DD216400A1 - CIRCUIT ARRANGEMENT FOR OPTIMIZING THE FEEDING REGULATOR PARAMETERS IN RADIO EROSION - Google Patents

Abstract

The invention relates to a circuit arrangement for on-line optimization of the feed controller parameters for spark erosion, with the adjustment of the process-dependent optima of the setpoint and the Reglerverstaerkung necessary to maintain an optimal machining gap feed control device maximum removal rate is realized with minimal possible wear, creating a significant improvement the machining accuracy and by increased process productivity reduced processing time or processing costs can be achieved. This is achieved by taking into account the stochastic character of the spark erosion process by means of statistical measurement methods and by optimizing the feed controller parameters in such a way that the probability density distributions of gap voltage amplitude and discharge time parameters approximate the deviation of the respectively most frequent measured values by changing the controller setpoint and the controller gain to optimum distributions is minimized by the corresponding arithmetic mean values and the associated measured value scatters are monitored for limit value exceeding, whereby the machining gap and discharge ratios optimized for the respective machining regime and the current process conditions are set, whereby a substantial increase of the process productivity and machining accuracy is achieved.

Description

Title of the invention

Circuit arrangement for optimizing the feed controller parameters during spark erosion.

Field of application of the invention

The invention relates to a circuit arrangement for the on-line optimization of parameters of the feed control system necessary for producing an optimum gap width, the working gap formed by the tool electrode and the workpiece as a function of the process conditions in spark erosive processing.

Characteristic of the known technical solutions

The spark erosion metal removal proceeds by the action of electrical discharges in a ötück formed by the tool electrode and tool ¹ and through which a flussigen dielectric gap. In order to ensure a continuous J ¹ rollover, compliance with an approximately constant gap width, in which discharges can take place, is unavoidable. Taking into account disturbance variables, such. B. gap width change by removal or variable dielectric strength of the dielectric in the gap, thus the tool electrode must be tracked during processing. This task is performed by a feed regulator in cooperation with a hydraulic or electrical

Thrust subsystem. The derivation of feed signals can be done in many ways and in part.

Circuit arrangements are known in which the mean value of the pulse voltage applied to the working gap or a voltage signal adequate to the effective ratio of the discharge is used as control variable in the feed control loop The process stability and the tendency of the process to erroneous discharges are mainly influenced by the feed control loop parameters. The feed controller parameters to be considered are the controller setpoint Xg and the controller gain Κ Machining gap influenced, the controller gain Kjj determines the speed of the correction of deviations and thus the process stability. Pure spark erosion, due to the large number of possible processing cases, has a large setting range of process parameters. As a result, depending on the processing regime, the optimum values for the controller solubility X "and the controller gain Kg vary within wide limits. In many practical applications of the EDM process, the feed controller parameters are adjusted by the plant operator according to empirical values and kept constant, as a result of which the processing result is decisively dependent on the skill and experience of the plant operator.

In order to achieve optimal processing results, an on-line optimization of various process parameters must be undertaken during processing.

Methods are known that incorporate feed control parameters in part in the on-line optimization of the spark erosion process. Sine known technical realization possibility! Minimizes the fluctuations of the effective relationship through changes

tion of the feed controller gain K _R. In this method, by Einzelimpulsabtastung considerable stress of the process computer and an increased computational time required for the MeßwertÜbernahme, whereby the method is only suitable for roughing.

Furthermore, a circuit arrangement is known in which an Isarithmenregelung the setpoint voltage Ug is made in connection with the pause time t _Q * This type of two-size control is very time-consuming as a pure search method for automatic Pindung the optimal operating point.

Furthermore, methods are known which maximize the feed stability ε by varying the regulator gain K _R. The disadvantage here is that in a variety of process regimes a maximum feed resistance is associated with increased arc trend, which i. generally results in an undesirable increase in tool electrode wear.

In various known circuit arrangements, the controller gain Κ «is determined by the front gap width a. For metrological detection of the gap width a but each work interruption is necessary.

Furthermore, a circuit arrangement is known, which adapts the controller gain to the respective process state. A high controller gain is set in the case of short-circuit and no-load conditions, with normal erosion this is reduced. This procedure does not guarantee the K _R adjustment to its optimum value for all machining cases.

