WO2017129093A1 - 一种基于共形几何代数的机械臂运动规划的形式化分析方法及系统 - Google Patents

一种基于共形几何代数的机械臂运动规划的形式化分析方法及系统 Download PDF

Info

Publication number
WO2017129093A1
WO2017129093A1 PCT/CN2017/072210 CN2017072210W WO2017129093A1 WO 2017129093 A1 WO2017129093 A1 WO 2017129093A1 CN 2017072210 W CN2017072210 W CN 2017072210W WO 2017129093 A1 WO2017129093 A1 WO 2017129093A1
Authority
WO
WIPO (PCT)
Prior art keywords
robot
geometric
model
motion
relationship
Prior art date
Application number
PCT/CN2017/072210
Other languages
English (en)
French (fr)
Inventor
施智平
马莎
关永
邵振洲
张倩颖
王瑞
李晓娟
李黎明
Original Assignee
首都师范大学
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 首都师范大学 filed Critical 首都师范大学
Priority to US15/780,138 priority Critical patent/US10650179B2/en
Publication of WO2017129093A1 publication Critical patent/WO2017129093A1/zh

Links

Images

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00Computer-aided design [CAD]
    • G06F30/30Circuit design
    • G06F30/32Circuit design at the digital level
    • G06F30/33Design verification, e.g. functional simulation or model checking
    • G06F30/3323Design verification, e.g. functional simulation or model checking using formal methods, e.g. equivalence checking or property checking
    • BPERFORMING OPERATIONS; TRANSPORTING
    • B25HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
    • B25JMANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
    • B25J19/00Accessories fitted to manipulators, e.g. for monitoring, for viewing; Safety devices combined with or specially adapted for use in connection with manipulators
    • B25J19/0025Means for supplying energy to the end effector
    • B25J19/0029Means for supplying energy to the end effector arranged within the different robot elements
    • BPERFORMING OPERATIONS; TRANSPORTING
    • B25HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
    • B25JMANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
    • B25J9/00Programme-controlled manipulators
    • B25J9/16Programme controls
    • B25J9/1602Programme controls characterised by the control system, structure, architecture
    • BPERFORMING OPERATIONS; TRANSPORTING
    • B25HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
    • B25JMANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
    • B25J9/00Programme-controlled manipulators
    • B25J9/16Programme controls
    • B25J9/1602Programme controls characterised by the control system, structure, architecture
    • B25J9/1605Simulation of manipulator lay-out, design, modelling of manipulator
    • BPERFORMING OPERATIONS; TRANSPORTING
    • B25HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
    • B25JMANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
    • B25J9/00Programme-controlled manipulators
    • B25J9/16Programme controls
    • B25J9/1656Programme controls characterised by programming, planning systems for manipulators
    • B25J9/1664Programme controls characterised by programming, planning systems for manipulators characterised by motion, path, trajectory planning
    • BPERFORMING OPERATIONS; TRANSPORTING
    • B25HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
    • B25JMANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
    • B25J9/00Programme-controlled manipulators
    • B25J9/16Programme controls
    • B25J9/1679Programme controls characterised by the tasks executed
    • B25J9/1682Dual arm manipulator; Coordination of several manipulators
    • GPHYSICS
    • G05CONTROLLING; REGULATING
    • G05BCONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B2219/00Program-control systems
    • G05B2219/30Nc systems
    • G05B2219/39Robotics, robotics to robotics hand
    • G05B2219/39079Solve inverse differential kinematics in closed, feedback loop, iterate
    • GPHYSICS
    • G05CONTROLLING; REGULATING
    • G05BCONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B2219/00Program-control systems
    • G05B2219/30Nc systems
    • G05B2219/40Robotics, robotics mapping to robotics vision
    • G05B2219/40371Control trajectory to avoid joint limit as well as obstacle collision
    • GPHYSICS
    • G05CONTROLLING; REGULATING
    • G05BCONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B2219/00Program-control systems
    • G05B2219/30Nc systems
    • G05B2219/40Robotics, robotics mapping to robotics vision
    • G05B2219/40519Motion, trajectory planning

