An adaptive damping matrix least squares singularity avoidance method for five-axis machine tools
By using the least squares method of adaptive damping matrix, singularities in five-axis machine tools are avoided in real time, solving the problem of singular phenomena that are difficult to avoid in existing technologies, and improving machining accuracy and surface quality.
Patent Information
- Application Number
- CN202310218670.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2023-03-09
- Publication Date
- 2024-11-15
- Estimated Expiration
- 2043-03-09
AI Technical Summary
Existing technologies struggle to avoid singular configurations in five-axis machine tools in real time, leading to reduced machining accuracy and surface quality damage. Current methods often focus on geometric approaches, which cannot effectively prevent singular phenomena.
The kinematics of a five-axis CNC machine tool are modeled using the least squares method with adaptive damping matrix. A damping diagonal matrix is added to improve the condition number of the Jacobian matrix, avoids singularities in real time, and ensures that the tool contact point position remains unchanged.
It achieves real-time avoidance of five-axis machine tool singularities under the constraint of tool axis direction error, improves machining accuracy and surface quality, and smoothly transitions between singular and non-singular regions.
Smart Images

Figure CN116304510B_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of high-end CNC machine tools and high-end equipment intelligent manufacturing, and in particular relates to an adaptive damping matrix least squares five-axis machine tool singularity avoidance method. Background Art
[0002] Compared with three-axis machine tools, five-axis machine tools have two more rotation axes, which can control the tool posture. Although the increased flexibility can be used to process more complex curved parts, at the same time, the processing path with free change of tool axis direction is very likely to pass through the singularity of the machine tool, resulting in significant degradation of the machine tool's motion and dynamic performance. Constrained by the motion performance of the machine tool's feed axis, this degradation will not only amplify the feed axis motion error, but also easily induce speed mutations, thereby reducing machining accuracy and damaging surface quality. Therefore, avoiding singularity is of great significance for five-axis CNC machining.
[0003] Existing document 1 “Singularity avoidance for five-axis machine tools through introducing geometrical constraints, Wan et al., International Journal of Machine Tools and Manufacture, 2018, 127: 1-13”, this document projects the three-dimensional “singular cone” that needs to be avoided into a “forbidden circle” on the two-dimensional P plane, and then obtains a non-singular path with the minimum error after optimized deformation by adjusting the projection of the B-spline curve control points of the machining path on the P plane; Existing document 2 “A gouge-free tool axis reorientation method for kinematics compliant avoidance of singularity in 5-axis machining, Sun et al., Journal of Manufacturing Science and Engineering, 2019, 141(5): 051010”, this document avoids the singularity of the five-axis machine tool by repositioning the tool axis direction and sequentially adjusting and optimizing the master and slave rotation axis paths. It can be seen that most of the existing technologies avoid the “singularity posture” from the perspective of “geometry” by modifying the tool path in the workpiece coordinate system, and cannot achieve real-time singularity avoidance in five-axis machining. Summary of the invention
[0004] The purpose of the present invention is to solve the problems in the prior art and propose an adaptive damping matrix least squares five-axis machine tool singularity avoidance method.
[0005] The present invention is realized by the following technical scheme. The present invention proposes an adaptive damping matrix least squares five-axis machine tool singularity avoidance method. The method firstly obtains the kinematic equation reflecting the relationship between the workspace WCS and the machine tool joint space MCS by kinematic modeling of the five-axis CNC machine tool, and then obtains the corresponding Jacobian matrix; secondly, the Jacobian matrix is divided into blocks to obtain a sub-Jacobian matrix reflecting the three-dimensional tool axis direction vector and the two-dimensional rotation axis position motion; then, in the process of solving the pseudo-inverse of the sub-Jacobian matrix to realize the inverse kinematic transformation from the workspace tool axis direction vector to the joint space rotation axis motion, a damping diagonal matrix is added to improve the condition number of the pseudo-inverse sub-Jacobian matrix, and the damping diagonal matrix is adaptively calculated under the collaborative constraints of the tool axis direction error and the condition number threshold; finally, according to the rotation axis motion, the motion of the three translation axes is solved under the principle of unchanged tool contact point position, so as to realize real-time avoidance of singularity within the tool axis direction offset tolerance range, and at the same time ensure that the machining tool contact point position remains unchanged.
