CN112859739B - Digital twin-driven multi-axis numerical control machine tool contour error suppression method - Google Patents
Digital twin-driven multi-axis numerical control machine tool contour error suppression method Download PDFInfo
- Publication number
- CN112859739B CN112859739B CN202110059640.XA CN202110059640A CN112859739B CN 112859739 B CN112859739 B CN 112859739B CN 202110059640 A CN202110059640 A CN 202110059640A CN 112859739 B CN112859739 B CN 112859739B
- Authority
- CN
- China
- Prior art keywords
- model
- solution
- speed
- digital twin
- contour error
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 75
- 230000001629 suppression Effects 0.000 title claims abstract description 23
- 230000008878 coupling Effects 0.000 claims abstract description 33
- 238000010168 coupling process Methods 0.000 claims abstract description 33
- 238000005859 coupling reaction Methods 0.000 claims abstract description 33
- 230000007246 mechanism Effects 0.000 claims abstract description 33
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 27
- 230000002457 bidirectional effect Effects 0.000 claims abstract description 14
- 238000010801 machine learning Methods 0.000 claims abstract description 13
- 238000004891 communication Methods 0.000 claims abstract description 9
- 238000005457 optimization Methods 0.000 claims abstract description 9
- 230000008447 perception Effects 0.000 claims abstract description 8
- 238000011217 control strategy Methods 0.000 claims abstract description 7
- 238000012360 testing method Methods 0.000 claims abstract description 6
- 230000008859 change Effects 0.000 claims description 22
- 230000008569 process Effects 0.000 claims description 19
- 238000013507 mapping Methods 0.000 claims description 17
- 230000006870 function Effects 0.000 claims description 15
- 238000012546 transfer Methods 0.000 claims description 14
- 230000004044 response Effects 0.000 claims description 13
- 230000001133 acceleration Effects 0.000 claims description 12
- 238000006073 displacement reaction Methods 0.000 claims description 10
- 239000011159 matrix material Substances 0.000 claims description 10
- 230000002441 reversible effect Effects 0.000 claims description 9
- 230000036461 convulsion Effects 0.000 claims description 8
- 230000009471 action Effects 0.000 claims description 7
- 238000013016 damping Methods 0.000 claims description 7
- 238000012549 training Methods 0.000 claims description 7
- 241000544061 Cuculus canorus Species 0.000 claims description 6
- 238000012545 processing Methods 0.000 claims description 6
- 238000012512 characterization method Methods 0.000 claims description 4
- 238000007405 data analysis Methods 0.000 claims description 3
- 238000013135 deep learning Methods 0.000 claims description 3
- 229920001971 elastomer Polymers 0.000 claims description 3
- 239000000806 elastomer Substances 0.000 claims description 3
- 238000002474 experimental method Methods 0.000 claims description 3
- 230000009467 reduction Effects 0.000 claims description 3
- 230000011218 segmentation Effects 0.000 claims description 3
- 239000000126 substance Substances 0.000 claims description 3
- 230000009466 transformation Effects 0.000 claims description 2
- 238000012417 linear regression Methods 0.000 claims 1
- 238000000513 principal component analysis Methods 0.000 claims 1
- 230000006399 behavior Effects 0.000 description 13
- 238000003754 machining Methods 0.000 description 9
- 238000010586 diagram Methods 0.000 description 5
- 230000001360 synchronised effect Effects 0.000 description 5
- 230000000694 effects Effects 0.000 description 4
- 230000004927 fusion Effects 0.000 description 3
- 230000005540 biological transmission Effects 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 230000009193 crawling Effects 0.000 description 2
- 230000003993 interaction Effects 0.000 description 2
- 238000012821 model calculation Methods 0.000 description 2
- 230000008054 signal transmission Effects 0.000 description 2
- 238000004088 simulation Methods 0.000 description 2
- 230000003068 static effect Effects 0.000 description 2
- 230000004888 barrier function Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 238000006880 cross-coupling reaction Methods 0.000 description 1
- 230000001186 cumulative effect Effects 0.000 description 1
- 230000009977 dual effect Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000009347 mechanical transmission Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000013139 quantization Methods 0.000 description 1
- 230000002787 reinforcement Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B19/00—Programme-control systems
- G05B19/02—Programme-control systems electric
- G05B19/18—Numerical control [NC], i.e. automatically operating machines, in particular machine tools, e.g. in a manufacturing environment, so as to execute positioning, movement or co-ordinated operations by means of programme data in numerical form
- G05B19/404—Numerical control [NC], i.e. automatically operating machines, in particular machine tools, e.g. in a manufacturing environment, so as to execute positioning, movement or co-ordinated operations by means of programme data in numerical form characterised by control arrangements for compensation, e.g. for backlash, overshoot, tool offset, tool wear, temperature, machine construction errors, load, inertia
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B2219/00—Program-control systems
- G05B2219/30—Nc systems
- G05B2219/35—Nc in input of data, input till input file format
- G05B2219/35408—Calculate new position data from actual data to compensate for contour error
Landscapes
- Engineering & Computer Science (AREA)
- Human Computer Interaction (AREA)
- Manufacturing & Machinery (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Numerical Control (AREA)
Abstract
The invention relates to the field of numerical control machines, and particularly discloses a contour error suppression method of a digital twin driven multi-axis numerical control machine, which comprises the following steps of S1, establishing a virtual model of a digital twin corresponding to a physical entity, and obtaining a time-varying coupling mechanism model of a multi-axis feeding system by adopting a multi-parameter gain scheduling control strategy based on a global task coordinate system; establishing a data driving model of the multi-axis feeding system by utilizing signal testing and machine learning; s2, synchronizing virtual and real, and establishing a bidirectional perception relationship between the digital twin and the physical entity through a communication protocol with strong compatibility; s3, dynamically estimating the contour error; and S4, suppressing contour error, and obtaining the optimal motion control parameters and the maximum limiting speed of the corresponding track by adopting a multi-objective optimization algorithm according to the pre-estimated model. By adopting the technical scheme of the invention, the contour error can be effectively reduced.
Description
Technical Field
The invention relates to the field of numerical control machine tools, in particular to a contour error suppression method of a digital twin-driven multi-axis numerical control machine tool.
Background
The digital twin is a simulation process integrating multidisciplinary, multi-physical quantity, multi-scale and multi-probability by fully utilizing data such as a physical model, sensor updating, operation history and the like, and mapping is completed in a virtual space, so that the full life cycle process of corresponding entity equipment is reflected. Digital twinning is an beyond-realistic concept that can be viewed as a digital mapping system of one or more important, interdependent equipment systems.
Based on a digital twin contour error suppression theory, a multi-axis feeding system is used as a carrier, and a closed-loop suppression mode of contour error modeling-prediction-control is constructed through high-fidelity digital twin modeling, virtual and real accurate synchronization, contour error high confidence dynamic estimation and contour error comprehensive suppression.
