A flying thorn crystal path planning method and device

Abstract

The present invention discloses a flying thorn crystal path planning method and device, including initializing the coordinates of multiple control points of the flying thorn crystal to generate multiple initial control point coordinates; constructing a kinematic constraint matrix and a control point basis function matrix based on the coordinates of each initial control point and preset massive transfer motion data; using a preset solution algorithm to determine the target control order based on the kinematic constraint matrix, the control point basis function matrix and the coordinates of each initial control point; determining the target Bezier motion curve of the flying thorn crystal based on the target control order, the kinematic constraint matrix and the control point basis function matrix; solving the technical problem that the existing flying thorn crystal path planning method causes the obtained Bezier path to be too complex.

Description

Technical Field

[0001] The present invention relates to the field of path optimization technology, and in particular to a flying thorn crystal path planning method and device. Background Art

[0002] With the continuous development of display technology, Mini / Micro LED has become a revolutionary technology in the new display field with its excellent display performance and longer service life. However, the mass production of Mini / Micro LED screens still faces a series of technical challenges such as substrate selection, inspection and repair. The most important technical difficulty is the mass transfer of chips. Mass transfer refers to the process of transferring Mini / Micro LED chips from the source substrate to the target circuit substrate. From the chip transfer method to the design of the transfer equipment's motion parameters and motion path, all will affect the chip transfer effect (positioning accuracy, response speed), thereby determining the yield rate and mass production efficiency of the finished Mini / Micro LED display.

[0003] While mass transfer devices based on flying spikes have been designed, detailed motion planning methods for these scenarios have yet to be fully explored. For these scenarios, the needle's trajectory must not only achieve smooth transitions and stable start / stop times, but also maintain the same velocity as the substrate during chip transfer to avoid scratching the chips.

[0004] Existing path planning methods in other fields mostly design motion paths based on polynomial curves. Bezier curves are widely used due to their high plasticity and path freedom. However, most planning methods can only optimize Bezier curves based on a fixed control order. When the motion requirements are too strict, the control order must be set to a large value, resulting in an overly complex Bezier path. Summary of the Invention

[0005] The present invention provides a flying sting crystal path planning method and device, which are used to solve the technical problem that the existing flying sting crystal path planning method causes the obtained Bezier path to be too complex.

[0006] A first aspect of the present invention provides a flying thorn crystal path planning method, comprising:

[0007] In response to the planning request, the coordinates of multiple control points of the flying thorn crystal are initialized to generate multiple initial control point coordinates;

[0008] Constructing a kinematic constraint matrix and a control point basis function matrix according to the coordinates of each of the initial control points and the preset massive transfer motion data;

[0009] Determining a target control order using a preset solution algorithm based on the kinematic constraint matrix, the control point basis function matrix, and the coordinates of each initial control point;

[0010] The target Bezier motion curve of the flying thorn crystal is determined according to the target control order, the kinematic constraint matrix and the control point basis function matrix.

[0011] Optionally, the preset mass transfer motion data includes a path motion starting point, a path motion end point, velocity vector restriction data, and upper and lower limit data of the motion space; the kinematic constraint matrix includes a constraint matrix on the left side of the kinematic equation, a constraint matrix on the right side of the kinematic equation, and a constraint matrix on the right side of the kinematic inequality; the control point basis function matrix includes an initial control point matrix and a basis function matrix corresponding to the initial control point matrix; the step of constructing the kinematic constraint matrix based on the coordinates of each of the initial control points and the preset mass transfer motion data includes:

[0012] Using the coordinates of each of the initial control points to construct an initial control point matrix;

[0013] Using the basis functions corresponding to the coordinates of each of the initial control points, a constraint matrix and a basis function matrix on the left side of the kinematic equation corresponding to the initial control point matrix are constructed;

[0014] According to the path motion starting point, path motion end point and velocity vector restriction data, the constraint matrix on the right side of the kinematic equation is constructed;

[0015] The constraint matrix on the right side of the kinematic inequality is constructed using the upper and lower limit data of the motion space.

[0016] Optionally, the step of determining the target control order according to the kinematic constraint matrix, the control point basis function matrix and the coordinates of each initial control point by using a preset solution algorithm includes:

[0017] Using the initial control point matrix as the initial order control point matrix;

[0018] Based on the order of the initial order control point matrix, a first optimized inequality constraint matrix is ​​constructed using the right-side constraint matrix of the kinematic inequality;

[0019] Constructing a first-order performance function according to the first optimized inequality constraint matrix, the right-side constraint matrix of the kinematic equation, the preset left-side constraint matrix of the kinematic inequality, the initial order control point matrix, and the left-side constraint matrix of the kinematic equation corresponding to the initial order control point matrix and a basis function matrix;

[0020] Solving the first-order performance function using the preset solving algorithm to determine a first optimization control point matrix and a first performance index;

[0021] Determining an increased-order matrix based on the order of the first optimized control point matrix;

[0022] Determining an intermediate order control point matrix according to the raised order matrix and the first optimized control point matrix;

[0023] Based on the order of the intermediate-order control point matrix, a second optimized inequality constraint matrix is ​​constructed using the right-side constraint matrix of the kinematic inequality;

[0024] Constructing a second-order performance function according to the second optimized inequality constraint matrix, the right-side constraint matrix of the kinematic equation, the left-side constraint matrix of the preset kinematic inequality, the intermediate-order control point matrix, and the left-side constraint matrix of the kinematic equation corresponding to the intermediate-order control point matrix and the basis function matrix;

[0025] Solving the second-order performance function using the preset solving algorithm to determine a second optimization control point matrix and a second performance index;

[0026] comparing the second performance indicator with the first performance indicator;

[0027] If the first performance index is less than or equal to the second performance index, the order of the first optimization control point matrix is ​​used as the target control order.

[0028] Optionally, it also includes:

[0029] If the first performance index is greater than the second performance index, taking the second optimization control point matrix as a new first optimization control point matrix and taking the second performance index as a new first performance index;

[0030] Determining a second performance indicator using the preset solving algorithm according to the new first optimization control point matrix, the right-side constraint matrix of the kinematic equation, the left-side constraint matrix of the preset kinematic inequality, and the order of the new first optimization control point matrix until the first performance indicator is less than or equal to the second performance indicator;

[0031] The order of the first optimized control point matrix determined when the first performance index is less than or equal to the second performance index is used as the target control order.

[0032] Optionally, the step of determining a target Bezier motion curve of the flying thorn crystal according to the target control order, the kinematic constraint matrix and the control point basis function matrix includes:

[0033] Constructing an initial Bezier curve using the initial control point matrix and a basis function matrix corresponding to the initial control point matrix;

[0034] Determining a target raised-order matrix based on the target control order and the order of the initial control point matrix;

[0035] Constructing a raised-order control point matrix according to the target raised-order matrix and the first optimized control point matrix;

[0036] Determining a multi-time point basis function matrix based on the column vectors in the raised-order control point matrix and the initial Bezier curve;

[0037] Based on the order of the initial control point matrix, an initial optimization inequality constraint matrix is ​​constructed using the right-side constraint matrix of the kinematic inequality;

[0038] Constructing a target order optimization performance function using the multi-time point basis function matrix, the initial control point matrix, the basis function matrix corresponding to the initial control point matrix, the left-side constraint matrix of the kinematic equation, the right-side constraint matrix of the kinematic equation, the left-side constraint matrix of the preset kinematic inequality, the target raised-order matrix, and the initial optimized inequality constraint matrix;

[0039] Solving the target order optimization performance function using the preset solving algorithm to determine the target control point matrix;

[0040] The target Bezier motion curve of the flying thorn crystal is determined according to the basis function matrix corresponding to the target control point matrix and the initial control point matrix.

[0041] Optionally, the target order optimization performance function is specifically:

[0042]

[0043] Among them, J(P n ) is the target order optimization performance function; P n is the initial control point matrix, which corresponds to the n-order control point; is a time derivative; B n (t) is the basis function matrix corresponding to the initial control point matrix, and the order of the matrix is ​​n; T is the transpose of the matrix; A eq (n) is the constraint matrix on the left side of the kinematic equation corresponding to the initial control point matrix, and the order of this matrix is ​​n; c eq A is the constraint matrix on the right side of the kinematic equation; ie is the constraint matrix on the left side of the preset kinematic inequality; is the target raised rank matrix; is the multi-time point basis function matrix; is the initial optimization inequality constraint matrix, the order of the matrix is ​​n; It is the preset upgrade fault tolerance distance.