[0006] Furthermore, the kinematic modeling of the five-axis CNC machine tool is specifically as follows:
[0007] First, the kinematic modeling of the five-axis CNC machine tool is carried out; taking the five-axis machine tool in the form of a B-C' turntable plus a swing head as an example, the kinematic equation reflecting the relationship between the workspace WCS and the machine tool joint space MCS is obtained:
[0008]
[0009] In formula (1), the coordinates of the B-axis and C-axis centers in the machine tool base coordinate system are (B x ,B y ,B x )、(C x ,C y ,C z ); the coordinate position of the workpiece center relative to the C axis is (D x ,D y ,D z ); the position and posture coordinates of the tool in the workpiece coordinate system are (T x ,T y ,T z ) and (i, j, k); the coordinates of each axis in the machine tool coordinate system are q x ,q y ,q z , b and c; the tool length is TL;
[0010] Derivative formula (1) yields:
[0011] dx=J·dq (2)
[0012] In formula (2), x = [T x ,Ty ,T z ,i,j,k] T is the tool pose vector in the workpiece coordinate system; q = [q x ,q y ,q z ,b,c] T is the feed axis coordinate vector of the machine tool joint space;
[0013] Further get the corresponding Jacobian matrix J:
[0014]
[0015] Furthermore, the damping matrix least squares singular avoidance is as follows:
[0016] First, the Jacobian matrix J is divided into blocks to obtain the vector x that reflects the knife axis o =[i,j,k] T and the joint space rotation feed axis coordinate vector q r =[b,c] T The sub-Jacobian matrix J of the relationship between r :
[0017] dx o =J r ·dq r (4)
[0018] In formula (4),
[0019]
[0020] Secondly, in order to avoid singularity, the feed axis coordinate vector q is rotated in the joint space. r In the process, add the damping matrix to the sub-Jacobian pseudo-inverse matrix
[0021]
[0022] In formula (6), dq r,damp is the coordinate vector of the rotation feed axis in the joint space after singularity avoidance; is the pseudo-inverse matrix of the damped sub-Jacobian.
[0023] Furthermore, the damping matrix is adaptively calculated as follows:
[0024] First, since the damping coefficient λ1 improves the pseudo-inverse matrix of the damped sub-Jacobian The condition number of is not affected much, and it will also cause tool direction deviation, so λ1 is set to 0. In this case, the damping matrix becomes
[0025] Secondly, given that the larger the condition number, the more serious the singularity phenomenon, in order to ensure the effect of avoiding singularity, an upper limit is set for the condition number of the damped sub-Jacobian pseudo-inverse matrix M, that is:
[0026] cond(M)≤T0 (7)
[0027] In formula (7), cond() is the condition number calculation function. To ensure that the condition number of the damped sub-Jacobi pseudo-inverse matrix M is limited to its upper limit threshold T0, the value range of the diagonal element θ2 in the diagonal damping matrix is:
[0028]
[0029] Next, in order to prevent the tool axis direction vector from being too large in deviation angle after avoiding singularity, the diagonal damping matrix is further restricted under the tool axis direction error constraint;
[0030] Assume that the maximum allowable value of the tool axis direction error is θ max , the value range of the diagonal element λ2 in the diagonal damping matrix is further restricted to:
[0031]
[0032] Combining equations (8) and (9), we get the value range of the diagonal element λ2:
[0033]
[0034] In formula (10), b min It is the minimum value of the B-axis motion and can be directly determined by the tool axis vector;
[0035] Afterwards, in order to solve the problem of smooth transition of the tool path in the singular area and the non-singular area, the adaptive rule is set for the value of the diagonal element λ2 as follows:
[0036]
[0037] In formula (11), is the coefficient of the exponential function in the adaptive rule, and T is the condition number of the sub-Jacobian matrix corresponding to the current tool position.
