based on a machine learning algorithm.
Similarly, the above-mentioned optimization schemes of the system may also respectively correspond to the optimization effects corresponding to the methods in the first embodiment, which are not described herein again.
Although the present application has been described in connection with specific features and embodiments thereof, it will be apparent that various modifications and combinations can be made without departing from the spirit and scope of the application. Accordingly, the specification and figures are merely exemplary illustrations of the application as defined in the appended claims and are to be construed as covering any and all modifications, variations, combinations, or equivalents that are within the scope of the application. It will be apparent to those skilled in the art that various modifications and variations can be made in the present application without departing from the scope of the application. Thus, if such modifications and variations of the present application fall within the scope of the present application and the equivalents thereof, the present application is intended to cover such modifications and variations.