[0044] A second aspect of the present invention provides a flying thorn crystal path planning device, comprising:

[0045] The response module is used to respond to the planning request, initialize the coordinates of multiple control points of the flying thorn crystal, and generate multiple initial control point coordinates;

[0046] A matrix construction module is used to construct a kinematic constraint matrix and a control point basis function matrix according to the coordinates of each of the initial control points and the preset massive transfer motion data;

[0047] A control order determination module is used to determine a target control order according to the kinematic constraint matrix, the control point basis function matrix and the coordinates of each initial control point using a preset solution algorithm;

[0048] The target path determination module is used to determine the target Bezier motion curve of the flying thorn crystal according to the target control order, the kinematic constraint matrix and the control point basis function matrix.

[0049] A third aspect of the present invention provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the flying spike path planning method as described in any one of the above items.

[0050] A fourth aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed, implements the steps of the flying spike crystal path planning method as described in any one of the above items.

[0051] A fifth aspect of the present invention provides a computer program product, which includes a computer program stored on a non-transitory computer-readable storage medium, and the computer program includes program instructions, wherein when the program instructions are executed by a computer, the computer executes the steps of the flying spike path planning method as described in any one of the above items.

[0052] It can be seen from the above technical solutions that the present invention has the following advantages:

[0053] The above scheme of the present invention provides a flying thorn crystal path planning method. When it is necessary to plan the path of the flying thorn crystal, the coordinates of multiple control points of the flying thorn crystal are first initialized to generate multiple initial control point coordinates; then, according to the coordinates of each initial control point and the preset massive transfer motion data, the kinematic constraint matrix and the control point basis function matrix are constructed; a preset solution algorithm is used to determine the target control order according to the kinematic constraint matrix, the control point basis function matrix and the coordinates of each initial control point; finally, the target Bezier motion curve of the flying thorn crystal is determined according to the target control order, the kinematic constraint matrix and the control point basis function matrix; based on the above scheme, a preset solution algorithm is used to determine the target control order according to the constructed kinematic constraint matrix, the control point basis function matrix and the initial control point coordinates obtained after initialization, and the target Bezier motion curve of the flying thorn crystal is determined in combination with the kinematic constraint matrix and the control point basis function matrix. There is no need to set a fixed control order in advance, and a low-order optimal Bezier curve can be obtained even under strict motion requirements, making the Bezier path more concise and easy to use. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0055] Figure 1 A flowchart of a flying thorn crystal path planning method provided by an embodiment of the present invention;

[0056] Figure 2 A schematic diagram of pre-setting a large amount of transferred motion data according to an embodiment of the present invention;

[0057] Figure 3 A diagram showing the displacement, velocity, and acceleration changes of the flying spike crystal in the X-axis direction provided by an embodiment of the present invention;

[0058] Figure 4 A diagram showing the displacement, velocity, and acceleration changes of a flying spike crystal in the Z-axis direction provided by an embodiment of the present invention;

[0059] Figure 5 A schematic diagram of the overall motion route, control points, and motion space of the flying thorn crystal provided by an embodiment of the present invention;

[0060] Figure 6 A flowchart of another flying thorn crystal path planning method provided by an embodiment of the present invention;

[0061] Figure 7A schematic diagram of the motion of a flying thorn crystal mass transfer provided by an embodiment of the present invention;

[0062] Figure 8 A schematic diagram of a flow chart of a flying thorn crystal path planning method provided in an embodiment of the present invention;

[0063] Figure 9 This is a structural block diagram of a flying thorn crystal path planning device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0064] The embodiments of the present invention provide a flying sting crystal path planning method and device, which are used to solve the technical problem that the existing flying sting crystal path planning method causes the obtained Bezier path to be too complex.

[0065] In order to make the purpose, features, and advantages of the present invention more obvious and easy to understand, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the embodiments described below 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 making creative work are within the scope of protection of the present invention.

[0066] See also Figure 1 , Figure 1 A flowchart of the steps of a flying thorn crystal path planning method provided by an embodiment of the present invention.

[0067] The present invention provides a flying thorn crystal path planning method, comprising:

[0068] Step 101: In response to a planning request, the coordinates of multiple control points of the flying thorn crystal are initialized to generate multiple initial control point coordinates.

[0069] It should be noted that, during the stage movement time, the flying thorn crystal will have corresponding multiple initial control point coordinates; wherein, the number of initial control point coordinates can be set as needed, and the present invention is not limited thereto.

[0070] Step 102: Construct a kinematic constraint matrix and a control point basis function matrix based on the coordinates of each initial control point and the preset massive transfer motion data.

[0071] The kinematic constraint matrix includes the constraint matrix on the left side of the kinematic equation, the constraint matrix on the right side of the kinematic equation, and the constraint matrix on the right side of the kinematic inequality.

[0072] The control point basis function matrix includes an initial control point matrix and a basis function matrix corresponding to the initial control point matrix.

[0073] Please note that Figure 2, preset massive transfer motion data including path motion starting point u0=[x0,z0] T 、Path motion end point u e =[x e ,z e ] T , velocity vector restriction data (i.e., the restriction of the velocity vector of the thorn crystal at the end point of the motion path v e =[x ve ,z ve ] T ), the upper and lower limit data of the motion space, the upper and lower limit data of the motion space include the lower limit u of the motion space m =[x m ,z m ] T and the upper limit u of the motion space M =[x M ,z M ] T , these data represent the range of restrictions on the spatial position of the transfer device during the movement process; among them, x0 represents the coordinate of the starting point of the movement in the X-axis direction, z0 represents the coordinate of the starting point of the movement in the Z-axis direction, x e Represents the coordinate of the end point of the movement in the X-axis direction, z e Represents the coordinate of the end point of the movement in the Z-axis direction, x m Represents the lower limit coordinate of the motion space in the X-axis direction, z m Represents the lower limit coordinate of the motion space in the Z-axis direction, x M Represents the upper limit coordinate of the motion space in the X-axis direction, z M Represents the upper limit coordinate of the motion space in the Z-axis direction, x ve Represents the limit value of the velocity vector of the thorn crystal on the X axis at the end point of the motion path, z ve Represents the limit value of the velocity vector of the thorn crystal on the Z axis at the end point of the motion path.

[0074] Furthermore, the preset massive transfer motion data also includes the j-order smooth start requirement of the motion process and the initial control order (representing the complexity of the Bessel motion path). The initial control order is the order of the initial control point matrix. The j-order smooth start requirement is defined as: at the starting point of the motion, the 1st to jth time derivative values ​​of the path are all 0.

[0075] Specifically, the coordinates of each initial control point are used to construct an initial control point matrix, which can be expressed as: P n =[p0,p1,…,p n ], where P n The initial control point coordinates p0, p1, ..., p nThe initial control point matrix is ​​composed of p0, p1 and p2, where p is the coordinate of the first initial control point, p is the coordinate of the second initial control point, and p is the coordinate of the second initial control point. n is the coordinate of the nth initial control point.

[0076] It is worth mentioning that according to the initial control point matrix and the basis function matrix corresponding to the initial control point matrix, the motion performance index function can be obtained, which represents the motion energy consumed by the motion path. The motion performance index function can be expressed as:

[0077]

[0078] Among them, J(P n ) is the motion performance index function; B n (t) is the basis function matrix corresponding to the initial control point matrix; is a time derivative; P n is the initial control point matrix; T is the matrix transpose.

[0079] Furthermore, the basis functions corresponding to the coordinates of each initial control point are used to construct the constraint matrix and basis function matrix on the left side of the kinematic equation corresponding to the initial control point matrix. The constraint matrix on the left side of the kinematic equation corresponding to the initial control point matrix can be expressed as:

[0080]

[0081] Among them, A eq (n) is the constraint matrix on the left side of the kinematic equation corresponding to the initial control point matrix of order n; is the jth time derivative; T is the transpose of the matrix; B n (t) is the basis function of the Berstein basis function (the basis function corresponding to the coordinates of each initial control point ) is an n-dimensional Berstein matrix (a basis function matrix of order n) composed of , which represents the time-dependent part of the Bezier motion curve, t∈[0,1].