[0038] Furthermore, after avoiding singularities, the feed axis coordinates are solved as follows:
[0039] First, obtain the rotation feed axis coordinate vector q in the joint space after the singularity avoidance of the current tool position point r,damp =(b damp ,c damp ) T Then, the actual tool axis vector x after singularity avoidance is calculated. o,damp =(idamp ,j damp ,k damp ) T :
[0040]
[0041] Secondly, by adjusting the position of the tool tip to ensure that the tool contact point remains unchanged, we have:
[0042]
[0043] In formula (13), (T x,damp ,T y,damp ,T z,damp ) is the coordinate of the tool tip after adjustment, and R is the tool radius;
[0044] Finally, through (T x,damp ,T y,damp ,T z,damp ),q r,damp and x o,damp The coordinate vector (x damp ,y damp ,z damp ),have:
[0045]
[0046] Rotate the feed axis coordinate b in the joint space after singularity avoidance damp ,c damp And the joint space rotation feed axis coordinate x after singularity avoidance damp ,y damp ,z damp By inputting it into the feed axis control system as the machining path, real-time five-axis singularity avoidance with tool axis direction error constraint and unchanged tool contact point can be achieved.
[0047] The beneficial effects of the present invention are:
[0048] The present invention proposes an adaptive damping matrix least squares singular avoidance method to avoid "singularity phenomenon" in a mathematical way. By dynamically modifying the Jacobian matrix used to convert the tool path in the workpiece coordinate system into the machine tool coordinate system, the five-axis machine tool tool path can be avoided in real time under the premise of tool axis direction error constraints and unchanged tool contact points. BRIEF DESCRIPTION OF THE DRAWINGS
[0049] Figure 1 The overall flow chart of the method of the present invention is as follows;
[0050] Figure 2 It is a "parabola" shaped machining path in the spatial rectangular coordinate system;
[0051] Figure 3 In order to avoid the singularity by using the method of the present invention, the speed comparison of each feed axis before and after is performed; wherein, the L axis represents the movement time, the unit is s, the M axis represents the speed, the unit is mm / s, and the N axis represents the angular velocity, the unit is rad / s;
[0052] Figure 4 In order to avoid singularity by using the method of the present invention, the acceleration of each feed axis before and after is compared; wherein, the L axis represents the movement time, the unit is s, and the M axis represents the acceleration, the unit is mm / s 2 , the N axis represents the angular acceleration, the unit is rad / s 2 ;
[0053] Figure 5 The method of the present invention is used to avoid the singular front and rear tool axis direction offset angle; wherein the L axis represents the movement time, the unit is s, and the M axis represents the offset angle, the unit is rad. DETAILED DESCRIPTION
[0054] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0055] The present invention aims to invent an adaptive damping matrix least squares five-axis machine tool singularity avoidance method for real-time singularity avoidance from a "mathematical" perspective. The method of the present invention does not directly modify the tool path in the workpiece coordinate system or change the fixture position of the workpiece to avoid singularities, but dynamically modifies the Jacobian matrix used to convert the tool path in the workpiece coordinate system into the machine tool coordinate system. In essence, most existing methods aim to avoid "singularity postures" geometrically, while the proposed method aims to avoid "singularity phenomena" mathematically. At the same time, the method of the present invention also takes into account the tool axis direction offset tolerance constraints and the smooth transition problem between singular areas and non-singular areas, and can adaptively modify the damping matrix, and finally ensure that the position of the tool contact before and after avoiding singularities remains unchanged.