The existing numerical control machine tool control methods based on digital twinning are few, and most of the existing numerical control machine tool control methods relate to digital twinning modeling methods of numerical control machine tools. At present, a control method of a numerical control machine tool based on digital twinning mainly focuses on realizing basic functions, simulating before machining, predicting the machining effect of a product and the like. As disclosed in the document with chinese patent publication No. CN 110865607 a, a five-axis numerical control machine tool control method based on digital twin is disclosed, the steps of the control method are: 1.) establishing a twin virtual model similar to the physical entity; 2) establishing a digital twin model of the five-axis numerical control machine tool; 3) establishing communication between a twin virtual model and a physical five-axis numerical control machine tool in virtual simulation software; 4) establishing a twin data processing and analyzing system; 5) and the five-axis numerical control machine tool and the twin virtual model keep twin information synchronization in real time. The method can perform real-time data synchronization between the twin virtual model and the entity five-axis numerical control machine tool, can perform simulation before machining and find a scheme for solving problems in time, can predict the machining effect of a product, and guides actual engineering operation; and in the machining process, the machining state can be monitored in real time and visually displayed, the problems of the machining link can be found and corrected in time, and the machining efficiency is improved.
However, in the above documents, the real-time states of the spindle rotation speed, the machine tool vibration effect, the temperature, the machining state and the mechanical change of the five-axis numerical control machine tool are mainly collected to achieve synchronization with the digital twin model thereof. However, for the product, the above document does not consider the influence of the dynamic characteristic parameters of the feed system (such as natural frequency and damping ratio), non-linear disturbances (friction, cutting and inertia forces) and profile parameters (curvature and corners), even the mutual coupling and influence relationship existing between the above influencing factors.
The research of the inventor finds that the dynamic characteristic parameters directly influence various controller parameters, particularly the controller parameters based on the model; non-linear interference can change the dynamic response of the system; the profile shape parameters are important constraints of the kinematic parameters, which in turn affect the dynamic response of the system and the controller parameters. In addition, during the operation of the multi-axis feeding system, the profile error and the influencing factors thereof are changed at any moment, the relationship between the profile error and the influencing factors is time-varying and dynamic, and the relationship is difficult to describe in a functional form.
Therefore, a method for suppressing the contour error of the numerical control machine tool driven by the digital twin drive, which can effectively reduce the contour error, is urgently needed.
Disclosure of Invention
The invention provides a contour error suppression method of a digital twin-driven multi-axis numerical control machine tool, which can effectively reduce contour errors.
In order to solve the technical problem, the present application provides the following technical solutions:
a contour error suppressing method for a multi-axis numerically-controlled machine tool driven by digital twin includes the following steps,
s1, establishing a virtual model of a digital twin body corresponding to a physical entity, and obtaining a time-varying coupling mechanism model of the multi-axis feeding system by adopting a multi-parameter gain scheduling control strategy based on a global task coordinate system; establishing a data driving model of the multi-axis feeding system by utilizing signal testing and machine learning;
s2, synchronizing virtual and real, and establishing a bidirectional perception relationship between the digital twin and the physical entity through a communication protocol with strong compatibility;
s3, contour error dynamic estimation: acquiring dynamic characteristic parameters, motion control parameters, nonlinear interference and contour shape parameters, wherein the dynamic characteristic parameters comprise natural frequency and damping ratio and are represented by Dy, the motion control parameters comprise various controller parameters and are represented by Co, the nonlinear interference comprises friction force, cutting force and inertia force and is represented by Di, the contour shape parameters comprise curvature and corners and are represented by Sh, and the kinematic parameters comprise speed, acceleration and jerk and are represented by Mo; performing order reduction characterization and generalization on the digital twin by adopting KPCA, Relief-X and LLE methods to obtain a dynamic mapping model between the contour error and various influence factors thereof, wherein CE is M (Dy, Co, Di, Sh and Mo); wherein CE represents the contour error, M represents the mapping relation between the contour error and the influencing factors thereof; training and precision correcting the dynamic mapping model by adopting Q-Learning and DRL algorithms to obtain an online estimation model of the contour error;
and S4, suppressing contour errors, obtaining optimal motion control parameters and the maximum track limiting speed corresponding to the optimal motion control parameters by adopting a multi-objective optimization algorithm according to the pre-estimated model, and performing interpolation control on the multi-axis feeding system by taking the maximum limiting speed and other conventional factors as constraints.
The working principle and the advantages of the scheme are as follows: in S1, the time-varying coupling mechanism model can embody the time-varying property and the axis mismatching property of the dynamic characteristic; but because the multi-shaft feeding system has non-linear influence factors of reverse clearance, friction force, inertia force and cutting force during the operation process, the factors belong to random and uncertain interference factors, and when the factors act, a time-varying coupling mechanism model is not enough to accurately express the internal mechanism and performance state of the multi-shaft feeding system. In the scheme, the high-fidelity digital twin body of the multi-shaft feeding system is realized by driving the model through data, so that the performance state and the internal influence factors of the physical entity are faithfully mapped.
In S2, the digital twin body obtains the multi-granularity information of the physical entity, and transmits various data information to the time-varying coupling mechanism model and the data driving model respectively through a communication protocol with strong compatibility, so as to provide a data basis for synchronous update, thereby implementing refined perception of the multi-granularity information of the physical space and the digital space.
In S3, the complex relation between the contour error and the influencing factors thereof can be accurately described, and the high efficiency of model calculation is ensured.
In S4, since the dynamic characteristics of the multi-axis feeding system in high-speed motion are changed from moment to moment, the motion control parameters must be time-varying, so as to ensure high precision and smoothness of motion. Therefore, the change of the motion control parameters and the influence of the change on the contour error are mainly considered, the optimal motion control parameters and the maximum track limiting speed corresponding to the optimal motion control parameters are obtained by adopting a multi-objective optimization algorithm, and then the maximum limiting speed and other conventional factors are used as constraints to carry out interpolation control on the multi-axis feeding system, so that the contour error suppression under the time variation of the motion control parameters is realized, and the contour error is effectively reduced.
Further, the optimal motion control parameters and the maximum track limiting speed corresponding to the optimal motion control parameters are solved in the following process, according to the pre-estimated model, the minimization of the contour error is taken as a target, the arch height error, the centripetal acceleration and the centripetal jerk are taken as constraints, and the cuckoo algorithm is adopted to carry out optimization in the feasible regions of the limiting speed and the motion control parameters.
Further, the solving process of the optimal motion control parameters and the maximum limiting speed of the corresponding track is as follows, firstly, determining an input item according to a cuckoo algorithm rule, and calculating according to an online pre-estimated model of the contour error to obtain an initial fitness value; then, for the situation that the iteration times do not reach the maximum, a group of old solutions are randomly selected, then new solutions are generated through Levy fli buckles, whether the new solutions are superior to the old solutions or not is judged, and the optimal solutions are selected; then, whether local search is carried out and a new solution is generated is determined by judging whether the solution abandoning factor is larger than the solution abandoning probability, and the optimal solution and the optimal fitness value are updated according to the new solution; and circulating the steps until a global optimal solution and an optimal fitness value are obtained, wherein the global optimal solution is the maximum limiting speed of a parameter curve and the optimal motion control parameter under the multi-constraint condition including the contour error.
Further, S5, based on the maximum limiting speed of the obtained parameter curve, adopting a constant speed segmentation algorithm of the parameter curve to segment the curve, taking the point with local maximum feeding speed or the area between break points on the parameter curve as a key area, and dividing the parameter curve into a plurality of segments according to the key area; determining the minimum feeding speed of each key area as the constant value feeding speed of the area, and obtaining the segmented curve constraint and planning parameter information required by speed planning; and then, considering that the processing time is shortest, performing speed planning by adopting a bidirectional scanning look-ahead algorithm to obtain a smooth speed curve under the limitation of geometric constraint, dynamic performance constraint, maximum limit speed and minimum speed fluctuation, and taking the smooth speed planning and the optimal motion control parameter corresponding to the parameter curve as interpolation control input.
