Ambiguous control technology applies mediumly in process of electric spark form cutting

Which Is Used To Ascertain The Trend Of Discharge Gap Change, is Presented.

And A Fuzzy Control Strtegy Is Introduced To Parctical Use, the Results Is Well.

The factor that 1 foreword of ▲ affects electric spark to machine a process is very much, random sex is very big, it is consequently in process of electric spark treatment, to assure to machine the precision of workpiece and exterior surface roughness, eliminate the hidden danger of burn workpiece, facilities of electric spark treatment should have what perfect quite from adaptive control system, make treatment of whole electric spark works in groove, in order to improve productivity and economy. Problem of core of pilot of electric spark form cutting is the control of condition of pair of electric spark discharge. But of traditional electric spark form cutting restrain adaptive control slightly from adaptive control plan (ACC) and optimal adaptive control (ACO) , because this kind controls strategy to try to use mathematical mathematical model to come,describe electric spark to machine this one complex nonlinear of system and much element when change system, existing consequently a lot of blemish. To get used to tall automation nobody are operated, those who reduce pair of handlers is skilled degree requirement, assure to machine quality, the article put forward a kind of new-style faintness to control strategy, obtained very good test result. Of condition of discharge of 2 electric spark monitor have a variety of is opposite of condition of the discharge in electric spark form cutting detect method, but at present the simplest also be the most commonly used it is the method that uses weaveform of differentiate clearance spark voltage, will judge the discharge condition in process of electric spark form cutting. Weaveform of electric spark spark voltage can be divided for 5 kinds of main types, if pursue,1 is shown. Graph 5 kinds of when clearance of form cutting of 1 electric spark discharges basic forms pursue 1 in, a is weaveform of open circuit voltage, b is weaveform of normal spark voltage, c does not stabilize weaveform of electric arc spark voltage for recoverable sex, d stabilizes weaveform of electric arc spark voltage to cannot restore burn sex, e is weaveform of short circuit voltage. Weaveform of spark voltage of 5 kinds of electric spark can be used at condition of discharge of differentiate electric spark. Commonly used method is the method that is worth Vref1, Vref2 through installing doorsill voltage, differentiate discharge condition for 3 area, it is open circuit when spark voltage is worth prep above Vref1 namely; When spark voltage the value is between Vref1 and Vref2 when for normal discharge area; Be worth when spark voltage under unusual discharge area is when Vref2. But, this kind of method gives discharge of instability electric arc inseparably. For condition of well and truly differentiate clearance discharge, put in high frequency signal considering voltage of instability electric arc, the article detects doorsill voltage law and high frequency signal law union rises. When Vref2 of ＜ of spark voltage U and when be being put in high frequency signal, discharge condition is condition of discharge of not stable electric arc; When Vref2 of ＜ of spark voltage U and when nonexistent high frequency signal, discharge condition is condition of discharge of stable electric arc or short circuit condition. Such dissociable a condition of 4 kinds of discharge, considering condition of discharge of stable electric arc and short circuit condition among them rise, them so add up is a kind. Take T(ms) to be analytic cycle, sampling frequency F(Hz) , e of ′ of your Td, Te, T, Ta is a discharge respectively the time that discharge of open circuit, normal discharge, flabby discharge of calm electric arc, stable electric arc and short circuit condition appear inside cycle. Inside each sampling cycle, make condition differentiate to collecting sample, it is open circuit, normal to in undertaking A of E of statistical τ D, τ , τ ′ E, τ is pulse of a voltage respectively, appear the sampling of discharge of discharge of not discharge, stable electric arc, stable electric arc and short circuit condition is worth a frequency. When undertaking clearance condition detects in project practice, the output that included power supply of electric spark pulse actually is in the condition of the phase between arteries and veins, right now sampling detects voltage value is 0, consider as short circuit, statistical result record arrives in τ A. The method that solves this one problem is to pass tally to record pulse several N, your Ta stabilizes the time of electric arc discharge and short circuit condition to appear actually inside cycle of a discharge, tt is between the arteries and veins that establishs power source of electric spark pulse, it is inside analytic cycle T, the time that status of 4 kinds of discharge appears is respectively: (　 of 　 of 1) 　 　 (　 of 2) 　 　 (　 of 　 of 3) 　 (4) 　 is analysing the total discharge time inside cycle: TT=T-N. 　 of 　 of TI 　 　 (because 5) 　 does not stabilize electric arc discharge condition is recoverable for normal discharge condition, cause discharge of stable electric arc possibly also, its are not big to machining the influence of the process. In be controlled actually, often allow to appear not to stabilize electric arc to discharge partly, but cannot appear too much, cause discharge of stable electric arc extremely easily otherwise, because this cannot regard its as completely harmful discharge. So the article makes analyse the time scale that unusual discharge condition appears inside periodic time to be in: 100% 　 of × of A)/tT of ′ of T of ∑ of E+ of ′ of T of Y=(η ∑ 　 　 　 (6) 　 among them η is one power worth, 0.