Definitions

  • the present application relates to a formal analysis method and system for mechanical arm motion planning based on conformal geometric algebra, and belongs to the field of computer science and technology.
  • Conformal Geometric Algebra is an advanced geometric representation and calculation system that provides a simple, intuitive and uniform homogeneous algebraic framework for classical geometry.
  • the CGA expresses the reference origin (e 0 ) and the infinity point (e ⁇ ) by adding two dimensions, so that the Euclidean space is embedded in the conformal space and given its inner product structure, which not only preserves the homogeneous space
  • the product's Grassmann structure makes the inner product have a clear geometric meaning that characterizes the basic metrics such as distance and angle.
  • CGA not only successfully solved the problem of how to use geometric language to complete geometric calculations, but also played a key role in solving the geometric problems of many high-tech fields such as engineering and computer science.
  • CGA is different from previous algebras in that it is not a number, but a geometry.
  • the robot research object is the geometric relationship formed by the system based on the basic geometry. Therefore, conformal geometric algebra is unique in robot research.
  • the purpose of the present application is to provide a formal analysis method and system for motion planning of a robot arm based on conformal geometric algebra, so as to solve the complicated calculation and inaccurate result when analyzing the motion planning of the robot arm in the prior art.
  • the problem is to provide a formal analysis method and system for motion planning of a robot arm based on conformal geometric algebra, so as to solve the complicated calculation and inaccurate result when analyzing the motion planning of the robot arm in the prior art.
  • a formal analysis method for motion planning of a robot arm based on conformal geometric algebra comprising:
  • Determining whether the motion logic relationship is established establishing an attribute indicating that the geometric relationship logic model satisfies the motion process to be verified or having the motion process to be verified, and not establishing that the geometric relationship logic model does not satisfy the waiting Verify the constraints of the motion process or do not have the properties of the motion process to be verified.
  • a formal analysis system for manipulator motion planning based on conformal geometric algebra which includes:
  • a parameter determination module for determining specific structural parameters and motion planning parameters of the robot
  • the motion planning parameters are based on the conformal geometric algebra theory to construct the corresponding geometric model for the basic components and motion planning constraints of the robot.
  • the constructed geometric model is described by high-order logic language;
  • the geometric relationship logic model building module of the robot motion process is used to formally model the motion process of the robot based on the constructed geometric model, and obtain a geometric relationship logic model of the motion process of the robot;
  • a logical proposition building module configured to combine the geometric relationship logic model to obtain a constraint of a motion process of the robot to be verified or a motion logic relationship corresponding to the attribute;
  • a proof module configured to verify whether the motion logic relationship is established, establishing a property indicating that the geometric relationship logic model satisfies the constraint of the motion process to be verified or having the motion process to be verified, and the logical relationship model indicating the geometric relationship is not established The constraint of the motion process to be verified or the property of the motion process to be verified is not satisfied.
  • a storage medium is provided, wherein the storage medium is for storing executable program code for performing the aforementioned conformal geometric algebra-based robot arm motion planning at runtime Formal analysis method.
  • an application is provided, wherein the application is for performing a formal analysis method of the aforementioned conformal geometric algebra-based robot arm motion planning at runtime.
  • an electronic device comprising:
  • processor a memory, a communication interface, and a bus
  • the processor, the memory, and the communication interface are connected by the bus and complete communication with each other;
  • the memory stores executable program code
  • the processor runs a program corresponding to the executable program code by reading executable program code stored in the memory for performing the foregoing formal analysis method of conformal geometric algebra-based robot arm motion planning .
  • the constraint relationship of the motion process to be verified or the motion logic relationship corresponding to the attribute is obtained according to the geometric relationship logic model, and verified. Whether the above motion logic relationship is established. Since the above method adopts a mathematical method for correctness verification, it is accurate and complete for the nature of verification. In addition, CGA can be used for geometric elements such as points, lines, faces, circles, spheres, and these geometric elements.
  • the unified modeling and processing of rotation and translation has a strong advantage in dealing with robot kinematics and motion planning problems, which can improve the dimensionality of problem solving, thereby simplifying the coupling in robot computing and thus reducing computational complexity.
  • FIG. 3 is a schematic diagram of calculation of a target circle Z t in the present application.
  • Figure 4 is a schematic diagram of calculation of the mechanical claw circle Z h in the present application.
  • FIG. 5 is a schematic diagram of the solution of the rotation operator and the translation operator in the present application.
  • the currently developed conformal geometric algebra system can be generalized to any dimension of space.
  • the main application is the 5-dimensional conformal space.
  • the conformal geometric algebra provides the Grassmann structure representing the geometry, the unity of the geometric transformation, and the representation geometry.
  • the amount of parentheses and invariant systems show significant advantages in geometric data processing and geometric calculations.
  • Different transformation expressions in the geometric algebra framework The and computational forms are uniform, and the geometrical and geometrical properties of the original geometric objects can be preserved after geometric spatial transformation.
  • the object of robot research is the geometric relationship formed by the system based on basic geometry. Unlike the previous algebra, the object of conformal geometric algebra is not a number, it is a geometry, so the conformal geometric algebra is unique in robot research.
  • Conformal geometric algebra can also construct the traditional methods of researching robots such as Lie group Lie algebra, Wu method, Clifford algebra and spiral vector, and apply it to the analysis and synthesis of institutions, which promotes the development of institutional science and forms a common
  • the theoretical method of sex reduces the complexity of institutional modeling and derivation calculation.
  • a simple mathematics can be obtained in the modeling of complex multi-degree-of-freedom parallel mechanism. model.
  • the geometry representation and construction under the CGA framework can be based on inner and outer products, respectively.
  • the geometric expression based on the outer product mainly reflects the mutual construction relationship between different levels of geometric shapes, while the inner product form expression can construct the corresponding parameter equation by measuring the parameters such as distance and angle.
  • the expression of the two can be converted by the dual operation between the inner and outer products.
  • Table 1 gives the geometry representations based on the inner product and based on the outer product.
  • the two representations are referred to by many literatures as standard representations and direct representations.
  • the standard inner product in the Euclidean space that is, the scalar product
  • e 1 , e 2 , and e 3 are unit orthogonal bases in the European three-dimensional space.
  • high-dimensional geometry is constructed mainly by the outer point " ⁇ " connecting the points ⁇ P i ⁇ on the geometry.
  • a ball can be expressed by four points on the ball.
  • the meaning of the outer product is different, indicating the intersection between the geometric bodies. For example, a circle can be formed by the intersection of two balls.
  • the embodiment of the present application provides a formal analysis method for motion planning of a robot arm based on conformal geometric algebra.
  • this method first, specific structural parameters and motion planning parameters of the robot are determined; The parameters, motion planning parameters, and the corresponding geometric models are constructed based on the conformal geometric algebra theory for the basic components and motion planning constraints of the robot.
  • the constructed geometric models are described by high-level logic language; and based on the geometric model
  • the robot's motion process is formally modeled, and the geometric relationship logic model of the robot's motion process is obtained.
  • the geometric relationship logic model is used to obtain the constraint of the motion process of the robot to be verified or the motion logic relationship corresponding to the attribute; finally verify the above motion Whether the logical relationship is established or not, the establishment indicates that the geometric model of the geometric relationship satisfies the constraint of the motion process to be verified or the attribute of the motion process to be verified, and the failure to establish indicates that the geometric model of the geometric relationship does not satisfy the constraint of the motion process to be verified or does not have to be verified.
  • Property procedures are used to obtain the constraint of the motion process of the robot to be verified or the motion logic relationship corresponding to the attribute.
  • the constraints or attributes of each movement process of the robot reflect to some extent the control, operation and control, operation and so on that can be realized by the robot during the movement.
  • the constraints or attributes of each motion process of the robot correspond to a motion logic relationship.
  • different robots have different geometric features, and different geometric features will affect the motion logic relationship. Therefore, when determining the motion logic relationship of each motion process constraint or attribute of the robot, it is necessary to combine the geometry of the robot.
  • the relational logic model is determined. For example, the two claw ends of a certain mechanical claw can be opened to 180 degrees, but due to the limitation of the working space, at a certain position in the space such as the position Q, the two claw ends can only be opened by 120 degrees.
  • the constraint of the motion process to be verified or the motion logic relationship corresponding to the property obtained by the robot can be understood as: the above-mentioned geometric relationship logic model, the motion process to be verified A logical proposition consisting of constraints or attributes.
  • the logic inference engine may be used to verify whether the motion logic relationship is established.
  • the motion process of the robot is formally modeled based on the constructed geometric model, and the geometric relationship logic model of the motion process of the robot is obtained until the logical logic inference engine is used to verify whether the motion logic relationship is established.
  • the formal model of CGA is established by HOL-Light, and the formal process of the CGA is formalized based on the established formal model of CGA, and the geometric relational logic model is obtained, and then the constraints or attributes of the motion process are verified. Modeling, to obtain the above-mentioned motion logic relationship, and finally based on the logic inference engine in the CGA formal model, to verify whether the formal model of the above-mentioned motion logic relationship is established.
  • FIG. 1 is a flow chart of formal modeling and verification of a robot grabbing object based on the conformal geometric algebra CGA method proposed in the present application.
  • the main idea of the robot operation based on the conformal geometric algebra CGA method is the CGA geometric motion transformation expression.
  • the transformed geometric object of the present application is a reference circle rather than a reference point, and the geometrical features of the mechanical claw and the grasped object are represented by circles.
  • the high-order logic theorem prover HOL-Light is used as a formalization tool.
  • HOL-Light is one of the most popular theorem pros and not only has a large research team and user group, but also contains a wealth of mathematical theorem libraries such as real number analysis library, transcendental function library, integral differential library, etc. and a series of efficient proof strategies. .
  • specific structural parameters and motion planning parameters of the robot component are determined.
  • the specific structural parameters may be: joint parameters, link parameters, reachable range, joint surface, etc.
  • the motion planning parameters may be translation vectors, rotations. Angle and so on; then construct corresponding geometric model based on conformal geometric algebra theory for robot components and motion planning constraints.
  • a point model can be constructed based on joint information
  • a line model can be constructed according to the link parameters
  • a ball model can be constructed according to the reachable range.
  • the surface model is constructed according to the joint surface, the rigid body motion model is constructed according to the motion planning parameters, etc.; the mechanical arm is grasped by the serial mechanical arm as the motion process to be verified, and the description information of the constraints and attributes of the motion process to be verified is determined, and HOL-Light is input. And verifying the motion process to be verified, for example, determining, by the attribute information of the motion process to be verified, whether the robot arm firmly grasps the grasped object, thereby achieving target verification.
  • the motion process of the robot includes not only the motion process of the robot arm grabbing the object, but also other motion processes, for example, the robot arm movement process of the robot, etc., the application only refers to the process of grasping the object.
  • the description is made by way of example and is not intended to limit the application.
  • the present application proposes a formal analysis method for motion planning of a robot arm based on conformal geometric algebra, which includes:
  • the corresponding geometric model is constructed for the basic components and motion planning constraints of the robot, and the geometric model established by the high-order logic language is used to form the basic geometric logic model system of the robot.
  • the establishment indicates that the model satisfies the constraint or has the attribute
  • the failure to establish indicates that the model does not satisfy the constraint or does not have the attribute.
  • the basic components of the robot are generally described by the specific structural parameters of the components
  • the motion planning constraints of the robot are generally described by the motion planning parameters of the motion planning, for which the robot is constructed.
  • the above specific structural parameters and motion planning parameters must be determined, so that the geometric models corresponding to the basic components and motion planning constraints can be constructed more clearly and effectively.
  • Conformal geometric algebra theory is the theoretical basis of the formal analysis method of robot motion planning.
  • the formal analysis method of robot motion planning can be applied to the general method of specific robot problem analysis.
  • the logical inference engine realizes the formal system of conformal geometric algebra and The logical expression of the formal analysis method of robot motion planning and the proof of reasoning calculus.
  • the geometric relationship logic model may be combined and the motion to be verified
  • the constraint of the motion process to be verified or the motion logic relationship corresponding to the attribute may refer to: refer to the constraint of the motion process to be verified or the logic formula corresponding to the attribute to obtain the above geometry.
  • the formal modeling of the motion planning process of the specific robot motion planning based on the basic geometric logic model system specifically includes the formal modeling of the motion planning process of the robot grabbing object.
  • the specific robot is a n-degree-of-freedom series mechanical arm, and the end includes a T-shaped mechanical claw.
  • the specific robot motion planning is the motion planning process of the robot arm grasping the object.
  • the above-mentioned step "describe the constraint or attribute of the motion process of the robot to be verified by the logic formula" is an attribute of the robot motion process that needs to be verified for the mechanical claw to achieve the object grasping success and grasping.
  • the geometric model constructed includes: a point model in which the robot joint is abstracted, a ball model in which the end of the robot joint can be abstracted, a surface model in which the auxiliary surface of the robot joint is abstracted, a line model formed by connecting the joint points of the robot, and the basic model of the robot
  • the constraint relationship of components is abstracted into The geometry intersection model, the geometric rotation model of the robot basic component rotation motion, the geometric pure translation model of the robot basic component translation motion abstraction, the geometric joint rigid motion model abstracted into the desired position of the robot joint end, and the robot basic component measurement relationship
  • the abstracted inter-geometry distance model and the geometrical angle model of the abstraction of the basic components of the robot is abstracted, a point model in which the robot joint is abstracted, a ball model in which the end of the robot joint can be abstracted, a surface model in which the auxiliary surface of the robot joint is abstracted, a line model formed by connecting the joint points of the robot, and the basic model of the robot
  • the geometry includes points, lines, faces, circles, spheres, point pairs, etc., which are used to represent the basic components of the robot.
  • the robot joint is abstracted into points, and the above mathematical expression is a standard expression of the point in the conformal geometric algebra.
  • the point_CGA function in this model represents the mapping from the 3D Euclidean space to the point in the 5D conformal space.
  • e 1 , e 2 , and e 3 are unit orthogonal groups in the European three-dimensional space.
  • s 1 , s 2 , and s 3 are coefficients
  • uppercase S is an expression of a point s from the Euclidean space to a point in the conformal space.
  • the input variable of the function point_CGA is a three-dimensional vector in Euclidean space, and its return value is a multiple vector in the five-dimensional conformal space.
  • the conformal geometric algebra can be constructed by the geometric algebra Cl 4,1 , its data type is defined as real ⁇ (4,1) multivector, where s$n represents the nth element of the multidimensional vector s, the mbasis function represents the basic scalar function, the dot function represents the dot product in Euclidean space, and the null_inf and null_zero functions represent the zero vector e 0 respectively And e ⁇ :
  • the point_CGA function in this model represents the mapping from the 3D Euclidean space to the point in the 5D conformal space.
  • the end-of-range range of the robot joint is abstracted into a sphere S with the joint as the radius of the spherical core.
  • the above mathematical expression is the standard expression of the sphere in the conformal geometric algebra.
  • the input variables p, r of the function sphere_CGA represent the center of the sphere and the radius, respectively, where the sphere can be represented by a three-dimensional vector in Euclidean space.
  • the auxiliary surface of the robot joint is abstracted into the surface ⁇ .
  • the above mathematical expression is the standard expression of the face in the conformal geometric algebra.
  • the input variables n, d of the function plane_CGA represent the distances of the auxiliary surface normal vector and the auxiliary surface to the origin, respectively, wherein the auxiliary surface normal vector can be represented by a three-dimensional vector in Euclidean space.
