Component mounting positioning method for high-precision SMT chip mounter

Abstract

The invention discloses a component mounting positioning method for a high-precision SMT chip mounter, and the method comprises the steps: carrying out the visual recognition of four Mark points, obtaining four groups of data relationships at different angles through the mutual combination of three adjacent points, carrying out the discrimination through an algorithm, reducing the sampling error, obtaining one of the most precise data as a reference point, and carrying out the positioning of the data. And judging errors in the previous data calculation process by using a PID (Proportion Integration Differentiation) adjustment mode to obtain an optimal calculation formula, substituting the position coordinates of all the components according to the optimal formula, and calculating the accurate positions of the projections of all the components in the gantry coordinate system of the SMT chip mounter. The visual sampling error is reduced, the positioning operation magnitude is reduced, the mechanical installation precision requirement is lowered, and the mechanical installation precision is improved.

Description

Technical Field

[0001] The invention relates to the technical field of control systems of high-speed SMT placement machines, and in particular to a component placement and positioning method for high-precision SMT placement machines. Background Art

[0002] With the development of technology, the size of PCB is getting smaller and smaller, and the size of components installed on PCB is also smaller, which puts higher and higher demands on SMT placement machines. At present, the placement accuracy of all SMT placement machines on the market is within ±0.035mm. The most important factor affecting placement accuracy is the accurate positioning of components, which requires that the center of each component is just near the center of the component package. The more accurate the positioning, the lower the requirements for visual accuracy and mechanical accuracy, so that the cost can be reduced accordingly.

[0003] Conventional PCB boards are set with 4 Mark points for visual positioning and recognition of placement machines. The PCB positioning algorithms on the market are generally classified into two-point positioning, three-point positioning and four-point positioning. Different methods require different numbers of Mark points.

[0004] Two-point positioning takes less time and is relatively convenient to operate. However, the fewer the points, the larger the measurement error introduced, which may eventually cause the positioning accuracy of components close to the sampling points to become increasingly poor as they move away from the sampling points.

[0005] There is little difference between the three-point positioning and four-point positioning algorithms. The three-point positioning algorithm is relatively simple, and the measurement accuracy of the four-point positioning can be compensated by the algorithm and corrected to reduce the error of visual sampling. Moreover, most positioning algorithms are based on the rectangular coordinate system, which leads to relatively high requirements for the mechanical installation accuracy of each SMT placement machine. It is also difficult to ensure the consistency of the mechanical installation accuracy of each SMT placement machine, and the installation difficulty and debugging of the SMT placement machine are very large. Summary of the invention

[0006] In view of the shortcomings of the prior art, the purpose of the present invention is to provide a component placement and positioning method for a high-precision SMT placement machine, so as to solve the problems of large positioning errors of components in the existing placement machines, component positioning offset due to mechanical wear after long-term use, high requirements for mechanical installation accuracy and poor positioning consistency of each machine.

[0007] In order to achieve the above object, the technical solution adopted by the present invention is:

[0008] The advantages of the present invention compared with the prior art are:

[0009] The present invention adopts an algorithm based on the Cartesian oblique coordinate system, identifies four Mark points, combines three adjacent points into four situations, obtains the most accurate set of data, and then obtains a calculation formula with the minimum error through the PID adjustment method to meet the requirements of SMT placement machine components for high-precision positioning. The present invention solves the technical problems of large positioning error of components by existing SMT placement machines, component positioning offset caused by mechanical wear after long-term use, and poor positioning consistency of different SMT placement machines through a new algorithm. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] Figure 1 The relationship between the PCB board of the present invention and the SMT placement machine gantry coordinate system is

[0011] Figure 2 This is a schematic diagram of three sampling points when the XY axis of the present invention is an acute angle.