1 　 of ＜ of 5 ＜ η takes optimal electric spark scale of unusual discharge time is Yd in form cutting, 　 of 　 of 　 of 　 of your Δ Y=Y-Yd 　 (analyse of 7) 　 devoir is periodic relatively before in analysing cycle, appear proportional change is time of unusual discharge condition: 　 of 　 of 　 of Y(t-1) 　 of Δ of Δ Y(t)- of D(Δ Y)/dt= (8) 　 analyses in state institute begins to occupy time scale to be inside cycle: 100% 　 of × of Q= ∑ Td/tT 　 　 　 (9) 　 takes institute of state of the open a way in form cutting of optimal electric spark to occupy time scale to be Qd, 　 of 　 of 　 of your Δ Q=Q-Qd 　 (analyse of 10) 　 devoir is periodic relatively before institute of state of the open a way in analysing cycle takes time proportional change is: 　 of 　 of 　 of 　 of Q(t-1) 　 of Δ of Δ Q(t)- of D(Δ Q)/dt= (11) 　 such, after each analysis cycle ends, the cost that passes Q)/dt of Δ of Q of Y)/dt of Δ of computational Δ Y, D(, Δ , D(, but the clearance discharge condition of form cutting of differentiate electric spark and its change trend. 3 faintness control strategy to tool electrode, when Δ Y ＞ 0 when, appeared to cross much harmful discharge state namely, controller should be given out to main shaft time cancel an order, increase discharge space in order to shun work by the possibility of burn, inspect Δ Y to be deviation right now, d(Δ Y)/dt changes for deviation; When Q ＞ 0 when, appeared to cross much open circuit state namely, controller should issue feed order to main shaft, reduce discharge gap in order to improve productivity and discharge efficiency, inspect Δ Q to be deviation right now, d(Δ Q)/dt changes for deviation. The VS of rotate speed deviation that chooses main shaft is control quantity, the process that can simplify the control of condition of discharge of clearance of form cutting of whole electric spark to be outputted for a 4 inputs so. Control a principle according to faintness, will input sendout model gelatinization above all. Be worth a description with 4 languages to Q of Δ Y, Δ , use 7 languages to be worth a description to Q)/dt of Δ of Y)/dt of Δ of VS, D(, D(, subject function is taken for triangular function form, if pursue,2 are shown. Graph 2 inputs, output is variable and subject spend function sketch map NB, negative big NM, the NS in losing, negative small Z, 0 PS, small PM, in the middle of PB, negative big faintness controls regulation to be as follows: (1) is like Δ Y for 0, and Δ Q is 0, criterion VS of servo output voltage is 0; (2) is like Δ Y small, and D(Δ Y)/dt big, criterion VS of servo output voltage is lost big; (3) is like Δ Y small, and in the middle of D(Δ Y)/dt, criterion VS of servo output voltage is lost big; (4) is like Δ Y small, and D(Δ Y)/dt small, criterion VS of servo output voltage is lost in; (5) is like Δ Y small, and D(Δ Y)/dt is 0, criterion VS of servo output voltage is lost small; (6) is like Δ Y small, and D(Δ Y)/dt is lost small, criterion VS of servo output voltage is lost small; (7) is like Δ Y small, and D(Δ Y)/dt is lost in, criterion VS of servo output voltage is lost small; (8) is like Δ Y small, and D(Δ Y)/dt is lost big, criterion VS of servo output voltage is lost small; (9) if in the middle of Δ Y, and D(Δ Y)/dt big, criterion VS of servo output voltage is lost big; (10) if in the middle of Δ Y, and in the middle of D(Δ Y)/dt, criterion VS of servo output voltage is lost big; (11) if in the middle of Δ Y, and D(Δ Y)/dt small, criterion VS of servo output voltage is lost big; (12) if in the middle of Δ Y, and D(Δ Y)/dt is 0, criterion VS of servo output voltage is lost in; (13) if in the middle of Δ Y, and D(Δ Y)/dt is lost small, criterion VS of servo output voltage is lost in; (14) if in the middle of Δ Y, and D(Δ Y)/dt is lost in, criterion VS of servo output voltage is lost in; (15) if in the middle of Δ Y, and D(Δ Y)/dt is lost big, criterion VS of servo output voltage is lost in; (16) is like Δ Y big, and D(Δ Y)/dt big, criterion VS of servo output voltage is lost big; (17) is like Δ Y big, and in the middle of D(Δ Y)/dt, criterion VS of servo output