The setting of optimum values for the controller setpoint Xg and the controller gain K _{R of} the feed control device as a function of the process conditions is indispensable for achieving a desired processing result with minimum processing time or minimum manufacturing costs. Since the feed control loop is subjected in the spark erosion of predominantly stochastic disturbances, the application of statistical measurement methods for the process identification to determine the optimum feed control parameters is essential.

''. . ' .. - 4 - ·. '' '

The circuits mentioned meet the ο. G. Demands not or rmt in part. Their application is largely limited to special processing regimes and therefore not universally valid.

Object of the invention

The aim of the invention is to avoid the Hachteile of the known technical solutions and to achieve by statistical signal processing on-line optimization of the parameters of the maintenance of an optimal gap width of the machining electrode formed by the tool electrode and workpiece machining feed control system, the IPunkenerosionsprozeß to keep in the optimum range and to avoid false discharges, whereby a substantial improvement in the machining accuracy and reducing the processing time or the manufacturing cost is achieved.

Presentation of the essence of the invention

Optimization of controller gain K «. and the controller setpoint value Sg of the feed control device of spark erosion systems for maximizing the removal rate and increasing the machining quality as a function of the respective process state, the object of the invention, which is achieved by a circuit arrangement for on-line optimization of the feed controller parameters during processing, by a process computer, with a sensor w (x) for the optional determination of the probability density distribution of the gap voltage amplitude w (u_) or the Y probability probability distribution of discharge time parameters w (t), a sensor E ^ for detecting the arithmetic mean value of the considered discharge time parameter and a Sensor u is connected to measure the average spark-ignition voltage ü _f , the optimum values for the controller setpoint value Xg ^ and the controller gain of the feedforward

be determined such that the sensor in the sensor w (x) determined most common voltage amplitude Uj. is compared with the mean Ptinkenbrennspannung CL · and brought into agreement by changing the control value 6 and sure that the optimal controller gain K _R ^ is determined such that the in the sensor w (x) determined probability density distribution of a discharge Time parameter w (t) is approximated by altering the gain of an optimal curve, where the deviation of the most frequent discharge time value t _n from the arithmetic mean ΐ becomes a minimum »

embodiment

The invention will be explained in more detail below using an exemplary embodiment. The accompanying drawings show:

Pig. 1: Diagrams for the influence of the controller setpoint and the controller gain on the probability of density distribution of the gap voltage and the relative ignition delay time, respectively, as the discharge time parameter used

2 shows a block diagram of the circuit arrangement for the on-line optimization of the pre-boost regulator parameters

InJ ^ ig. 1 shows under a the principle dependence of the probability density distribution of the gap voltage w (u_) on the controller setpoint value X ₃ ; Curve (II) indicates the optimum process state, maximum removal rate with minimal wear.

Pig. 1 b represents the equivalent relationship between the probability density distribution of the ignition delay time w (t _D ) and the controller gain K ^. In turn, curve (II) describes optimal removal ratios. The progressions shown are independent of the processing regime.

 ι "· · · · · · 6 · ·

ι ',.' '. \ .. '... ,

Di & e exploitation of these relationships for the Parameteroptimieitjuhg in a process control loop for the spark-erosive lowering Fig. 2 in! Principle representation.

shown is a pulse generator 1, which emits the electrical pulses to the working gap 2, which is traversed by a liquid dielectric from the reservoir ^ The feed control and adjusting device 4 performs the tool electrode during the removal process and ensures the time required for periodic flashover gap width in Work gap The gap voltage curve u _a (t) provided by the pulse generator 1 and adjusting at the working gap 2 is determined in the sensor 5 for the selective detection of the probability density variations of the gap voltage amplitude w (u _& ) or the ignition delay w (tjj) in the sensor 6 for determination the average relative delay time I 1 and in the sensor 7 for the measurement of the average ignition voltage 5 1. evaluated. The sensors 6 and 7 also need to decouple the pulse break the Generätorlakt U. ₁

All obtained measuring signals w (u _a ), w (t _D ), ΐ _β , u _f are fed via the process computer peripheral unit 8 to the microcomputer 9, which in turn determines the strategy of the measured value acquisition via control and address lines. The microcomputer calculates the optimum feed controller parameters Xg ^ and ^K n _0D ^ 'which are output via the process computer peripheral unit 8 to the feed regulator 4 by comparison operations of the recorded measurement values with each other and determined in technological studies regime independent optimal probability density gradients.