[0012] Figure 3 This is the Mark point sampling flow chart of the present invention

[0013] Figure 4 The best sampling data calculation flow chart of the present invention is

[0014] Figure 5 This is the PID calculation flow chart of the optimal XY axis angle of the present invention

[0015] Figure 6 This is the overall flow chart of the component positioning method of the present invention DETAILED DESCRIPTION

[0016] A component placement positioning method for a high-precision SMT placement machine. Figures 1 to 6 As shown, the following steps are included:

[0017] Step 1: Obtain the absolute coordinates of the four Mark points of the PCB board in the SMT placement machine gantry coordinate system and the relative coordinates of the PCB board according to the visual recognition circle center. First select the absolute coordinates of three adjacent Mark points (a, b, c) (a x ,a y )(b x ,b y )(c x ,c y ) and relative coordinates (a' x ,a' y )(b' x ,b' y )(c' x ,c' y ) and compare them to get the deflection angle of the XY axis of the relative coordinate system of the PCB board relative to the gantry coordinate system of the SMT placement machine, including the X-axis angle and the Y-axis angle.

[0018] The calculation formulas for the X-axis angle and the Y-axis angle are

[0019] X-axis angle = arctan(c y -b y / c x -b x )-arctan(c' y -b' y / c' x -b' x )

[0020]

[0021] Step 2: According to the obtained angle values ​​of the X-axis angle and the Y-axis angle, one of the sides with the smaller angle value is selected as the parallel side, the angle relationship between the board entry track and the XY axis and whether the XY axis is in a vertical relationship are obtained, and the current angle values ​​of the X-axis angle and the Y-axis angle are recorded;

[0022] Step 3: If the current XY axis recorded in step 2 is a vertical relationship, then the mutual distances abs_len_ab, abs_len_bc and abs_len_ac of the three Mark points (a, b, c) in the SMT placement machine gantry coordinate system (absolute coordinate system) are calculated according to the distance formula, as well as the mutual distances rel_len_ab, rel_len_bc and rel_len_ac of the corresponding three Mark points (a, b, c) in the relative coordinate system.

[0023] The distance formula is

[0024]

[0025] Step 4: If the current XY axis recorded in step 2 is not in a vertical relationship, connect the three Mark points (a, b, c) in step 1 to get a triangle. According to the angle formula of the three sides, the three angles ∠abc, ∠bac and ∠acb of the triangle in the SMT placement machine gantry coordinate system and the three angles ∠ABC, ∠BAC and ∠ACB in the relative coordinate system are obtained.

[0026] The formula for finding the angle from the three sides is:

[0027]

[0028] Step 5: Divide the lengths of the corresponding sides and the corresponding angles of the absolute coordinates and the relative coordinates one by one to obtain the proportional coefficients of the three side lengths and angles. By comparing the proportional coefficients, the XY axis angle θ of the absolute coordinate axis can be obtained.

[0029] Step 6: According to the distance formula of the oblique coordinate system, the distance between the three mark points (a, b, c) in the SMT placement machine gantry coordinate system is recalculated, and then compared with the values ​​obtained in step 1, the corrected angle is β.

[0030] The distance formula of the oblique coordinate system is

[0031]

[0032] Step 7: Identify the fourth MARK point d, and obtain the actual distances from point d to the three Mark points (a, b, c) based on the known coordinates. Use the scale factor obtained in step 5 as the scaling factor to project the three Mark points (a, b, c) onto the XY axis plane of the SMT placement machine gantry coordinate system to obtain three projection points. Calculate the distances len_da, len_db, and len_dc from point d to the three projection points.

[0033] Step 8: Substitute the distances len_da, len_db, and len_dc obtained in step 7 into the following length formula to calculate the projection coordinates of point d on the XY axis plane in the SMT placement machine gantry coordinate system.

