DE102015000291A1 - Method for controlling a processing machine - Google Patents

Abstract

The invention relates to a method for controlling a processing machine, in which at least one tool (1) arranged movably on the processing machine is automatically moved from a predetermined starting point (SP) to at least one predetermined end point (EP) within a working space (R). According to the invention, a tool path (B) is determined within the working space (R) along which the tool (1) is moved without collision in a minimum time from the starting point (SP) to the end point (EP), the tool path (B) being determined by means of an evolutionary algorithm is determined.

Description

The invention relates to a method for controlling a processing machine, in which at least one tool movably arranged on the processing machine is automatically moved from a predetermined starting point to at least one predetermined end point within a working space.

Efficiency of manufacturing processes is the focus of today's production. In order to further reduce a production time for a component, it is necessary to minimize non-value-adding activities of a processing machine, in particular of a machine tool. A starting point for the reduction of non-value-adding actions in a machine tool is a minimization of technical non-productive times. In this case, a trajectory or tool path of a tool and a spindle of the processing machine plays a very large role.

For this purpose, various methods are known from the prior art. Such a method for controlling a processing machine describes the DE 10 2013 021 653 A1 wherein at least one tool movably arranged on the processing machine is automatically moved from a predetermined starting point to at least one predetermined end point within a working space. Within the working space, a movement space for free movement of the tool is specified. That is, the control of the processing machine is designed such that the movement of the tool within the working space does not follow a predefined trajectory, but is described only within the movement space and on the basis of target coordinates.

The invention is based on the object to provide a comparison with the prior art improved method for controlling a processing machine.

The object is achieved by a method having the features specified in claim 1.

Advantageous embodiments of the invention are the subject of the dependent claims.

In a method for controlling a processing machine, at least one tool movably arranged on the processing machine is automatically moved from a predetermined starting point to at least one predetermined end point within a working space.

According to the invention, a tool path is determined within the working space, along which the tool is moved without collision in a minimum time from the starting point to the end point, the tool path being determined by means of an evolutionary algorithm.

Such an evolutionary algorithm or genetic algorithm forms a class of stochastic, metaheuristic optimization methods whose operation is inspired by the evolution of natural living beings. Based on nature, solution candidates for a specific problem are artificially evolved.

By means of the method according to the invention, due to the determination of the tool path along which the tool is movable collision-free in a minimum time from the starting point to the end point, it is possible to minimize non-value-adding activities of the processing machine and thus to minimize cycle times during the machining of workpieces. As a result, a manufacturing time for a component is reduced, so that an efficiency of manufacturing processes is increased. Due to the use of the evolutionary algorithm for determining the tool path, no complex and costly simulation software for motion optimization of the processing machine is required, since the performed by the evolutionary algorithm optimization of the tool path directly in a control of the machine during commissioning or operation of the machine fully automatic executable. Furthermore, an optimization potential of the tool path is exhausted by the evolutionary algorithm in comparison to the prior art to a greater extent.

Embodiments of the invention are explained in more detail below with reference to drawings.

Showing:

1 schematically a flowchart of an evolutionary algorithm for determining a tool path, and

2 schematically a working space with a workpiece in a sectional view and a predetermined tool path for a tool of a processing machine.

Corresponding parts are provided in all figures with the same reference numerals.

In 1 is a flowchart of an evolutionary algorithm for determining an in 2 shown tool path B of a tool 1 a processing machine, not shown, according to a possible embodiment of a method according to the invention. 2 shows a working space R with a workpiece 2 in a sectional view and the predetermined tool path B for the tool 1 which is in particular a so-called spindle or comprises these.

The tool 1 is in the illustrated embodiment for producing holes X1, X2 within the workpiece 2 is provided and is starting from a first hole X1, which forms a starting point SP of the tool path B, moved along the tool path B to an end point EP on which a second bore X2 is generated. Between the starting point SP and the end point EP is a Störgeometrie 3 arranged or formed, on which the tool 1 passed without collision.

In order to minimize a manufacturing time for a component, it is necessary to minimize non-value-added activities of the processing machine. In this case non-value-added actions of the machine, in the illustrated embodiment, a collision-free passing of the tool 1 at the Störgeometrie 3 To minimize, by means of an optimization of the tool path B, also referred to as trajectory, technical non-productive times are minimized.

For this purpose, it is provided that within the working space R a tool path B is determined, along which the tool 1 is moved collision-free in a minimum time from the starting point SP to the end point EP, wherein the tool path B is determined by means of an evolutionary algorithm. In this case, the algorithm is used to search for two reference points P1, P2, which make it possible to determine the interference geometry 3 time-optimal and avoid collision, with the basis of the starting point SP, the bases P1, P2 and the end point EP, the tool path B in the form of a polynomial (spline English) is determined.