Further, the input items include: the population size N, the problem dimension D, the solution abandoning probability Pa, the upper/lower critical value L/U and the MaxGen at the moment of maximum iteration are calculated according to the information, the fitness value of the initial solution is calculated, and the optimal solution X is updatedbestAnd an optimal fitness value fminThen judging whether the iteration times t meet the maximum iteration times, if not, outputting a global optimal solution GXbestAnd an optimal fitness value Gfmin(ii) a If so, randomly selecting a solution, generating a new solution through Levy flight, judging whether the new solution is superior to the old solution, and if so, replacing the old solution with the new solution; if not, keeping the previous generation old solution; then judging whether the random factor R is greater than the solution abandoning probability Pa, if so, abandoning the solution and locally searching to generate a new solution to obtain an updated current optimal solution XbestAnd an optimal fitness value fminThen judging whether the iteration times t meet the maximum iteration times again, and circulating; if not, obtaining an updated current optimal solution XbestAnd an optimal fitness value fminAnd then judging whether the iteration times t meet the maximum iteration times again, and circulating.
Further, the time-varying coupling mechanism model is obtained by adopting the Darbel theorem and Laplace transform to obtain a multi-degree-of-freedom rigid-flexible coupling transfer function model of the feeding system,
wherein, GrigidRepresenting a rigid body transfer function matrix, Gflcx，kRepresenting a kth order elastomer transfer function matrix; through an identification experiment, a time-varying rigid-flexible coupling transfer function matrix of the feeding system is estimated by adopting a least square method and an orthogonal polynomial curve fitting method,
wherein the content of the first and second substances,
respectively representing displacement, speed and load quality, all of which are time-varying variables, then obtaining a time-varying coupling mechanism model of the multi-axis feeding system by adopting a multi-parameter gain scheduling control strategy based on a global task coordinate system, and representing the time-varying coupling mechanism model by using a state space model as follows:
further, the data-driven model is obtained by: the servo driver acquires the displacement, the speed and the acceleration of each shaft, and a machine learning method is adopted to represent the influence relation of the reverse clearance on the pose change of the multi-shaft feeding system; aiming at the uncertainty of the inertial force and the cutting force, position and force data are collected by means of a servo driver and an external sensor, and a machine learning method is adopted to represent the dynamic response of the multi-axis feeding system under the action of the inertial force and the cutting force; the method comprises the steps of collecting displacement, speed and torque aiming at the jump of a reverse gap and friction, representing dead zones and creeping phenomena in the operation process of the multi-axis feeding system by adopting a Gaussian process regression method, carrying out big data analysis on various collected data, representing the pose, dynamic response, jump and the mutual influence relation of the pose, the dynamic response and the jump of the multi-axis feeding system under the action of nonlinear external interference by adopting a deep learning method, and obtaining a data driving model.
In the scheme, in order to avoid frequent fluctuation of the feeding speed, the minimum feeding speed of each key area is determined as the constant value feeding speed of the area, and segmented curve constraint and planning parameter information required by speed planning, such as a starting point or an end point parameter, an estimated length, and the maximum feeding speed of the starting point or the end point parameter, are obtained. And then, considering that the processing time is shortest, performing speed planning by adopting a bidirectional scanning look-ahead algorithm to obtain a smooth speed curve under the limitation of geometric constraint, dynamic performance constraint, maximum limit speed and minimum speed fluctuation, and taking the smooth speed plan corresponding to the parameter curve and the optimal motion control parameter as interpolation control input to realize the comprehensive suppression of the profile error of the multi-axis feeding system.
Drawings
FIG. 1 is a logic diagram of a mechanism-data hybrid driven digital twin body multi-field and multi-dimensional fusion modeling in a digital twin driven multi-axis numerical control machine tool contour error suppression method in a second embodiment;
FIG. 2 is a logic diagram of virtual and real precise synchronization of a multi-axis feeding system in a second embodiment of a contour error suppression method for a multi-axis numerical control machine driven by a digital twin;
FIG. 3 is a schematic diagram of an online estimation model of a contour error in a second embodiment of a digital twin-driven multi-axis numerical control machine tool contour error suppression method;
FIG. 4 is a logic diagram of a solving process of the optimal motion control parameters and the maximum limiting speed in the second embodiment of the numerical twin driven multi-axis numerical control machine tool contour error suppression method;
fig. 5 is a schematic diagram of an interpolation algorithm in a second embodiment of a digital twin-driven multi-axis numerical control machine tool contour error suppression method.
Detailed Description
The following is further detailed by way of specific embodiments:
example one
The contour error suppression method of the numerical twin driven multi-axis numerical control machine tool comprises the following steps:
s1, establishing a virtual model of the digital twin corresponding to the physical entity, namely establishing the digital twin modeling of the multi-shaft feeding system;
s2, carrying out accurate virtual-real synchronization, namely virtual-real synchronization, of the multi-axis feeding system, and establishing a bidirectional perception relationship between the digital twin and a physical entity through a communication protocol with strong compatibility;
s3, dynamically predicting contour error, namely dynamically predicting contour error, and acquiring dynamic characteristic parameters, motion control parameters, nonlinear interference and contour shape parameters, wherein the dynamic characteristic parameters comprise natural frequency and damping ratio and are represented by Dy, the motion control parameters comprise various controller parameters and are represented by co, the nonlinear interference comprises friction force, cutting force and inertia force and is represented by Di, the contour shape parameters comprise curvature and corner and are represented by Sh, and the kinematic parameters comprise speed, acceleration and jerk and are represented by Mo; performing order reduction characterization and generalization on the digital twin by adopting KPCA, Relief-X and LLE methods to obtain a dynamic mapping model between the contour error and various influence factors thereof, wherein CE is M (Dy, Co, Di, Sh and Mo); wherein CE represents the contour error, M represents the mapping relation between the contour error and the influencing factors thereof; training and correcting the precision of the dynamic mapping model by adopting Q-Learning and DRL algorithms to obtain an online estimation model of the contour error;
s4, carrying out comprehensive suppression on contour errors, namely contour error suppression, obtaining optimal motion control parameters and the maximum limiting speed of the corresponding track by adopting a multi-objective optimization algorithm according to the pre-estimated model, and then carrying out interpolation control on the multi-axis feeding system by taking the maximum limiting speed and other conventional factors as constraints.
Example two
The difference between this embodiment and the first embodiment is that in this embodiment, S1-S4 will be described in addition specifically.
In the digital twin modeling of the multi-axis feeding system of S1, considering that the multi-axis feeding system has the characteristic of multi-attribute cross coupling such as structure, mechanics, control, communication and the like, the digital twin modeling needs to relate to modeling technologies in multiple fields of dynamics, system identification, motion control, signal testing, machine learning, network communication, computers and the like. Meanwhile, in the operation process of the multi-axis feeding system, due to the changes of pose, speed and acceleration and the influence of nonlinear interference, the dynamic characteristics of the multi-axis feeding system are obvious in time variation, mismatching between axes and uncertainty, and the dual driving of mechanisms and data needs to be considered in the digital twin body modeling. Therefore, aiming at the above characteristics of the multi-axis feeding system, a mechanism-data hybrid driven digital twin multi-field and multi-dimensional fusion modeling method is proposed, as shown in fig. 1.