[0034] The length formula is

[0035] len_da 2 =(x d -x a ) 2 +(y d -y a ) 2 -2*(x d -x a )*(y d -y a )*cosβ

[0036] len_db 2 =(x d -x b ) 2 +(y d -y b ) 2 -2*(x d -x b )*(y d -y b )*cosβ

[0037] len_dc 2 =(x d -x c ) 2 +(y d -yc ) 2 -2*(x d -x c )*(y d -y c )*cosβ;

[0038] Step 9: Based on the three-term relationship in step 8, compare the coordinates of point d with the coordinates sampled by vision to obtain the error coefficient in the first sampling case. The calculation formula of the error coefficient is:

[0039] Error coefficient = (sample X / calculated X-sample Y / calculated Y);

[0040] Step 10: Repeat the above steps to obtain error coefficients under 4 different sampling conditions, sort the 4 error coefficients, and select the sampling data under the sampling condition with the smallest error;

[0041] Step 11: Determine whether the optimal error coefficient is 0. If it is not 0, perform the angle recalibration algorithm and calculate the error angle β' according to the angle formula. The angle formula is:

[0042] β' = arccos (error coefficient).

[0043] Step 12: Establish a PID relationship to determine the impact of the error value of angle β on the entire calculation result.

[0044] deltaPara = sample X - calculation X

[0045] Integral=Σ(currentβ′-preβ′)

[0046] derivative=(currentβ'-preβ') / deltaPara

[0047] Angle correction = P*β'+I*Integral+D*derivative+β

[0048] Step 13: Substitute the angle correction result into step 8 to obtain the calculated coordinates of point d, thereby obtaining the error coefficient of step 9, and determine the optimal XY axis angle by comparing the change in the error coefficient;

[0049] Step 14: After PID calculation, the corresponding angle in the minimum error coefficient is obtained, and the angle is substituted into step 8. Then, all the corresponding component positions in the BOM table of the PCB are calculated one by one and stored in a new table waiting to be called.

[0050] The present invention visually identifies 4 Mark points, first combines 3 adjacent points to obtain 4 groups of data relationships at different angles, then uses algorithm identification to reduce sampling errors, and obtains one of the most accurate data as a reference point, and then uses PID adjustment to determine the error in the previous data calculation process to obtain the best calculation formula, and then substitutes all component position coordinates according to the best formula to obtain the accurate position of all components under the coordinate projection of the SMT placement machine. The present invention reduces the error of visual sampling, reduces the magnitude of positioning operations, and reduces and improves the mechanical installation accuracy requirements.

[0051] The present invention reduces the mechanical installation accuracy of the SMT placement machine, reduces the error of visual sampling and reduces the magnitude of positioning calculations, while improving the positioning accuracy of components by conventional SMT placement machines, and has the following technical advantages.

[0052] 1. After multiple sampling and comparison, the vision system determines whether the size of the Mark points recognized each time is consistent, reducing the error introduced by inaccurate sampling and reducing the difficulty of operation for staff.

[0053] 2. The mixed calculation methods of rectangular coordinate system and oblique coordinate system are adopted. Whether the XY axis of the SMT placement machine is right angle or not, the correct data can be obtained, which reduces the difficulty of installation for mechanical assemblers and reduces the cost of recalibrating the axis after long-term use of the equipment.

[0054] 3. By using data to calibrate each other, several sets of coordinates are added for algorithm correction relative to the two-point and three-point positioning methods, which increases the accuracy of the data and improves the fault tolerance of the algorithm.

[0055] 4. Repeat the verification of the algorithm results, use PID to further correct the angle, confirm that the obtained data is the best result calculated in the current sampling coordinates, and further verify the stability and accuracy of the algorithm.