[0082] The construction process of the basis function matrix corresponding to the initial control point matrix can be expressed as:

[0083]

[0084] Among them, B n (t) is the basis function matrix of order n; t is the stage motion time; is the basis function corresponding to the coordinates of the first initial control point; is the basis function corresponding to the coordinates of the second initial control point; is the basis function corresponding to the coordinates of the n+1th initial control point; is the basis function corresponding to the coordinates of the i+1th initial control point; ti is the stage motion time corresponding to the coordinates of the i+1th initial control point.

[0085] Furthermore, according to the path motion starting point, path motion end point and velocity vector restriction data, a constraint matrix on the right side of the kinematic equation is constructed, wherein the constraint matrix on the right side of the kinematic equation can be expressed as:

[0086]

[0087] Among them, c eq is the constraint matrix on the right side of the kinematic equation; x0 is the coordinate of the starting point of the motion in the X-axis direction; z0 is the coordinate of the starting point of the motion in the Z-axis direction; x e is the coordinate of the end point of the movement in the X-axis direction; z e is the coordinate of the end point of the movement in the Z-axis direction; ve The limit value of the velocity vector of the thorn crystal at the end point of the motion path on the X axis; ze The limit value of the velocity vector of the thorn crystal on the Z axis at the end point of the motion path.

[0088] Furthermore, the right-side constraint matrix of the kinematic inequality is constructed through the upper and lower limit data of the motion space; wherein, the right-side constraint matrix of the kinematic inequality can be expressed as: c ie =[x M -x m z M -z m ] T , where c ie is the constraint matrix on the right side of the kinematic inequality, x M is the upper limit coordinate of the motion space in the X-axis direction, x m is the lower limit coordinate of the motion space in the X-axis direction, z M is the upper limit coordinate of the motion space in the Z-axis direction, z m It is the lower limit coordinate of the motion space in the Z-axis direction.

[0089] Step 103: Using a preset solution algorithm, the target control order is determined according to the kinematic constraint matrix, the control point basis function matrix and the coordinates of each initial control point.

[0090] Specifically, the initial control point matrix is ​​used as the initial order control point matrix, that is, in the process of matrix order upgrading, the initial order control point matrix is ​​initialized with the initial control point matrix to obtain the initial order control point matrix, wherein the initial order control point matrix can be expressed as: P k =[p0,p1,…,p k ], P kis the initial order control point matrix of order k, p0 is the first control point in the initial order control point matrix, p1 is the second control point in the initial order control point matrix, p k is the kth control point in the initial order control point matrix.

[0091] Furthermore, based on the order of the initial order control point matrix, the right-hand side constraint matrix of the kinematic inequality is used to construct the first optimization inequality constraint matrix; wherein, the first optimization inequality constraint matrix can be expressed as: is the first optimization inequality constraint matrix composed of k+1 kinematic inequality right-hand constraint matrices. The order of this matrix is ​​k, representing p k The corresponding optimization inequality constraint matrix), c ie is the constraint matrix on the right side of the kinematic inequality.

[0092] Furthermore, P k All control points in the kinematic inequality are restricted by the mechanical energy of the right-side constraint matrix. The first-order performance function is constructed according to the first optimized inequality constraint matrix, the right-side constraint matrix of the kinematic equation, the preset left-side constraint matrix of the kinematic inequality, the initial order control point matrix, and the left-side constraint matrix of the kinematic equation corresponding to the initial order control point matrix, and the basis function matrix. By solving the first-order performance function, P is found under the condition that it does not exceed the range of motion and meets the 1-j order smooth start requirements. k Make the motion performance index function reach the minimum value; the first-order performance function can be expressed as:

[0093]

[0094] Among them, J(P k 0 is the first-order performance function; P k is the initial order control point matrix; is a time derivative; B k (t0 is the basis function matrix corresponding to the initial order control point matrix, the order of the matrix is ​​k; T is the matrix transpose; A eq (k) is the constraint matrix on the left side of the kinematic equation corresponding to the initial order control point matrix, and the order of this matrix is ​​k; c eq A is the constraint matrix on the right side of the kinematic equation; ie The constraint matrix on the left side of the preset kinematic inequality can be expressed as is the first optimization inequality constraint matrix, and the order of the matrix is ​​k.

[0095] It is worth mentioning that according to the above-mentioned principle of using the basis functions corresponding to the coordinates of each initial control point to construct the constraint matrix and basis function matrix on the left side of the kinematic equation corresponding to the initial control point matrix, the constraint matrix and basis function matrix on the left side of the kinematic equation corresponding to the control point matrix are composed of the basis functions corresponding to the coordinates of the control points. Therefore, the constraint matrix and basis function matrix on the left side of the kinematic equation corresponding to the initial order control point matrix can be obtained based on the principle of this step; among which, the basis function matrix corresponding to the initial order control point matrix can be expressed as:

[0096]

[0097] Among them, B k (t) is the basis function matrix corresponding to the initial order control point matrix, the order of this matrix is ​​k, representing the k-dimensional Berstein matrix; is the first basis function in the basis function matrix corresponding to the initial order control point matrix; is the second basis function in the basis function matrix corresponding to the initial order control point matrix; is the kth basis function in the basis function matrix corresponding to the initial order control point matrix; T is the matrix transpose.

[0098] Furthermore, a preset solution algorithm is used to solve the first-order performance function to determine the first optimization control point matrix and the first performance index; wherein the preset solution algorithm is a projection gradient algorithm, a penalty function method, a multiplier method and other solution methods; by solving the above k-order optimization problem (first-order performance function), the k-order optimal control point matrix is ​​obtained (first optimization control point matrix) and k-order suboptimal performance index J k (the first performance indicator), i.e.

[0099] Furthermore, based on the order of the first optimized control point matrix, a boosted order matrix is ​​determined; wherein the boosted order matrix is ​​a k+1 dimensional boosted order matrix having k+1 rows and k+2 columns, which represents the initial order control point matrix P k The calculation process of increasing the number of control points from k+1 to k+2; when constructing the intermediate-order control point matrix, it is necessary to traverse and calculate the elements in each row and column of the raised-order matrix. The calculation formula of the element located in the x+1th row and y+1th column of the raised-order matrix, that is, the construction process in the raised-order matrix, can be expressed as:

[0100]

[0101] in, is the element in the x+1th row and y+1th column of the raised-order matrix; x is the xth row of the raised-order matrix; y is the yth column of the raised-order matrix; k is the order of the first optimization control point matrix.

[0102] Furthermore, according to the raised order matrix and the first optimized control point matrix, an intermediate order control point matrix is ​​determined. The intermediate order control point matrix is ​​a k+1-dimensional control point matrix obtained after the first optimized control point matrix is ​​raised in order, and consists of a plurality of control points p0, p1, ..., p k+1 The construction process of the intermediate order control point matrix can be expressed as: P k+1 is the intermediate order control point matrix, is the first optimized control point matrix, is a raised rank matrix.

[0103] Furthermore, based on the order of the intermediate-order control point matrix, the right-hand side constraint matrix of the kinematic inequality is used to construct the second optimization inequality constraint matrix. The second optimization inequality constraint matrix is ​​a k+1-order optimization inequality constraint matrix, which can be expressed as The second optimization inequality constraint matrix consists of k+2 kinematic inequality right-hand constraint matrices, representing the intermediate order control point matrix P k+1 All control points in the equation are constrained using inequality constraints c ie Make restrictions.

[0104] Furthermore, a second-order performance function is constructed based on the second optimized inequality constraint matrix, the right-side constraint matrix of the kinematic equation, the left-side constraint matrix of the preset kinematic inequality, the intermediate-order control point matrix, and the left-side constraint matrix of the kinematic equation corresponding to the intermediate-order control point matrix, and the basis function matrix; wherein, regarding the intermediate-order control point matrix P k+1 The specific form of the k+1 order optimization problem is as follows, that is, the second-order performance function can be expressed as:

[0105]

[0106] Among them, J(P k+1 ) is the second-order performance function; P k+1 is the intermediate order control point matrix; is a time derivative; B k+1 (t) is the basis function matrix corresponding to the intermediate order control point matrix, the order of this matrix is ​​k+1; T is the matrix transpose; A eq (k+10 is the constraint matrix on the left side of the kinematic equation corresponding to the intermediate order control point matrix, and the order of this matrix is ​​k+1; c eq A is the constraint matrix on the right side of the kinematic equation; ie The constraint matrix on the left side of the preset kinematic inequality can be expressed as is the second optimization inequality constraint matrix, and the order of this matrix is ​​k+1.