The principle of determination of the optimum controller setpoint ^ soot lies basis that a particular controller setpoint ^ "3opt ^s * ^s P ^recnen ^ ^e optimal gap width is then set if the mean value of the punk burning voltage u ^ corresponds approximately to the gap voltage amplitude with the maximum frequency and additionally the measured value dispersion b does not exceed a minimum permissible Jfert ^ ₀₁ , _en2 · a smaller dispersion φ "^. i? Grenz ^deu * ^{et on erhlöll} * ^e lichtbogentendenz out. The most common gap voltage amplitude U _n and the measured value dispersion

is measured by the process computer 9 from that with the w (x) sensor 5; Probability density distribution of the gap voltage, the average spark ignition voltage n ^ with the ü ^ sensor ermiti. The optimization of the controller setpoint value Xg starts from a starting value Xg ₀ ^> Xg _opt as follows

Wach Entering the w. (Ac) sensor 5 metrologically detected distribution of the probability density of the gap voltage is detected in the process computer 9, the most common gap voltage amplitude. (If several (> »1), approximately equal frequency maxima occur, this indicates an increased feed instability. Consequently, in this case the controller gain K _R must first be optimized. Otherwise, after the input of the ^ value from the sensor 7 into the Process computer 9 the most common voltage value Uj ₁ compared with the arithmetic mean of the spark ignition voltage u ^ and ^ e after the sign of the deviation, a gradual increase b · w .Reduction of the controller setpoint caused so long until the deviation of the two voltage values becomes a minimum then the Meßwertstreuung Transmed and at ^^^ Grena brought about an increase of the controller setpoint and the cycle started again. Otherwise the optimization cycle is finished and need only be repeated at longer intervals with the former optimum value as starting value Xg ₀ : «Xs ₀₀ ^.

The process-dependent optimal setting of the controller gain Kj ₁ is realized by analyzing the probability density distribution of the ignition delay time w (t _D ) with the w (x) sensor 5. Characteristic of a prefixed stability optimally chosen over K _R is that the distribution curve w (t ^) approximates that of a half-meridional distribution

A · ...... · - ... · '

! W (t _D ) = K ₁ * t _D ² * exp (-t _D ² / K ₂ ) (1)

with the most common value * ie * "/ ^Κ 2

equivalent. Another condition is in turn the setting of the minimum possible deviation of the most common of the mean Zündverzugszeitwertes. To avoid process degenerations in the start cycle, the start value K _Ro «c K _Ro ^ is set during the first pass. By statistical evaluation, the measured distribution curve is analyzed in the control computer 9 and compared with the optimal curve. If there is a deviation of the measured from the nominal curve, this deviation is compared with that of the previous cycle. As long as the difference of the deviations becomes smaller, the previous direction of adjustment can be maintained, otherwise the setting direction is changed. The circuit arrangement shown in FIG. 2 thus ensures the automatic setting of optimal values for the controller setpoint and the controller gain of the feed control loop, which ensures the maintenance of optimum discharge conditions Episode has.

Claims (2)

invention claim Circuit arrangement for optimizing the feed controller parameter, on-line optimization, in the spark erosion of the work gap required for setting the optimum gap width of the working gap formed by the machine electrode and the workpiece: ί ', Thrust control device as a function of the process conditions in the EDM machining with a pulse generator, characterized in that by a process computer (9), the rjiit with a sensor w (x) (5) for selectively determining the probability of density distribution of the gap voltage amplitude w (u_) or the probability density distribution of discharge time parameters w (t), a sensor f _D (6) for detecting the arithmetic mean value of the considered discharge time parameter and a sensor u ^ (7) for measuring the average spark voltage u _{f is} interconnected, the optimum values for the controller setpoint 3Cg _op ^ uncl the controller gain of yorschubeinrichtung be determined such that in the sensor w (x) (5) determined most common voltage amplitude Uj. is compared with the instantaneous spark-burning voltage u ^ and brought into conformity by changing the measured value deviation and the optimum controller gain is again determined such that the probability density distribution of a sensor determined in the sensor w (x) (5) Discharge Zeitpärame- feva w (t) is approximated by changing the controller gain of an optimal curve in which the deviation of the most common discharge time t ^ from the arithmetic mean ΐ becomes a minimum. gierzu 2 sheets drawing