voltage is lost big; (18) is like Δ Y big, and D(Δ Y)/dt small, criterion VS of servo output voltage is lost big; (19) is like Δ Y big, and D(Δ Y)/dt is 0, criterion VS of servo output voltage is lost big; (20) is like Δ Y big, and D(Δ Y)/dt is lost small, criterion VS of servo output voltage is lost big; (21) is like Δ Y big, and D(Δ Y)/dt is lost in, criterion VS of servo output voltage is lost big; (22) is like Δ Y big, and D(Δ Y)/dt is lost big, criterion VS of servo output voltage is lost big; (23) is like Δ Q small, and D(Δ Q)/dt big, criterion servo outputs voltage VS big; (24) is like Δ Q small, and in the middle of D(Δ Q)/dt, criterion servo outputs voltage VS big; Wait for 43 regulation. Regulation narrates before reaching by regulation of ambiguous gather operation 1: (In the type on 12) 　 " . " , " × " , " ∩ " express aggregate respectively synthesis, Cartesian indigestion, make 3 kinds of operation. But be in actually, Δ Y and Δ Q are calculate after that through be being measured actually computational certain numerical value, because this is not ambiguous, the grade number that their subject case numerical value measures at place besides correspondence is 1 beyond, the others is 0. Such type (the operation of 12) can simplify greatly. Might as well the Y that set Δ is I class (I=0, 1, 2... , 12) , the D(Δ Y)/dt that by Δ Y computation gives is J class (J=0, 1, 2... , 12) , Δ Q is M class (M=0, 1, 2... , 12) , the D(Δ Q)/dt that by Δ Q computation gives is N class (N=0, 1, 2... , 12) , criterion type (12) can simplify for: (Y(i) of Δ of Z of the μ in the type on 13) 　 is I of Y of Δ of ambiguous gather Z subject case of the element is numeric, q(m) of μ Z Δ is M of Q of Δ of ambiguous gather Z subject case of the element is numeric. Can beg with same method piece 2, 3, ... VS ～ 43. It is I class to Δ Y, d(Δ Y)/dt is J class, Δ Q is M class, the control quantity that D(Δ Q)/dt is N class should be: (14) 　 is an ambiguous gather, by the choice principle of subject function maximum or increase advantageous position average principle is attainable an affirmatory control measures VS* . Above operation can use the computer to finish conveniently, to value of different I, J, M, N, computation gives different VS* . Reach finally be based on VS*(13 of group of a 4 dimension, 13, 13, the faintness of 13) dominates a list. In applying actually, the basis detects and can get through computation I, J, M, N is worth, check a watch to get the value of VS*(i, J, M, N) , output after that, such circular move back and forth go down, realized the ambiguous control to process of electric spark form cutting thereby. 4 experiments test and verify uses this faintness to control strategy to undertake controlling to process of electric spark form cutting, tool electrode is black lead electrode, workpiece material is steel makings, strong fluid condition develops fluid to time, exterior surface roughness is 10 μ M, tool electrode appearance is wedge. Machine pump uses Nantong power especially mechanical limited company and institute of technology of numerical control of treatment of report of Nanjing aerospace university are newest the forming machine of numerical control electric spark of development. Through the experiment, reach if the graph is shown 3 times,the relation that machines deepness and handling time pursues. Graph 3 treatment deepness and handling time relation pursue by on the graph is visible, ambiguous control relatively ACC/ACO more agree with the control of process of electric spark form cutting, avoiding discharge of harmful electric spark effectively to be opposite workpiece burn while, make machine efficiency to get rising significantly. The ambiguous control method that 5 conclusion use article place to put forward undertakes controlling to the main shaft speed of electric spark forming machine, can obtain taller than method of traditional ACC/ACO control productivity, system rash club sex is better, performance is steadier, the ACC/ACO that can replace a tradition completely controls a method. CNC Milling CNC Machining