DE102009039988A1 - Travels for painting robots - Google Patents

Abstract

Method and system for determining travel paths of painting robots, in particular for electronics production. By means of mixed integer linear programming (Mixed Integer Linear Programming, MILP), a surface to be painted is broken down into horizontally and vertically painted areas and strips in a hierarchical approach, which are advantageously moved by a painting robot.

Description

The invention relates to a method for determining travel distances of painting robots, in particular for electronics production.

In electronics manufacturing assemblies (for example consisting of a printed circuit board and components) can be sprayed with a protective lacquer to protect the PCB from physical and chemical influences. By this coating, an increase in robustness against external influences (moisture, pollution, harmful gases, etc.) is achieved. The painting usually takes place after the assembly and testing of an assembly. A coating of an assembly can be done on one side or on both sides. Subassemblies can be painted partially or completely depending on the requirements.

This painting process is usually carried out by a painting machine, which moves with a robot arm to which a nozzle is attached, over the circuit board and painted the desired faces of the plate in strips. For this, the travel paths of the robotic arm and the placement of the paint strips must be determined. With a manual input in the robot programming, the error probabilities and the time required are very high, furthermore, the quality of the painting process is not always ensured. During a manual input, the paint strips and the travel paths of the painting robot are entered by an employee using an image of the printed circuit board. For this purpose, the coordinates are first measured on the image, then the employee enters the coordinates in a software that later controls the painting machine. The created painting program is tested on the machine. If the program was faulty (collisions with components, etc.), it will be manually corrected and retested afterwards.

From the German patent application DE 10 2007 041 424 A1 are a method and a system for the optimized planning of complex production sequences in large-scale plant operations, especially for the steel industry, based on methods of Mixed Integer Linear Programming (MILP) known.

The object of the present invention is to provide a method and a system for optimally determining travel paths for painting robots, in particular for electronics production.

• a1) Bestimmung von Parametern einer zu verfahrenden und zu lackierenden Leiterplatte;
• a2) Bestimmung von Parametern für den Lackierroboter;
• a3) Festlegen von Sicherheits- und Nebenbedingungen für das Verfahren und Lackieren der Leiterplatte;
• b) Modellieren der zu lackierenden Flächen der Leiterplatte als eine Vereinigung von Rechtecken, wobei die Kanten der Rechtecke parallel zu den Achsen eines kartesischen Koordinatensystems der Leiterplatte angeordnet sind;
• c) Generieren von Lackierrichtungen für das Verfahren der zu lackierenden Rechtecke durch Gemischt Ganzzahlige Lineare Programmierung (Mixed Integer Linear Programming, MILP),

wobei die zu lackierende Fläche in horizontal und in vertikal zu lackierende Bereiche zerlegt wird;

• d) Generieren von Lackstreifen für jeden zusammenhängenden zu lackierenden Bereich je Lackierrichtung durch Gemischt Ganzzahlige Lineare Programmierung (Mixed Integer Linear Programming). Wenn eine Lackiermaschine oder ein Lackierroboter keine Optimierungssoftware hat, müssen die Fahrtwege des Roboterarms und die Bestimmung der Lackstreifen manuell bestimmt werden. Bei einer Manuellen Eingabe sind Fehlerwahrscheinlichkeiten und der Zeitaufwand für die Eingabe sehr hoch. Ein automatisch erstelltes, optimiertes Lackierprogramm ermöglicht es, die Fehlermöglichkeiten (z. B. Verkleben von Löchern durch den Lack, Lackspritzer auf frei zubleibenden Flächen, Hängenbleiben des Roboterarms an Bauelementen)) zu eliminieren, den Planungsaufwand drastisch zu reduzieren und den gesamten Lackierprozess effizienter und schneller zu machen. Mit der Gemischt Ganzzahligen Linearen Programmierung (Mixed Integer Linear Programming, MILP) lassen sich Softwareprogramme zur optimalen Bestimmung von Lackstreifen und zur Berechnung von optimalen Fahrtwegen des Roboterarms beschreiben.

