CN105259791A - Machining parameter optimization method based on general cutting energy consumption model - Google Patents

Abstract

The invention provides a machining parameter optimization method based on a general cutting energy consumption model and relates to a machining parameter optimization method, particularly a machining parameter optimization method in consideration of the cutting energy consumption of a machine tool. In the prior art, due to the adoption of a conventional cutting method that lays emphasis on the machining efficiency, the machine tool is too large in energy consumption. In order to solve the above problem, the method comprises the steps of firstly, analyzing the energy consumption characteristics during the milling, turning and drilling process, establishing a general cutting energy consumption model Ei=SECi*Vi+Pairi*delta tairi for each feed, determining the total cutting energy consumption during the milling, turning and drilling process, conducting the normalized treatment on a time function tw and an overall value of the cutting energy consumption during the part machining process to obtain the time function tw and the overall value of the cutting energy consumption during the part machining process after the normalized treatment, with the weighted sum of the time function tw and the overall value of the cutting energy consumption during the part machining process after the normalized treatment as an optimization objective, solving the optimization objective based on the improved genetic algorithm to obtain optimal cutting parameters. The above method is applied to the optimization of machining parameters.

Description

A kind of machining parameters optimization method based on general-cutting energy consumption model

Technical field

The present invention relates to a kind of machining parameters optimization method, be specifically related to a kind of machining parameters optimization method considering machine cut energy consumption.

Background technology

Cutting data in rational selection machining not only can improve the efficiency of cut, and effectively can reduce the power consumption of polymer processing of lathe, reduces machine tooling process further to the negative effect of environment.

At present, the research of the optimization aspect of machine cut consumption focuses mostly in the research of machine tooling efficiency, considers that the research of lathe energy consumption is relatively less.And in engineering practice, digital control processing personnel determine the cutting data of lathe in process according to the processing experience of reality usually, and these processing experiences are often from lathe producer and cutter Cheng Chan producer or determine according to the actual processing of oneself.The energy consumption that the guidance that such processing mode often lacks theory causes lathe work in-process to produce is too much.

In addition, also have some to focus on the cutting working method of machine cut energy consumption at present, and these consider the job operation of machine cut energy consumption often to have ignored the efficiency of machine tooling.

Summary of the invention

The problem that the lathe generation energy consumption that the present invention causes in order to the cutting working method solving existing emphasis working (machining) efficiency is too much.

Based on a machining parameters optimization method for general-cutting energy consumption model, comprise the following steps:

Energy consumption characteristics in step 1, analysis milling, turning, drilling process, set up the general-cutting energy consumption model of above-mentioned processing mode:

E＝SEC·V+P _air△t _air(1)

Each feed is then had

E _i＝SEC _i·V _i+P _airi·△t _airi(2)

Wherein, E cuts energy consumption; V is the volume removing material; P _airit is sky cutting power; △ t _airthat process time is cut in cut-in without ball; Footmark i is the sequence number of feed, E _i, SEC _i, V _i, P _airi, △ t _airibe respectively E, SEC, V, P that i-th feed is corresponding _air, t _air;

$SEC=\frac{P_{normal}}{MRR}=k_1\frac{n}{MRR}+k_2{\mathrm{MRR}}^{k_3}+k_4\frac{1}{MRR}---\left(3\right)$

Wherein, P _normalcut stage power; MRR is material removing rate; k ₁it is the constant coefficient that experiment obtains; N is the speed of mainshaft; k ₂it is power constant coefficient relevant with machine tool type in working angles; k ₃it is constant relevant with machine tool type in working angles; k ₄=P _standby+ P _fluid+ a is the constant coefficient in working angles; P _standbyit is lathe standby power; P _fluidit is cutting fluid consumed power; A is the power constant that experiment obtains;

P _air＝P _standby+P _fluid+k ₁n+a+k ₅f+b

F is feed rate, k ₅, b is feeding power of motor constant coefficient;

Step 2, determine to cut total energy consumption in milling, turning, drilling process according to formula (2):

Wherein, E _alwaysfor cutting total energy consumption; M is the feed number of times in process;

Step 3, set up the function of time t of part process _w, respectively by the function of time t of part process _wwith cutting total energy consumption E _alwaysbe normalized, obtain the function of time t of the part process after normalized _w ^*with cutting total energy consumption

Step 4, with the function of time t of the part process after normalized _w ^*with cutting total energy consumption weighted sum be optimization aim, adopt Revised genetic algorithum to solve optimization aim, obtain optimum cutting parameter.