  • the above mathematical expression is an expression of a line L * composed of any two points in the conformal geometric algebra, and can be used for the representation of lines or other auxiliary lines formed by joint points in the robot, wherein A and B respectively represent the conformal geometric space. Two points.
  • the model is used to abstract the links between the two joints into a line.
  • the above expression is a direct expression of the line in the conformal geometric algebra.
  • the input variables a and b of the function line_direct_CGA represent two points on the line, respectively, and the data type is real ⁇ (4,1)multivector, where the function outer represents the outer product in the geometric algebra.
  • the operation, the outer product operation can realize the construction from low-dimensional geometry to high-dimensional geometry, that is, the expansion operation.
  • the model utilizes an outer product to implement an intersection operation representing the geometry of the basic components of the robot.
  • the input variable of the function MEET represents any geometry to be intersected, and the data type is real ⁇ (4,1)multivector.
  • the constraint relationship includes the joint on the connecting rod (ie, the point is on the line), the connecting rod on the auxiliary surface where the joint is located (the line is on the surface), and the like.
  • the model uses a geometric product to implement a geometric rotation transformation, and R is a rotation operator in a conformal geometric algebra.
  • Is R geometric trans, L represents a rotation axis, [Phi] is the rotation angle, o representation of the expression before rotating geometry, o
  • Rotated denotes an input variable expressions rotated geometry
  • function rotation_CGA of t, l respectively represent rotation pairs Angle and rotation axis, this function implements the function of the rotation operator.
  • the input variables x, t, and l of the function pure_rotationed_CGA represent the geometry, the rotation angle, and the rotation axis.
  • the return value of the function is the rotated geometry representation, where the function reversion represents the geometric inverse.
  • the input variables x, t of the function pure_translationed_CGA represent the geometry, the translation vector, and the return value of the function is the geometric representation of the translation.
  • the model uses the geometric product to realize the geometric rigid body transformation, where R and T are the rotation operator and the translation operator respectively.
  • R and T are the rotation operator and the translation operator respectively.
  • M is the Motor operator.
  • o is the expression of the geometry before the rigid body motion
  • o rigid_body_motion represents the geometric representation of the motion
  • the input variables x, a, l, t of the function motor_CGA represent the geometry
  • rotation angle, rotation axis, translation Vector this function implements the function of the motor operator.
  • the return value of the function rigid_body_motion is represented by the geometry after the rigid body motion.
  • the expression indicates that the geometry is first translated and then rotated, but the motor operator has the combination, and the rotation operator and the translation operator can be exchanged.
  • a and B are the expressions of point a and point b in the conformal space respectively.
  • the above model is the point distance expression in the conformal geometric algebra, where dist(a , b) represents the two-point distance, which can be expressed by the inner product of geometric algebra.
  • the dotted line distance, the penalty distance, and the ball distance also have specific expressions in the conformal geometric algebra, which can be used in the specific calculation in the robot problem.
  • the geometry angle model abstracted by the angle of the basic components of the robot is used to calculate the angle ⁇ of the basic components of the robot.
  • the model is an expression of the angle ⁇ of the geometry in the conformal geometric algebra.
  • the input variables o 1 and o 2 of the function vector_angles_CGA respectively represent the geometry of the angle to be found, which can be a line or a surface, and the data type is real ⁇ (4, 1) multivector, where the function inner represents the left-contraction operation in the geometric algebra, the $$ ⁇ represents the magnitude of the scalar of the 0-order subspace, and the function mult_norm functions as the modulo operation of the multiple vector.
  • This model can be used to calculate the angle between the basic components of the robot, o 1 * , o 2 * , Represents the duality of o 1 and o 2 respectively.
  • the method may specifically include:
  • the motion process of the robot arm may include various forms.
  • the following is a formalization of the motion planning of the manipulator based on the conformal geometric algebra provided by the embodiment of the present application.
  • the analytical method is described in detail.
  • the specific robot in the method is an n-degree-of-freedom series mechanical arm, and the end includes a T-shaped mechanical claw;
  • the specific robot motion planning in the method is a motion planning process of the robot arm grasping the object
  • the property of the motion process to be verified in the method is that the mechanical claw achieves successful grasping of the object and grasps it.
  • FIG. 2 is a flow chart showing the formal modeling of the motion planning process for a robot to grab an object in the present application. As shown in Figure 2, it includes:
  • Step 30 Extract feature points of the object and the mechanical claw respectively
  • Step 31 Calculate a target circle where the grabbed position of the object is located
  • the positional orientation of the object is obtained by four feature points (x1, x2, x3, x4) of the edge of the object, where x1, x2, and x3 are any three points on the bottom edge of the object, and x4 is an object. Any point on the top edge.
  • a reference circle consisting of three points at the bottom of the object Can be obtained from the direct expression of the circle:
  • Ic is the pseudo-scalar of the conformal geometric algebra.
  • the pseudo-scalar can realize the dual operation, and the dual operation realizes the mutual conversion of the two representation methods of the geometry.
  • the target circle Z t should be the bottom reference circle to translate in the - ⁇ b direction.
  • the corresponding translation operator T is:
  • x1 x2 x3 x4 represents the four feature points of the object
  • the HOL type is real ⁇ 3, which represents the three-dimensional Euclidean vector
  • the feature points are embedded into the five-dimensional total by the three-dimensional Euclidean space through the function point_CGA In the shape space.
  • the target circle formed by the object being grasped is realized by the pure translation function pure_translationed_CGA.
  • the target circle calculated in this step can be understood as the aforementioned first target circle, since the mechanical claw of the robot picks up the grasped object, and generally the claw end of the mechanical grip contacts the object.
  • Step 32 Calculate the circle where the mechanical claw is located
  • the auxiliary ball S h calls the ball to represent the standard function of sphere_CGA, (ph:real ⁇ 3) and (r:real) respectively represent the center of the auxiliary ball P h and the radius r, and the auxiliary face where the mechanical claw is located
  • the direct representation function of the call surface, plane_direct_CGA, is implemented. (a:real ⁇ 3) and (b:real ⁇ 3) represent the two points on the gripper.
  • the dual operator uses the DUAL function.
  • Step 33a Calculate a translation operator of the mechanical claw grasping object, the translation operator including a translation axis and a translation length;
  • the translation length should be the center distance of the target circle Z t and the circle Z h where the mechanical claw is located, and the translation axis is a straight line passing through the two centers.
  • the translation axis Since the translation axis is determined by two centers, the translation axis can be calculated by a direct expression.
  • center_point_pt x1 x2 x3 x4 (circle_zt x1 x2 x3 x4)*null_inf*(circle_zt x1 x2 x3 x4)
  • Step 33b calculating a rotation operator of the mechanical claw grasping the object, including a rotation axis and a rotation angle;
  • the mechanical axis captures the axis of rotation of the object to meet several constraints:
  • the angle of rotation is easy to know as the angle between the two axes, which is obtained by the angle formula:
  • the solution of the rotation angle ⁇ is calculated using the CGA angle formula, which is implemented by the function vector_angles_CGA.
  • Step 34 Calculate the new target position of the mechanical claw
  • R and T are rotation operators and translation operators, respectively.
  • the new target position of the mechanical claw can be calculated by first rotating and then shifting.
  • the rigid body operator can be calculated as:
  • Z 'h is a new target position of the gripper, For the geometry of R, For the geometry of T.
  • the circle z h formed by the mechanical claw surface is first rotated by the pure rotation function pure_rotationed_CGA, and the input variable of the function is the rotated geometry z h , the rotation angle ⁇ and the rotation axis, ie (circle_zh ph rab), (rotation_angle x1 x2 x3 x4ph rab) and (--DUAL(dual_rotation_axis x1 x2 x3 x4 ph rab)), and then the rotated circle
  • the translational motion is done by the pure translation function pure_translationed_CGA, the function input variable is the rotated geometry Translation vector dl T .
  • the main content of the above code involves the conformal geometric algebra CGA geometry representation, the inter-geometry distance feature and the formalization of the geometry motion transformation.
  • the mechanical claw circle Z h is stepped closer to the target circle Z t of the object position, and finally the mechanical claw successfully grasps the object and grasps a certain geometric constraint relationship.
  • the constraint relationship is that the new position of the circle Z h of the mechanical claw plane should finally coincide with the position of the target circle Z t of the object, that is, the expression should satisfy an equal relationship in the conformal geometric algebra CGA.
  • the geometric constraint relationship of the target verification can be established in HOL-light:
  • the method for formal modeling and verification of the motion planning of the robot based on the CGA in the solution provided by the above various embodiments can be loaded and used in the HOL-Light.
  • the biggest difficulty is that the theorem proving technology requires a lot of human-computer interaction, the workload is large, and the time is long.
  • the modeling process emphasizes the user's familiarity with the CGA theoretical knowledge of conformal geometric algebra.
  • the verification process requires strict thinking. And a certain logical reasoning experience.
  • the constraint relationship of the motion process to be verified or the motion logic relationship corresponding to the attribute is obtained according to the geometric relationship logic model, and Verify that the above motion logic relationship is true. Since the above method adopts a mathematical method for correctness verification, it is accurate and complete for the nature of verification.
  • CGA can uniformly model and process the geometric elements such as points, lines, faces, circles, spheres, and the rotation and translation of these geometric elements. It has a strong advantage in dealing with robot kinematics and motion planning problems, and can improve Solve the dimensionality of the problem, simplifying the coupling in robotic calculations and reducing computational complexity.
  • the verification accuracy can be improved while reducing the computational complexity, and not only the advantages of the CGA and the formalized methods are fully utilized.
  • the combination of the two strengthens each other's advantages.
  • the embodiment of the present application further provides a formal analysis system for the manipulator motion planning based on the conformal geometric algebra.
  • the system includes:
  • a parameter determination module for determining specific structural parameters and motion planning parameters of the robot
  • a basic geometric logic model building module of the robot configured to construct a corresponding geometric model according to the specific structural parameter, the motion planning parameter, and the basic component and motion planning constraint of the robot based on the conformal geometric algebra theory, wherein the constructed geometric model Geometric models are described in higher-order logic languages;
  • the geometric relationship logic model building module of the robot motion process is used to formally model the motion process of the robot based on the constructed geometric model, and obtain a geometric relationship logic model of the motion process of the robot;
  • a logical proposition building module configured to combine the geometric relationship logic model to obtain a constraint of a motion process of the robot to be verified or a motion logic relationship corresponding to the attribute;
  • a proof module configured to verify whether the motion logic relationship is established, establishing a property indicating that the geometric relationship logic model satisfies the constraint of the motion process to be verified or having the motion process to be verified, and the logical relationship model indicating the geometric relationship is not established The constraint of the motion process to be verified or the property of the motion process to be verified is not satisfied.
  • the constraint relationship of the motion process to be verified or the motion logic relationship corresponding to the attribute is obtained according to the geometric relationship logic model, and verified. Whether the above motion logic relationship is established. Since the above method adopts a mathematical method for correctness verification, it is accurate and complete for the nature of verification. In addition, CGA can uniformly model and process the geometric elements such as points, lines, faces, circles, spheres, and the rotation and translation of these geometric elements. It has a strong advantage in dealing with robot kinematics and motion planning problems, and can improve Solve the dimensionality of the problem, simplifying the coupling in robotic calculations and reducing computational complexity.
  • the verification accuracy can be improved while reducing the computational complexity, and not only the advantages of the CGA and the formalized mode are fully exerted, The combination of the two strengthens each other's advantages.
  • the present application also provides a storage medium, wherein the storage medium is for storing executable program code for performing a conformal geometric algebra described herein at runtime
  • the formal analysis method for the manipulator motion planning includes:
  • the robot's motion process is formally modeled, and the geometric relationship logic model of the robot's motion process is obtained.
  • Determining whether the motion logic relationship is established establishing an attribute indicating that the geometric relationship logic model satisfies the motion process to be verified or having the motion process to be verified, and not establishing that the geometric relationship logic model does not satisfy the waiting Verify the constraints of the motion process or do not have the properties of the motion process to be verified.
  • the constraint relationship of the motion process to be verified or the motion logic relationship corresponding to the attribute is obtained according to the geometric relationship logic model, and Verify that the above motion logic relationship is true. Since the above method adopts a mathematical method for correctness verification, it is accurate and complete for the nature of verification. In addition, CGA can uniformly model and process the geometric elements such as points, lines, faces, circles, spheres, and the rotation and translation of these geometric elements. It has a strong advantage in dealing with robot kinematics and motion planning problems, and can improve Solve the dimensionality of the problem, simplifying the coupling in robotic calculations and reducing computational complexity.
  • the verification accuracy can be improved while reducing the computational complexity, and not only the advantages of the CGA and the formalized mode are fully exerted, The combination of the two strengthens each other's advantages.
  • the present application further provides an application, wherein the application is configured to perform a formal analysis method of a conformal geometric algebra-based robot arm motion planning as described in the present application, wherein A formal analysis method for motion planning of a robot arm based on conformal geometric algebra described in the present application, comprising:
  • the robot's motion process is formally modeled, and the geometric relationship logic model of the robot's motion process is obtained.
  • Determining whether the motion logic relationship is established establishing an attribute indicating that the geometric relationship logic model satisfies the motion process to be verified or having the motion process to be verified, and not establishing that the geometric relationship logic model does not satisfy the waiting Verify the constraints of the motion process or do not have the properties of the motion process to be verified.
  • the geometric relationship logic model is modeled in a formal manner, and the constraint relationship of the motion process to be verified or the motion logic relationship corresponding to the attribute is obtained according to the geometric relationship logic model, and the above-mentioned motion logic relationship is verified. Whether it is established. Since the above method adopts a mathematical method for correctness verification, it is accurate and complete for the nature of verification. In addition, CGA can uniformly model and process the geometric elements such as points, lines, faces, circles, spheres, and the rotation and translation of these geometric elements. It has a strong advantage in dealing with robot kinematics and motion planning problems, and can improve Solve the dimensionality of the problem, simplifying the coupling in robotic calculations and reducing computational complexity.
  • the verification accuracy can be improved while reducing the computational complexity, and not only the advantages of the CGA and the formalized mode are fully exerted, The combination of the two strengthens each other's advantages.
  • an electronic device including:
  • processor a memory, a communication interface, and a bus
  • the processor, the memory, and the communication interface are connected by the bus and complete communication with each other;
  • the memory stores executable program code
  • the processor runs a program corresponding to the executable program code by reading executable program code stored in the memory for performing a conformal geometric algebra-based robotic arm as described herein
  • a formal analysis method for motion planning wherein a formal analysis method for motion planning of a robot arm based on conformal geometric algebra described in the present application includes:
  • the robot's motion process is formally modeled, and the geometric relationship logic model of the robot's motion process is obtained.
  • Determining whether the motion logic relationship is established establishing an attribute indicating that the geometric relationship logic model satisfies the motion process to be verified or having the motion process to be verified, and not establishing that the geometric relationship logic model does not satisfy the waiting Verify the constraints of the motion process or do not have the properties of the motion process to be verified.
  • the geometric relationship model After the electronic device provided by the embodiment is modeled in a formal manner to obtain a geometric relationship logical model, the geometric relationship model obtains a constraint of the motion process to be verified or a motion logic relationship corresponding to the attribute, and verifies whether the motion logic relationship is established. . Since the above method adopts a mathematical method for correctness verification, it is accurate and complete for the nature of verification. In addition, CGA can uniformly model and process the geometric elements such as points, lines, faces, circles, spheres, and the rotation and translation of these geometric elements. It has a strong advantage in dealing with robot kinematics and motion planning problems, and can improve Solve the dimensionality of the problem, simplifying the coupling in robotic calculations and reducing computational complexity.
  • the verification accuracy can be improved while reducing the computational complexity, and not only the advantages of the CGA and the formalized mode are fully exerted, The combination of the two strengthens each other's advantages.
  • the description is relatively simple, and the relevant parts can be referred to the description of the method embodiment.