The object is achieved by a method for determining travel distances of painting robots, in particular for electronics production, comprising the following steps:

• a1) determination of parameters of a circuit board to be traversed and painted;
• a2) determination of parameters for the painting robot;
• a3) defining safety and ancillary conditions for the process and painting of the printed circuit board;
• b) modeling the surfaces of the printed circuit board to be painted as a union of rectangles, the edges of the rectangles being arranged parallel to the axes of a Cartesian coordinate system of the printed circuit board;
• c) Generating painting directions for the method of the rectangles to be painted by Mixed Integer Linear Programming (MILP),

wherein the surface to be painted is divided into horizontally and vertically to be painted areas;

• d) Generating paint strips for each contiguous area to be painted per coating direction by Mixed Integer Linear Programming. If a painting machine or a painting robot has no optimization software, the travel distances of the robot arm and the determination of the paint strips must be determined manually. With a manual input, error probabilities and the time required for the input are very high. An automatically created, optimized coating program makes it possible to eliminate the possibility of errors (eg sticking of holes through the paint, paint spatter on free surfaces, sticking of the robot arm to components), drastically reducing the planning effort and making the entire painting process more efficient and efficient to make faster. Mixed Integer Linear Programming (MILP) software can be used to describe software programs for the optimal determination of paint strips and the calculation of optimum travel distances of the robot arm.

A first advantageous embodiment of the invention is that the step c further comprises:
Creating a grid of the circuit board so that each cell of the circuit board is either completely in a surface to be painted, or outside; and
Minimize the sums of lattice edge lengths according to one of the following conditions:
the two cells on either side of the grid edge belong to the area to be painted, but the painting directions of both cells are different, or
only one of the two cells on both sides of the grating edge belongs to the surface to be painted, and the painting direction and the edge direction of these cells are different. The method is based on a hierarchical approach and determines step-by-step optimal solutions for the respective problem descriptions. Thus, in a top-down approach mixed integer linear programs and heuristics for problem solving can be created.

A further advantageous embodiment of the invention is that the step d further comprises:
Producing a list of all horizontally and all vertical candidates for a paint strip selection, the paint strips being aligned either at the corner points of the corresponding area to be painted or on other already generated paint strips, and permissible with respect to the constraints; and
Selecting Strips to Be Painted by Mixed Integer Linear Programming, where at least a predetermined percentage of the printed circuit board area can be covered by the selected paint strips, minimizing overlapping stripes and reducing the total run time required to travel through the selected paint strips , is minimized. First a very large list of possible stripes is created and then with the MILP a part of it is selected as a solution. All surfaces of the problem instance (PCB to be painted) are modeled as a union of rectangles with the edges of the rectangles parallel to the axes of the coordinate system. If a surface does not meet these requirements (eg, a cuboid component rotated 45 °), it is converted at a given resolution into an area whose edges are horizontal or vertical. This simplifies the program development for problem solving.

A further advantageous embodiment of the invention is that step d) further comprises:
Shortening a strip if it does not reduce the covered area, or
Lengthening a strip, as a result of which other overlapping strips are shortened and the overall cycle time becomes smaller. This increases the efficiency of the program created with respect to the travel paths of the painting robot.

A further advantageous embodiment of the invention is that wherein step d) further comprises:
Moving a strip if it increases the area covered, or keeps the covered area the same, but reduces the overlaps. This also increases the efficiency of the program created with respect to the travel paths of the painting robot.

A further advantageous embodiment of the invention is that step d) further comprises:
Adding a strip, if it can shorten or eliminate other strips, and the throughput time is smaller. This also increases the efficiency of the program created with respect to the travel paths of the painting robot.

A further advantageous embodiment of the invention lies in determining the shortest paths between all edge points of different paint strips, taking into account the acceleration and braking distance of the painting robot. So z. For example, between the stripes (a → b) and stripes (c → d) four shortest paths are calculated (a → c, a → d, b → c, b → d). This is done for all possible pairs of stripes. Geometric conditions of the printed circuit board and the painting robot are taken into account, thereby calculating a time-optimized travel path for the robot arm between two points. Obstacles are taken into account.

A further advantageous embodiment of the invention is in determining the travel paths of the painting robot, wherein the strips are placed in an order that avoid spraying and wherein the sum of the travel times between the strips is minimized. This minimizes the travel distances of the robot and increases the quality of the painting process.

The object is further achieved by a computer-readable medium (CD, floppy disk, memory card, USB stick, etc.) or computer program product comprising instructions which, when executed on a suitable computer, cause the computer to carry out the method according to the invention. This facilitates the flexibility of the use as well as the distribution and commercial distribution of the method according to the invention.