[0056] The present invention proposes an adaptive damping matrix least squares five-axis machine tool singularity avoidance method. The method firstly performs kinematic modeling on the five-axis CNC machine tool to obtain a kinematic equation reflecting the relationship between a workspace WCS and a machine tool joint space MCS, and then obtains a corresponding Jacobian matrix; secondly, the Jacobian matrix is divided into blocks to obtain a sub-Jacobian matrix reflecting a three-dimensional tool axis direction vector and a two-dimensional rotation axis position motion; then, in the process of solving the pseudo-inverse of the sub-Jacobian matrix to realize the inverse kinematic transformation from the workspace tool axis direction vector to the joint space rotation axis motion, a damping diagonal matrix is added to improve the condition number of the pseudo-inverse sub-Jacobian matrix, and the damping diagonal matrix is adaptively calculated under the collaborative constraints of the tool axis direction error and the condition number threshold; finally, according to the rotation axis motion, the motion of the three translation axes is solved under the principle of unchanged tool contact point position, so as to realize real-time singularity avoidance within the tool axis direction offset tolerance range, and at the same time ensure that the machining tool contact point position remains unchanged.
[0057] The kinematic modeling of the five-axis CNC machine tool is specifically as follows:
[0058] First, the kinematic modeling of the five-axis CNC machine tool is carried out; taking the five-axis machine tool in the form of a B-C' turntable plus a swing head as an example, the kinematic equation reflecting the relationship between the workspace WCS and the machine tool joint space MCS is obtained:
[0059]
[0060] In formula (1), the coordinates of the B-axis and C-axis centers in the machine tool base coordinate system are (B x ,B y ,B z )、(C x ,C y ,C z ); the coordinate position of the workpiece center relative to the C axis is (D x ,D y ,D z ); the position and posture coordinates of the tool in the workpiece coordinate system are (T x ,T y ,T z ) and (i, j, k); the coordinates of each axis in the machine tool coordinate system are q x ,q y ,q z , b and c; the tool length is TL;
[0061] Derivative formula (1) yields:
[0062] dx=J·dq (2)
[0063] In formula (2), x = [T x ,T y ,T z,i,j,k] T is the tool pose vector in the workpiece coordinate system; q = [q x ,q y ,q z ,b,c] T is the feed axis coordinate vector of the machine tool joint space;
[0064] Further get the corresponding Jacobian matrix J:
[0065]
[0066] The damping matrix least squares singular avoidance is as follows:
[0067] First, the Jacobian matrix J is divided into blocks to obtain the vector x that reflects the knife axis o =[i,j,k] T and the joint space rotation feed axis coordinate vector q r =[b,c] T The sub-Jacobian matrix J of the relationship between r :
[0068] dx o =J r ·dq r (4)
[0069] In formula (4),
[0070]
[0071] Secondly, in order to avoid singularity, the feed axis coordinate vector q is rotated in the joint space. r In the process, add the damping matrix to the sub-Jacobian pseudo-inverse matrix
[0072]
[0073] In formula (6), dq r,damp is the coordinate vector of the rotation feed axis in the joint space after singularity avoidance; is the pseudo-inverse matrix of the damped sub-Jacobian.
[0074] The damping matrix is adaptively calculated as follows:
[0075] First, since the damping coefficient λ1 improves the pseudo-inverse matrix of the damped sub-Jacobian The condition number of is not affected much, and it will also cause tool direction deviation, so λ1 is set to 0. In this case, the damping matrix becomes
[0076] Secondly, given that the larger the condition number, the more serious the singularity phenomenon, in order to ensure the effect of avoiding singularity, an upper limit is set for the condition number of the damped sub-Jacobian pseudo-inverse matrix M, that is:
[0077] cond(M)≤T0 (7)
[0078] In formula (7), cond() is the condition number calculation function. To ensure that the condition number of the damped sub-Jacobi pseudo-inverse matrix M is limited to its upper limit threshold T0, the value range of the diagonal element λ2 in the diagonal damping matrix is:
[0079]
[0080] Next, in order to prevent the tool axis direction vector from being too large in deviation angle after avoiding singularity, the diagonal damping matrix is further restricted under the tool axis direction error constraint;
[0081] Assume that the maximum allowable value of the tool axis direction error is θ max , the value range of the diagonal element λ2 in the diagonal damping matrix is further restricted to:
[0082]
[0083] Combining equations (8) and (9), we get the value range of the diagonal element λ2:
[0084]
[0085] In formula (10), b min It is the minimum value of the B-axis motion and can be directly determined by the tool axis vector;
[0086] Afterwards, in order to solve the problem of smooth transition of the tool path in the singular area and the non-singular area, the adaptive rule is set for the value of the diagonal element λ2 as follows:
[0087]
[0088] In formula (11), is the coefficient of the exponential function in the adaptive rule, and T is the condition number of the sub-Jacobian matrix corresponding to the current tool position.