Landscapes

  • Engineering & Computer Science (AREA)
  • Robotics (AREA)
  • Mechanical Engineering (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Hardware Design (AREA)
  • Physics & Mathematics (AREA)
  • Automation & Control Theory (AREA)
  • Evolutionary Computation (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Geometry (AREA)
  • Manipulator (AREA)
  • Numerical Control (AREA)

Abstract

一种基于共形几何代数的机械臂运动规划的形式化分析方法和系统。所述方法包括:确定机器人的具体结构参数和运动规划参数;基于共形几何代数理论对机器人的基本构件、运动规划约束构建对应的几何模型,其中,所购建几何模型以高阶逻辑语言进行描述;基于所构建几何模型对机器人的运动过程进行形式化建模,得到机器人的运动过程的几何关系逻辑模型;获得机器人的待验证运动过程的约束或属性对应的运动逻辑关系;验证所述运动逻辑关系是否成立。应用本方法进行分析,提高了验证精度,降低了计算复杂度。

Description

一种基于共形几何代数的机械臂运动规划的形式化分析方法及系统
本申请要求于2016年1月27日提交中国专利局、申请号为201610055938.2发明名称为“一种基于共形几何代数的机械臂运动规划的形式化分析方法及系统”的中国专利申请的优先权,其全部内容通过引用结合在本申请中。
技术领域
本申请涉及一种基于共形几何代数的机械臂运动规划的形式化分析方法及系统,属于计算机科学技术领域。
背景技术
共形几何代数(CGA,Conformal Geometric Algebra)作为一种先进的几何表示和计算系统,它为经典几何提供了简洁、直观和统一的齐性代数框架。CGA通过增加两个维度来表示基准原点(e0)和无穷远点(e),使得欧氏空间嵌入到共形空间中,并赋予其闵氏内积结构,不仅保留了齐次空间中外积的Grassmann结构、更使得内积具备了表征距离、角度等基本度量的明确几何意义。CGA不仅成功地解决了如何运用几何语言完成几何计算的问题,还对许多高科技领域如工程学和计算机科学的几何问题的解决起了关键性的作用。在机器人学中,CGA与以往的代数不同,它的运算对象不是数字,而是几何体。而机器人研究对象是基于基本几何体建立的系统所形成的几何关系。因此,共形几何代数在机器人研究中具有独特性。
目前利用CGA为数学工具进行建模和计算分析时,传统上使用基于计算机的数值计算分析和计算机代数系统(CASs)如Maple、CLUCalc、Gaalop等,然而这两种方法均不能完全保证结果的正确性和精确性。由于计算的迭代次数受限于计算机内存和浮点数限制,数值计算分析并不能完全保证结果的精确性;而CASs提供的符号方法虽然利用核心算法可以精确推导出符号表达式的解,但是对庞大的符号集进行运算的算法并没有经过验证,并且在边界条件的处理上存在短板,所得到的结果仍然可能存在问题。
另外,近几十年来,形式化方法在很多领域中都取得了巨大进步,基础研究的进展加上技术进步的推动,使新方法和新工具不断出现并逐步完善成 为一种成熟的高可靠验证技术。它的主要思想是根据数学理论来证明所设计的系统满足系统的规范或具有所期望的性质。与人工笔纸分析和上述传统方法相比,形式化方法可根据数学逻辑的严密性提高发现微小而关键的早期设计错误的机率。
由于利用CGA为数学工具进行建模和计算分析时会出现上述问题,而对CGA理论进行形式化分析是一种理想的防止出现上述问题的方式,又由于CGA在机器人研究中具有独特性,因此,在机器人研究过程中,如何将CGA与形式化方式相结合是一个亟待解决的技术问题。
发明内容
有鉴于此,本申请的目的在于提供一种基于共形几何代数的机械臂运动规划的形式化分析方法及系统,以解决现有技术中对机械臂运动规划进行分析时计算复杂、结果不精确的问题。
根据本申请一方面,其提供了一种基于共形几何代数的机械臂运动规划的形式化分析方法,其包括:
确定机器人的具体结构参数和运动规划参数;
根据所述具体结构参数、所述运动规划参数,并基于共形几何代数理论对机器人的基本构件、运动规划约束构建对应的几何模型,其中,所构建的几何模型以高阶逻辑语言进行描述;基于所构建几何模型对机器人的运动过程进行形式化建模,得到机器人的运动过程的几何关系逻辑模型;
结合所述几何关系逻辑模型,获得机器人的待验证运动过程的约束或属性对应的运动逻辑关系;
验证所述运动逻辑关系是否成立,成立表明所述几何关系逻辑模型满足所述待验证运动过程的约束或者具备所述待验证运动过程的属性,不成立表明所述几何关系逻辑模型不满足所述待验证运动过程的约束或者不具备所述待验证运动过程的属性。
根据本申请另一方面,提供了一种基于共形几何代数的机械臂运动规划的形式化分析系统,其包括:
参数确定模块,用于确定机器人的具体结构参数和运动规划参数;
机器人基本几何逻辑模型建立模块,用于根据所述具体结构参数、所述 运动规划参数,并基于共形几何代数理论对机器人的基本构件、运动规划约束构建对应的几何模型,其中,所构建的几何模型以高阶逻辑语言进行描述;
机器人运动过程几何关系逻辑模型建立模块,用于基于所构建几何模型对机器人的运动过程进行形式化建模,得到机器人的运动过程的几何关系逻辑模型;
逻辑命题构成模块,用于结合所述几何关系逻辑模型,获得机器人的待验证运动过程的约束或属性对应的运动逻辑关系;
证明模块,用于验证所述运动逻辑关系是否成立,成立表明所述几何关系逻辑模型满足所述待验证运动过程的约束或者具备所述待验证运动过程的属性,不成立表明所述几何关系逻辑模型不满足所述待验证运动过程的约束或者不具备所述待验证运动过程的属性。
根据本申请另一方面,提供了一种存储介质,其中,该存储介质用于存储可执行程序代码,所述可执行程序代码用于在运行时执行前述基于共形几何代数的机械臂运动规划的形式化分析方法。
根据本申请另一方面,提供了一种应用程序,其中,该应用程序用于在运行时执行前述基于共形几何代数的机械臂运动规划的形式化分析方法。
根据本申请另一方面,提供了一种电子设备,包括:
处理器、存储器、通信接口和总线;
所述处理器、所述存储器和所述通信接口通过所述总线连接并完成相互间的通信;。
所述存储器存储可执行程序代码;
所述处理器通过读取所述存储器中存储的可执行程序代码来运行与所述可执行程序代码对应的程序,以用于执行前述基于共形几何代数的机械臂运动规划的形式化分析方法。
和传统方法不同,本申请实施例提供的技术方案中以形式化方式建模得到几何关系逻辑模型后,根据该几何关系逻辑模型获得待验证运动过程的约束或属性对应的运动逻辑关系,并验证上述运动逻辑关系是否成立。由于上述方式中是采用数学方法进行正确性验证的,针对验证的性质而言是精确和完备的。另外,CGA可以对点、线、面、圆、球等几何元素以及这些几何元素 的旋转和平移进行统一建模和处理,在处理机器人运动学和运动规划问题上具有很强的优势,能够提高解决问题的维数,从而简化机器人计算中的耦合,进而降低计算复杂度。综合上述两方面,应用本申请实施例提供的技术方案对机器人的机械臂运动规划进行分析时,能够在提高验证精度的同时降低计算复杂度,不仅充分发挥了CGA和形式化方式各自的优势,两者相结合更是互相强化了各自的优势。
附图说明
为了更清楚地说明本申请实施例和现有技术的技术方案,下面对实施例和现有技术中所需要使用的附图作简单地介绍,显而易见地,下面描述中的附图仅仅是本申请的一些实施例,对于本领域普通技术人员来讲,在不付出创造性劳动的前提下,还可以根据这些附图获得其他的附图。
图1为本申请中基于CGA的机器人抓取物体的形式化建模流程图;
图2为本申请中机器人抓取物体算法的验证流程图;
图3为本申请中目标圆Zt的计算示意图;
图4为本申请中机械爪圆Zh的计算示意图;
图5为本申请中旋转算子和平移算子的求解示意图。
具体实施方式
为使本申请的目的、技术方案、及优点更加清楚明白,以下参照附图并举实施例,对本申请进一步详细说明。显然,所描述的实施例仅仅是本申请一部分实施例,而不是全部的实施例。基于本申请中的实施例,本领域普通技术人员在没有作出创造性劳动前提下所获得的所有其他实施例,都属于本申请保护的范围。
以下将参照附图更详细地描述本申请的各种实施例。在各个附图中,相同的元件采用相同或类似的附图标记来表示。为了清楚起见,附图中的各个部分没有按比例绘制。
目前发展的共形几何代数系统,理论上可以推广到空间的任意维数,主要应用是5维共形空间,共形几何代数提供了表示几何体的Grassmann结构、表示几何变换的统一作用、表示几何量的括号系统和不变量系统,在几何数据处理和几何计算方面表现出显著优势。在几何代数框架下,不同变换表达 和运算形式具有统一性,经几何空间变换后,仍可保持原始几何对象的几何意义和几何特性。机器人研究对象是基于基本几何体建立的系统所形成的几何关系,而与以往的代数不同,共形几何代数的运算对象不是数字,是几何体,所以共形几何代数在机器人研究中具有独特性。共形几何代数还可将研究机器人的传统方法如李群李代数、吴方法、Clifford代数、螺旋矢量进行统一构造,运用到机构的分析和综合中,推动了机构学的发展,形成了具有普遍性的理论方法,降低了机构学建模和推导计算的复杂度,通过扩展维数将空间的各种几何元素表达为球体坐标,在复杂多自由度并联机构建模中可以得到比较简单的数学模型。
CGA框架下的几何体表达和构建可分别基于内积和外积进行。基于外积的几何体表达主要反映不同层次几何形体间相互构建关系,而内积形式表达则可通过距离、角度等度量表征参数构建相应的参数方程。对于任意k阶片积A,其基于外积和基于内积构建的参数方程分别为X∧A=0,X·A=0。两者的表达可以通过内、外积间的对偶运算进行相互转换。表1给出基于内积和基于外积构建的几何体表达,两种表示方式被很多文献分别称作标准表达(standard representation)和直接表达(direct representation)。
表1CGA基本几何体表示
Figure PCTCN2017072210-appb-000001
表1中x、n标记为粗体,表示欧氏三维空间中的向量:
x=x1e1+x2e2+x3e3
另外,表1中
Figure PCTCN2017072210-appb-000002
为欧氏空间中的标准内积即点积(scalar product),可以用多维向量库中的点积函数描述,e1、e2、e3是欧式三维空间中的单位正交基。直接表达法中主要通过外积“∧”连接几何体上的点{Pi}来构建出高维 的几何体,例如一个球可以用球上的四个点来表达。在标准表达法中外积的含义不同,表示几何体间的相交,例如一个圆可以由两个球相交而成。
具体的,本申请实施例提供了一种基于共形几何代数的机械臂运动规划的形式化分析方法,这一方法中,首先,确定机器人的具体结构参数和运动规划参数;然后根据上述具体结构参数、运动规划参数,并基于共形几何代数理论对机器人的基本构件、运动规划约束构建对应的几何模型,其中,所构建的几何模型以高阶逻辑语言进行描述;并基于所构建几何模型对机器人的运动过程进行形式化建模,得到机器人的运动过程的几何关系逻辑模型;再结合该几何关系逻辑模型,获得机器人的待验证运动过程的约束或属性对应的运动逻辑关系;最后验证上述运动逻辑关系是否成立,成立表明上述几何关系逻辑模型满足待验证运动过程的约束或者具备待验证运动过程的属性,不成立表明上述几何关系逻辑模型不满足待验证运动过程的约束或者不具备待验证运动过程的属性。
值得一提的是,机器人的每一运动过程的约束或者属性,在一定程度上反映了机器人的在该运动过程中能够实现的控制、操作以及不能实现的控制、操作等等,为此可以认为机器人的每一运动过程的约束或者属性是与一种运动逻辑关系相对应的。另外,不同的机器人具有不同的几何特征,不同几何特征又会对运动逻辑关系带来影响,因此,确定机器人的每一运动过程的约束或属性对应的运动逻辑关系时,需结合该机器人的几何关系逻辑模型进行确定。例如,某一机械爪的两个爪端之间原本可以打开成180度,但由于作业空间的限制,在空间的某个位置如位置Q,该两个爪端之间只能打开120度,否则机械爪将与该作业空间内的其他物体相碰撞,那么,此时的120度就是该机械爪在该位置Q的一种约束,也就是说,该机械爪在该作业空间移动时,如果移动到位置Q,则两个爪端之间打开的度数必须由180度调整为120度,否则将发生碰撞事故。
再者从数学的角度来说,结合几何关系逻辑模型,获得的机器人的待验证运动过程的约束或属性对应的运动逻辑关系时,可以理解为:由上述几何关系逻辑模型、待验证运动过程的约束或属性组成的逻辑命题。
具体的,验证上述运动逻辑关系是否成立时,可以利用逻辑推理引擎验证上述运动逻辑关系是否成立。
在本申请的一种实现方式中,基于所构建几何模型对机器人的运动过程进行形式化建模,得到机器人的运动过程的几何关系逻辑模型,直至利用上述逻辑推理引擎验证上述运动逻辑关系是否成立的过程中,可以引入高阶逻辑证明工具HOL-Light。具体的,利用HOL-Light建立CGA的形式化模型,并基于所建立的CGA的形式化模型对运动过程进行形式化建模,得到几何关系逻辑模型,进而对待验证运动过程的约束或者属性进行形式化建模,得到上述运动逻辑关系,最后基于CGA形式化模型中的逻辑推理引擎,对上述运动逻辑关系的形式化模型是否成立进行验证。
图1是本申请提出的基于共形几何代数CGA方法对机器人抓取物体的形式化建模与验证流程图,基于共形几何代数CGA方法对机器人操作的主要思想是CGA的几何体运动变换表达,但是本申请的变换几何对象是参考圆而不是参考点,机械爪和被抓取物体的几何特征都用圆来表示。本申请中,使用高阶逻辑定理证明器HOL-Light作为形式化工具。HOL-Light是最流行的定理证明器之一,不仅拥有庞大的研究团队和用户群,而且包含丰富的数学定理库如实数分析库、超越函数库、积分微分库等和一系列高效的证明策略。
具体的,首先确定机器人构件的具体结构参数以及运动规划参数,其中,具体结构参数可以是:关节参数、连杆参数、可达范围、关节所在面等等,运动规划参数可以是平移向量、旋转角度等等;然后基于共形几何代数理论对机器人构件、运动规划约束构建对应的几何模型,例如,可以根据关节信息构建点模型,根据连杆参数构建线模型,根据可达范围构建球模型,根据关节所在面构建面模型,根据运动规划参数构建刚体运动模型等等;以串联机械臂抓取物体作为待验证运动过程,确定该待验证运动过程的约束、属性等描述信息,输入HOL-Light,对该待验证运动过程进行验证,如通过该待验证运动过程的属性信息确定机械臂是否牢固抓取被抓取物体,从而实现目标验证。
需要说明的是,机器人的运动过程不仅仅包括机器人的机械臂抓取物体的运动过程,还可以包括其他运动过程,例如,机器人的机械臂移动过程等等,本申请仅仅以抓取物体的过程为例进行说明,并不对本申请构成限定。
如图1所示,本申请提出了一种基于共形几何代数的机械臂运动规划的形式化分析方法,其包括:
基于共形几何代数理论对机器人的基本构件、运动规划约束构建对应的几何模型,再用高阶逻辑语言描述所建立的几何模型,形成机器人的基本几何逻辑模型系统;
确定机器人的具体结构参数和运动规划参数;
基于基本几何逻辑模型系统对具体机器人运动过程的形式化建模,得到具体机器人运动过程的几何关系逻辑模型;
用逻辑公式描述需要验证的机器人运动过程的约束或属性;
将具体机器人运动过程的所述几何关系逻辑模型和待验证的运动过程的约束或属性组成一个逻辑命题;
利用逻辑推理引擎证明所述逻辑命题是否成立,成立表明所述模型满足所述的约束或者具备所述的属性,不成立表明所述模型不满足所述的约束或者不具备所述的属性。
本领域内技术人员可以理解的是,机器人的基本构件一般是通过这些构件的具体结构参数进行描述的,机器人的运动规划约束一般是通过运动规划的运动规划参数描述的,为此在构建机器人的基本构建、运动规划约束对应的几何模型之前,需先确定上述具体结构参数和运动规划参数,这样才能够更佳清楚、有效的构建基本构件、运动规划约束对应的几何模型。
鉴于上述情况,在构建对应的几何模型时,通常已经确定了机器人的具体结构参数和运动规划参数。
共形几何代数理论是机器人运动规划的形式化分析方法的理论基础,机器人运动规划的形式化分析方法可以针对具体机器人问题分析的普遍方法,通过逻辑推理引擎实现对共形几何代数形式化系统和机器人运动规划的形式化分析方法的逻辑表达与证明的推理演算。
具体的,基于共形几何代数理论对机器人的基本构件、运动规划约束构件对应的几何模型后,可能会得到多个几何模型,这些几何模型之间一般存在关联关系,为此,我们可以理解为这些几何模型能够形成机器人的基本几何逻辑模型系统,这样后续构建运动过程的几何关系逻辑模型时,可以相应的理解为是依据上述基本几何逻辑模型系统进行建模,得到的上述几何关系逻辑模型。