The object is further achieved by a system for carrying out the method according to the invention. The system can be realized by commercially available painting robots or painting machines with conventional controls. The program creation can take place offline textually (in a programming language: eg C, C ++) and / or CAD-supported.

An embodiment of the invention is illustrated in the drawing and will be explained below.

Showing:

1 a schematic representation of a painting process on a printed circuit board,

2 a schematic representation of a circuit board, with horizontal and vertical areas to be painted, and

3 a section of the circuit board of 2 with paint strips to be painted horizontally and vertically, as well as an associated travel path for a painting robot.

1 shows a schematic representation of a painting process on a printed circuit board LP1 by a painting robot R. In electronics production assemblies with components that are attached by soldering, plugging or adhesive technology on a printed circuit board LP1, usually provided with a paint to their robustness to the outside To increase influences. An assembly can be completely or partially painted. In 1 is shown how the paint is applied through a nozzle of a painting robot R. The robot R thereby moves the circuit board LP1 and sprayed with the nozzle to be painted areas on the circuit board LP1. Depending on the type of paint used, the curing takes place at room temperature, in a drying oven or z. B. by UV light. Decisive for the quality and durability of the paint finish is the paint coat density within the defined areas to be painted. Often this results in the requirement that during lacquering each lacquer spot may only be painted once. This must be taken into account when determining the travel of the robot R. The nozzle for painting can usually be moved, rotated, opened or closed.

Furthermore, there must be no damage to components, the PCB LP1 or the paint shop R. Collisions of the nozzle with components, the circuit board LP1 or with the paint shop R must therefore not take place.

• • der Roboterarm mit keinem Bauelement zusammenstößt,
• • ein (vorgegebener) hoher Prozentsatz der zu lackierenden Flächen lackiert wird,
• • vorgegebene Flächen (zum Beispiel Löcher, Kontaktflächen) vom Kopf nicht überfahren werden, um Lacktropfen zu vermeiden,
• • die Lackstreifen so lackiert werden, dass „Spritzen” vermieden wird und
• • die für den Lackiervorgang benötigte Zeit minimiert wird.

For a printed circuit board LP1, the placement of the paint strips to be applied and the travel paths of the robot arm R are to be determined so that:

• The robot arm does not collide with any component,
• A (predetermined) high percentage of the surfaces to be painted is painted,
• • do not run over specified areas (for example, holes, contact surfaces) from the head to avoid paint drops,
• • the lacquer strips are painted so that "spraying" is avoided and
• • the time required for the painting process is minimized.

The whole problem is solved hierarchically. The approach described in this invention, mixed integer linear programming, graph algorithms and heuristics, the paint strips and the travel paths of the robot arm are calculated.

First, the problem instance (determining the paint-free surfaces, determining the Lackierflächen, height of the components, safety distances, etc.) prepared for the optimization, then general paint directions are determined on the paint surface and then strips are generated. These stripes are optimized with local search. Shortest paths between all pairs of strip endpoints are calculated and the strips are placed in an optimized order.

The coating machine can not only pull paint strips, but they can stop over the circuit board, the nozzle open briefly and thus produce a so-called paint spot. These spots are modeled as rectangles and treated in the following description analogously with regular paint strips.

Input of the procedure:

• Description of the circuit board:

• • PCB surface as polygon
• • Structured as cuboid components with size, posi tion and rotation
• • as cylinder modeled components with diam water and position
• • Milling contours, modeled as rectangles
• • Holes with center and diameter
• • Contact surfaces, modeled as circular surfaces
• • area to be painted, described as the quantity of polygons
• • List of components that must not be painted

• Description of the machine:

• • Velocity and acceleration data of the robot arm
• • maximum acceleration and braking distances before and after a painting strip
• • Bleeding of the paint for stripes without acceleration and braking distance
• • Diameter of the spray nozzle
• • maximum Z travel height of the robot arm
• • Sizes and parameters of paint spots

• Security and constraints:

• • Minimum distance of the head from the edges of the PCB
• • minimum distance of the head from components that must not be painted
• • minimum distance of the head from the contact surfaces
• • minimum distance of the head from the holes
• • minimum distance in XY direction between the head and a high component
• • minimum distance in the Z direction between the head and a component
• • minimum percentage of paint surface to be covered with paint

2 shows a schematic representation of a circuit board LP2 with horizontal HB1-HB4 and vertical VB1-VB5 areas to be painted. The areas to be painted are formulated as MILP, with the aim of obtaining long stripes, as they are cheaper to paint.