The time-varying coupling mechanism model is the core foundation of the digital twin body, and is modeled from multiple dimensions such as a geometric layer, a physical layer, a behavior layer and a regular layer in order to ensure high fidelity of the digital twin body. The structural attributes of the multi-axis feeding system mainly comprise form and position dimensions, assembly relationship, relative motion relationship and the like, and are expressed in a geometric layer. The mechanical properties of the multi-axis feeding system mainly include statics, dynamics and the like, and are expressed in a physical layer. Aiming at the time-varying property of the dynamic characteristic of the multi-axis feeding system, the influence of rigid-flexible coupling vibration and electromechanical coupling rigidity is considered, the multi-degree-of-freedom rigid-flexible coupling transfer function model of the feeding system can be obtained by adopting the Dalnbell's theorem and Laplace transformation, and the multi-degree-of-freedom rigid-flexible coupling transfer function model is uniformly expressed by the following formula
Wherein Grigid represents a rigid body transfer function matrix, and Gflex, k represents a k-th order elastomer transfer function matrix. Through identification experiments, a time-varying rigid-flexible coupling transfer function matrix of the feeding system is estimated by adopting a least square method, an orthogonal polynomial curve fitting method and the like, and is uniformly represented by the following formula
Wherein the content of the first and second substances,
respectively representing displacement, velocity and load mass, all time varying variables. For the time-varying coupling mechanism model, the physical layer is a deep-level expression based on the geometric layer.
The control attribute of the multi-axis feeding system is used for motion and positioning control of each axis, common control algorithms comprise PID control, sliding mode control, robust control and the like, in order to reflect the time-varying property and the mismatching property between axes of dynamic characteristics, a multi-parameter gain scheduling control strategy based on a global task coordinate system is adopted and is expressed at a behavior layer to obtain a time-varying coupling mechanism model of the multi-axis feeding system, and a state space model is used for uniformly expressing the time-varying coupling mechanism model
Similarly, the behavior layer is a further deep expression of the integrated gain scheduling control strategy based on the time-varying coupling transfer function matrix of the physical layer. In order to represent the internal rules of the multi-axis feeding system, the rules need to be expressed from a rule layer, and the change rules of the structural parameters, the dynamic parameters, the controller parameters and the like of the multi-axis feeding system are reflected.
The multi-axis feeding system has nonlinear influence factors such as reverse clearance, friction force, inertia force, cutting force and the like in the operation process, the factors belong to random and uncertain interference factors, and when the factors act, a time-varying coupling mechanism model is not enough to accurately express the internal mechanism, the performance state and the like of the multi-axis feeding system. Therefore, it is necessary to establish a data-driven model of the multi-axis feeding system by using methods such as signal testing and machine learning, and similarly, the data-driven model is expressed from multiple dimensions such as geometry, physics, behavior, and rules.
In a geometric layer, aiming at the jump change of the reverse clearance of the feeding system, data such as displacement, speed, acceleration and the like of each axis are acquired by a servo driver, and the influence relation of the reverse clearance on the pose change of the multi-axis feeding system is represented by adopting a machine learning method. In a physical layer, aiming at the uncertainty of the inertial force and the cutting force, position and force data are collected by a servo driver and an external sensor, and a machine learning method is adopted to represent the dynamic response of the multi-axis feeding system under the action of the inertial force and the cutting force. In a behavior layer, data such as displacement, speed, torque and the like are collected aiming at the jump of a reverse gap and friction, and a Gaussian process regression method is adopted to represent the jump phenomena such as dead zones, crawling and the like in the operation process of the multi-axis feeding system. In a rule layer, the pose, the dynamic response, the jump and the influence relation among the pose, the dynamic response and the jump of the multi-axis feeding system under the action of nonlinear external interference are represented by carrying out big data analysis on various acquired data and adopting a deep learning method. In the data-driven model, the logical way of layer-by-layer depth is also followed from the geometric layer to the rule layer.
In order to ensure data interaction between the digital twin and the physical entity and between each component of the digital twin, a signal interface model needs to be established. Aiming at the signal transmission process among a control system, a servo drive, a mechanical transmission and sensors of a multi-axis feeding system, a relation interface between an instruction signal and a position parameter is emphasized and established in a geometric layer, a relation interface between a sensor signal and a dynamic characteristic parameter is emphasized and established in a physical layer, a protocol interface for transmitting a physical space-digital space signal is emphasized and established in an action layer, and signal transmission formats, decoding rules and the like between virtual-real, virtual-virtual and real are emphasized and established between real and real in a rule layer.
Considering that a time-varying coupling mechanism model, a data driving model and a signal interface model respectively express a multi-axis feeding system from different subject fields, based on a universal semantic representation method, a unified modeling language is adopted, and the time-varying coupling mechanism model, the data driving model and the signal interface model are respectively integrated and expressed from a geometric layer, a physical layer, a behavior layer, a rule layer and the like to obtain a high-fidelity digital twin body of the multi-axis feeding system so as to faithfully map the performance state of a physical entity and the internal influence factors thereof.
In short, in the embodiment, the digital twin is obtained by respectively establishing the time-varying coupling mechanism model, the data driving model and the signal interface model, and then integrating the models. The whole method for establishing the digital twin body is a multi-field and multi-dimension fusion modeling method of the digital twin body driven by mechanism-data mixing. Signal testing and machine learning are the methods employed to build data-driven models.
In the virtual and real accurate synchronization of the S2 multi-axis feeding system, the data interaction between the physical entity and the digital twin is a precondition guarantee for realizing the virtual and real synchronization of the multi-axis feeding system. Data information of the multi-axis feeding system is mainly transmitted through a numerical control system, a servo driver, an external sensor and the like, and the sampling frequency, the magnitude, the type and the like of the data information are different, so the data information is called multi-granularity information. In addition, the multi-granularity information transmission of the multi-axis feeding system has the characteristics of cross protocols and cross interfaces. Therefore, based on a bus, a network port and a serial port transmission mechanism, communication protocols with strong compatibility such as Modbus, OPC-UA, MTCONNECT and NCLink are adopted to establish a bidirectional sensing relationship between the digital twin and the physical entity, as shown in the bidirectional sensing part of FIG. 2. The digital twin body senses the multi-granularity information of the physical entity through the signal interface, and transmits various data information to the time-varying coupling mechanism model and the data driving model respectively, so as to provide a synchronously updated data base for the data information, and further realize the refined sensing of the multi-granularity information of the physical space and the digital space.
The virtual-real synchronization of the multi-axis feeding system not only needs the synchronization of the pose, but also needs the synchronization of the characteristics and the performance, even the synchronization of the logic and the law, so that based on the composition characteristics of the established digital twin, a synchronization mechanism between a physical entity and the digital twin is established from multiple dimensions such as geometry, physics, behavior, rules and the like, and further the virtual-real accurate synchronization of the multi-axis feeding system is realized. As shown in the virtual and real synchronization mechanism part of fig. 2, in the geometric layer, the object entity transmits instruction data such as NC codes to the instruction signal interface through the numerical control system, and drives the digital twin body to complete real-time change of displacement, speed, acceleration and jerk, and at the same time, the digital twin body feeds back the changed pose data to the servo driver, thereby realizing virtual and real synchronization of the pose.