[0107] It is worth mentioning that the principles of the matrix construction steps for the constraint matrix and basis function matrix on the left side of the kinematic equation corresponding to the intermediate order control point matrix are consistent with the principles of the matrix construction steps for the constraint matrix and basis function matrix on the left side of the kinematic equation corresponding to the initial order control point matrix, and the present invention will not elaborate on them in detail; among them, the basis function matrix corresponding to the intermediate order control point matrix can be expressed as:

[0108]

[0109] Among them, B k+1 (t) is the basis function matrix corresponding to the intermediate order control point matrix, the order of this matrix is ​​k+1, representing the k+1 dimensional Berstein matrix; is the first basis function in the basis function matrix corresponding to the intermediate order control point matrix; is the second basis function in the basis function matrix corresponding to the intermediate order control point matrix; is the k+2th basis function in the basis function matrix corresponding to the intermediate-order control point matrix; T is the matrix transpose.

[0110] Furthermore, under the condition of not exceeding the range of motion and meeting the requirements of 1~j-order smooth start, find P k+1 In order to make the motion performance index function reach the minimum value, the preset solution algorithm is used to solve the second-order performance function to determine the second optimal control point matrix and the second performance index. That is, by solving the above k+1 order optimization problem (second-order performance function), the k+1 order optimal control point matrix (second optimal control point matrix) is obtained. ) and the k+1-order suboptimal performance index (second performance index), which can be expressed as

[0111] Further, the second performance index and the first performance index are compared; if the first performance index is less than or equal to the second performance index, the order of the first optimization control point matrix is ​​used as the target control order; if the first performance index is greater than the second performance index, the second optimization control point matrix is ​​used as the new first optimization control point matrix, the second performance index is used as the new first performance index, and the step of determining the upgraded matrix based on the order of the first optimization control point matrix is ​​jumped to execution until the first performance index is less than or equal to the second performance index; the order of the first optimization control point matrix determined when the first performance index is less than or equal to the second performance index is used as the target control order.

[0112] Step 104: Determine the target Bezier motion curve of the flying thorn crystal according to the target control order, the kinematic constraint matrix and the control point basis function matrix.

[0113] Specifically, the initial control point matrix and the basis function matrix corresponding to the initial control point matrix are used to construct the initial Bezier curve; wherein the initial Bezier curve can be expressed as:

[0114]

[0115] Among them, f(t) is the initial Bezier curve; p i The coordinates of the i+1th initial control point corresponding to the initial Bezier curve; is the i+1th basis function in the basis function matrix corresponding to the initial control point matrix; P n is the initial control point matrix; B n (t) is the basis function matrix corresponding to the initial control point matrix; t is the stage motion time;

[0116] Furthermore, based on the target control order and the order of the initial control point matrix, a target raised order matrix is ​​determined; the target raised order matrix is ​​an n+1 dimensional raised order matrix with n+1 rows and m+1 columns, which represents the initial control point matrix P n The calculation process of increasing the number of control points from n+1 to m+1 requires traversing the elements of each row and column in the matrix when constructing the matrix. The calculation formula of the element at the x+1th row and y+1th column, that is, the construction process of the target raised-order matrix, can be expressed as:

[0117]

[0118] in, is the element in the x+1th row and y+1th column of the target raised-order matrix; n is the order of the initial control point matrix; m is the target control order; x is the xth row of the target raised-order matrix; y is the yth column of the target raised-order matrix.

[0119] Furthermore, a raised-order control point matrix is ​​constructed according to the target raised-order matrix and the first optimized control point matrix; wherein the first optimized control point matrix used to construct the raised-order control point matrix is ​​the first optimized control point matrix obtained in the first matrix order raising process, that is, the first optimized control point matrix obtained by solving the first-order performance function based on the initial-order control point matrix; the raised-order control point matrix can be expressed as: Q m is the raised-order control point matrix, is the first optimized control point matrix, is the target raised rank matrix.

[0120] Furthermore, based on the column vectors in the raised-order control point matrix and the initial Bezier curve, the multi-time point basis function matrix is ​​determined. The construction process of the multi-time point basis function matrix can be expressed as:

[0121]

[0122] in, is the multi-time point basis function matrix (multi-time point Berstein matrix); is the first basis function matrix in the multi-time point basis function matrix; is the m+1th basis function matrix in the multi-time point basis function matrix; are multiple time points corresponding to the multi-time point basis function; is the vector 2-norm; q j is the column vector of the jth column in the ascending control point matrix; f(t) is the initial Bezier curve; is the second basis function matrix in the multi-time point basis function matrix.

[0123] Furthermore, based on the order of the initial control point matrix, the right-hand constraint matrix of the kinematic inequality is used to construct the initial optimization inequality constraint matrix; wherein, the initial optimization inequality constraint matrix represents the initial control point matrix p n All control points in the equation are constrained using inequality constraints c ie The restriction can be expressed as: is the initial optimization inequality constraint matrix composed of n+1 right-hand constraint matrices of kinematic inequalities. The order of this matrix is ​​n, representing the n-order optimization inequality constraint matrix. c ie is the constraint matrix on the right side of the kinematic inequality.

[0124] Furthermore, the target order optimization performance function is constructed by using the multi-time point basis function matrix, the initial control point matrix, the basis function matrix corresponding to the initial control point matrix, the left-side constraint matrix of the kinematic equation, the right-side constraint matrix of the kinematic equation, the left-side constraint matrix of the preset kinematic inequality, the target order matrix, and the initial optimization inequality constraint matrix. The target order optimization performance function can be expressed as:

[0125]

[0126] Among them, J(P n ) is the target order optimization performance function; P n is the initial control point matrix, the order of the matrix is ​​n; is a time derivative; B n (t) is the basis function matrix corresponding to the initial control point matrix, and the order of the matrix is ​​n; T is the transpose of the matrix; A eq (n) is the constraint matrix on the left side of the kinematic equation corresponding to the initial control point matrix, and the order of this matrix is ​​n; c eq A is the constraint matrix on the right side of the kinematic equation; ieis the constraint matrix on the left side of the preset kinematic inequality; is the target raised rank matrix; is the multi-time point basis function matrix; is the initial optimization inequality constraint matrix, the order of the matrix is ​​n; To preset the tolerance distance for the order of improvement, set the value between [0,1], which represents the maximum distance between the Bezier curve and the constraint boundary during optimization, to avoid excessive optimization leading to violation of the constraints.

[0127] Furthermore, by solving the above m-order optimization problem (target order optimization performance function), the optimal control point matrix (target control point matrix) is obtained. ), that is, the preset solution algorithm is used to solve the target order optimization performance function and determine the target control point matrix. The process can be expressed as:

[0128] Further, see Figure 3-Figure 5 , according to the basis function matrix corresponding to the target control point matrix and the initial control point matrix, the target Bezier motion curve (motion path) of the flying thorn crystal is determined; the target Bezier motion curve obtained by the above-mentioned step of temporarily mapping the low-order control points to high-order control points for optimization can meet the stringent motion requirements and make the motion path more concise and easy to use; wherein, the target Bezier motion curve can be expressed as: t∈[0,1], is the target Bezier motion curve, is the target control point matrix, B n (t) is the basis function matrix corresponding to the initial control point matrix, and t is the stage motion time.

[0129] In an embodiment of the present invention, the present invention provides a flying thorn crystal path planning method. When it is necessary to plan the path of the flying thorn crystal, the coordinates of multiple control points of the flying thorn crystal are first initialized to generate multiple initial control point coordinates; then, according to the coordinates of each initial control point and the preset massive transfer motion data, a kinematic constraint matrix and a control point basis function matrix are constructed; a preset solution algorithm is used to determine the target control order according to the kinematic constraint matrix, the control point basis function matrix and the coordinates of each initial control point; finally, the target Bezier motion curve of the flying thorn crystal is determined according to the target control order, the kinematic constraint matrix and the control point basis function matrix; based on the above scheme, a preset solution algorithm is used to determine the target control order according to the constructed kinematic constraint matrix, the control point basis function matrix and the initial control point coordinates obtained after initialization, and the process of determining the target Bezier motion curve of the flying thorn crystal in combination with the kinematic constraint matrix and the control point basis function matrix does not require setting a fixed control order in advance, and a low-order optimal Bezier curve can be obtained even under strict motion requirements, making the Bezier path more concise and easy to use.