[0089] After avoiding singularities, the feed axis coordinates are solved as follows:
[0090] First, obtain the rotation feed axis coordinate vector q in the joint space after the singularity avoidance of the current tool position point r,damp =(b damp ,c damp ) T Then, the actual tool axis vector x after singularity avoidance is calculated. o,damp =(i damp ,j damp,k damp ) T :
[0091]
[0092] Secondly, by adjusting the position of the tool tip to ensure that the tool contact point remains unchanged, we have:
[0093]
[0094] In formula (13), (T x,damp ,T y,damp ,T z,damp ) is the coordinate of the tool tip after adjustment, and R is the tool radius;
[0095] Finally, through (T x,damp ,T y,damp ,T z,damp ),q r,damp and x o,damp The coordinate vector (x damp ,y damp ,z damp ),have:
[0096]
[0097] Rotate the feed axis coordinate b in the joint space after singularity avoidance damp ,c damp And the joint space rotation feed axis coordinate x after singularity avoidance damp ,y damp ,z damp By inputting it into the feed axis control system as the machining path, real-time five-axis singularity avoidance with tool axis direction error constraint and unchanged tool contact point can be achieved.
[0098] Example
[0099] Attached Figure 1 The overall flow chart of the method is shown in Figure 2 It is a "parabola" machining path in the spatial rectangular coordinate system. Figure 2 As an example of the machining path shown in the figure, the feed speed of the tool tip is set to 20 mm / s, and the tool position x is obtained by interpolation. Figure 1 Describe the specific implementation process in detail.
[0100] The first step is the kinematic modeling of five-axis CNC machine tools
[0101] First, the kinematic modeling of the five-axis CNC machine tool is carried out; the five-axis machine tool in the form of a B-C' turntable plus a swing head is used as the model, the tool length is TL=50mm, and the kinematic equation is established as follows:
[0102]
[0103] Further get the corresponding Jacobian matrix J:
[0104]
[0105] The second step is the damping matrix least squares singular avoidance
[0106] Divide the Jacobian matrix J into blocks to obtain the sub-Jacobian matrix J r :
[0107]
[0108] Secondly, to avoid singularity, add a damping matrix have:
[0109]
[0110] The third step is adaptive calculation of the damping matrix
[0111] First, set λ1 to 0, and the damping matrix becomes Secondly, the maximum allowable value of the tool axis direction error is θ max =0.01rad, the condition number of the damped sub-Jacobi pseudo-inverse matrix M is limited to its upper threshold T0 = 30, then:
[0112]
[0113] And b min It can be directly obtained from the knife position x: min =0.007rad, then the value of the diagonal element λ2 is:
[0114]
[0115] Secondly, the value of the Jacobian matrix corresponding to each tool position point can be determined by the movement of the rotating axis. When the tool position point is in the non-singular region, the condition number is T<<30. At this time, the value of λ2 is very small, and the impact on the original processing path is negligible. When the tool position point enters the singular region, the condition number T increases rapidly, and the value of λ2 changes accordingly, and the transition in the middle is smooth, which can timely adjust the condition number of the damped sub-Jacobian pseudo-inverse matrix M.