The evolutionary algorithm is a so-called genetic algorithm, which is based on a so-called mountaineering algorithm. Such a mountaineering algorithm is understood to be a simple, heuristic optimization method in which a potential solution to a given problem is progressively improved.

Here the algorithm is divided into three phases. In a first phase, boundary conditions are specified in a first step S1 of the algorithm. In this case, it is possible for a user to vary given boundary conditions or to accept them unchanged. In particular, the user has the option of setting limit values of the interpolation points P1, P2. From this it follows how far a respective support point P1, P2 may be moved three-dimensionally in space. Decisive for the choice of limit values is a geometry and arrangement of the interference geometry 3 to make a collision of the tool 1 to avoid this.

Furthermore, it is a consideration of a geometry of the tool 1 and possibly further the workspace R limiting structures required.

In addition, it is possible for the user to influence or define a so-called starting population of the algorithm. By means of the starting population, it is determined how many pairs of bases P1, P2 are randomly generated at the beginning within the prescribed limits.

Furthermore, an initial step size is automatically and / or by the user when using a distance function on a definition set of a value landscape for generating the tool path B determined by at least one parameter. Also, a reduction factor is set automatically and / or by the user by which the step size is reduced after each pass of an optimization.

In addition, a minimum increment is automatically and / or specified by the user, the achievement of which represents a termination criterion.

After the specification of the boundary conditions in the first step S1, in a second phase of the algorithm in a second step S2, random starting values are generated for each two support points P1, P2 for a plurality of polynomial trains in the workspace R, along which the tool 1 collision-free from the starting point SP to the end point EP is movable.

In a third step S3, all random polynomial trains are then traversed within the scope of an optimization, and the polynomial train with the associated pair of interpolation points P1, P2 is determined and selected from the plurality of polynomial trains, which selects the shortest time for the collision-free movement of the tool 1 from the starting point SP to the end point EP. The remaining polynomial trains and their associated interpolation points P1, P2 are discarded in a fourth step S4.

Subsequently, in a fifth step S5, one of the interpolation points P1, P2 is selected for an alternating variation of coordinates of the two interpolation points P1, P2 in the working space R.

In a subsequent sixth step S6, a first axis is selected, on which the selected interpolation point P1, P2 for optimizing the Polynomial train is moved before it is checked in a seventh step S7, whether the given in the first step S1 boundary conditions are violated by such a shift of the support point P1, P2.

If the boundary conditions are not violated (represented by a path N1), the selected interpolation point P1, P2 is shifted in an eighth method step S8 by the selected step along the selected axis in the working space R, before it is checked in a ninth step S9 whether carried out a shift optimization of the polynomial train.

If there is no temporal optimization by the displacement (represented by a path N2), a displacement direction on the selected axis is changed in a tenth step S10. The change in the direction of displacement in step S10 also takes place if the boundary conditions are violated in the seventh step S7 (represented by a path J1).

Even after the change of the shift direction in the tenth step S10, it is checked in an eleventh step S11 whether the boundary conditions specified in the first step S1 are violated. If this is the case (represented by a path J2), in a twelfth step S12 a selection of another axis takes place, along which the selected interpolation point P1, P2 is displaced within the working space R.

On the other hand, if the boundary conditions are not violated, a shift of the selected interpolation point P1, P2 by the selected increment is performed in a thirteenth step S13 and the polynomial train is executed or traversed.

The thirteenth step S13 is also executed without executing steps S10 and S11 immediately following the ninth step S9 (represented by a path J3) when the polynomial train has been temporally optimized by the shift performed in the eighth step S8.

Subsequent to the thirteenth step S13, it is checked in a fourteenth step S14 whether a temporal optimization of the polynomial train takes place due to the displacement carried out. If this is the case (represented by a path J4), the steps S13 and S14 are repeated until it is determined in the fourteenth step S14 that it is no longer possible to optimize the polynomial drag by shifting along the selected axis (represented by a path N3) ).

In this case, it is checked in a fifteenth step S15 whether a temporal optimization by selection of another axis, along which the selected support point P1, P2 is moved within the working space R, is feasible. If this is the case (represented by a path J5), the sixth step S6 is performed again.

If this is not the case (represented by a path N4), it is checked in a sixteenth step S16 whether other axes have already been considered. If other axes have not yet been viewed (represented by a path N5), the twelfth step S12 is executed again.