[0130] See also Figure 6 , Figure 6 A flowchart of the steps of another flying spike crystal path planning method provided by an embodiment of the present invention.

[0131] The present invention provides a flying thorn crystal path planning method, comprising:

[0132] Step 601: In response to a planning request, the coordinates of multiple control points of the flying thorn crystal are initialized to generate multiple initial control point coordinates.

[0133] In this embodiment, the specific implementation process of step 601 is similar to that of step 101 and will not be repeated here.

[0134] Step 602: Use the coordinates of each initial control point to construct an initial control point matrix.

[0135] It should be noted that the initial control point matrix can be expressed as: P n =[p0,p1,…,p n ], where P n The initial control point coordinates p0, p1, ..., p n The initial control point matrix is ​​composed of p0, p1 and p2, where p is the coordinate of the first initial control point, p is the coordinate of the second initial control point, and p is the coordinate of the second initial control point. n is the coordinate of the nth initial control point.

[0136] Step 603: Using the basis functions corresponding to the coordinates of each initial control point, construct the constraint matrix and basis function matrix on the left side of the kinematic equation corresponding to the initial control point matrix.

[0137] The preset massive transfer motion data includes the path motion starting point, the path motion end point, the velocity vector restriction data, and the upper and lower limit data of the motion space; the kinematic constraint matrix includes the constraint matrix on the left side of the kinematic equation, the constraint matrix on the right side of the kinematic equation, and the constraint matrix on the right side of the kinematic inequality; the control point basis function matrix includes the initial control point matrix and the basis function matrix corresponding to the initial control point matrix.

[0138] Step 604: Construct a constraint matrix on the right side of the kinematic equation according to the path motion starting point, path motion end point and velocity vector restriction data.

[0139] It should be noted that the constraint matrix on the right side of the kinematic equation can be expressed as:

[0140]

[0141] Among them, c eq is the constraint matrix on the right side of the kinematic equation; x0 is the coordinate of the starting point of the motion in the X-axis direction; z0 is the coordinate of the starting point of the motion in the Z-axis direction; x e is the coordinate of the end point of the movement in the X-axis direction; z e is the coordinate of the end point of the movement in the Z-axis direction; ve The limit value of the velocity vector of the thorn crystal at the end point of the motion path on the X axis; ze The limit value of the velocity vector of the thorn crystal on the Z axis at the end point of the motion path.

[0142] Step 605: Construct the right-side constraint matrix of the kinematic inequality using the upper and lower limit data of the motion space.

[0143] It should be noted that the right-side constraint matrix of the kinematic inequality is constructed through the upper and lower limit data of the motion space. The right-side constraint matrix of the kinematic inequality can be expressed as: c ie =[x M -x m z M -z m ] T , where c ie is the constraint matrix on the right side of the kinematic inequality, x M is the upper limit coordinate of the motion space in the X-axis direction, x m is the lower limit coordinate of the motion space in the X-axis direction, z M is the upper limit coordinate of the motion space in the Z-axis direction, z m It is the lower limit coordinate of the motion space in the Z-axis direction.

[0144] Step 606: Determine the target control order using a preset solution algorithm based on the kinematic constraint matrix, the control point basis function matrix, and the coordinates of each initial control point.

[0145] Furthermore, step 606 may include the following sub-steps:

[0146] S61, using the initial control point matrix as the initial order control point matrix;

[0147] S62. Based on the order of the initial order control point matrix, a first optimization inequality constraint matrix is ​​constructed using the right-side constraint matrix of the kinematic inequality;

[0148] S63, constructing a first-order performance function according to the first optimized inequality constraint matrix, the right-side constraint matrix of the kinematic equation, the preset left-side constraint matrix of the kinematic inequality, the initial order control point matrix, the left-side constraint matrix of the kinematic equation corresponding to the initial order control point matrix, and the basis function matrix;

[0149] S64, using a preset solution algorithm to solve the first-order performance function to determine a first optimization control point matrix and a first performance index;

[0150] S65. Determine an increased-order matrix based on the order of the first optimized control point matrix;

[0151] S66. Determine an intermediate-order control point matrix based on the raised-order matrix and the first optimized control point matrix;

[0152] S67. Based on the order of the intermediate-order control point matrix, a second optimization inequality constraint matrix is ​​constructed using the right-hand side constraint matrix of the kinematic inequality;

[0153] S68. Construct a second-order performance function according to the second optimized inequality constraint matrix, the right-side constraint matrix of the kinematic equation, the left-side constraint matrix of the preset kinematic inequality, the intermediate-order control point matrix, the left-side constraint matrix of the kinematic equation corresponding to the intermediate-order control point matrix, and the basis function matrix;

[0154] S69, using a preset solution algorithm to solve the second-order performance function to determine a second optimization control point matrix and a second performance index;

[0155] S610, comparing the second performance indicator with the first performance indicator;

[0156] S611: If the first performance index is less than or equal to the second performance index, the order of the first optimized control point matrix is ​​used as the target control order.

[0157] Optionally, it also includes:

[0158] If the first performance index is greater than the second performance index, the second optimized control point matrix is ​​used as a new first optimized control point matrix, and the second performance index is used as a new first performance index;

[0159] Determining a second performance indicator using a preset solving algorithm according to the new first optimization control point matrix, the right-side constraint matrix of the kinematic equation, the left-side constraint matrix of the preset kinematic inequality, and the order of the new first optimization control point matrix until the first performance indicator is less than or equal to the second performance indicator;

[0160] The order of the first optimized control point matrix determined when the first performance index is less than or equal to the second performance index is used as the target control order.

[0161] It should be noted that if the first performance indicator is greater than the second performance indicator, the second optimization control point matrix is ​​used as the new first optimization control point matrix and the second performance indicator is used as the new first performance indicator; and the process jumps to S65 until the first performance indicator is less than or equal to the second performance indicator; the order of the first optimization control point matrix determined when the first performance indicator is less than or equal to the second performance indicator is used as the target control order.

[0162] Step 607: Determine the target Bezier motion curve of the flying thorn crystal according to the target control order, the kinematic constraint matrix and the control point basis function matrix.

[0163] Furthermore, step 607 may include the following sub-steps:

[0164] S71. Construct an initial Bezier curve using the initial control point matrix and the basis function matrix corresponding to the initial control point matrix;

[0165] S72. Determine a target raised-order matrix based on the target control order and the order of the initial control point matrix;

[0166] S73, constructing a raised-order control point matrix according to the target raised-order matrix and the first optimized control point matrix;

[0167] S74. Determine a multi-time point basis function matrix based on the column vectors in the raised-order control point matrix and the initial Bezier curve;

[0168] S75. Based on the order of the initial control point matrix, the right-hand side constraint matrix of the kinematic inequality is used to construct the initial optimization inequality constraint matrix;

[0169] S76. Construct a target order optimization performance function using a multi-time point basis function matrix, an initial control point matrix, a basis function matrix corresponding to the initial control point matrix, a left-side constraint matrix of a kinematic equation, a right-side constraint matrix of a kinematic equation, a left-side constraint matrix of a preset kinematic inequality, a target order matrix, and an initial optimization inequality constraint matrix;

[0170] S77, using a preset solution algorithm to solve the target order optimization performance function and determine the target control point matrix;

[0171] S78. Determine the target Bezier motion curve of the flying thorn crystal according to the basis function matrix corresponding to the target control point matrix and the initial control point matrix.

[0172] It should be noted that the target order optimization performance function is specifically:

[0173]

[0174] Among them, J(P n ) is the target order optimization performance function; P n is the initial control point matrix, which corresponds to the n-order control point; is a time derivative; B n 9t) is the basis function matrix corresponding to the initial control point matrix, and the order of the matrix is ​​n; T is the transpose of the matrix; A eq 9n) is the constraint matrix on the left side of the kinematic equation corresponding to the initial control point matrix, and the order of this matrix is ​​n; c eq A is the constraint matrix on the right side of the kinematic equation; ie is the constraint matrix on the left side of the preset kinematic inequality; is the target raised rank matrix; is the multi-time point basis function matrix; is the initial optimization inequality constraint matrix, the order of the matrix is ​​n; It is the preset upgrade fault tolerance distance.