本领域内技术人员可以理解的是,从数学角度来讲,机器人的运动过程的约束或者属性可以通过逻辑公式进行描述,这样可以使得机器人的运动过程所蕴藏的逻辑关系能够更加直观的展现出来,也更加有利于获得运动过程的约束或者属性对应的运动逻辑关系。
鉴于此,在本申请的一种实现方式中,结合几何关系逻辑模型,获得机器人的待验证运动过程的约束或者属性对应的运动逻辑关系时,可以结合上述几何关系逻辑模型,并在待验证运动过程的约束或者属性对应的逻辑公式的基础上,获得待验证运动过程的约束或者属性对应的运动逻辑关系,或者称之为:参照待验证运动过程的约束或者属性对应的逻辑公式,获得上述几何关系逻辑模型、待验证运动过程的约束或属性组成的逻辑命题。
所述基于基本几何逻辑模型系统对具体机器人运动规划的运动规划过程形式化建模具体包括对机器人抓取物体的运动规划过程形式化建模。
优选地,上述步骤“确定机器人的具体结构参数和运动规划参数”中,具体机器人为n自由度串联机械臂,末端包括一个T字形机械爪。
优选地,上述步骤“基于基本几何逻辑模型系统对具体机器人运动过程的形式化建模,得到具体机器人运动过程的几何关系逻辑模型”中,具体机器人运动规划为机器臂抓取物体的运动规划过程。
优选地,上述步骤“用逻辑公式描述需要验证的机器人运动过程的约束或属性”中需要验证的机器人运动过程的属性为机械爪实现物体抓取成功并且抓牢。
所构建的几何模型包括:机器人关节抽象化成的点模型、机器人关节末端可达范围抽象化成的球模型、机器人关节所在辅助面抽象化成的面模型、机器人关节点连接构成的线模型、将机器人基本构件的约束关系抽象化成的 几何体求交模型、机器人基本构件旋转运动抽象化成的几何体纯旋转模型、机器人基本构件平移运动抽象化成的几何体纯平移模型、机器人关节末端到达期望位置抽象化成的几何体刚体运动模型、机器人基本构件度量关系抽象化成的几何体间距离模型、机器人基本构件夹角抽象化成的几何体夹角模型。
下面对上述各个模型进行详细介绍。其中,几何体包括点、线、面、圆、球、点对等,用于表示机器人的基本构件。
机器人关节抽象化成的点模型:
Figure PCTCN2017072210-appb-000003
上述模型可以在HOL-Light中进行形式化,具体表现为:
Figure PCTCN2017072210-appb-000004
该模型中将机器人关节抽象化成点,上述数学表达式是共形几何代数中点的标准表达式。该模型中point_CGA函数功能表示从三维欧氏空间到五维共形空间中点的映射关系。
上述模型表达式中小写s=s1e1+s2e2+s3e3表示欧氏三维空间中的点,e1、e2、e3是欧式三维空间中的单位正交基,s1、s2、s3为系数,大写S为点s从欧氏空间映射到共形空间中的点的表达式。函数point_CGA的输入变量为欧氏空间中三维向量,其返回值为五维共形空间中的多重矢量,由于共形几何代数可由几何代数Cl4,1构造而成,其数据类型定义为real^(4,1)multivector,其中s$n表示多维向量s的第n个元素,mbasis函数表示基本片积函数,dot函数表示欧氏空间中的点积,null_inf和null_zero函数分别表示零矢量e0和e
Figure PCTCN2017072210-appb-000005
e=e-+e+
其中e0表示原点,e表示无穷远点,利用e、e0取代e+、e-能够更紧凑的表示共形空间中的点。本申请将基矢量e+,e-定义为mbasis{4}和mbasis{5},表示CGA空间中第四个基矢量和第五个基矢量。形式化定义如下:
|-null_zero=(&1/&2)%(mbasis{5}-mbasis{4})/\null_inf=(mbasis{5})+(mbasis{4})
该模型中point_CGA函数功能表示从三维欧氏空间到五维共形空间中点的映射关系。
机器人关节末端可达范围抽象化成的球模型:
Figure PCTCN2017072210-appb-000006
上述模型可以在HOL-Light中进行形式化,具体表现为:
Figure PCTCN2017072210-appb-000007
将机器人关节末端可达范围抽象化成以关节为球心连杆为半径的球S,上述数学表达式是共形几何代数中球的标准表达式。函数sphere_CGA的输入变量p、r分别表示球心和半径,其中球心可以用欧氏空间中三维向量表示。
机器人关节所在辅助面抽象化成的面模型:π=n+de
上述模型可以在HOL-Light中进行形式化,具体表现为:
将机器人关节所在辅助面抽象化为面π,上述数学表达式是共形几何代数中面的标准表达式。函数plane_CGA的输入变量n、d分别表示辅助面法向量和辅助面到原点的距离,其中辅助面法向量可以用欧氏空间中三维向量表示。
机器人关节点连接构成的线模型:L*=A∧B∧e
上述模型可以在HOL-Light中进行形式化,具体表现为:
Figure PCTCN2017072210-appb-000009
上述数学表达式是共形几何代数中任意两点构成的线L*的表达式,可用于机器人中关节点构成的线或者其他辅助线的表示,其中A、B分别表示共形几何空间中的两个点。所述模型用于将两关节间的连杆抽象为线。上述表达式为共形几何代数中线的直接表达式,函数line_direct_CGA的输入变量a、b分别表示线上的两点,数据类型为real^(4,1)multivector,其中函数outer表示几何代数中外积运算,外积运算可实现由低维几何体向高维几何体的构建,即扩维运算。
将机器人基本构件的约束关系抽象化成的几何体求交模型o:o=o1∧o2∧K∧on
上述模型可以在HOL-Light中进行形式化,具体表现为:
|-MEET o1 o2 … on=o1outer o2outer … on
其中,oi表示代表第i个机器人基本构件的几何体,i=1,2,…,n;
所述模型利用外积实现代表机器人基本构件的几何体的求交运算。函数MEET的输入变量表示待求交的任意几何体,数据类型为real^(4,1)multivector。约束关系包括关节在连杆上(即点在线上)、连杆在关节所在辅助面上(线在面上)等。
机器人基本构件旋转运动抽象化成的几何体纯旋转模型:
Figure PCTCN2017072210-appb-000010
上述模型可以在HOL-Light中进行形式化,具体表现为:
|-rotation_CGA t l=cos(t/&2)%mbasis{}-l*(sin(t/&2)%mbasis{}
|-pure_rotationed_CGA x t l=(rotation_CGA t l)*x*(reversion(rotation_CGA t l))
所述模型利用几何积实现几何体旋转变换,R为共形几何代数中的旋转算子,
Figure PCTCN2017072210-appb-000011
为R的几何反,L表示的旋转轴,φ是旋转角度,o表示旋转前几何体的表达形式,orotated表示旋转后几何体的表达形式,函数rotation_CGA的输入变量t、l分别表示旋转副的旋转角度与旋转轴,该函数实现了旋转算子的功能。函数pure_rotationed_CGA的输入变量x、t、l表示几何体、旋转角度和旋转轴,该函数的返回值为旋转后的几何体表示,其中函数reversion表示几何反。
机器人基本构件平移运动抽象化成的几何体纯平移模型:
Figure PCTCN2017072210-appb-000012
上述模型可以在HOL-Light中进行形式化,具体表现为:
|-Translation_CGA t=mbasis{}-(&1/&2)%(t*null_inf)
|-pure_translationed_CGA x t=(Translation_CGA t)*x*(reversion(Translation_CGA t))
所述模型利用几何积实现几何体平移变换,T为共形几何代数中的平移算子,
Figure PCTCN2017072210-appb-000013
为T的几何反,其中t=t1e1+t2e2+t3e3是平移向量,表示平移的方向和长度,t1、t2、t3为系数,o表示平移前几何体的表达形式,otranslated表示平移 后几何体的表达形式,函数Translation_CGA的输入变量t是平移向量,表示平移的方向和长度,该函数实现了平移算子的功能。函数pure_translationed_CGA的输入变量x、t表示几何体、平移向量,该函数的返回值为平移后的几何体表示。
所述机器人关节末端到达期望位置抽象化成的几何体刚体运动模型:
Figure PCTCN2017072210-appb-000014
上述模型可以在HOL-Light中进行形式化,具体表现为:
|-motor_CGA a l t=rotation_CGA a l*Translation_CGA t
|-rigid_body_motion x a l t=(motor_CGA a l t)*x*(reversion(motor_CGA a l t))
该模型利用几何积实现几何体刚体变换,其中R、T分别是旋转算子和平移算子,两者用几何积运算连接,M为Motor算子,
Figure PCTCN2017072210-appb-000015
为M的几何反,o为经过刚体运动前几何体的表达形式,origid_body_motion表示运动后的几何体表达形式,函数motor_CGA的输入变量x、a、l、t分别表示几何体、旋转角度、旋转轴、平移向量,该函数实现了马达算子的功能。函数rigid_body_motion的返回值为经过刚体运动后的几何体表示,该表达式表示几何体是先平移后旋转,但马达算子具有结合性,旋转算子和平移算子是可以交换的。
机器人基本构件度量关系抽象化成的几何体间距离模型:
Figure PCTCN2017072210-appb-000016
Figure PCTCN2017072210-appb-000017
其中a、b表示三维欧氏空间中任意两点,A、B分别为点a、点b在共形空间中的表达式,上述模型为共形几何代数中点点距离表达式,其中dist(a,b)表示两点距离,可用几何代数内积表达.其中点线距离、点球距离、球球距离在共形几何代数中也有具体的表达式,可用于机器人问题中具体计算时使用。
机器人基本构件夹角抽象化成的几何体夹角模型:
Figure PCTCN2017072210-appb-000018
上述模型可以在HOL-Light中进行形式化,具体表现为:
Figure PCTCN2017072210-appb-000019
机器人基本构件夹角抽象化成的几何体夹角模型用于计算机器人基本构件夹角θ。
该模型为共形几何代数中几何体夹角θ的表达式,函数vector_angles_CGA的输入变量o1、o2分别表示待求夹角的几何体,可以是线或面等,数据类型为real^(4,1)multivector,其中函数inner表示几何代数中左缩积运算,$${}表示取0阶子空间即标量的大小,函数mult_norm功能为多重矢量的求模运算。该模型可用于机器人基本构件之间夹角的计算,o1 *,o2 *,
Figure PCTCN2017072210-appb-000020
分别表示o1和o2的对偶。
在本申请的一种实现方式中,基于所构建几何模型对机器人的运动过程进行形式化建模时,具体可以包括:
基于所构建几何模型,并结合形式化方式计算机械爪爪端所在的第一目标圆;基于所构建几何模型、机械爪上的特征点,并结合形式化方式计算机械爪所在的第二目标圆;根据所述第一目标圆和第二目标圆,并结合形式化方式计算机械爪的平移算子;根据所述第一目标圆和第二目标圆,并结合形式化方式计算机械爪抓取物体的旋转算子;根据所述平移算子和旋转算子并结合形式化方式,计算机械爪新的目标位置,实现对机器人的运动过程的形式化建模。
机器人机械臂的运动过程可以包含多种形式,为了更加清楚的对本申请实施例进行说明,下面以机器人抓取物体为例对本申请实施例提供的基于共形几何代数的机械臂运动规划的形式化分析方法进行详细描述。
优选地,所述方法中具体机器人为n自由度串联机械臂,末端包括一个T字形机械爪;
所述方法中具体机器人运动规划为机器臂抓取物体的运动规划过程;
所述方法中待验证的运动过程的属性为机械爪实现物体抓取成功并且抓牢。
图2示出了本申请中对机器人抓取物体的运动规划过程形式化建模的流程图。如图2所示,其包括:
步骤30:分别提取物体和机械爪的特征点;
对应于图2中的S30A和S30B。
步骤31:计算物体的被抓取位置所在的目标圆;
对应于图2中的S31。
如图3和图4所示,通过物体边缘的四个特征点(x1,x2,x3,x4)得到物体的位置方位,其中x1,x2,x3为物体底部边沿上任意三点,x4为物体顶部边沿上任意一点。由物体底部三点构成的参考圆
Figure PCTCN2017072210-appb-000021
可通过圆的直接表达式得到:
Figure PCTCN2017072210-appb-000022
则物体底部参考圆
Figure PCTCN2017072210-appb-000023
所在的面πb为:
Figure PCTCN2017072210-appb-000024
其中,Ic为共形几何代数的伪标量,通过伪标量是可实现对偶运算,对偶运算实现几何体两种表示方法的相互转换。
由于方便设定物体被抓取的位置是中部,所以目标圆Zt应该是底部参考圆沿-πb方向平移
Figure PCTCN2017072210-appb-000025
长度后的圆,则相应的平移算子T为:
Figure PCTCN2017072210-appb-000026
则物体被抓取的部位所形成的目标圆Zt为:
Figure PCTCN2017072210-appb-000027
Figure PCTCN2017072210-appb-000028
为T的几何反。
其中,
Figure PCTCN2017072210-appb-000029
上述计算过程在HOL-Light中进行形式化:
let dual_circle_zb x1 x2 x3=(circle_direct_CGA(point_CGA x1)(point_CGA x2)(point_CGA x3))
let plane_b x1 x2 x3=((dual_circle_zb x1 x2 x3)outer null_inf)*pseudo
let circle_zt x1 x2 x3 x4=pure_translationed_CGA(--DUAL(dual_circle_zb x1 x2 x3))(--(&1/&2)%((plane_b x1 x2 x3)inner(point_CGA x4)))
其中,物体底部参考圆
Figure PCTCN2017072210-appb-000030
调用圆的直接表示函数circle_direct_CGA,x1 x2 x3 x4表示物体的四个特征点,HOL类型均是real^3,表示三维欧氏向量,通过函数point_CGA使特征点由三维欧氏空间嵌入到五维共形空间中。物体被抓取的部位所形成的目标圆通过纯平移函数pure_translationed_CGA实现。
需要说明的是,由于机器人的机械爪在抓取被抓取物体时,一般是机械抓的爪端接触物体,所以本步骤中计算得到的目标圆可以理解为前述的第一目标圆。
步骤32:计算机械爪所在圆;
对应于图2中的S32。
如图5所示,已知机械爪所在圆的圆心Ph、半径ρ、机械爪上两特征点a,b可计算出机械爪所在圆的位置,首先构造机械爪所在球的位置,通过球的标准表达式得到:
Figure PCTCN2017072210-appb-000031
然后通过圆心Ph、机械爪上两特征点a,b计算机械爪所在面πh *
Figure PCTCN2017072210-appb-000032
机械爪所在球Sh和所在面πh *相交得到机械爪所在圆:
Zh=Sh∧πh*
上述计算过程在HOL-Light中进行如下形式化:
let sphere_Sh ph r=sphere_CGA ph r
let dual_plane_pih ph a b=plane_direct_CGA(point_CGA ph)(point_CGA a)(point_CGA b)
let circle_zh ph r a b=(sphere_Sh ph r)outer(--DUAL(dual_plane_pih ph a b))
其中,辅助球Sh调用球的标准表示函数sphere_CGA实现,(ph:real^3)和(r:real)分别表示辅助球的圆心Ph和半径r,机械爪所在的辅助面
Figure PCTCN2017072210-appb-000033
调用面的 直接表示函数plane_direct_CGA实现,(a:real^3)和(b:real^3)分别表示机械爪上两点。对偶算子用DUAL函数。
步骤33a:计算机械爪抓取物体的平移算子,所述平移算子包括平移轴和平移长度;