The present invention has following beneficial effect:

The cutting parameter adopting the present invention to obtain carries out cut, has taken into account working (machining) efficiency and lathe energy consumption.Compare the cutting working method that existing is focused on working (machining) efficiency, the energy consumption that the cutting parameter adopting the present invention to obtain carries out cut generation can reduce about 20%.Compare the cutting working method that existing is focused on machine cut energy consumption, the working (machining) efficiency that the cutting parameter adopting the present invention to obtain carries out cut can improve about 37%.

Accompanying drawing explanation

Fig. 1 is process flow diagram of the present invention;

Fig. 2 is embodiment process power;

Fig. 3 is SEC and rotating speed, material removing rate graph of a relation;

Fig. 4 is the tax right-value optimization flow chart of steps of genetic algorithm;

Fig. 5 is championship-roulette wheel selection process figure; Wherein, Fig. 5-(a) is that initial population is individual, the population at individual of Fig. 5-(b) for algorithm of tournament selection and after eliminating, and Fig. 5-(c) is the population at individual after roulette selection;

Fig. 6 is target convergence curve map.

Embodiment

Embodiment one: composition graphs 1 illustrates present embodiment,

Based on a machining parameters optimization method for general-cutting energy consumption model, comprise the following steps:

Energy consumption characteristics in step 1, analysis milling, turning, drilling process, set up the general-cutting energy consumption model of above-mentioned processing mode:

E＝SEC·V+P _air△t _air(1)

Each feed is then had

E _i＝SEC _i·V _i+P _airi·△t _airi(2)

Wherein, E cuts energy consumption; V is the volume removing material; P _airit is sky cutting power; △ t _airthat process time is cut in cut-in without ball; Footmark i is the sequence number of feed, E _i, SEC _i, V _i, P _airi, △ t _airibe respectively E, SEC, V, P that i-th feed is corresponding _air, t _air;

$SEC=\frac{P_{normal}}{MRR}=k_1\frac{n}{MRR}+k_2{\mathrm{MRR}}^{k_3}+k_4\frac{1}{MRR}---\left(3\right)$

Wherein, P _normalcut stage power; MRR is material removing rate; k ₁it is the constant coefficient that experiment obtains; N is the speed of mainshaft; k ₂it is power constant coefficient relevant with machine tool type in working angles; k ₃it is constant relevant with machine tool type in working angles; k ₄=P _standby+ P _fluid+ a is the constant coefficient in working angles; P _standbyit is lathe standby power; P _fluidit is cutting fluid consumed power; A is the power constant that experiment obtains;

P _air＝P _standby+P _fluid+k ₁n+a+k ₅f+b

F is feed rate, k ₅, b is feeding power of motor constant coefficient;

Step 2, determine to cut total energy consumption in milling, turning, drilling process according to formula (2):

Wherein, E _alwaysfor cutting total energy consumption; M is the feed number of times in process;

Step 3, set up the function of time t of part process _w, respectively by the function of time t of part process _wwith cutting total energy consumption E _alwaysbe normalized, obtain the function of time t of the part process after normalized _w ^*with cutting total energy consumption

Step 4, with the function of time t of the part process after normalized _w ^*with cutting total energy consumption weighted sum be optimization aim, adopt Revised genetic algorithum to solve optimization aim, obtain optimum cutting parameter.

Embodiment two: the function of time t setting up part process described in present embodiment step 3 _was follows:

$t_w=\frac{\mathrm{\π}dV}{1000V_c{f}_t{\mathrm{Za}}_p{a}_e}+\frac{t_{ct}{{\mathrm{\πdVV}}_c}^{x-1}{a_p}^{y-1}{f_t}^{u-1}{a_e}^{w-1}{Z}^{q-1}}{1000C_T}+t_{ot}---\left(5\right)$

In formula: d is tool diameter; Z is number of teeth; t _ctit is the time that tool changing is once consumed; C _tcoefficient, relevant to workpiece material, machining condition and cutter itself; X, y, u, w, q are index, represent the influence degree of each milling usage to tool life respectively; t _otit is the non-cutting time outside exchanging knives process; V _cit is cutting speed; f _tit is feed engagement; a _pit is axial cutting depth; a _eit is the radial cutting degree of depth.

Other steps are identical with embodiment one with parameter.