[0175] For example, B1. Assume that the number of the coordinates of the multiple control points of the flying thorn crystal is 8, and set the motion information of this mass transfer, that is, preset the mass transfer motion data, including the motion starting point u0 = [0,1] of the path T , movement end point u e =[3,0] T , the lower limit u of the motion space m =[-1,-1] T and upper limit u M =[4,2] T , the motion process is required to maintain a 2nd order smooth start, and the velocity vector in the uniform pricking stage is [1,-1] T In particular, the initial control order n is set to 7, that is, the order of the initial control point matrix is ​​7.

[0176] Furthermore, a kinematic constraint matrix is ​​constructed based on the set motion information, including the constraint matrix A on the left side of the kinematic equation eq (k), the constraint matrix A on the left side of the kinematic inequality ie , the constraint matrix c on the right side of the kinematic equation eq 、The right-hand constraint matrix c of the kinematic inequality ie , the matrix expressions are as follows:

[0177]

[0178] Furthermore, B2. sets the initial Bezier curve to in, P7=[p0,p1,…,p7], and then construct the motion performance index function J(P7) as Set the optimization iteration count to k=7 and set the order tolerance distance

[0179] Furthermore, B3. Construct a k-order optimization inequality constraint matrix It is expressed as Based on the above basic construction, the control point matrix P k The k-order optimization problem of , and solve it to get the k-order optimal control point matrix And the k-order suboptimal performance index J k ; About the control point matrix P k The k-order optimization problem is specifically:

[0180]

[0181] Among them, J(P k ) is the control point matrix P k k-order optimization problem of P k is the k-order control point matrix; is the first time derivative; B k (t0 is the basis function matrix corresponding to the k-order control point matrix, the order of the matrix is ​​k-order; T is the matrix transpose; A eq (k) is the constraint matrix on the left side of the kinematic equation corresponding to the k-order control point matrix, and the order of this matrix is ​​k; c eq A is the constraint matrix on the right side of the kinematic equation; ie The constraint matrix on the left side of the preset kinematic inequality can be expressed as k-order optimization inequality constraint matrix.

[0182] Furthermore, B4. constructs a k+1 order raised control point matrix P k+1 , which can be expressed as: Construct the k+1 order optimization inequality constraint matrix, which can be expressed as; Construct the control point matrix P k+1 The k+1 order optimization problem is solved and the k+1 order optimal control point matrix is ​​obtained. And the k+1-order suboptimal performance index J k+1 , control point matrix P k+1 The k+1 order optimization problem is specifically:

[0183]

[0184] Among them, J(P k+1 ) is the control point matrix P k+1 k+1 order optimization problem; P k+1 is the k+1 order control point matrix; is the first time derivative; B k+1 (t) is the basis function matrix corresponding to the k+1 order control point matrix, and the order of this matrix is ​​k+1; T is the matrix transpose; A eq (k+1) is the constraint matrix on the left side of the kinematic equation corresponding to the k+1 order control point matrix, and the order of this matrix is ​​k+1; c eq A is the constraint matrix on the right side of the kinematic equation; ie The constraint matrix on the left side of the preset kinematic inequality can be expressed as Optimize the k+1-order inequality constraint matrix.

[0185] Furthermore, B5. If J k <J k+1 , then let k = k + 1, and re-execute steps B3 to B4, otherwise the optimal order is m = k. The optimal order is m = 14, and the initial control point matrix P is re-constructed. n The m-order optimization problem:

[0186]

[0187] in, is the 7th-order optimization inequality constraint matrix, specifically is the multi-time point Berstein matrix, specifically:

[0188]

[0189] in, is the vector 2-norm; is the multiple time points corresponding to the multi-time point basis function matrix; q j is the raised-order control point matrix Q 14 The column vector of the jth column in the ascending control point matrix Q 14 It can be expressed as:

[0190]

[0191] Furthermore, by solving the above initial control point matrix P n The m-order optimization problem is obtained, and the optimal control point matrix is ​​obtained. And the optimal Bezier motion curve Specifically:

[0192]

[0193] It is worth mentioning that see Figure 7 According to the calculation and optimization process of the above example, a low-order optimal Bezier curve that meets the stringent motion requirements is obtained, which makes the motion path simpler and easier to use, and allows the flying thorn crystal to perform flying thorn crystal mass transfer according to the obtained optimal Bezier motion curve. Compared with the PnP process (Pick-and-Place), the flying thorn crystal transfer process can achieve smooth movement of the needle at the start and uniform motion during thorn crystal, which greatly reduces the vibration of the working platform. Moreover, the source substrate and the target substrate only need to move synchronously at a uniform speed to achieve chip positioning throughout the thorn crystal process, which simplifies the positioning process, improves the transfer efficiency, and has high application potential. Among them, the calculation and optimization processes of the above examples are all run on the computing software MATLAB.

[0194] As a comparison of technical effects, we can refer to the existing technology. At present, traditional chip transfer technology mostly adopts precise pick-and-release (PnP) method, including source substrate positioning, chip picking, chip positioning, chip mounting and other important processes. With the increasing demand for chips in the display industry, the traditional PnP process can no longer meet the high transfer rate requirements of Mini / Micro LED chip transfer. In contrast, the "flying thorn crystal" transfer technology, which is improved based on the traditional needle-piercing transfer technology, transforms the static positioning thinking into dynamic positioning thinking. The LED chip, source substrate (blue film) and target substrate maintain a synchronized feed speed at the same vertical position. The transfer component moves above the chip. After the thorn needle head on the transfer component maintains the same lateral movement speed as the chip, it pushes the chip downward and peels it off to the target substrate, thereby realizing the dynamic transfer of the chip.

[0195] Furthermore, chip transfer involves the positioning of the source substrate, target substrate, and transfer components. Therefore, all chip transfer processes must consider transfer performance, working environment, and other conditions, and rationally design the motion path, speed, and acceleration of the moving components (i.e., path optimization). For the flying needle transfer process, the motion path of the needle tip must not only have smooth transitions and meet the requirements of smooth start / stop, but also ensure that the transfer process maintains the same moving speed as the source substrate to avoid scratching the chip.

[0196] Common path optimization algorithms include line-circle interpolation and spiral interpolation. These algorithms connect several given path points using basic shapes like straight lines, circles, and spirals to create a complete motion path. However, these curves often suffer from shortcomings such as path discontinuity and unstable switching. Bezier curves, also polynomial curves, are designed using control points. Because control points don't need to be strictly constrained on the motion path, Bezier curves offer greater scalability and path freedom, making Bezier curve-based path optimization a mainstream path optimization algorithm. However, the order (complexity) of a Bezier curve is directly related to the number of control points. For more stringent motion requirements, too few control points reduces the curve's design freedom; too many control points can make the resulting curve overly complex and difficult to apply. Therefore, while optimizing Bezier curves, it's also necessary to further optimize the control order (order-raising optimization) to obtain the optimal motion path with minimal curve complexity. Currently, no research has been conducted on Bezier curve order-raising optimization algorithms specifically for mass transfer technology.

[0197] For the above questions, please refer to Figure 8 The present invention proposes a flying thorn crystal path planning method, which calculates a stable, low-jitter Bezier motion path according to the motion requirements of chip mass transfer, thereby improving the chip transfer efficiency while ensuring the thorn crystal accuracy; specifically, A1. Set the motion information of this mass transfer and construct the kinematic constraint matrix based on the set motion information; A2. Set the initial Bezier curve f n (t), construct the initial control point matrix P n , construct the motion performance index function J(P n ), set the optimization iteration count to k = n, set the upgrade fault tolerance distance A3. Constructing the inequality constraint matrix for order optimization Construct the control point matrix P k Solve the k-order optimization problem and obtain the k-order optimal control point matrix And the k-order suboptimal performance index J k ; A4. Construct k+1 order ascending control point matrix P k+1 , construct k+1 order optimization inequality constraint matrix Construct the control point matrix P k+1 Solve the k+1 order optimization problem and obtain the k+1 order optimal control point matrix And the k+1-order suboptimal performance index J k+1 ; A5. If J k <J k+1 , then set k = k + 1 and re-execute steps A3 to A4; otherwise, the optimal order m = k is obtained. Reconstruct the initial control point matrix P nThe m-order optimization problem is solved to obtain the optimal control point matrix Obtaining the optimal Bezier motion curve based on the optimal control point matrix