对应于图2中的S33。
已知平移长度应该为目标圆Zt和机械爪所在圆Zh的圆心距离,而平移轴是过两圆心的直线。首先计算目标圆Zt的圆心:
Pt=ZteZt
由于平移轴由两圆心确定,可通过直接表达式计算平移轴
Figure PCTCN2017072210-appb-000034
Figure PCTCN2017072210-appb-000035
相应地,平移长度为d:
Figure PCTCN2017072210-appb-000036
上述计算过程在HOL-Light中进行形式化:
let center_point_pt x1 x2 x3 x4=(circle_zt x1 x2 x3 x4)*null_inf*(circle_zt x1 x2 x3 x4)
let dual_translation_axis ph x1 x2 x3 x4=line_direct_CGA(point_CGA ph)(center_point_pt x1 x2 x3 x4)
let distance ph x1 x2 x3 x4=--sqrt(&2*((point_CGA ph)inner(center_point_pt x1 x2 x3 x4))$${})
步骤33b:计算机械爪抓取物体的旋转算子,包括旋转轴和旋转角度;
对应于图2中的S33。
机械爪抓取物体的旋转轴满足几点约束:
1.过机械爪所在圆的圆心Ph
2.在由两圆的轴线确定的面上
所以,先计算目标圆Zt和机械爪所在圆Zh的两轴线
Figure PCTCN2017072210-appb-000037
Figure PCTCN2017072210-appb-000038
Figure PCTCN2017072210-appb-000039
Figure PCTCN2017072210-appb-000040
由两轴线确定的面
Figure PCTCN2017072210-appb-000041
为:
Figure PCTCN2017072210-appb-000042
则旋转轴计算为:
Figure PCTCN2017072210-appb-000043
而旋转角度易知为两轴线的夹角,由夹角公式得:
Figure PCTCN2017072210-appb-000044
上述计算过程在HOL-Light中进行形式化:
let dual_lh ph r a b=(circle_zh ph r a b)outer null_inf
let dual_lt x1 x2 x3 x4=(circle_zt x1 x2 x3 x4)outer null_inf
let dual_plane_th x1 x2 x3 x4ph r a b=(dual_lt x1 x2 x3 x4)outer((dual_lh ph r a b)*(null_zero outer null_inf))
let dual_rotation_axis x1 x2 x3 x4ph r a b=(point_CGA ph)outer(--DUAL(dual_plane_th x1 x2 x3 x4ph r a b))outer null_inf
let rotation_angle x1 x2 x3 x4ph r a b=vector_angles_CGA(dual_lh ph r a b)(dual_lt x1 x2 x3 x4)
其中,旋转角度θ的求解则使用CGA夹角公式来计算,由函数vector_angles_CGA实现.
步骤34:计算机械爪新的目标位置;
对应于图2中的S34。
通过步骤33和步骤34计算得到的平移轴
Figure PCTCN2017072210-appb-000045
平移长度d、旋转轴
Figure PCTCN2017072210-appb-000046
旋转角度θ,可以得到机械爪的旋转算子R和平移算子T:
Figure PCTCN2017072210-appb-000047
其中,R和T分别为旋转算子和平移算子,
Figure PCTCN2017072210-appb-000048
机械爪新的目标位置可以先旋转后平移计算得到,利用刚体算子可计算为:
Figure PCTCN2017072210-appb-000049
其中,Z'h为机械爪新的目标位置,
Figure PCTCN2017072210-appb-000050
为R的几何反,
Figure PCTCN2017072210-appb-000051
为T的几何反.
最终可以得到机械爪新的目标位置,在HOL-Light中进行形式化:
let circle_zh_new x1 x2 x3 x4ph r a b=pure_translationed_CGA
(pure_rotationed_CGA(circle_zh ph r a b)(rotation_angle x1 x2 x3 x4 ph r a b)(--DUAL(dual_rotation_axis x1 x2 x3 x4 ph r a b)))
((distance ph r a b x1 x2 x3 x4)%(--DUAL(dual_translation_axis ph r a b x1 x2 x3 x4)))
其中,由于机械爪新的目标位置是通过先旋转后平移得到,所以先将机械爪面所构成的圆zh通过纯旋转函数pure_rotationed_CGA进行旋转运动,该函数的输入变量是被旋转的几何体zh、旋转角度θ和旋转轴,即(circle_zh ph r a b)、(rotation_angle x1 x2 x3 x4ph r a b)和(--DUAL(dual_rotation_axis x1 x2 x3 x4 ph r a b)),再将旋转后的圆
Figure PCTCN2017072210-appb-000052
通过纯平移函数pure_translationed_CGA完成平移运动,该函数输入变量是旋转的几何体
Figure PCTCN2017072210-appb-000053
平移向量dlT
上述代码主要内容涉及共形几何代数CGA几何体表示、几何体间距离特征和几何体运动变换的形式化。机械爪圆Zh一步步靠近物体位置所在目标圆Zt,最终机械爪成功抓取到物体并且抓牢必须满足一定的几何约束关系。上述算法中,该约束关系为机械爪面所在圆Zh新的位置最终应该与物体所在目标圆Zt的位置吻合,即在共形几何代数CGA中表达式应该满足相等的关系。可以在HOL-light中建立目标验证这层几何约束关系:
Figure PCTCN2017072210-appb-000054
需要说明的是,前述描述中所涉及的“物体”可以理解为“被抓取物体”。
另外,本步骤中计算得到新的目标位置后,还需计算新的目标位置形成的圆,并判断计算得到的圆与前述S31中计算得到的目标圆是否相等,若相 等,则说明机械抓抓取物体成功。该判断是否与目标圆相等对应于图2中的S35。
上述各个实施例提供的方案中关于基于CGA对机器人的运动规划进行形式化建模与验证的方法,在HOL-Light中可以加载使用。其中最大难点是定理证明技术需要大量的人机交互,工作量较大,时间耗费较长,建模过程中强调使用者对共形几何代数CGA理论知识的熟识度,验证过程中要求严密的思维和一定的逻辑推理经验。
另外和传统方法不同,上述各个实施例提供的技术方案中以形式化方式建模得到几何关系逻辑模型后,根据该几何关系逻辑模型获得待验证运动过程的约束或属性对应的运动逻辑关系,并验证上述运动逻辑关系是否成立。由于上述方式中是采用数学方法进行正确性验证的,针对验证的性质而言是精确和完备的。另外,CGA可以对点、线、面、圆、球等几何元素以及这些几何元素的旋转和平移进行统一建模和处理,在处理机器人运动学和运动规划问题上具有很强的优势,能够提高解决问题的维数,从而简化机器人计算中的耦合,进而降低计算复杂度。综合上述两方面,应用上述各个实施例提供的技术方案对机器人的机械臂运动规划进行分析时,能够在提高验证精度的同时降低计算复杂度,不仅充分发挥了CGA和形式化方式各自的优势,两者相结合更是互相强化了各自的优势。
与上述一种基于共形几何代数的机械臂运动规划的形式化分析方法相对应,本申请实施例还提供了一种基于共形几何代数的机械臂运动规划的形式化分析系统。
具体的,该系统包括:
参数确定模块,用于确定机器人的具体结构参数和运动规划参数;
机器人基本几何逻辑模型建立模块,用于根据所述具体结构参数、所述运动规划参数,并基于共形几何代数理论对机器人的基本构件、运动规划约束构建对应的几何模型,其中,所构建的几何模型以高阶逻辑语言进行描述;
机器人运动过程几何关系逻辑模型建立模块,用于基于所构建几何模型对机器人的运动过程进行形式化建模,得到机器人的运动过程的几何关系逻辑模型;
逻辑命题构成模块,用于结合所述几何关系逻辑模型,获得机器人的待验证运动过程的约束或属性对应的运动逻辑关系;
证明模块,用于验证所述运动逻辑关系是否成立,成立表明所述几何关系逻辑模型满足所述待验证运动过程的约束或者具备所述待验证运动过程的属性,不成立表明所述几何关系逻辑模型不满足所述待验证运动过程的约束或者不具备所述待验证运动过程的属性。
另外和传统方法不同,本实施例提供的技术方案中以形式化方式建模得到几何关系逻辑模型后,根据该几何关系逻辑模型获得待验证运动过程的约束或属性对应的运动逻辑关系,并验证上述运动逻辑关系是否成立。由于上述方式中是采用数学方法进行正确性验证的,针对验证的性质而言是精确和完备的。另外,CGA可以对点、线、面、圆、球等几何元素以及这些几何元素的旋转和平移进行统一建模和处理,在处理机器人运动学和运动规划问题上具有很强的优势,能够提高解决问题的维数,从而简化机器人计算中的耦合,进而降低计算复杂度。综合上述两方面,应用本实施例提供的技术方案对机器人的机械臂运动规划进行分析时,能够在提高验证精度的同时降低计算复杂度,不仅充分发挥了CGA和形式化方式各自的优势,两者相结合更是互相强化了各自的优势。
相应的,本申请还提供了一种存储介质,其中,该存储介质用于存储可执行程序代码,所述可执行程序代码用于在运行时执行本申请所述的一种基于共形几何代数的机械臂运动规划的形式化分析方法,其中,本申请所述的一种基于共形几何代数的机械臂运动规划的形式化分析方法,其包括:
确定机器人的具体结构参数和运动规划参数;
根据所述具体结构参数、所述运动规划参数,并基于共形几何代数理论对机器人的基本构件、运动规划约束构建对应的几何模型,其中,所构建的几何模型以高阶逻辑语言进行描述;
基于所构建几何模型对机器人的运动过程进行形式化建模,得到机器人的运动过程的几何关系逻辑模型;
结合所述几何关系逻辑模型,获得机器人的待验证运动过程的约束或属性对应的运动逻辑关系;
验证所述运动逻辑关系是否成立,成立表明所述几何关系逻辑模型满足所述待验证运动过程的约束或者具备所述待验证运动过程的属性,不成立表明所述几何关系逻辑模型不满足所述待验证运动过程的约束或者不具备所述待验证运动过程的属性。
通过执行本实施例提供的存储介质中存储的程序代码,以形式化方式建模得到几何关系逻辑模型后,根据该几何关系逻辑模型获得待验证运动过程的约束或属性对应的运动逻辑关系,并验证上述运动逻辑关系是否成立。由于上述方式中是采用数学方法进行正确性验证的,针对验证的性质而言是精确和完备的。另外,CGA可以对点、线、面、圆、球等几何元素以及这些几何元素的旋转和平移进行统一建模和处理,在处理机器人运动学和运动规划问题上具有很强的优势,能够提高解决问题的维数,从而简化机器人计算中的耦合,进而降低计算复杂度。综合上述两方面,应用本实施例提供的技术方案对机器人的机械臂运动规划进行分析时,能够在提高验证精度的同时降低计算复杂度,不仅充分发挥了CGA和形式化方式各自的优势,两者相结合更是互相强化了各自的优势。
相应的,本申请还提供了一种应用程序,其中,该应用程序用于在运行时执行如本申请所述的一种基于共形几何代数的机械臂运动规划的形式化分析方法,其中,本申请所述的一种基于共形几何代数的机械臂运动规划的形式化分析方法,其包括:
确定机器人的具体结构参数和运动规划参数;
根据所述具体结构参数、所述运动规划参数,并基于共形几何代数理论对机器人的基本构件、运动规划约束构建对应的几何模型,其中,所构建的几何模型以高阶逻辑语言进行描述;
基于所构建几何模型对机器人的运动过程进行形式化建模,得到机器人的运动过程的几何关系逻辑模型;
结合所述几何关系逻辑模型,获得机器人的待验证运动过程的约束或属性对应的运动逻辑关系;
验证所述运动逻辑关系是否成立,成立表明所述几何关系逻辑模型满足所述待验证运动过程的约束或者具备所述待验证运动过程的属性,不成立表明所述几何关系逻辑模型不满足所述待验证运动过程的约束或者不具备所述待验证运动过程的属性。
通过执行本实施例提供的应用程序,以形式化方式建模得到几何关系逻辑模型后,根据该几何关系逻辑模型获得待验证运动过程的约束或属性对应的运动逻辑关系,并验证上述运动逻辑关系是否成立。由于上述方式中是采用数学方法进行正确性验证的,针对验证的性质而言是精确和完备的。另外,CGA可以对点、线、面、圆、球等几何元素以及这些几何元素的旋转和平移进行统一建模和处理,在处理机器人运动学和运动规划问题上具有很强的优势,能够提高解决问题的维数,从而简化机器人计算中的耦合,进而降低计算复杂度。综合上述两方面,应用本实施例提供的技术方案对机器人的机械臂运动规划进行分析时,能够在提高验证精度的同时降低计算复杂度,不仅充分发挥了CGA和形式化方式各自的优势,两者相结合更是互相强化了各自的优势。
相应的,本申请还提供了一种电子设备,包括:
处理器、存储器、通信接口和总线;
所述处理器、所述存储器和所述通信接口通过所述总线连接并完成相互间的通信;
所述存储器存储可执行程序代码;
所述处理器通过读取所述存储器中存储的可执行程序代码来运行与所述可执行程序代码对应的程序,以用于执行如本申请所述的一种基于共形几何代数的机械臂运动规划的形式化分析方法,其中,本申请所述的一种基于共形几何代数的机械臂运动规划的形式化分析方法,其包括:
确定机器人的具体结构参数和运动规划参数;
根据所述具体结构参数、所述运动规划参数,并基于共形几何代数理论对机器人的基本构件、运动规划约束构建对应的几何模型,其中,所构建的几何模型以高阶逻辑语言进行描述;
基于所构建几何模型对机器人的运动过程进行形式化建模,得到机器人的运动过程的几何关系逻辑模型;
结合所述几何关系逻辑模型,获得机器人的待验证运动过程的约束或属性对应的运动逻辑关系;
验证所述运动逻辑关系是否成立,成立表明所述几何关系逻辑模型满足所述待验证运动过程的约束或者具备所述待验证运动过程的属性,不成立表明所述几何关系逻辑模型不满足所述待验证运动过程的约束或者不具备所述待验证运动过程的属性。
本实施例提供的电子设备,以形式化方式建模得到几何关系逻辑模型后,根据该几何关系逻辑模型获得待验证运动过程的约束或属性对应的运动逻辑关系,并验证上述运动逻辑关系是否成立。由于上述方式中是采用数学方法进行正确性验证的,针对验证的性质而言是精确和完备的。另外,CGA可以对点、线、面、圆、球等几何元素以及这些几何元素的旋转和平移进行统一建模和处理,在处理机器人运动学和运动规划问题上具有很强的优势,能够提高解决问题的维数,从而简化机器人计算中的耦合,进而降低计算复杂度。综合上述两方面,应用本实施例提供的技术方案对机器人的机械臂运动规划进行分析时,能够在提高验证精度的同时降低计算复杂度,不仅充分发挥了CGA和形式化方式各自的优势,两者相结合更是互相强化了各自的优势。
对于系统、存储介质、应用程序、电子设备实施例而言,由于其基本相似于方法实施例,所以描述的比较简单,相关之处参见方法实施例的部分说明即可。
需要说明的是,在本文中,诸如第一和第二等之类的关系术语仅仅用来将一个实体或者操作与另一个实体或操作区分开来,而不一定要求或者暗示这些实体或操作之间存在任何这种实际的关系或者顺序。而且,术语“包括”、“包含”或者其任何其他变体意在涵盖非排他性的包含,从而使得包括一系 列要素的过程、方法、物品或者设备不仅包括那些要素,而且还包括没有明确列出的其他要素,或者是还包括为这种过程、方法、物品或者设备所固有的要素。在没有更多限制的情况下,由语句“包括一个……”限定的要素,并不排除在包括所述要素的过程、方法、物品或者设备中还存在另外的相同要素。
本说明书中的各个实施例均采用相关的方式描述,各个实施例之间相同相似的部分互相参见即可,每个实施例重点说明的都是与其他实施例的不同之处。尤其,对于装置实施例而言,由于其基本相似于方法实施例,所以描述的比较简单,相关之处参见方法实施例的部分说明即可。
本领域普通技术人员可以理解实现上述方法实施方式中的全部或部分步骤是可以通过程序来指令相关的硬件来完成,所述的程序可以存储于计算机可读取存储介质中,这里所称得的存储介质,如:ROM/RAM、磁碟、光盘等。
以上所述仅为本申请的较佳实施例而已,并不用以限制本申请,凡在本申请的精神和原则之内,所做的任何修改、等同替换、改进等,均应包含在本申请保护的范围之内。