In order to accurately represent the virtual-real synchronization of the physical layer, the behavior layer and the rule layer, slow time T is introduced on the basis of standard time T, and T is more than or equal to T and less than or equal to 3T. The method comprises the steps of considering the characteristics of dynamic characteristics of a physical entity, parameters of a controller and dynamic evolution of a performance state, respectively aiming at the synchronization requirements of a physical layer, a behavior layer and a rule layer, extracting key characterization parameters based on a built digital twin, building a digital twin reconfiguration model under a slow time scale, and fusing physical data to drive the digital twin to realize synchronization with the physical entity under the slow time scale in the physical layer, the behavior layer and the rule layer.
As shown in the virtual-real synchronization mechanism part of fig. 2, in the physical layer, stress, strain, natural frequency, damping ratio and the like are extracted as key parameters, the physical entity synchronous external sensor transmits a force, vibration and other time domain signals and a converted frequency domain signal to a time-frequency domain signal interface, and drives the reconstruction of a twin model and the dynamic update of the key parameters in the slow time scale of the physical layer, so as to realize the virtual-real synchronization of static and dynamic characteristics, wherein T is usually 2-3T. In a behavior layer, controller parameters, friction, crawling and the like are extracted as key parameters, a physical entity transmits current, torque and the like to a corresponding signal interface through a servo driver, and twin model reconstruction and real-time updating of the key parameters under a slow time scale of the behavior layer are driven, so that virtual and real synchronization of control performance and jump phenomena is realized, and T is usually 1-2T. In a rule layer, pose, dynamic characteristics, performance states and the like are extracted as key parameters, various real-time data, historical data and the like of a physical entity are fused, online learning is adopted to carry out model correction on a twin body reconstruction model in the rule layer, dynamic updating of the key parameters is driven, and therefore virtual-real synchronization of evolution rules is achieved, and T is usually 1-2T.
Based on a bidirectional perception relation of a digital-physical space and a virtual-real synchronization mechanism of different dimensions, association and classification processing are performed on multi-granularity information of the physical space and the digital space from multiple dimensions of a geometric layer, a physical layer, a behavior layer, a rule layer and the like of a digital twin body by machine learning algorithms such as Apriori, C4.5, KNN and the like, and then specific association or homogeneous data are mined and fused by a multiple regression method, a stepwise regression method and the like to obtain effective data in each dimension. On the basis, according to a theoretical calculation formula of performance state parameters, parameters such as dynamic characteristics (rigidity, damping and natural frequency), following errors and contour errors of the multi-axis feeding system are quantized and assigned, and finally, the virtual and real accurate synchronization of the digital twin body and the physical entity under the drive of data is achieved.
In short, the virtual-real synchronization in this embodiment mainly includes three parts: bidirectional perception relationship, virtual and real synchronization mechanism and performance state quantization. Among them, the most important is the virtual-real synchronization mechanism, and the second is the bidirectional perception relationship.
In the S3 high confidence dynamic estimation of the profile error, in the operation process of a multi-axis feeding system, a plurality of influencing factors are applied to the profile error, wherein the dynamic characteristic parameters of the feeding system mainly comprise natural frequency and damping ratio, which are uniformly expressed by Dy, the motion control parameters mainly comprise various controller parameters, which are uniformly expressed by Co, the nonlinear disturbance mainly comprises friction force, cutting force, inertia force and the like, which are uniformly expressed by Di, the profile shape parameters mainly comprise curvature, corner and the like, which are uniformly expressed by Sh, and the kinematic parameters mainly comprise speed, acceleration, jerk and the like, which are uniformly expressed by Mo. Meanwhile, mutual coupling and influence relationship exist among various influencing factors, such as: the dynamic characteristic parameters directly affect various controller parameters, especially model-based controller parameters; non-linear interference can change the dynamic response of the system; the profile shape parameters are important constraints of the kinematic parameters, which in turn affect the dynamic response of the system and the controller parameters. In addition, during the operation of the multi-shaft feeding system, the profile error and the influencing factors thereof are changed at any time, the relationship between the profile error and the influencing factors is time-varying and dynamic, and the relationship is difficult to describe in a functional form.
In order to accurately describe the complex relation between the contour error and the influence factors thereof and ensure the high efficiency of model calculation, the digital twin body is subjected to reduced order representation and generalization by adopting methods such as KPCA, Relief-X, LLE and the like to obtain a dynamic mapping model between the contour error and various influence factors thereof, and the dynamic mapping model is represented by the following formula
CE＝M(Dy，Co，Di，Sh，Mo) (4)
Where CE represents the profile error, and M represents the mapping relationship between the profile error and its influencing factors, as shown in fig. 3. And (3) fusing the measured data, the soft sensing data, the calculation data, the instruction data, the mechanism formula and the like, and training and correcting the precision of the dynamic mapping model by adopting strong Learning algorithms such as Q-Learning and DRL (dry Learning) to obtain the online estimation model of the contour error.
The multi-granularity information used for training the pre-estimation model mainly comprises real-time data and historical data, the real-time data is the key point that the contour error online pre-estimation model can keep the dynamic change characteristic, and the accuracy of the pre-estimation model depends on the support of the historical data. Aiming at the characteristic that the real-time data volume is small and is not enough to ensure the accuracy of the model, sample data is expanded by adopting a Bayes method based on a large amount of historical data. Aiming at the characteristic that historical data cannot support online estimation precision, a data set in the operation process of the multi-axis feeding system is updated and learned based on rules, training samples with large deviation are subjected to sample training again, and timeliness of the data is guaranteed, so that continuous high confidence estimation of the profile error of the multi-axis feeding system is achieved.
In S4 contour error comprehensive suppression, as one of the main means of contour error suppression, the interpolation algorithm plans the trajectory speed mainly in consideration of geometric constraint, kinematic constraint, contour error constraint, and the like, and usually assumes that the motion control parameters are unchanged. However, the dynamic characteristics of the multi-axis feeding system in high-speed motion are changed at all times, so that the motion control parameters are required to be time-varying, and the high precision and the smoothness of the motion can be ensured. Therefore, the change of the motion control parameters and the influence of the change on the contour error are mainly considered, the optimal motion control parameters and the maximum track limiting speed corresponding to the optimal motion control parameters are obtained by adopting a multi-objective optimization algorithm, and then the maximum limiting speed and other conventional factors are used as constraints to carry out interpolation control on the multi-axis feeding system, so that the contour error suppression under the time variation of the motion control parameters is realized.
The solving process of the maximum limiting speed of the optimal motion control parameter and the parameter curve is shown in fig. 4, based on a dynamic mapping model between the contour error and various influence factors thereof, the minimization of the contour error is taken as a target, the bow-height error, the centripetal acceleration, the centripetal jerk and the like are taken as constraints, the cuckoo algorithm is adopted to carry out optimization in feasible domains of the limiting speed and the motion control parameter, wherein L/U refers to the upper and lower critical values of the optimal motion control parameter and the maximum limiting speed, the fitness value f refers to the contour error, and the solution X refers to the optimal motion control parameter and the maximum limiting speed.
Firstly, determining an input item according to experience and a cuckoo algorithm rule, and calculating to obtain an initial fitness value according to a contour error online estimation model. And then, for the condition that the iteration times do not reach the maximum, randomly selecting a group of solutions, generating a new solution through Levy flight, further judging whether the new solution is superior to the old solution, and selecting a better solution. And determining whether to perform local search and generate a new solution by judging whether the solution abandoning factor is greater than the solution abandoning probability, and updating the optimal solution and the optimal fitness value according to the new solution. And circulating the steps until a global optimal solution and an optimal fitness value are obtained, wherein the global optimal solution is the maximum limiting speed of a parameter curve and the optimal motion control parameter under the multi-constraint condition including the contour error.