[0198] In summary, the present invention aims at the motion requirements of the flying thorn crystal mass transfer technology, constructs and sets a kinematic constraint matrix that meets the requirements; for different control orders, constructs corresponding order Bezier curves and motion performance index functions, constructs Bezier curve optimization problems of different orders, solves and obtains the optimal control point matrix of the corresponding order, and obtains the optimal control order after multiple iterations; based on the order-raising calculation, the initial order control point matrix is ​​mapped, and the constraint conditions are reconstructed according to the order-raising fault tolerance distance to obtain the optimal order Bezier curve optimization problem, and the order-raising optimal control point matrix and the order-raising optimal Bezier curve are solved. At the same time, the existing path optimization methods based on line circle and spiral curve interpolation must pass through the path points to be optimized. The scalability and degree of freedom of the motion path are small, and the motion path is not necessarily smooth, and the path switching is unstable. Compared with the above-mentioned technologies, the motion path in the present invention does not necessarily need to pass through the control points to be optimized. Therefore, the motion path has the advantages of multi-order continuity of expression, highly smooth contour, high degree of freedom, and high scalability, which ensures the stability of the chip mass transfer process. Secondly, other existing Bezier path optimization methods can only optimize the Bezier curve based on a set fixed control order. When the motion requirements are too strict, the control order can only be set to a larger value, resulting in a more complex and difficult to apply high-order Bezier path. In contrast, the present invention temporarily maps low-order control points to high-order control points for optimization, and obtains a low-order optimal Bezier curve that meets stringent motion requirements, making the motion path more concise and easy to use.

[0199] In an embodiment of the present invention, the present invention provides a flying thorn crystal path planning method. When it is necessary to plan the path of the flying thorn crystal, the coordinates of multiple control points of the flying thorn crystal are first initialized to generate multiple initial control point coordinates; then, according to the coordinates of each initial control point and the preset massive transfer motion data, a kinematic constraint matrix and a control point basis function matrix are constructed; a preset solution algorithm is used to determine the target control order according to the kinematic constraint matrix, the control point basis function matrix and the coordinates of each initial control point; finally, the target Bezier motion curve of the flying thorn crystal is determined according to the target control order, the kinematic constraint matrix and the control point basis function matrix; based on the above scheme, a preset solution algorithm is used to determine the target control order according to the constructed kinematic constraint matrix, the control point basis function matrix and the initial control point coordinates obtained after initialization, and the process of determining the target Bezier motion curve of the flying thorn crystal in combination with the kinematic constraint matrix and the control point basis function matrix does not require setting a fixed control order in advance, and a low-order optimal Bezier curve can be obtained even under strict motion requirements, making the Bezier path more concise and easy to use.

[0200] See also Figure 9 , Figure 9 This is a structural block diagram of a flying thorn crystal path planning device provided by an embodiment of the present invention.

[0201] The present invention provides a flying thorn crystal path planning device, comprising:

[0202] The response module 901 is used to respond to the planning request, initialize the coordinates of multiple control points of the flying thorn crystal, and generate multiple initial control point coordinates;

[0203] A matrix construction module 902 is used to construct a kinematic constraint matrix and a control point basis function matrix according to the coordinates of each initial control point and the preset massive transfer motion data;

[0204] The control order determination module 903 is used to determine the target control order according to the kinematic constraint matrix, the control point basis function matrix and the coordinates of each initial control point using a preset solution algorithm;

[0205] The target path determination module 904 is used to determine the target Bezier motion curve of the flying thorn crystal according to the target control order, the kinematic constraint matrix and the control point basis function matrix.

[0206] Furthermore, the preset mass transfer motion data includes a path motion starting point, a path motion end point, velocity vector restriction data, and upper and lower limit data of the motion space; the kinematic constraint matrix includes a constraint matrix on the left side of the kinematic equation, a constraint matrix on the right side of the kinematic equation, and a constraint matrix on the right side of the kinematic inequality; the control point basis function matrix includes an initial control point matrix and a basis function matrix corresponding to the initial control point matrix; and the matrix construction module 902 is specifically used to:

[0207] Using the coordinates of each initial control point, construct the initial control point matrix;

[0208] Using the basis functions corresponding to the coordinates of each initial control point, the constraint matrix and basis function matrix on the left side of the kinematic equation corresponding to the initial control point matrix are constructed;

[0209] According to the path motion starting point, path motion end point and velocity vector restriction data, the constraint matrix on the right side of the kinematic equation is constructed;

[0210] The constraint matrix on the right side of the kinematic inequality is constructed using the upper and lower limit data of the motion space.

[0211] Furthermore, the control order determination module 903 is specifically configured to:

[0212] The initial control point matrix is ​​used as the initial order control point matrix;

[0213] Based on the order of the initial order control point matrix, the first optimization inequality constraint matrix is ​​constructed using the right-hand side constraint matrix of the kinematic inequality;

[0214] Constructing a first-order performance function according to the first optimization inequality constraint matrix, the right-side constraint matrix of the kinematic equation, the left-side constraint matrix of the preset kinematic inequality, the initial order control point matrix, the left-side constraint matrix of the kinematic equation corresponding to the initial order control point matrix, and the basis function matrix;

[0215] Solving the first-order performance function using a preset solving algorithm to determine a first optimization control point matrix and a first performance index;

[0216] Determining an increased-order matrix based on the order of the first optimized control point matrix;

[0217] Determine an intermediate order control point matrix according to the raised order matrix and the first optimized control point matrix;

[0218] Based on the order of the intermediate-order control point matrix, the second optimization inequality constraint matrix is ​​constructed using the right-hand side constraint matrix of the kinematic inequality;

[0219] Constructing a second-order performance function according to the second optimization inequality constraint matrix, the right-side constraint matrix of the kinematic equation, the left-side constraint matrix of the preset kinematic inequality, the intermediate-order control point matrix, the left-side constraint matrix of the kinematic equation corresponding to the intermediate-order control point matrix, and the basis function matrix;

[0220] Solving the second-order performance function using a preset solving algorithm to determine a second optimization control point matrix and a second performance index;

[0221] comparing the second performance indicator with the first performance indicator;

[0222] If the first performance index is less than or equal to the second performance index, the order of the first optimized control point matrix is ​​used as the target control order.

[0223] Optionally, it also includes:

[0224] A first module is configured to use the second optimized control point matrix as a new first optimized control point matrix and the second performance index as a new first performance index if the first performance index is greater than the second performance index;

[0225] The second module is configured to determine a second performance indicator using a preset solving algorithm according to the new first optimization control point matrix, the right-side constraint matrix of the kinematic equation, the left-side constraint matrix of the preset kinematic inequality, and the order of the new first optimization control point matrix, until the first performance indicator is less than or equal to the second performance indicator;

[0226] The third module is used to use the order of the first optimization control point matrix determined when the first performance index is less than or equal to the second performance index as the target control order.

[0227] Furthermore, the target path determination module 904 is specifically configured to:

[0228] The initial control point matrix and the basis function matrix corresponding to the initial control point matrix are used to construct the initial Bezier curve;

[0229] Determine the target boosted order matrix based on the target control order and the order of the initial control point matrix;

[0230] Constructing a raised-order control point matrix according to the target raised-order matrix and the first optimized control point matrix;

[0231] Determine the multi-time point basis function matrix based on the column vectors in the raised-order control point matrix and the initial Bezier curve;

[0232] Based on the order of the initial control point matrix, the right-hand side constraint matrix of the kinematic inequality is used to construct the initial optimization inequality constraint matrix;

[0233] The target order optimization performance function is constructed using the multi-time point basis function matrix, the initial control point matrix, the basis function matrix corresponding to the initial control point matrix, the left-side constraint matrix of the kinematic equation, the right-side constraint matrix of the kinematic equation, the left-side constraint matrix of the preset kinematic inequality, the target order matrix, and the initial optimization inequality constraint matrix.

[0234] Use the preset solution algorithm to solve the target order optimization performance function and determine the target control point matrix;

[0235] The target Bezier motion curve of the flying thorn crystal is determined according to the basis function matrix corresponding to the target control point matrix and the initial control point matrix.

[0236] Optionally, the target order optimizes the performance function, specifically:

[0237]

[0238] Among them, J(P n ) is the target order optimization performance function; P n is the initial control point matrix, which corresponds to the n-order control point; is a time derivative; B n (t) is the basis function matrix corresponding to the initial control point matrix, and the order of the matrix is ​​n; T is the transpose of the matrix; A eq (n) is the constraint matrix on the left side of the kinematic equation corresponding to the initial control point matrix, and the order of this matrix is ​​n; c eq A is the constraint matrix on the right side of the kinematic equation; ie is the constraint matrix on the left side of the preset kinematic inequality; is the target raised rank matrix; is the multi-time point basis function matrix; is the initial optimization inequality constraint matrix, the order of the matrix is ​​n; It is the preset upgrade fault tolerance distance.