[0116] Step 4: Solve the feed axis coordinates after avoiding singularities
[0117] First, calculate the actual tool axis vector x after singularity avoidance o,damp =(i damp ,j damp ,k damp ) T:
[0118]
[0119] Secondly, take the radius of the ball-end tool as R = 5mm, and adjust the position of the tool tip to ensure that the tool contact point remains unchanged, and we have:
[0120]
[0121] Finally, through (T x,damp ,T y,damp ,T z,damp ),q r,damp and x o,damp The coordinate vector (x damp ,y damp ,z damp ),have:
[0122]
[0123] Rotate the feed axis coordinate b in the joint space after singularity avoidance damp ,c damp And the joint space rotation feed axis coordinate x after singularity avoidance damp ,y damp ,z damp By inputting it into the feed axis control system as the machining path, real-time five-axis singularity avoidance with tool axis direction error constraint and unchanged tool contact point can be achieved.
[0124] Attached Figure 3 The figure shows the speed comparison of each feed axis before and after avoiding singularity in the above steps; where the L axis represents the motion time in seconds, the M axis represents the speed in mm / s, and the N axis represents the angular velocity in rad / s; Figure 3 It can be seen that before and after avoiding the singularity, the velocity mutation in the singular area is significantly improved, and the transition between the singular area and the non-singular area is smooth. The maximum speeds of the translation axis and the rotation axis are reduced from 3094 mm / s and 111.8 rad / s to 90.08 mm / s and 4 rad / s, respectively.
[0125] Attached Figure 4 The figure shows the acceleration comparison of each feed axis before and after the above steps are used to avoid singularity; where the L axis represents the motion time in seconds, and the M axis represents the acceleration in mm / s 2 , the N axis represents the angular acceleration, the unit is rad / s 2 ; Before and after avoiding the singularity, the acceleration mutation in the singular area is significantly improved, and the transition between the singular area and the non-singular area is smooth. The maximum acceleration of the translation axis and the rotation axis are 382000mm / s respectively 2 and 8157rad / s2 Down to 1484 mm / s 2 and 57.27 rad / s 2 .
[0126] Attached Figure 5 The figure shows the use of the above steps to avoid the singular front and rear tool axis direction offset angle; where the L axis represents the movement time, the unit is s, and the M axis represents the offset angle, the unit is rad; it can be seen that the avoidance of the singular front and rear tool axis direction offset angle is less than the maximum allowable value of the tool axis direction error of 0.01rad. Figure 3 , Attachment Figure 4 and attached Figure 5 It can be seen that the method of the present invention is effective.
[0127] The above is a detailed introduction to the adaptive damping matrix least squares five-axis machine tool singularity avoidance method proposed in the present invention. This article uses specific examples to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea; at the same time, for general technical personnel in this field, according to the idea of the present invention, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present invention.
Claims
1. An adaptive damping matrix least squares five-axis machine tool singularity avoidance method, characterized by: The method firstly obtains the kinematic equation reflecting the relationship between the workspace WCS and the machine tool joint space MCS by kinematic modeling of the five-axis CNC machine tool, and then obtains the corresponding Jacobian matrix; Secondly, the Jacobian matrix is divided into blocks to obtain a sub-Jacobian matrix that reflects the three-dimensional tool axis direction vector and the two-dimensional rotation axis position motion; then, in the process of solving the pseudo-inverse of the sub-Jacobian matrix to realize the inverse kinematic transformation from the workspace tool axis direction vector to the joint space rotation axis motion, a damping diagonal matrix is added to improve the condition number of the pseudo-inverse sub-Jacobian matrix. The damping diagonal matrix is adaptively calculated under the collaborative constraints of the tool axis direction error and the condition number threshold; finally, according to the rotation axis motion, the motion of the three translation axes is solved under the principle of unchanged tool contact point position, so as to achieve real-time avoidance of singularities within the tool axis direction offset tolerance range, while ensuring that the machining tool contact point position remains unchanged; The kinematic modeling of the five-axis CNC machine tool is specifically as follows: First, the kinematic modeling of the five-axis CNC machine tool is carried