Claims

1. A component placement positioning method for a high-precision SMT placement machine, characterized in that: The following steps are included: Step 1: Obtain the absolute coordinates of the four Mark points of the PCB board in the SMT placement machine gantry coordinate system and the relative coordinates of the PCB board according to the visual recognition circle center. First select the absolute coordinates of three adjacent Mark points (a, b, c) (a x ,a y )(b x ,b y )(c x ,c y ) and relative coordinates (a' x ,a' y )(b' x ,b' y )(c' x ,c' y ) and compare them to get the deflection angle of the XY axis of the relative coordinate system of the PCB board relative to the gantry coordinate system of the SMT placement machine, including the X-axis angle and the Y-axis angle. The calculation formulas for the X-axis angle and the Y-axis angle are X-axis angle = arctan(c y -b y / c x -b x )-arctan(c' y -b' y / c' x -b' x ) Step 2: According to the obtained angle values ​​of the X-axis angle and the Y-axis angle, one of the sides with the smaller angle value is selected as the parallel side, the angle relationship between the board entry track and the XY axis and whether the XY axis is in a vertical relationship are obtained, and the current angle values ​​of the X-axis angle and the Y-axis angle are recorded; Step 3: If the current XY axis recorded in step 2 is a vertical relationship, then the mutual distances abs_len_ab, abs_len_bc and abs_len_ac of the three Mark points (a, b, c) in the SMT placement machine gantry coordinate system are calculated according to the distance formula, as well as the mutual distances rel_len_ab, rel_len_bc and rel_len_ac of the corresponding three Mark points (a, b, c) in the PCB relative coordinate system. The distance formula is Step 4: If the current XY axis recorded in step 2 is not in a vertical relationship, connect the three Mark points (a, b, c) in step 1 to get a triangle. According to the angle formula of the three sides, the three angles ∠abc, ∠bac and ∠acb of the triangle in the SMT placement machine gantry coordinate system and the three angles ∠ABC, ∠BAC and ∠ACB in the relative coordinate system are obtained. The formula for finding the angle from the three sides is: Step 5: Divide the lengths of the corresponding sides and the corresponding angles of the absolute coordinates and the relative coordinates one by one to obtain the proportional coefficients of the three side lengths and angles, and obtain the XY axis angle θ of the absolute coordinate axis by comparing the proportional coefficients; Step 6: According to the distance formula of the oblique coordinate system, the distance between the three mark points (a, b, c) in the SMT placement machine gantry coordinate system is recalculated, and then compared with the values ​​obtained in step 1, the corrected angle is β. The distance formula of the oblique coordinate system is Step 7: Identify the fourth MARK point d, and obtain the actual distances from point d to the three Mark points (a, b, c) based on the known coordinates. Use the scale factor obtained in step 5 as the scaling factor to project the three Mark points (a, b, c) onto the XY axis plane of the SMT placement machine gantry coordinate system to obtain three projection points. Calculate the distances len_da, len_db, and len_dc from point d to the three projection points. Step 8: Substitute the distances len_da, len_db, and len_dc obtained in step 7 into the following length formula to calculate the projection coordinates of point d on the XY axis plane in the SMT placement machine gantry coordinate system. The length formula is len_da 2 (x) d -x a ) 2 +(and d -and a ) 2 -2*(x d -x a )*(and d -and a )*cosβ len_db 2 =(x d -x b ) 2 +(y d -y b ) 2 -2*(x d -x b )*(y d -y b )*cosβ len_dc 2 (x) d -x c ) 2 +(and d -and c ) 2 -2*(x d -x c )*(and d -and c )*cosβ; Step 9: Based on the three-term relationship in step 8, compare the coordinates of point d with the coordinates sampled by vision to obtain the error coefficient in the first sampling case. The calculation formula of the error coefficient is: Error coefficient = (sample X / calculated X-sample Y / calculated Y); Step 10: Repeat the above steps to obtain error coefficients under 4 different sampling conditions, sort the 4 error coefficients, and select the sampling data under the sampling condition with the smallest error; Step 11: Determine whether the optimal error coefficient is 0. If it is not 0, perform the angle recalibration algorithm and calculate the error angle β' according to the angle formula. The angle formula is β'=arccos (error coefficient) Step 12: Establish a PID relationship to determine the impact of the error value of angle β on the entire calculation result. deltaPara = sample X - calculation X Integral=∑(currentβ′-preβ′) derivative=(currentβ'-preβ') / deltaPara Angle correction = P*β'+I*Integral+D*derivative+β Step 13: Substitute the angle correction result into step 8 to obtain the calculated coordinates of point d, thereby obtaining the error coefficient of step 9, and determine the optimal XY axis angle by comparing the change in the error coefficient; Step 14: After PID calculation, the corresponding angle in the minimum error coefficient is obtained, and the angle is substituted into step 8. Then, all the corresponding component positions in the PCB BOM table are calculated one by one and stored in a new table waiting to be called.

Images