All faces of the problem instance are modeled as a union of rectangles with the edges of the rectangles parallel to the axes of the coordinate system. If a surface does not meet these requirements (for example, a cuboid component rotated 45 °), it is converted at a given resolution into an area whose edges are horizontal or vertical.

• • Lackfreie Fläche: die Fläche darf mit Lack nicht in Berührung kommen. Hier werden die Fräskonturen, Bohrungen, Kontaktflächen, nicht lackierbaren Bauelemente und die Seiten der Leiterplatte mit den gegebenen Sicherheitsabständen berücksichtigt.
• • Lackfläche: die im Input gegebene zu lackierende Fläche abzüglich der lackfreien und physikalisch unerreichbaren Flächen

The following areas are generated from the input data:

• • Paint-free surface: the surface must not come into contact with paint. Here, the milling contours, holes, contact surfaces, non-paintable components and the sides of the circuit board with the given safety distances are taken into account.
• • Paint area: the area to be painted given in the input minus the paint-free and physically unreachable areas

For each point above the board, the minimum height at which the head is above the point is calculated. Components and safety distances are taken into account.

• • Der Input ist das Gitter der Lackfläche als 0–1 Matrix repräsentiert (1: Zelle gehört zur Lackfläche, 0: Zelle gehört nicht zur Lackfläche).
• • Für jede Zelle, die zur Lackfläche gehört, wird eine 0–1 Variable definiert (1: wird horizontal lackiert, 0: wird vertikal lackiert).
• • Manche Variablen können auf Grund physikalischer Bedingungen fixiert werden.
• • Zielfunktion: Minimieren die Summe der Längen der Gitterkanten, für die eine der folgenden Bedingungen gilt:
• • Die zwei Zellen zu beiden Seiten der Kante gehören zur Lackfläche, aber die Lackrichtungen beider Zellen sind unterschiedlich.
• • Nur eine der zwei Zellen zu beiden Seiten der Kante gehört zur Lackfläche, und die Lackrichtung und die Kantenrichtung dieser Zelle sind unterschiedlich.

In a first phase, the goal is to determine general paint directions (horizontal or vertical) for parts (or areas) of the paint surface. In this phase, no paint strips are determined, but only the direction of the stripes. A grid is created so that each cell is either completely in the paint surface, or outside. A Mixed Integer Linear Program (MILP) is created as follows:

• • The input is the lattice of the lacquer surface represented as 0-1 matrix (1: cell belongs to the lacquer surface, 0: cell does not belong to the lacquer surface).
• • For each cell that belongs to the paint surface, a 0-1 variable is defined (1: is painted horizontally, 0: is painted vertically).
• • Some variables can be fixed due to physical conditions.
• • Objective function: Minimize the sum of the lengths of the lattice edges for which one of the following conditions applies:
• • The two cells on both sides of the edge belong to the paint surface, but the paint directions of both cells are different.
• • Only one of the two cells on either side of the edge belongs to the paint surface, and the paint direction and edge direction of this cell are different.

This MILP separates the paint surface into areas that are horizontal HB1-HB4 and areas that are to be painted vertically VB1-VB5. The aiming function guarantees that the directions are chosen so that the areas can be painted with a few long stripes.

There are various commercial or open source solvers (solution software) for solving MILPs, eg. SYMPHONY MILP solver, LP Solv, SCIP etc.

Further shows 2 a section A (shown by a dashed rectangle) of the circuit board LP2, the in 3 is explained in more detail.

3 shows a section A 'of the section A of the circuit board LP2 of 2 equivalent. The section A 'comprises the horizontal area HB1 and the vertical area VB4 of FIG 2 , In 3 the horizontal area HB1 'corresponds to the horizontal area HB1 of FIG 2 and the vertical area VB4 'is the vertical area VB4 of FIG 2 , For the horizontal area HB1 ', the paint strips HS1-HS7 to be painted horizontally and the vertical area VB4' of the paint strips VS1 to be painted vertically are calculated by MILP. The corresponding travel path VW calculated by heuristics (eg greedy heuristics, local search heuristics) is represented by a painting robot for painting the paint strips HS1-HS7 and VS1 is represented by a line to be traversed and is based on the shortest path algorithm Dijkstra.