out; for the five-axis machine tool in the form of a B-C' turntable plus a swing head, the kinematic equation reflecting the relationship between the workspace WCS and the machine tool joint space MCS is: In formula (1), the coordinates of the B-axis and C-axis centers in the machine tool base coordinate system are (B x , B y , B z )、(C x , C y , C z ); the coordinate position of the workpiece center relative to the C axis is (D x , D y , D z ); the position and posture coordinates of the tool in the workpiece coordinate system are (T x , T y , T z ) and (i, j, k); the coordinates of each axis in the machine tool coordinate system are q x ,q y ,q z , b and c; the tool length is TL; Derivative formula (1) yields: dx=J·dq (2) In formula (2), x = [T x , T y , T z , i, j, k] T is the tool pose vector in the workpiece coordinate system; q = [q x ,q y ,q z , b, c] T is the feed axis coordinate vector of the machine tool joint space; Further get the corresponding Jacobian matrix J:
2. The method according to claim 1, characterized in that The damping matrix least squares singular avoidance is as follows: First, the Jacobian matrix J is divided into blocks to obtain the vector x that reflects the knife axis o =[i, j, k] T and the joint space rotation feed axis coordinate vector q r =[b,c] T The sub-Jacobian matrix J of the relationship between r : dx o =J r ·dq r (4) In formula (4), Secondly, in order to avoid singularity, the feed axis coordinate vector q is rotated in the joint space. r In the process, add the damping matrix to the sub-Jacobian pseudo-inverse matrix In formula (6), dq r,damp is the coordinate vector of the rotation feed axis in the joint space after singularity avoidance; is the pseudo-inverse matrix of the damped sub-Jacobian.
3. The method according to claim 2, characterized in that The damping matrix is adaptively calculated as follows: First, since the damping coefficient λ1 improves the pseudo-inverse matrix of the damped sub-Jacobian The condition number of is not affected much, and it will also cause tool direction deviation, so λ1 is set to 0. In this case, the damping matrix becomes Secondly, given that the larger the condition number, the more serious the singularity phenomenon, in order to ensure the effect of avoiding singularity, an upper limit is set for the condition number of the damped sub-Jacobian pseudo-inverse matrix M, that is: cond(M)≤T0 (7) In formula (7), cond() is the condition number calculation function. To ensure that the condition number of the damped sub-Jacobi pseudo-inverse matrix M is limited to its upper limit threshold T0, the value range of the diagonal element λ2 in the diagonal damping matrix is: Next, in order to prevent the tool axis direction vector from being too large in deviation angle after avoiding singularity, the diagonal damping matrix is further restricted under the tool axis direction error constraint; Assume that the maximum allowable value of the tool axis direction error is θ max , the value range of the diagonal element λ2 in the diagonal damping matrix is further restricted to: Combining equations (8) and (9), we get the value range of the diagonal element λ2: In formula (10), b min It is the minimum value of the B-axis motion and can be directly determined by the tool axis vector; Afterwards, in order to solve the problem of smooth transition of the tool path in the singular area and the non-singular area, the adaptive rule is set for the value of the diagonal element λ2 as follows: In formula (11), is the coefficient of the exponential function in the adaptive rule, and T is the condition number of the sub-Jacobian matrix corresponding to the current tool position.
4. The method according to claim 3, characterized in that After avoiding singularities, the feed axis coordinates are solved as follows: First, obtain the rotation feed axis coordinate vector q in the joint space after the singularity avoidance of the current tool position point r,damp =(b damp , c damp ) T Then, the actual tool axis vector x after singularity avoidance is calculated. o,damp =(i damp , j damp , k damp ) T : Secondly, by adjusting the position of the tool tip to ensure that the tool contact point remains unchanged, we have: In formula (13), (T x,damp , T y,damp , T z,damp ) is the coordinate of the tool tip after adjustment, and R is the tool radius; Finally, through (T x,damp , T y,damp , T z,damp ),q r,damp and x o,damp The coordinate vector (x damp ,y damp , z damp ),have: Rotate the feed axis coordinate b in the joint space after singularity avoidance damp, c damp And the joint space rotation feed axis coordinate x after singularity avoidance damp ,y damp , z damp By inputting it into the feed axis control system as the machining path, real-time five-axis singularity avoidance with tool axis direction error constraint and unchanged tool contact point can be achieved.