Based on the maximum limiting speed of the obtained parameter curve, a constant speed segmentation algorithm of the parameter curve is adopted to segment the curve, as shown in fig. 5. The region between the points or breakpoints with local maximum feed speed on the parameter curve is taken as a key region, and the parameter curve is divided into a plurality of segments according to the key region. In order to avoid frequent fluctuation of the feeding speed, the minimum feeding speed of each key area is determined as the constant value feeding speed of the area, and segmented curve constraint and planning parameter information required by speed planning, such as a starting point or an end point parameter, an estimated length, a starting point or an end point parameter, a maximum feeding speed and the like, are obtained. Then, considering that the processing time is shortest, a bidirectional scanning look-ahead algorithm is adopted to perform speed planning, and a smooth speed curve under the limitation of geometric constraint, dynamic performance constraint, maximum limit speed and minimum speed fluctuation is obtained, as shown in fig. 5. And taking the smooth speed plan corresponding to the parameter curve and the optimal motion control parameter as interpolation control input to realize the comprehensive suppression of the profile error of the multi-axis feeding system.
In short, the dynamic estimation of the profile error in the embodiment mainly includes three steps, 1, obtaining a dynamic mapping model through digital twin reduced order representation, 2, obtaining an online estimation model through reinforcement learning, and 3, guaranteeing timeliness of data through data sample expansion and update learning.
EXAMPLE III
Compared with the first embodiment, the difference is that in S2, a bidirectional sensing relationship between the digital twin and the physical entity is established, then after a digital twin reconfiguration model under a slow time scale is constructed, the digital twin and the physical entity are respectively scanned with T as a period, and when the digital twin is controlled to change in one period, synchronization is performed from an imaginary to a real; after the physical entity is controlled to change in a period, the physical entity is synchronized from real to virtual in the period; and after the physical entity is controlled to be subjected to uncontrolled change in a period, pausing real-to-virtual synchronization in the period, repeating scanning the physical entity in the period, carrying out uncontrolled change identification, and finally carrying out real-to-virtual synchronization before scanning the digital twin body in the next period.
This is because in the existing operation logic, the change of the digital twin is mainly the human operation, and the related change is controlled. But the physical entity essentially comprises the relevant parameters such as the structural characteristics and the motion characteristics of the real product. And collecting relevant parameters through corresponding sensors within an acceptable range, and carrying out relevant modeling based on the parameters. However, in actual use, the product is easily affected by the external environment, and some uncontrolled changes are generated. These uncontrolled variations, if accumulated, result in virtual-real out-of-sync. However, if synchronization is performed at any time, the mechanism of virtual-real synchronization becomes problematic, that is, when the two are not consistent, who should be the calibration target. In this embodiment, the change of the physical entity is divided into controlled and uncontrolled, and the controlled change is synchronized in the scanning period. Uncontrolled changes (like cumulative errors, etc.) are detected by the relevant sensors, relevant identification is performed, and then the digital twins are synchronized before scanning the digital twins in the next cycle, so that the situation of inconsistency between the two can be avoided. Simultaneously with directly all compare with the prior art that physical entity regarded as the calibration object, this embodiment can be through the mode of distinguishing at this cycle and next cycle, conveniently carry out effectual sign on the one hand, on the other hand can guarantee the promptness. Even compared with a mode of directly carrying out synchronization from real to virtual in the same period, the embodiment can avoid the situation that the synchronization fails due to the conflict between the real and virtual.
The above are merely examples of the present invention, and the present invention is not limited to the field related to this embodiment, and the common general knowledge of the known specific structures and characteristics in the schemes is not described herein too much, and those skilled in the art can know all the common technical knowledge in the technical field before the application date or the priority date, can know all the prior art in this field, and have the ability to apply the conventional experimental means before this date, and those skilled in the art can combine their own ability to perfect and implement the scheme, and some typical known structures or known methods should not become barriers to the implementation of the present invention by those skilled in the art in light of the teaching provided in the present application. It should be noted that, for those skilled in the art, without departing from the structure of the present invention, several changes and modifications can be made, which should also be regarded as the protection scope of the present invention, and these will not affect the effect of the implementation of the present invention and the practicability of the patent. The scope of the claims of the present application shall be determined by the contents of the claims, and the description of the embodiments and the like in the specification shall be used to explain the contents of the claims.
Claims (7)
1. A contour error suppression method of a digital twin driven multi-axis numerical control machine tool is characterized by comprising the following steps: comprises the following steps of (a) carrying out,
s1, establishing a virtual model of a digital twin body corresponding to a physical entity, and obtaining a time-varying coupling mechanism model of the multi-axis feeding system by adopting a multi-parameter gain scheduling control strategy based on a global task coordinate system; establishing a data driving model of the multi-axis feeding system by utilizing signal testing and machine learning;
s2, synchronizing virtual and real, and establishing a bidirectional perception relationship between the digital twin body and the physical entity through a communication protocol with compatibility;
s3, dynamically estimating contour errors, and acquiring dynamic characteristic parameters, motion control parameters, nonlinear interference and contour shape parameters, wherein the dynamic characteristic parameters comprise natural frequency and damping ratio and are represented by Dy, the motion control parameters comprise various controller parameters and are represented by Co, the nonlinear interference comprises friction force, cutting force and inertia force and is represented by Di, the contour shape parameters comprise curvature and corners and are represented by Sh, and the kinematic parameters comprise speed, acceleration and jerk and are represented by Mo; performing order reduction characterization and generalization on the digital twin by adopting KPCA (kernel principal component analysis), Relief and LLE (Linear regression with edge) methods to obtain a dynamic mapping model between the contour error and various influence factors thereof, wherein CE is M (Dy, Co, Di, Sh and Mo); wherein CE represents the contour error, M represents the mapping relation between the contour error and the influencing factors thereof; training and correcting the precision of the dynamic mapping model by adopting Q-Learning and DRL algorithms to obtain an online estimation model of the contour error;
and S4, suppressing contour errors, obtaining optimal motion control parameters and the maximum track limiting speed corresponding to the optimal motion control parameters by adopting a multi-objective optimization algorithm according to the pre-estimated model, and performing interpolation control on the multi-axis feeding system by taking the maximum limiting speed and other conventional factors as constraints.
2. The digital twin driven multi-axis numerical control machine tool contour error suppressing method according to claim 1, characterized in that: the optimal motion control parameter and the maximum track limiting speed corresponding to the optimal motion control parameter are solved in the following flow, according to the pre-estimated model, the minimization of the contour error is taken as a target, the arch height error, the centripetal acceleration and the centripetal jerk are taken as constraints, and the cuckoo algorithm is adopted to carry out optimization in the feasible region of the limiting speed and the motion control parameter.
3. The digital twin driven multi-axis numerical control machine tool contour error suppressing method according to claim 2, characterized in that: the solving process of the optimal motion control parameters and the maximum track limiting speed corresponding to the optimal motion control parameters comprises the following steps of firstly, determining an input item according to a cuckoo algorithm rule, and calculating according to an online estimation model of a contour error to obtain an initial fitness value; then, for the situation that the iteration times do not reach the maximum, a group of old solutions are randomly selected, then a new solution is generated through Levy flight, whether the new solution is superior to the old solution or not is judged, and a better solution is selected; then, whether the solution abandoning factor is larger than the solution abandoning probability is judged, whether local search is carried out or not is determined, a new solution is generated, and the optimal solution and the optimal fitness value are updated according to the new solution; and circulating the steps until a global optimal solution and an optimal fitness value are obtained, wherein the global optimal solution is the maximum limiting speed of a parameter curve and the optimal motion control parameter under the multi-constraint condition including the contour error.