Subsequent to the phase of determining the horizontal and vertical regions, paint strips are generated for each contiguous region of the paint surface HB1 ', VB4' in which the paint direction is the same.

• • Eine Liste von allen möglichen Streifen, die in die gegebene Richtung laufen, wird erstellt. Diese Streifen sind entweder an den Eckpunkten des Lackflächenbereichs oder an anderen schon generierten Streifen ausgerichtet. Alle generierten Streifen sind bezüglich der Nebenbedingungen zulässig. Die Zahl der erzeugten Streifen ist viel größer als die Zahl der Streifen in der endgültigen Lösung.
• • Mit Hilfe einer MILP-Formulierung werden aus der oben erzeugten Menge diejenigen Streifen HS1–HS7, VS1 ausgewählt, die zur Lösung gehören,
• • jedem möglichen Streifen ist eine 0–1 Variable zugeordnet (1: Streifen gehört zur Lösung, 0: Streifen gehört nicht zur Lösung),
• • wobei mindestens der vorgeschriebene Prozentsatz der Fläche mit den ausgewählten Streifen bedeckt werden soll,
• • wobei mehrfache Überdeckung von Flächen (überlappende Streifen) möglichst vermieden werden soll,
• • wobei das Ziel die Minimierung der Gesamtzeit ist, die zum Durchfahren der ausgewählte Streifen nötig ist.

The algorithm works as follows:

• • A list of all possible stripes running in the given direction is created. These stripes are aligned either at the corners of the paint area or other already generated stripes. All generated stripes are permissible with respect to the constraints. The number of generated stripes is a lot greater than the number of strips in the final solution.
• • With the aid of a MILP formulation, those strips HS1-HS7, VS1 which belong to the solution are selected from the quantity produced above,
• • every possible stripe is assigned a 0-1 variable (1: stripe belongs to the solution, 0: stripe does not belong to the solution),
• Where at least the prescribed percentage of the area is to be covered with the selected strips,
• • where multiple overlapping of surfaces (overlapping stripes) should be avoided as far as possible,
• Where the goal is to minimize the total time needed to travel through the selected strips.

The solution of this MILP is also calculated with commercially available MILP solver (eg SCIP). The solution obtained can be improved with local search heuristics. The selected amount of strip is changed with the aim of minimizing the time required to travel through the strips.

• • Ein Streifen kann verkürzt werden, wenn dadurch die bedeckte Fläche nicht verkleinert wird.
• • Ein Streifen kann verlängert werden, wenn dadurch andere überlappende Streifen verkürzt werden können und die Gesamtdurchlaufzeit kleiner wird.
• • Ein Streifen kann bewegt werden, wenn dadurch
• • die bedeckte Fläche vergrößert wird, oder
• • die bedeckte Fläche gleich bleibt, aber die Überlappungen reduziert werden.
• • Ein zulässiger Streifen kann hinzugefügt werden, wenn dadurch andere Streifen verkürzt oder eliminiert werden können, und die Durchlaufzeit kleiner wird.

The following local modification steps are evaluated or taken into account in an optimization:

• • A strip can be shortened if it does not reduce the covered area.
• • A strip can be extended if it can shorten other overlapping strips and reduce the overall cycle time.
• • A strip can be moved if by doing so
• • the covered area is increased, or
• • the covered area stays the same, but the overlaps are reduced.
• • A permissible strip can be added if it can shorten or eliminate other strips and reduce the cycle time.

For very long strips, it may happen that the entire strip must be painted from a higher head position to avoid collision with components. Since the quality of the painting is better when the head moves lower, these strips are broken into smaller strips and only the partial strips on the large components are painted with a high head position.

Calculating Shortest Paths Between Strip Endpoints Another phase of the procedure is to calculate the shortest paths between strip endpoints. For each stripe, a start position, an end position and a ride height are calculated. The start and end position is usually not at the ends of the strips, because sufficient acceleration and braking distances improve the paint quality. These acceleration and braking distances are determined by the geometric conditions.