On the other hand, if all the other axes have already been considered (represented by a path J6), a seventeenth step S17 determines whether a temporal optimization of the polynomial pull can be achieved by selecting the respective not yet considered support point P1, P2 and its displacement along axes in working space R. is. If this is the case (represented by a path J7), the fifth step S5 is carried out for the respectively remaining support point P1, P2.

If the result of the seventeenth step S17 is negative (represented by a path N6), it is checked in an eighteenth step S18 whether all the interpolation points P1, P2 have already been considered. If this is not the case (represented by a path N7), the selection of a remaining support point P1, P2 takes place in a nineteenth step S19 and then the execution of the sixth step S6.

On the other hand, if all interpolation points P1, P2 have already been considered, the increment is reduced by the reduction factor in a twentieth step S20.

In a 21st step S21, it is checked whether the termination criterion, i. H. reaching the predetermined minimum increment, is present. If this is not the case (represented by a path N8), the fifth step S5 is performed again with a reduced step size. On the other hand, if the abort criterion is reached (represented by a path J8), the temporal optimization of the polynomial train is ended in a 22nd step S22.

The algorithm is thus executed in such a way that, beginning with an axis, a coordinate of the selected interpolation point P1, P2 is shifted by the initial increment. If a temporal improvement can be achieved by traversing the polynomial train, then the coordinate is shifted again by one section in the following step. This happens until the achieved time increases again or a boundary condition is violated. If a deterioration has already been achieved in the first step, the displacement of the coordinate of the point P1, P2 in the other direction is carried out on the same axis. This move the Coordinate along one axis is then also performed for the remaining other two axes of the working space R. If no temporal optimization of the polynomial train could be achieved, the remaining interpolation point P1, P2 is varied. But if a temporal improvement was achieved for one of the axles, a further variation of the other two axles takes place.

Should an improvement be achieved by the variation of the further interpolation point P1, P2, the interpolation point P1, P2 tested first must also be varied again. However, if no time improvement could be achieved for all three axes in the working space R of the two interpolation points P1, P2, the reduction of the step size follows the previously defined reduction factor and the shifting of the coordinates of the interpolation points P1, P2 starts again. If the step size reaches the abort criterion, the optimization is run through for the last time and ends as soon as a further reduction would have to take place.

After the optimization carried out automatically during operation of the processing machine or during its startup, in a third phase of the algorithm, the interpolation points P1, P2 of the polynomial traction are transmitted with the minimum time to a control unit (not shown) for controlling the processing machine. The control unit, which is in particular part of the processing machine, controls it on the basis of the support points P1, P2 such that the tool 1 along the tool path B is moved without collision in the minimum time from the starting point SP to the end point EP.

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 102013021653 A1 [0003]

Claims (8)

Method for controlling a processing machine, in which at least one tool movably arranged on the processing machine ( 1 ) is automatically moved from a predetermined starting point (SP) to at least one predetermined end point (EP) within a working space (R), characterized in that within the working space (R) a tool path (B) is determined, along which the tool ( 1 ) is moved without collision in a minimum time from the starting point (SP) to the end point (EP), the tool path (B) being determined by means of an evolutionary algorithm. A method according to claim 1, characterized in that the tool path (B) is automatically determined during operation of the processing machine. A method according to claim 1 or 2, characterized in that the tool path (B) is determined in the form of a polynomial. Method according to Claim 3, characterized in that boundary conditions are specified in a first phase of the algorithm, boundary conditions stipulating at least limit values for interpolation points (P1, P2) of at least one polynomial traction. Method according to Claim 4, characterized in that a number of polynomial trains are determined as a function of the interpolation points (P1, P2), along which the tool ( 1 ) is collision-free from the starting point (SP) to the end point (EP) is movable. A method according to claim 5, characterized in that in a second phase of the algorithm of the plurality of polynomial trains the polynomial train is selected, which is the shortest time for collision-free movement of the tool ( 1 ) from the starting point (SP) to the end point (EP). Method according to Claim 6, characterized in that, in the second phase of the algorithm, at least one of the interpolation points (P1, P2) of the selected polynomial train is displaced three-dimensionally in the working space (R) until the time in which the tool ( 1 ) is collision-free from the starting point (SP) to the end point (EP), is minimized. A method according to claim 7, characterized in that the support points (P1, P2) of the polynomial with the minimum time are transmitted to a control unit for controlling the processing machine, which controls the processing machine based on the support points (P1, P2) such that the tool ( 1 ) is moved along the tool path (B) without collision in the minimum time from the starting point (SP) to the end point (EP).

Images