Claims (22)

  1. 一种基于共形几何代数的机械臂运动规划的形式化分析方法,其包括:
    确定机器人的具体结构参数和运动规划参数;
    根据所述具体结构参数、所述运动规划参数,并基于共形几何代数理论对机器人的基本构件、运动规划约束构建对应的几何模型,其中,所构建的几何模型以高阶逻辑语言进行描述;
    基于所构建几何模型对机器人的运动过程进行形式化建模,得到机器人的运动过程的几何关系逻辑模型;
    结合所述几何关系逻辑模型,获得机器人的待验证运动过程的约束或属性对应的运动逻辑关系;
    验证所述运动逻辑关系是否成立,成立表明所述几何关系逻辑模型满足所述待验证运动过程的约束或者具备所述待验证运动过程的属性,不成立表明所述几何关系逻辑模型不满足所述待验证运动过程的约束或者不具备所述待验证运动过程的属性。
  2. 如权利要求1所述的方法,其中,所构建的几何模型包括:机器人关节抽象化成的点模型、机器人关节末端可达范围抽象化成的球模型、机器人关节所在辅助面抽象化成的面模型、机器人关节点连接构成的线模型、将机器人基本构件的约束关系抽象化成的几何体求交模型、机器人基本构件旋转运动抽象化成的几何体纯旋转模型、机器人基本构件平移运动抽象化成的几何体纯平移模型、机器人关节末端到达期望位置抽象化成的几何体刚体运动模型、机器人基本构件度量关系抽象化成的几何体间距离模型、机器人基本构件夹角抽象化成的几何体夹角模型。
  3. 如权利要求2所述的方法,其中,所述机器人关节抽象化成的点模型如下表示:
    Figure PCTCN2017072210-appb-100001
    Figure PCTCN2017072210-appb-100002
    e=e-+e+
    其中,小写s=s1e1+s2e2+s3e3表示欧氏三维空间中的点,e1、e2、e3是欧式三维空间中的单位正交基,s1、s2、s3为系数,大写S为点s从欧氏空间映射到共形空间中的点的表达式;e0表示原点,e表示无穷远点,e+、e-分别为共形几何空间中的第四个基矢量和第五个基矢量。
  4. 如权利要求2所述的方法,其中,所述机器人关节末端可达范围抽象化成的球模型用于将机器人关节末端可达范围抽象化为以关节为球心以两关节间的连杆为半径的球S,如下表示:
    Figure PCTCN2017072210-appb-100003
    其中,P、r分别表示球心和半径,e表示无穷远点。
  5. 如权利要求2所述的方法,其中,所述机器人关节所在辅助面抽象化成的面模型用于将机器人关节所在辅助面抽象为面π,如下表示:
    π=n+de
    其中,n、d分别表示辅助面法向量和辅助面到原点的距离,e表示无穷远点。
  6. 如权利要求2所述的方法,其中,所述机器人关节点连接构成的线模型用于将两关节间的连杆抽象为线L*,如下表示:
    L*=A∧B∧e
    其中,A和B分别表示两关节所代表的点,e表示无穷远点。
  7. 如权利要求2所述的方法,其中,所述将机器人基本构件的约束关系抽象化成的几何体求交模型o,如下表示:
    o=o1∧o2∧…∧on
    其中,oi表示代表第i个机器人基本构件的几何体,i=1,2,…,n。
  8. 如权利要求2所述的方法,其中,所述机器人基本构件旋转运动抽象化成的几何体纯旋转模型如下表示:
    Figure PCTCN2017072210-appb-100004
    其中,R为共形几何代数中的旋转算子,
    Figure PCTCN2017072210-appb-100005
    为R的几何反,L表示旋转轴,φ是旋转角度,o表示旋转前几何体的表达形式,orotated表示旋转后几何体的表达形式。
  9. 如权利要求2所述的方法,其中,所述机器人基本构件平移运动抽象化成的几何体纯平移模型如下表示:
    Figure PCTCN2017072210-appb-100006
    其中,T为共形几何代数中的平移算子,
    Figure PCTCN2017072210-appb-100007
    为T的几何反,其中t=t1e1+t2e2+t3e3是平移向量,表示平移的方向和长度,o表示平移前几何体的表达形式,otranslated表示平移后几何体的表达形式,e表示无穷远点。
  10. 如权利要求2所述的方法,其中,所述机器人关节末端到达期望位置抽象化成的几何体刚体运动模型如下表示:
    Figure PCTCN2017072210-appb-100008
    其中,R、T分别是旋转算子和平移算子,M为Motor算子,
    Figure PCTCN2017072210-appb-100009
    为M的几何反,o为经过刚体运动前几何体的表达形式,origid_body_motion表示经过刚体运动后的几何体表达形式。
  11. 如权利要求2所述的方法,其中,所述机器人基本构件度量关系抽象化成的几何体间距离模型如下表示:
    Figure PCTCN2017072210-appb-100010
    其中,a、b表示三维欧氏空间中任意两点,A、B分别为点a、点b在共形空间中的表达式。
  12. 如权利要求2所述的方法,其中,所述机器人基本构件夹角抽象化成的几何体夹角模型用于计算机器人基本构件夹角θ,如下表示:
    Figure PCTCN2017072210-appb-100011
    其中,o1、o2分别表示待求夹角的几何体,o1 *,o2 *,
    Figure PCTCN2017072210-appb-100012
    分别表示o1和o2的对偶。
  13. 如权利要求1-12任一项所述的方法,其中,所述基于所构建几何模型对机器人的运动过程进行形式化建模,具体包括:
    步骤31:基于所构建几何模型,并结合形式化方式计算机械爪爪端所在的第一目标圆;
    步骤32:基于所构建几何模型、机械爪上的特征点,并结合形式化方式计算机械爪所在的第二目标圆;
    步骤33:根据所述第一目标圆和第二目标圆,并结合形式化方式计算机械爪的平移算子;
    步骤34:根据所述第一目标圆和第二目标圆,并结合形式化方式计算机械爪抓取物体的旋转算子;
    步骤35:根据所述平移算子和旋转算子并结合形式化方式,计算机械爪新的目标位置,实现对机器人的运动过程的形式化建模。
  14. 如权利要求13所述的方法,机器人的运动过程包括:机器人的抓取物体的运动过程;
    其中,步骤31中第一目标圆通过如下方式计算得到:
    通过被抓取物体边缘的四个特征点x1、x2、x3、x4得到被抓取物体的位置方位,其中x1、x2、x3为被抓取物体底部边沿上任意三点,x4为被抓取物体顶部边沿上任意一点,由被抓取物体底部三点构成的参考圆
    Figure PCTCN2017072210-appb-100013
    通过圆的下述直接表达式得到:
    Figure PCTCN2017072210-appb-100014
    则被抓取物体底部参考圆
    Figure PCTCN2017072210-appb-100015
    所在的面πb如下得到:
    Figure PCTCN2017072210-appb-100016
    其中,Ic为共形几何代数的伪标量;e表示无穷远点;
    被抓取物体的被抓取位置所在的第三目标圆Zt为底部参考圆沿-πb方向平移
    Figure PCTCN2017072210-appb-100017
    长度后的圆,则相应的平移算子T为:
    Figure PCTCN2017072210-appb-100018
    则通过形式化方式计算得到所述第三目标圆Zt为:
    Figure PCTCN2017072210-appb-100019
    其中,所述
    Figure PCTCN2017072210-appb-100020
    为T的几何反;
    将所述第三目标圆确定为所述第一目标圆。
  15. 如权利要求14所述的方法,其中,步骤32中所述机械抓所在第二目标圆如下计算:
    已知机械爪所在圆的圆心Ph、半径ρ、机械爪上两特征点a、b,计算出机械爪所在圆的位置,首先构造机械爪所在球的位置,通过球的标准表达式得到:
    Figure PCTCN2017072210-appb-100021
    其中,Sh为所述机械爪所在的球;
    根据机械爪上两特征点a、b计算机械爪所在面πh *
    Figure PCTCN2017072210-appb-100022
    通过形式化方式,计算机械爪所在球Sh和所在面
    Figure PCTCN2017072210-appb-100023
    相交得到机械爪所在第二目标圆Zh
    Zh=Sh∧πh*。
  16. 如权利要求15所述的方法,其中,所述步骤33中机械爪的平移算子包括平移轴和平移长度,具体如下计算:
    首先计算所述第一目标圆Zt的圆心:
    Pt=ZteZt
    通过形式化方式及直接表达式计算平移轴
    Figure PCTCN2017072210-appb-100024
    Figure PCTCN2017072210-appb-100025
    通过形式化方式,计算平移长度d:
    Figure PCTCN2017072210-appb-100026
  17. 如权利要求16所述的方法,其中,步骤34中机械爪的旋转算子包括旋转轴和旋转角度,具体如下计算:
    计算所述第一目标圆Zt和所述第二目标圆Zh的两轴线
    Figure PCTCN2017072210-appb-100027
    Figure PCTCN2017072210-appb-100028
    Figure PCTCN2017072210-appb-100029
    获取两轴线确定的面
    Figure PCTCN2017072210-appb-100030
    为:
    Figure PCTCN2017072210-appb-100031
    其中,所述e0为表示原点,e为表示无穷远点;
    通过形式化方式,计算旋转轴
    Figure PCTCN2017072210-appb-100032
    Figure PCTCN2017072210-appb-100033
    通过形式化方式,计算旋转角度:
    Figure PCTCN2017072210-appb-100034
  18. 如权利要求17所述的方法,其中,步骤35中机械爪新的目标位置如下计算:
    通过步骤33和步骤34计算得到的平移轴
    Figure PCTCN2017072210-appb-100035
    平移长度d、旋转轴
    Figure PCTCN2017072210-appb-100036
    旋转角度θ,得到机械爪抓取物体运动的旋转算子和平移算子:
    Figure PCTCN2017072210-appb-100037
    其中,R和T分别为旋转算子和平移算子,
    Figure PCTCN2017072210-appb-100038
    通过形式化方式,机械爪新的目标位置如下计算:
    Figure PCTCN2017072210-appb-100039
    其中,Z'h为机械爪新的目标位置,
    Figure PCTCN2017072210-appb-100040
    为R的几何反,
    Figure PCTCN2017072210-appb-100041
    为T的几何反。
  19. 一种基于共形几何代数的机械臂运动规划的形式化分析系统,其包括:
    参数确定模块,用于确定机器人的具体结构参数和运动规划参数;
    机器人基本几何逻辑模型建立模块,用于根据所述具体结构参数、所述运动规划参数,并基于共形几何代数理论对机器人的基本构件、运动规划约束构建对应的几何模型,其中,所构建的几何模型以高阶逻辑语言进行描述;
    机器人运动过程几何关系逻辑模型建立模块,用于基于所构建几何模型对机器人的运动过程进行形式化建模,得到机器人的运动过程的几何关系逻辑模型;
    逻辑命题构成模块,用于结合所述几何关系逻辑模型,获得机器人的待验证运动过程的约束或属性对应的运动逻辑关系;
    证明模块,用于验证所述运动逻辑关系是否成立,成立表明所述几何关系逻辑模型满足所述待验证运动过程的约束或者具备所述待验证运动过程的属性,不成立表明所述几何关系逻辑模型不满足所述待验证运动过程的约束或者不具备所述待验证运动过程的属性。
  20. 一种存储介质,其特征在于,所述存储介质用于存储可执行程序代码,所述可执行程序代码用于在运行时执行如权利要求1-18任一项所述的一种基于共形几何代数的机械臂运动规划的形式化分析方法。
  21. 一种应用程序,其特征在于,所述应用程序用于在运行时执行如权利要求1-18任一项所述的一种基于共形几何代数的机械臂运动规划的形式化分析方法。
  22. 一种电子设备,包括:
    处理器、存储器、通信接口和总线;
    所述处理器、所述存储器和所述通信接口通过所述总线连接并完成相互间的通信;
    所述存储器存储可执行程序代码;
    所述处理器通过读取所述存储器中存储的可执行程序代码来运行与所述可执行程序代码对应的程序,以用于执行如权利要求1-18任一项所述的一种基于共形几何代数的机械臂运动规划的形式化分析方法。
PCT/CN2017/072210 2016-01-27 2017-01-23 一种基于共形几何代数的机械臂运动规划的形式化分析方法及系统 WO2017129093A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US15/780,138 US10650179B2 (en) 2016-01-27 2017-01-23 Method and system for formally analyzing the motion planning of a robotic arm based on conformal geometric algebra