4. The digital twin driven multi-axis numerical control machine tool contour error suppressing method according to claim 3, characterized in that: s5, based on the maximum limiting speed of the obtained parameter curve, adopting a constant speed segmentation algorithm of the parameter curve to segment the curve, taking the point with local maximum feeding speed or the area between break points on the parameter curve as a key area, and dividing the parameter curve into a plurality of segments according to the key area; determining the minimum feeding speed of each key area as the constant value feeding speed of the area, and obtaining the segmented curve constraint and planning parameter information required by speed planning; and then, considering that the processing time is shortest, performing speed planning by adopting a bidirectional scanning look-ahead algorithm to obtain a smooth speed curve under the limitation of geometric constraint, dynamic performance constraint, maximum limit speed and minimum speed fluctuation, and taking the smooth speed planning and the optimal motion control parameter corresponding to the parameter curve as interpolation control input.
5. The digital twin driven multi-axis numerical control machine tool contour error suppressing method according to claim 4, characterized in that: the input items include: population size N, problem dimension D, solution abandoning probability Pa, upper/lower critical value L/U and maximum iteration number MaxGen, calculating the fitness value of the initial solution according to the information, and updating the optimal solution XbestAnd an optimal fitness value fminThen judging whether the iteration times t meet the maximum iteration times, if so, outputting a global optimal solution GXbestAnd an optimal fitness value Gfmin(ii) a If not, randomly selecting a solution, generating a new solution through Levy flight, then judging whether the new solution is superior to the old solution, and if so, replacing the old solution of the previous generation with the new solution; if not, keeping the previous generation old solution; then judging whether the random factor R is greater than the solution abandoning probability Pa, if so, abandoning the solution and locally searching to generate a new solution to obtain an updated current optimal solution XbestAnd an optimal fitness value fminThen judging whether the iteration times t meet the maximum iteration times again, and circulating; if not, obtaining an updated current optimal solution XbestAnd an optimal fitness value fminAnd then judging whether the iteration times t meet the maximum iteration times again, and circulating.
6. The digital twin driven multi-axis numerical control machine tool contour error suppressing method according to claim 5, characterized in that: the time-varying coupling mechanism model is obtained by adopting the Dalenberg theorem and Laplace transformation to obtain a multi-degree-of-freedom rigid-flexible coupling transfer function model of the feeding system,
wherein G isrigidRepresenting a rigid body transfer function matrix, GflexK represents a kth order elastomer transfer function matrix; through an identification experiment, a time-varying rigid-flexible coupling transfer function matrix of the feeding system is estimated by adopting a least square method and an orthogonal polynomial curve fitting method,
wherein the content of the first and second substances,
respectively representing displacement, speed and load quality, all of which are time-varying variables, then obtaining a time-varying coupling mechanism model of the multi-axis feeding system by adopting a multi-parameter gain scheduling control strategy based on a global task coordinate system, and representing the time-varying coupling mechanism model by using a state space model as follows:
7. the digital twin driven multi-axis numerical control machine tool contour error suppressing method according to claim 6, characterized in that: the data-driven model is obtained by: the servo driver acquires the displacement, the speed and the acceleration of each shaft, and a machine learning method is adopted to represent the influence relation of the reverse clearance on the pose change of the multi-shaft feeding system; aiming at the uncertainty of the inertial force and the cutting force, position and force data are collected by means of a servo driver and an external sensor, and a machine learning method is adopted to represent the dynamic response of the multi-axis feeding system under the action of the inertial force and the cutting force; the method comprises the steps of collecting displacement, speed and torque aiming at the jump of a reverse gap and friction, representing dead zones and creeping phenomena in the operation process of the multi-axis feeding system by adopting a Gaussian process regression method, carrying out big data analysis on various collected data, representing the pose, dynamic response, jump and the mutual influence relation of the pose, the dynamic response and the jump of the multi-axis feeding system under the action of nonlinear external interference by adopting a deep learning method, and obtaining a data driving model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110059640.XA CN112859739B (en) | 2021-01-15 | 2021-01-15 | Digital twin-driven multi-axis numerical control machine tool contour error suppression method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110059640.XA CN112859739B (en) | 2021-01-15 | 2021-01-15 | Digital twin-driven multi-axis numerical control machine tool contour error suppression method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112859739A CN112859739A (en) | 2021-05-28 |
| CN112859739B true CN112859739B (en) | 2022-07-01 |
Family
ID=76005936
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110059640.XA Active CN112859739B (en) | 2021-01-15 | 2021-01-15 | Digital twin-driven multi-axis numerical control machine tool contour error suppression method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112859739B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20230058974A1 (en) * | 2021-08-18 | 2023-02-23 | General Electric Company | Vulnerability-driven cyberattack protection system and method for industrial assets |
Families Citing this family (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113468724B (en) * | 2021-06-09 | 2023-01-17 | 中国民航大学 | Digital twin system simulation method and device for airport aircraft landing guidance |
| CN113486520B (en) * | 2021-07-08 | 2023-02-21 | 内蒙古民族大学 | Digital twinning-based dynamic regulation and control method, system and device for cutting technological parameters |
| CN113504768B (en) * | 2021-08-05 | 2022-06-03 | 东华大学 | High-precision product digital twin computability method for assembly quality prediction |
| CN114035425B (en) * | 2021-11-05 | 2022-07-26 | 广东工业大学 | Virtual and real state synchronization method and system for unmanned aerial vehicle digital twin system |
| CN114115131B (en) * | 2021-12-08 | 2023-02-03 | 中国科学院大学 | Time spline curve fitting and interpolation method applied to five-axis numerical control machine tool |
| CN114564880B (en) * | 2022-01-26 | 2022-11-25 | 南京理工大学 | Method for constructing digital twin module in additive manufacturing process |
| CN114609917B (en) * | 2022-05-11 | 2022-08-05 | 曜石机器人(上海)有限公司 | Servo driver and servo system based on digital twin technology |
| CN114880930B (en) * | 2022-05-11 | 2023-08-15 | 北京半人科技有限公司 | Online learning-based entity equipment digital twin modeling simulation prediction method |
| CN115576267B (en) * | 2022-11-07 | 2023-07-07 | 河北工业大学 | Wheel hub machining dimension error correction method based on digital twin |
| CN116679621B (en) * | 2023-07-31 | 2023-10-17 | 南京德克威尔自动化有限公司 | Multi-axis synchronous motion control system and control method thereof |