Embodiment three: described in present embodiment step 3 by the function of time t of part process _wthe process be normalized is as follows:

${t_w}^{*}=\frac{t_w(V_c,f_t,a_p,a_e)-t_{w\min }}{t_{wmax}-t_{w\min }}---\left(6\right)$

In formula: t _w ^*for the function of time of the part process after normalized; t _wminand t _wmaxonly to the minimum and maximal value that process time is optimized.

Other steps are identical with embodiment one or two with parameter.

Embodiment four: total energy consumption E will be cut described in present embodiment step 3 _alwaysthe process be normalized is as follows:

In formula: for the cutting total energy consumption after normalized; E _{total min}and E _{total max}only to the minimum and maximal value that power consumption of polymer processing is optimized.

Other steps are identical with one of embodiment one to three with parameter.

Embodiment five: the optimization aim described in present embodiment step 4 is as follows:

In formula: w ₁and w ₂weight coefficient, and w ₁+ w ₂=1.

In the process adopting Revised genetic algorithum to solve optimization aim,

Random the selecting individuality of roulette selection strategy, is conducive to the randomness keeping population at individual, avoids genetic algorithm in optimizing process to occur locally optimal solution.Inferior solution in population at individual is eliminated by algorithm of tournament selection strategy, ensure that the optimization efficiency of genetic algorithm.For above two kinds of selection strategies, propose a kind of system of selection of championship-roulette, both can ensure the search efficiency of algorithm, the diversity of population can be improved again.First be by algorithm of tournament selection method, the inferior solution in population at individual is eliminated, the validity that the setting impact heredity of mortality is distributed in the process of eliminating.If mortality is too high, the major part individuality selected will be eliminated, and so genetic algorithm is just probably absorbed in local convergence; If the setting of mortality is too low, so the selection strategy of championship-roulette will be substantially identical with the selection strategy of roulette.

So adopt championship-roulette selection method to carry out population select operation, the concrete steps of championship-roulette selection method are as follows:

Step 4.1, set the quantity of population as h, calculate each kind of group unit C _jrelative adaptability degrees R _j, j={1,2 ..., h};

Step 4.2, relative adaptability degrees R according to each unit _jto C _jsort, setting mortality, eliminates the unit that relative adaptability degrees is low;

Step 4.3, [0, R] the interior random generation h group random number of scope, according to the method for roulette, be with the individuality of the corresponding relative adaptability degrees of random number produced the individuality chosen, wherein R be not eliminated in step 4.2 individual relative adaptability degrees and.

Embodiment:

For NC milling, with HaasVF-2 numerical control machining center for research object, to considering that the NC Machining Process of energy consumption carries out parameter optimization.Use the two meshing golden milling cutters (Φ 16mm) that cutter is TiN coating, workpiece material is 45# steel, and dimensions is 100mm × 55mm × 40mm.Optimizing process adopts the processing technology of slabbing, and milling depth is 5mm, and material removal amount is 27500mm ³.

Based on a machining parameters optimization method for general-cutting energy consumption model, comprise the following steps:

1) analyze energy consumption characteristics in milling, turning and drilling process, set up the general-cutting energy consumption model of above-mentioned three kinds of processing modes:

As shown in Figure 2, the powertrace of numerically-controlled machine process reflects lathe and changes in the energy consumption in different process segments.When lathe is in holding state, because the power of lathe tends towards stability, performance number remains unchanged substantially, therefore can think that the standby power of lathe is fixed value.Lathe fast feed from holding state has reached the Working position of specifying, and now main shaft does not also start to rotate, and due to the fast feed campaign of the feed shaft of lathe, machine power can produce larger pulse.After this, machine tool chief axis starts to rotate, and main shaft rotates suddenly and can produce output pulses, and specify after rotating speed when main shaft reaches subsequently, numerically-controlled machine is in free-runing operation state, and the size of the changed power now caused because of main axis is relevant with the rotating speed of main shaft.After cutting fluid system is started working, there is certain change and tend towards stability in power, therefore the change of power that cutting fluid system work causes can be considered as fixed value.When lathe enter cut-in without ball cut the stage time, the feed motion of feed shaft can cause the change of machine power, because the power of the servomotor of feed system is lower, thus cuts the change of stage machine power and not obvious in the cut-in without ball of lathe.Lathe, in working angles, can cause the change of main shaft cutting force during Tool in Cutting workpiece, increase the load of spindle motor, powertrace now outstanding obvious.After machine tooling terminates, get back to holding state.Can find out that lathe changes five parts in the energy consumption that the energy consumption in the stage of cutting is caused by standby energy consumption, main shaft idle running energy consumption, cutting fluid system energy consumption, feed system energy consumption and cut substantially and forms from the analysis of machine power curve.