For each pair of strip endpoints, the shortest path for the head is calculated using an implementation of the Dijkstra algorithm.

For this purpose, a graph is constructed, which has all vertices of the paint surfaces, components and paint-free surfaces and the grid points of a grid of given resolution as a node. This node set is duplicated at various relevant ride heights. Edges are introduced for each node pair if the edge does not traverse a paint-free surface. On the edges a cost function is defined based on the edge length, the acceleration and the maximum speed of the head.

Calculation of the routes

• • kein „Spritzen” vorkommt: Wenn der Kopf beim Streifenziehen eine so genannte „trocken-nass Kante” durchquert, d. h. von einem noch nicht lackierten Bereich auf einen schon lackierten Bereich fährt, können, Lacktropfen auf andere Teile der Leiterplatte spritzen. Dadurch könnten lackfreie Flächen durch Lacktropfen bedeckt werden, was die Leiterplatte unbrauchbar macht. Durch sorgfältiges Wählen der Reihenfolge von Lackstreifen kann dieses Spritzen vermieden werden, und
• • die Summe der Fahrtzeiten zwischen den Streifen minimiert wird.

In this phase, the stripes are placed in a sequence so that:

• • No "spraying" occurs: When the head passes through a so-called "dry-wet edge" during stripping, ie when it moves from an area that has not yet been painted to an already painted area, paint droplets can splash onto other parts of the printed circuit board. As a result, paint-free surfaces could be covered by paint drops, which makes the circuit board unusable. By carefully choosing the order of paint strips this spraying can be avoided, and
• • the sum of the travel times between the strips is minimized.

• • Ein Teilintervall der Streifen wird aus der Reihenfolge gelöscht, und die Streifen werden einzeln wieder an einer Position hinzugefügt, wo weiterhin kein Spritzen verursacht wird und die Fahrtzeit nach dem Einfügen minimal ist. Wenn alle gelöschten Streifen wieder hinzugefügt werden können und die Gesamtzeit besser geworden ist, wird die neue Lösung übernommen.
• • Ein Teilintervall der Streifen wird aus der Reihenfolge gelöscht und umgekehrt wieder an der gleichen Stelle eingefügt. Wenn die neue Reihenfolge kein Spritzen verursacht und die Gesamtzeit besser geworden ist, wird die neue Lösung übernommen.

To calculate the order a local search heuristic is used. First, a startup sequence is generated in which no spraying occurs. This starting solution is improved by the following modification steps:

• • A sub-interval of the stripes is deleted from the sequence, and the stripes are added one by one at a position where no spraying is still caused and the travel time after insertion is minimum. If all deleted strips can be added again and the total time has improved, the new solution is adopted.
• • A sub-interval of the strips is deleted from the sequence and vice versa inserted again in the same place. If the new order does not cause splashing and the overall time has improved, the new solution will be adopted.

• • allgemeine Lackierrichtungen werden mit einer globale Optimierungsmethode (in fast allem Fällen) optimal bestimmt,
• • die Auswahl der Streifen in der Erstlösung wird global optimiert,
• • spezielle lokale Verbesserungen liefern eine Lösung, die aus elektrotechnischer Sicht optimal ist,
• • kürzeste Wege zwischen Punkten werden optimal bestimmt,
• • die Reihenfolgeoptimierung vermeidet Spritzen und minimiert Fahrtzeiten gleichzeitig

The solution includes the following ideas:

• • general painting directions are optimally determined with a global optimization method (in almost all cases),
• • the selection of stripes in the first solution is optimized globally,
• • Special local improvements provide a solution that is optimal from an electrical point of view
• • shortest paths between points are optimally determined
• • Sequence optimization avoids spattering and minimizes travel times at the same time

• • globaler Optimierungsansatz für die Berechnung der Bereiche und Streifen,
• • fehlerfreie Lösung (im Vergleich zur manuellen Planung),
• • Planungsaufwand wird erheblich reduziert,
• • der Lackiervorgang der Maschine ist wesentlich schneller, d. h. bessere Qualität der Lackierung,
• • Einsatz von Üblichen MILP-Solvern möglich.

Benefits of the automated approach based on MILP:

• • global optimization approach for the calculation of areas and stripes,
• • error-free solution (compared to manual planning),
• • Planning effort is significantly reduced
• • The painting process of the machine is much faster, ie better quality of the paint,
• • Use of common MILP solvers possible.