| CN117021118B (en) * | 2023-10-08 | 2023-12-15 | 中北大学 | Dynamic compensation method for digital twin track error of parallel robot |
| CN117330507B (en) * | 2023-10-12 | 2024-04-05 | 苏州星帆华镭光电科技有限公司 | Remote test control method for handheld laser instrument |
| CN117644431B (en) * | 2024-01-29 | 2024-04-02 | 南京航空航天大学 | CNC machine tool machining quality analysis method and system based on digital twin model |
| CN117908464A (en) * | 2024-03-20 | 2024-04-19 | 山东大学 | Machine tool contour error prediction method, system, device and readable storage medium |
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108107841A (en) * | 2017-12-26 | 2018-06-01 | 山东大学 | A kind of twin modeling method of numerically-controlled machine tool number |
| EP3367185A1 (en) * | 2017-02-28 | 2018-08-29 | Siemens Aktiengesellschaft | Drive system and its evaluation |
| CN108919760A (en) * | 2018-07-05 | 2018-11-30 | 长安大学 | A kind of intelligent workshop autonomy production process dynamic linkage control method twin based on number |
| WO2019055576A1 (en) * | 2017-09-12 | 2019-03-21 | Arconic Inc. | Systems and methods for performing calibration in additive manufacture |
| CN110704974A (en) * | 2019-09-30 | 2020-01-17 | 江苏科技大学 | Modeling and using method of process model based on digital twin drive |
| CN110865607A (en) * | 2019-11-07 | 2020-03-06 | 天津大学 | Five-axis numerical control machine tool control method based on digital twinning |
| CN110968038A (en) * | 2019-12-18 | 2020-04-07 | 大连理工大学 | Numerical control machine tool feed shaft thermal error monitoring method based on digital twinning |
| EP3696622A1 (en) * | 2019-02-14 | 2020-08-19 | Rockwell Automation Technologies, Inc. | Ai extensions and intelligent model validation for an industrial digital twin |
| CN111859566A (en) * | 2020-07-17 | 2020-10-30 | 重庆大学 | Surface roughness stabilizing method based on digital twinning |
| CN111964575A (en) * | 2020-07-06 | 2020-11-20 | 北京卫星制造厂有限公司 | Digital twin modeling method for milling of mobile robot |
| CN112162519A (en) * | 2020-10-21 | 2021-01-01 | 桂林电子科技大学 | Compound machine tool digital twin monitoring system |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109270899B (en) * | 2018-09-03 | 2020-12-25 | 江苏科技大学 | Digital twin-based marine diesel engine heavy part manufacturing process control method |
| US10921794B2 (en) * | 2019-04-10 | 2021-02-16 | Guangdong University Of Technology | Parallel control method based on multi-period differential sampling and digital twinning technologies |
| CN111752243B (en) * | 2020-06-12 | 2021-10-15 | 中国电子产品可靠性与环境试验研究所（（工业和信息化部电子第五研究所）（中国赛宝实验室）） | Production line reliability testing method and device, computer equipment and storage medium |
-
2021
- 2021-01-15 CN CN202110059640.XA patent/CN112859739B/en active Active
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP3367185A1 (en) * | 2017-02-28 | 2018-08-29 | Siemens Aktiengesellschaft | Drive system and its evaluation |
| WO2019055576A1 (en) * | 2017-09-12 | 2019-03-21 | Arconic Inc. | Systems and methods for performing calibration in additive manufacture |
| CN108107841A (en) * | 2017-12-26 | 2018-06-01 | 山东大学 | A kind of twin modeling method of numerically-controlled machine tool number |
| CN108919760A (en) * | 2018-07-05 | 2018-11-30 | 长安大学 | A kind of intelligent workshop autonomy production process dynamic linkage control method twin based on number |
| EP3696622A1 (en) * | 2019-02-14 | 2020-08-19 | Rockwell Automation Technologies, Inc. | Ai extensions and intelligent model validation for an industrial digital twin |
| CN110704974A (en) * | 2019-09-30 | 2020-01-17 | 江苏科技大学 | Modeling and using method of process model based on digital twin drive |
| CN110865607A (en) * | 2019-11-07 | 2020-03-06 | 天津大学 | Five-axis numerical control machine tool control method based on digital twinning |
| CN110968038A (en) * | 2019-12-18 | 2020-04-07 | 大连理工大学 | Numerical control machine tool feed shaft thermal error monitoring method based on digital twinning |
| CN111964575A (en) * | 2020-07-06 | 2020-11-20 | 北京卫星制造厂有限公司 | Digital twin modeling method for milling of mobile robot |
| CN111859566A (en) * | 2020-07-17 | 2020-10-30 | 重庆大学 | Surface roughness stabilizing method based on digital twinning |
| CN112162519A (en) * | 2020-10-21 | 2021-01-01 | 桂林电子科技大学 | Compound machine tool digital twin monitoring system |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20230058974A1 (en) * | 2021-08-18 | 2023-02-23 | General Electric Company | Vulnerability-driven cyberattack protection system and method for industrial assets |
| US11880464B2 (en) * | 2021-08-18 | 2024-01-23 | General Electric Company | Vulnerability-driven cyberattack protection system and method for industrial assets |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112859739A (en) | 2021-05-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112859739B (en) | Digital twin-driven multi-axis numerical control machine tool contour error suppression method | |
| Tong et al. | Real-time machining data application and service based on IMT digital twin | |
| CN108107841B (en) | Numerical twin modeling method of numerical control machine tool | |
| CN112904805B (en) | Digital twin modeling and virtual-real synchronization method for multi-axis numerical control machine tool | |
| Kadir et al. | Towards high-fidelity machining simulation | |
| Helu et al. | A standards-based approach for linking as-planned to as-fabricated product data | |
| Li et al. | A flexible manufacturing assembly system with deep reinforcement learning | |
| Brugnaro et al. | Adaptive robotic carving: training methods for the integration of material performances in timber manufacturing | |
| Finkeldey et al. | Real-time prediction of process forces in milling operations using synchronized data fusion of simulation and sensor data | |
| CN108803499A (en) | control device and machine learning device | |
| CN113518999A (en) | Semanteme-based production facility optimizing device with interpretability | |
| CN112416021B (en) | Learning-based path tracking prediction control method for rotor unmanned aerial vehicle | |
| Böttjer et al. | A review of unit level digital twin applications in the manufacturing industry | |
| Liu et al. | Predicting the position-dependent dynamics of machine tools using progressive network | |
| Schnelle et al. | Adaptive nonlinear model predictive control design of a flexible-link manipulator with uncertain parameters | |
| Gunal | Simulation for the better: The future in industry 4.0 | |
| Ganser et al. | DPART-A DIGITAL TWIN FRAMEWORK FOR THE MACHINING DOMAIN. | |
| WO2019100505A1 (en) | Computer numerical control method and system | |
| Mehta et al. | Smart manufacturing: state-of-the-art review in context of conventional and modern manufacturing process modeling, monitoring and control | |
| CN111103846A (en) | Numerical control machine tool state prediction method based on time sequence | |
| Plessing et al. | AI-Based Surface Roughness Prediction Model for Automated CAM-Planning Optimization | |
| Jiang et al. | Application of an optimized grey system model on 5-Axis CNC machine tool thermal error modeling | |
| Edoimioya et al. | Vibration compensation of delta 3D printer with position-varying dynamics using filtered B-splines | |
| Klingel et al. | Comparison and Application of Multi-Rate Methods for Real-Time Simulations of Production Systems | |
| Seidelson | Automation 4.0: The impact on manufacturing and engineering |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20240105 Address after: Building 2, 2nd Floor, No. 9 Juxing Road, Xianshuigu Town, Jinnan District, Tianjin, 300000 RMB Patentee after: TIANJIN TIANSEN INTELLIGENT EQUIPMENT Co.,Ltd. Address before: 409 Guangrong Road, Beichen District, Tianjin Patentee before: TIANJIN University OF COMMERCE |