Energy consumption in machine cut process can be divided into two parts, and a part is fixing energy consumption, and another part changes with the change of machine cut situation, meets following relational expression

P＝P _idle+k·MRR(9-1)

In formula: MRR---machine tooling material removing rate,

P---the power in working angles,

P _idle---idle capacity,

K---the empirical parameter relevant with cutting parameter and machine tool type.

The people such as Gutowski think that the power with material removing rate change in machine tooling process is all proportional with material removing rate, and namely k is constant.But the cut of reality shows, k is not constant, and it is relevant with cutting parameter and machine tool type, and its relation is as follows.

$k=B_0{\mathrm{MRR}}^{B_1}---(9-2)$

In formula: B0, B1---the parameter relevant with machine tool type.

As shown in Figure 2, cut before lathe idle capacity P _idlebe made up of standby power, main shaft idle capacity and cutting fluid consumed power, computing formula is as follows:

P _idle＝P _standby+P _spindle+P _fluid(9-3)

In formula: P _idle---idle capacity,

P _standby---lathe standby power,

P _spindle---main shaft idle capacity,

P _fluid---cutting fluid consumed power.

In numerically-controlled machine process, consider feed power P _feedproportion shared in machine cut process is very little, therefore can not consider feed power in the modeling process in machine cut stage.By analysis before, the standby power P of lathe _standbywith cutting fluid system consumed power P _fluidcan be considered fixed value.Actual processing experience shows, main shaft of numerical control machine tool idle capacity P _spindleand there is direct relation between the speed of mainshaft and machine tool lubrication situation, machine tool chief axis idle capacity P _spindleand be approximately linear relation between the speed of mainshaft, by as shown in the formula calculating.

P _spindle＝k ₁n+a(9-4)

In formula: n---the speed of mainshaft,

K ₁---the constant coefficient that experiment obtains,

A---the power constant that experiment obtains.

Main shaft idle capacity formula (9-4) is substituted into (9-3) obtain

P _idle＝P _standby+k ₁n+a+P _fluid(9-5)

The power module of lathe in working angles can be obtained by formula (9-1), (9-2) and (9-5)

$P_{normal}=P_{s\tan dby}+P_{fluid}+k_{1}n+a+k_2{\mathrm{MRR}}^{k_3+1}---(9-6)$

In formula: k ₂---power constant coefficient relevant with machine tool type in working angles,

K ₃---constant relevant with machine tool type in working angles.

Above formula is with obtaining divided by MRR

$SEC=\frac{P_{normal}}{MRR}=k_1\frac{n}{MRR}+k_4\frac{1}{MRR}+k_2{\mathrm{MRR}}^{k_3}---(9-7)$

In formula: k ₄=P _standby+ P _fluid+ a---the constant coefficient in working angles.

Therefore, the energy consumption in working angles can by following formulae discovery

E _normal＝SEC·V(9-8)

In formula: E _normal---the energy consumption in working angles,

V---remove the volume of material.

For empty working angles, the power of empty working angles can be made up of four parts, comprises standby power, main shaft idle capacity, cutting fluid system power and feed system power.As shown in Figure 2, stage power model is cut in the cut-in without ball can set up thus.

P _air＝P _standby+P _fluid+P _spindle+P _feed(9-9)

P _feed＝k ₅f+b(9-10)

In formula: f---feed rate,

K ₅, b---feeding power of motor constant coefficient.

By can the have leisure model of cutting power of formula (9-9) and formula (9-10) be then

P _air＝P _standby+P _fluid+k ₁n+a+k ₅f+b(9-11)

P _air＝k ₁n+k ₅f+c(9-12)

In formula: c=P _standby+ P _fluid+ a+b---the constant term in empty working angles.

Energy consumption calculation then in empty working angles is as follows

E _air＝P _air△t _air(9-13)

In formula: E _air---empty working angles energy consumption,

△ t _air---process time is cut in cut-in without ball.

It is as follows by the energy consumption formulas in stage can be cut above

E＝E _normal+E _air＝SEC·V+P _air△t _air(9-14)

It should be noted that above-mentioned energy consumption model is applicable to turning, milling and drilling process through experimental verification, the present invention, for NC milling, is described in detail to the optimization method proposed.The energy consumption modeling of this example adopts 4 factor 3 level, 27 groups of orthogonal determination machined parameters.Scope and the level of cutting parameter are as shown in table 1.Stochastic choice wherein 22 groups of experimental datas is used for energy consumption modeling, and other 5 groups of experimental datas carry out Accuracy Verification to energy consumption model.Numerical control milling experiment energy consumption is as shown in table 2.