Method and system for determining the travel distances of painting robots, in particular for electronics production. By Mixed Integer Linear Programming (MILP), in a hierarchical approach a surface to be painted is separated into horizontally and vertically to be painted areas and strips, which are advantageously moved by a painting robot.

reference numeral

R

Painting robots

LP1, LP2

circuit board

A, A '

Detail of a circuit board

HB1-HB4, HB1 '

Horizontal painting area

VB1-VB5, VB4 '

Vertical painting area

VW

traverse

HS1-HS7

Horizontal paint strip

VS1

Vertical paint strip

QUOTES INCLUDE IN THE DESCRIPTION

This list of the documents listed by the applicant has been generated automatically and is included solely for the better information of the reader. The list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions.

Cited patent literature

• DE 102007041424 A1 [0004]

Claims (10)

Method for determining travel distances (VW) of painting robots (R), in particular for electronics production, comprising the following steps: a1) determination of parameters of a printed circuit board (LP1, LP2) to be traversed and painted; a2) determination of parameters for the painting robot (R); a3) defining safety and ancillary conditions for the process and painting of the PCB (LP1, LP2); b) modeling the surfaces of the printed circuit board (LP1, LP2) to be painted as a union of rectangles, the edges of the rectangles being arranged parallel to the axes of a Cartesian coordinate system of the printed circuit board (LP1, LP2); c) Generating painting directions for the method of the rectangles to be painted by Mixed Integer Linear Programming, the area to be painted being separated into areas to be painted horizontally (HB1-HB4, HB1 ') and vertically (VB1-VB5, VB4'); d) Generating paint strips (HS1-HS7, VS1) for each contiguous area to be painted (HB1-HB4, HB1 ', VB1-VB5, VB4') per coating direction by Mixed Integer Linear Programming. The method of claim 1, wherein step c further comprises: Creating a grid of the printed circuit board (LP1, LP2) so that each cell of the printed circuit board (LP1, LP2) is either completely in a surface to be painted or outside; and Minimize the sums of lattice edge lengths according to one of the following conditions: the two cells on either side of the grid edge belong to the area to be painted, but the painting directions of both cells are different, or only one of the two cells on both sides of the grating edge belongs to the surface to be painted, and the painting direction and the edge direction of these cells are different. The method of claim 1, wherein step d further comprises: Creating a list of all horizontally (HS1-HS7) and of all vertical candidates (VS1) for a paint strip selection, wherein the paint strips (HS1-HS7, VS1) either at the vertices of the corresponding area to be painted (HB1-HB4, HB1 ' , VB1-VB5, VB4 ') or other already produced paint strips (HS1-HS7, VS1), and are permissible with regard to the secondary conditions; and Selecting Strips to Be Painted by Mixed Integer Linear Programming, where at least a given percentage of the printed circuit board area (LP1, LP2) can be covered with the selected resist stripe (HS1-HS7, VS1), minimizing overlapping stripes and the total run time required to travel through the selected paint strips is minimized. The method of claim 1 or 3, wherein step d) further comprises: Shortening a strip (HS1-HS7, VS1), if it does not reduce the covered area, or Lengthening of a strip (HS1-HS7, VS1), as a result of which other overlapping strips are shortened and the overall cycle time becomes smaller. The method of claim 1 or 3, wherein step d) further comprises: Moving a strip if it increases the area covered, or keeps the covered area the same, but reduces the overlaps. The method of claim 1 or 3, wherein step d) further comprises: Adding a strip, if it can shorten or eliminate other strips (HS1-HS7, VS1), and the cycle time is smaller. The method of claim 1, further comprising: Determining the shortest paths between all edge points of different coating strips (HS1-HS7, VS1) taking into account the acceleration and braking distance of the painting robot (R). The method of claim 1, further comprising: Determining the travel paths (VW) of the painting robot (R), wherein the strips (HS1-HS7, VS1) are placed in an order that avoid spraying and wherein the sum of the travel times between the strips (HS1-HS7, VS1) is minimized , A computer readable medium or computer program product comprising instructions which, when executed on a suitable computer, cause the computer to carry out the method of any one of claims 1 to 8. System for carrying out a method according to one of claims 1 to 8.

Images