[0239] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described devices and modules can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0240] An embodiment of the present invention further provides a computer device comprising a memory and a processor, wherein a computer program is stored in the memory; when the computer program is executed by the processor, the processor executes the steps of the flying spike path planning method as described in any of the above embodiments.

[0241] An embodiment of the present invention further provides a computer-readable storage medium having a computer program / instruction stored thereon. When the computer program / instruction is executed by a processor, the steps of the flying spike path planning method according to any of the above embodiments are implemented.

[0242] An embodiment of the present invention further provides a computer program product, including a computer program / instruction, which, when executed by a processor, implements the steps of the flying spike path planning method as described in any of the above embodiments.

[0243] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of units is only a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0244] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0245] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that the technical solutions described in the above embodiments can still be modified, or some of the technical features thereof can be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A flying thorn crystal path planning method, characterized in that: include: In response to the planning request, the coordinates of multiple control points of the flying thorn crystal are initialized to generate multiple initial control point coordinates; Constructing a kinematic constraint matrix and a control point basis function matrix according to the coordinates of each of the initial control points and the preset massive transfer motion data; Determining a target control order using a preset solution algorithm based on the kinematic constraint matrix, the control point basis function matrix, and the coordinates of each initial control point; Determining a target Bezier motion curve of the flying thorn crystal according to the target control order, the kinematic constraint matrix and the control point basis function matrix; The preset mass transfer motion data includes the path motion starting point, the path motion end point, the velocity vector restriction data, and the upper and lower limit data of the motion space; the kinematic constraint matrix includes the left-side constraint matrix of the kinematic equation, the right-side constraint matrix of the kinematic equation, and the right-side constraint matrix of the kinematic inequality; the control point basis function matrix includes the initial control point matrix and the basis function matrix corresponding to the initial control point matrix; The step of constructing a kinematic constraint matrix based on the coordinates of each of the initial control points and the preset massive transfer motion data includes: Using the coordinates of each of the initial control points to construct an initial control point matrix; Using the basis functions corresponding to the coordinates of each of the initial control points, a constraint matrix and a basis function matrix on the left side of the kinematic equation corresponding to the initial control point matrix are constructed; According to the path motion starting point, path motion end point and velocity vector restriction data, the constraint matrix on the right side of the kinematic equation is constructed; The constraint matrix on the right side of the kinematic inequality is constructed using the upper and lower limit data of the motion space.

2. The flying thorn crystal path planning method according to claim 1, characterized in that: The step of using a preset solution algorithm to determine the target control order according to the kinematic constraint matrix, the control point basis function matrix and the coordinates of each initial control point includes: Using the initial control point matrix as the initial order control point matrix; Based on the order of the initial order control point matrix, a first optimized inequality constraint matrix is ​​constructed using the right-side constraint matrix of the kinematic inequality; Constructing a first-order performance function according to the first optimized inequality constraint matrix, the right-side constraint matrix of the kinematic equation, the preset left-side constraint matrix of the kinematic inequality, the initial order control point matrix, and the left-side constraint matrix of the kinematic equation corresponding to the initial order control point matrix and a basis function matrix; Solving the first-order performance function using the preset solving algorithm to determine a first optimization control point matrix and a first performance index; Determining an increased-order matrix based on the order of the first optimized control point matrix; Determining an intermediate order control point matrix according to the raised order matrix and the first optimized control point matrix; Based on the order of the intermediate-order control point matrix, a second optimized inequality constraint matrix is ​​constructed using the right-side constraint matrix of the kinematic inequality; Constructing a second-order performance function according to the second optimized inequality constraint matrix, the right-side constraint matrix of the kinematic equation, the left-side constraint matrix of the preset kinematic inequality, the intermediate-order control point matrix, and the left-side constraint matrix of the kinematic equation corresponding to the intermediate-order control point matrix and the basis function matrix; Solving the second-order performance function using the preset solving algorithm to determine a second optimization control point matrix and a second performance index; comparing the second performance indicator with the first performance indicator; If the first performance index is less than or equal to the second performance index, the order of the first optimization control point matrix is ​​used as the target control order.

3. The flying thorn crystal path planning method according to claim 2, characterized in that: Also includes: If the first performance index is greater than the second performance index, taking the second optimization control point matrix as a new first optimization control point matrix and taking the second performance index as a new first performance index; Determining a second performance indicator using the preset solving algorithm according to the new first optimization control point matrix, the right-side constraint matrix of the kinematic equation, the left-side constraint matrix of the preset kinematic inequality, and the order of the new first optimization control point matrix until the first performance indicator is less than or equal to the second performance indicator; The order of the first optimized control point matrix determined when the first performance index is less than or equal to the second performance index is used as the target control order.

4. The flying thorn crystal path planning method according to claim 1, characterized in that: The step of determining the target Bezier motion curve of the flying thorn crystal according to the target control order, the kinematic constraint matrix and the control point basis function matrix includes: Constructing an initial Bezier curve using the initial control point matrix and a basis function matrix corresponding to the initial control point matrix; Determining a target raised-order matrix based on the target control order and the order of the initial control point matrix; Constructing a raised-order control point matrix according to the target raised-order matrix and the first optimized control point matrix; Determining a multi-time point basis function matrix based on the column vectors in the raised-order control point matrix and the initial Bezier curve; Based on the order of the initial control point matrix, an initial optimization inequality constraint matrix is ​​constructed using the right-side constraint matrix of the kinematic inequality; Constructing a target order optimization performance function using the multi-time point basis function matrix, the initial control point matrix, the basis function matrix corresponding to the initial control point matrix, the left-side constraint matrix of the kinematic equation, the right-side constraint matrix of the kinematic equation, the left-side constraint matrix of the preset kinematic inequality, the target raised-order matrix, and the initial optimized inequality constraint matrix; Solving the target order optimization performance function using the preset solving algorithm to determine the target control point matrix; The target Bezier motion curve of the flying thorn crystal is determined according to the basis function matrix corresponding to the target control point matrix and the initial control point matrix.

5. The flying thorn crystal path planning method according to claim 4, characterized in that: The target order optimization performance function is specifically: ； in, Optimize the performance function for the target order; is the initial control point matrix, the order of the matrix is ​​n, that is, it corresponds to n-order control points; Take the time derivative for once; is the basis function matrix corresponding to the initial control point matrix, the order of the matrix is ​​n; T is the transpose of the matrix; is the constraint matrix on the left side of the kinematic equation corresponding to the initial control point matrix, and the order of this matrix is ​​n; is the constraint matrix on the right side of the kinematic equation; is the constraint matrix on the left side of the preset kinematic inequality; is the target raised rank matrix; is the multi-time point basis function matrix; is the initial optimization inequality constraint matrix, the order of the matrix is ​​n; It is the preset upgrade fault tolerance distance.

6. A flying thorn crystal path planning device, applied to the flying thorn crystal path planning method according to claim 1, characterized in that: include: The response module is used to respond to the planning request, initialize the coordinates of multiple control points of the flying thorn crystal, and generate multiple initial control point coordinates; A matrix construction module is used to construct a kinematic constraint matrix and a control point basis function matrix according to the coordinates of each of the initial control points and the preset massive transfer motion data; A control order determination module is used to determine a target control order according to the kinematic constraint matrix, the control point basis function matrix and the coordinates of each initial control point using a preset solution algorithm; The target path determination module is used to determine the target Bezier motion curve of the flying thorn crystal according to the target control order, the kinematic constraint matrix and the control point basis function matrix.

7. A computer device, characterized in that: It includes a memory and a processor, wherein a computer program is stored in the memory, and when the computer program is executed by the processor, the processor executes the steps of the flying spike crystal path planning method according to any one of claims 1 to 5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed, the flying spike path planning method according to any one of claims 1 to 5 is implemented.

9. A computer program product, characterized in that The computer program product includes a computer program stored on a non-transitory computer-readable storage medium, and the computer program includes program instructions, wherein when the program instructions are executed by a computer, the computer is caused to execute the flying spike path planning method according to any one of claims 1 to 5.

Images