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201610055938.2 2016-01-27
CN201610055938.2A CN105760576A (zh) 2016-01-27 2016-01-27 一种基于共形几何代数的机械臂运动规划的形式化分析方法及系统

Publications (1)

Publication Number Publication Date
WO2017129093A1 true WO2017129093A1 (zh) 2017-08-03

Family

ID=56342629

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2017/072210 WO2017129093A1 (zh) 2016-01-27 2017-01-23 一种基于共形几何代数的机械臂运动规划的形式化分析方法及系统

Country Status (3)

Country Link
US (1) US10650179B2 (zh)
CN (2) CN105760576A (zh)
WO (1) WO2017129093A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113021353A (zh) * 2021-03-26 2021-06-25 福州大学 机器人碰撞检测方法
CN115798655A (zh) * 2023-01-25 2023-03-14 江苏新恒基特种装备股份有限公司 一种通过相关参数对锻件进行建模的方法

Families Citing this family (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105760576A (zh) 2016-01-27 2016-07-13 首都师范大学 一种基于共形几何代数的机械臂运动规划的形式化分析方法及系统
CN108393928B (zh) * 2018-01-26 2020-06-30 南京理工大学 柔性机器人机械臂接触碰撞的多刚体-有限元混合分析方法
CN109623810B (zh) * 2018-11-26 2022-04-22 南京航空航天大学 一种机器人平滑的时间最优轨迹规划的方法
CN110125942B (zh) * 2019-06-21 2022-06-10 上海工程技术大学 一种用于移动型蛇形机械臂的平面轨迹跟踪方法
CN110276803B (zh) * 2019-06-28 2021-07-20 首都师范大学 摄像机位姿估计形式化方法、装置、电子设备及存储介质
CN110781596B (zh) * 2019-10-28 2023-03-31 首都师范大学 平面并联机构形式化验证的方法及装置
CN110795856B (zh) * 2019-11-04 2023-04-14 首都师范大学 机械臂稳定性形式化分析方法、装置、设备及存储介质
CN113580142B (zh) * 2021-08-20 2023-10-13 天津大学 基于旋量理论的机器人无冗余几何误差模型解析建模方法

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101244561A (zh) * 2008-03-21 2008-08-20 哈尔滨工程大学 蒸汽发生器检修机械臂安装过程运动学逆解方法
CN103901898A (zh) * 2014-03-28 2014-07-02 哈尔滨工程大学 一种多自由度机器人的逆运动学通用求解方法
US20140277741A1 (en) * 2013-03-15 2014-09-18 Samsung Electronics Co., Ltd. Robots and methods of controlling the same
CN104772773A (zh) * 2015-05-08 2015-07-15 首都师范大学 一种机械臂运动学形式化分析方法
CN105760576A (zh) * 2016-01-27 2016-07-13 首都师范大学 一种基于共形几何代数的机械臂运动规划的形式化分析方法及系统

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US10274325B2 (en) * 2016-11-01 2019-04-30 Brain Corporation Systems and methods for robotic mapping
US10001780B2 (en) * 2016-11-02 2018-06-19 Brain Corporation Systems and methods for dynamic route planning in autonomous navigation

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101244561A (zh) * 2008-03-21 2008-08-20 哈尔滨工程大学 蒸汽发生器检修机械臂安装过程运动学逆解方法
US20140277741A1 (en) * 2013-03-15 2014-09-18 Samsung Electronics Co., Ltd. Robots and methods of controlling the same
CN103901898A (zh) * 2014-03-28 2014-07-02 哈尔滨工程大学 一种多自由度机器人的逆运动学通用求解方法
CN104772773A (zh) * 2015-05-08 2015-07-15 首都师范大学 一种机械臂运动学形式化分析方法
CN105760576A (zh) * 2016-01-27 2016-07-13 首都师范大学 一种基于共形几何代数的机械臂运动规划的形式化分析方法及系统

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
LI, QICAI: "Kinematics Study of Reconfigurable Robot Based on Conformal Geometric Algebra", ELECTRONIC TECHNOLOGY & INFORMATION SCIENCE , CHINA MASTER'S THESES FULL-TEXT DATABASE, 15 August 2015 (2015-08-15), pages 1, 8 and 9 *

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113021353A (zh) * 2021-03-26 2021-06-25 福州大学 机器人碰撞检测方法
CN115798655A (zh) * 2023-01-25 2023-03-14 江苏新恒基特种装备股份有限公司 一种通过相关参数对锻件进行建模的方法

Also Published As

Publication number Publication date
US10650179B2 (en) 2020-05-12
CN106777486A (zh) 2017-05-31
CN105760576A (zh) 2016-07-13
US20180365349A1 (en) 2018-12-20

Similar Documents

Publication Publication Date Title
WO2017129093A1 (zh) 一种基于共形几何代数的机械臂运动规划的形式化分析方法及系统
Wu et al. Finding the kinematic base frame of a robot by hand-eye calibration using 3D position data
CN102637158B (zh) 一种六自由度串联机器人运动学逆解的求解方法
Liu et al. Grasping posture of humanoid manipulator based on target shape analysis and force closure
Patel et al. Manipulator performance measures-a comprehensive literature survey
Schegg et al. Review on generic methods for mechanical modeling, simulation and control of soft robots
Zhang et al. A computationally efficient method for hand–eye calibration
CN112276953B (zh) 连续型超冗余机械臂的臂型协同规划方法、终端设备和存储介质
Ma et al. Higher-order logic formalization of conformal geometric algebra and its application in verifying a robotic manipulation algorithm
Ottenhaus et al. Visuo-haptic grasping of unknown objects based on gaussian process implicit surfaces and deep learning
Wang et al. Optimal robot-world and hand-eye calibration with rotation and translation coupling
Zhang et al. Wsrender: A workspace analysis and visualization toolbox for robotic manipulator design and verification
Ruan et al. Closed-form Minkowski sums of convex bodies with smooth positively curved boundaries
Zheng et al. Generalized distance between compact convex sets: Algorithms and applications
Wu et al. Globally optimal symbolic hand-eye calibration
Wang et al. Manipulation planning with soft constraints by randomized exploration of the composite configuration space
Zhao et al. A learning-based multiscale modelling approach to real-time serial manipulator kinematics simulation
Dai et al. Certified polyhedral decompositions of collision-free configuration space
CN111429481A (zh) 一种基于自适应表达的目标追踪方法、装置及终端
Bezawada et al. Shape reconstruction of soft manipulators using vision and imu feedback
Xing et al. Reconstruction based hand-eye calibration using arbitrary objects
Dong et al. A review of robotic grasp detection technology
Bai et al. Calibration method based on models and least-squares support vector regression enhancing robot position accuracy
Wu et al. Geometric interpretation of the general POE model for a serial-link robot via conversion into DH parameterization
Valsamos et al. Introduction of the high performance area measure for the evaluation of metamorphic manipulator anatomies

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 17743708

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 17743708

Country of ref document: EP

Kind code of ref document: A1