The scope of table 1 cutting parameter and level

Table 2 numerical control milling experiment energy consumption data

Carry out matching according to energy consumption modeling experiment data to obtain numerical control milling SEC model and cut-in without ball to cut power module as follows:

$SEC=3.485\frac{n}{MRR}+951.6\frac{1}{MRR}+38.15{\mathrm{MRR}}^{-0.5493}$

P _air＝3.2436n+0.933f+988.5

SEC and rotating speed, material removing rate relational model are as shown in Figure 3.

2) the general-cutting energy consumption model set up is verified, ensure the forecasting accuracy of energy consumption model.The energy consumption testing experimental data in cutting stage and accuracy as shown in table 3.Experimental result shows, the accuracy of numerical control milling energy consumption model is more than 97%.

Table 3 cuts stage energy consumption checking

3) objective function and the constraint of numerical control processing technology parameter optimization is set up according to general energy consumption model.

In NC milling, cutting speed V is mainly to machine tooling energy consumption and the maximum parameter of machine tooling time effects _c, feed engagement f _t, axial cutting-in a _pand radial cutting-in a _e, therefore select these four groups of cutting datas to be optimized variable.

According to restrictive condition and the experience value of HaasVF2 numerical control machining center and machine tool, the span of four parameters is

$s.t.\left\{\begin{array}{c}30m/min\≤V_c\≤100m/min\\ 0.025mm/t\≤f_t\≤0.075mm/t\\ 4mm\≤a_e\≤12mm\\ 0.5mm\≤a_p\≤1.5mm\end{array}\right.$

After determining the span of cutting data, before to machine tooling process optimization, need to encode to cutting data.Binary coding mode is adopted to encode to four kinds of cutting datas, in cataloged procedure, each optimized variable can arrange the binary code of certain numerical digit, then is carried out by the binary code of four optimized variables splicing gene required in composition heredity, forms " chromosome " string.The available following genomic constitution of solution produced at random represents:

Assuming that cutting speed V in optimizing _cscope be [V _cmin, V _cmax], there is binary code S _acorresponding a bit x ₁, then there is following relational expression:

$V_c=V_{cmin}+\frac{x_1}{2^a-1}(V_{cmax}-V_{cmin})---(9-15)$

Assuming that the f of feed engagement in optimizing _tscope be [f _tmin, f _tmax], use binary code S _bthe binary number x of corresponding b position ₂, then there is following relational expression:

$f_t=f_{tmin}+\frac{x_2}{2^b-1}(f_{tmax}-f_{tmin})---(9-16)$

Assuming that axial cutting-in a when optimizing _pscope be [a _pmin, a _pmax], there is binary code S _ccorresponding c bit x ₃, then there is following relational expression:

$a_p=a_{pmin}+\frac{x_3}{2^c-1}(a_{pmax}-a_{pmin})---(9-17)$

In like manner, assuming that radial cutting-in a when optimizing _escope be [a _emin, a _emax], there is binary code S _dcorresponding d bit x ₄, then there is following relational expression:

$a_e=a_{emin}+\frac{x_4}{2^d-1}(a_{emax}-a_{emin})---(9-18)$

Cutting speed V _c, feed engagement f _t, axial cutting-in a _pand radial cutting-in a _ethe population at individual chromosome of these four optimized variables also can be expressed as S _as _bs _cs _d.

Energy consumption of the present invention mainly comprises sky working angles energy consumption and actual cut process energy consumption.The energy consumption model obtaining digital control processing according to energy consumption modeling is before

In formula: m is the feed number of times in process.

Material removing rate MRR, speed of mainshaft n and speed of feed f in formula and the relation of milling usage are respectively

$MRR=a_p{a}_{e}f=\frac{1000v_c{f}_t{\mathrm{Za}}_p{a}_e}{\mathrm{\π}d}---(9-20)$

$n=\frac{V_c\·1000}{\mathrm{\π}d}---(9-21)$

f＝nf _tZ(9-22)

Most high efficiency is that the number of parts of the minimal time consumed or the processing of production of units time processing each part is weighed at most.Time of a certain operation in part process comprises machine cut process time, non-cutting time and tool change time.If the value of the cutting data selected is larger, in process, the wearing and tearing of cutter will be more serious, and the life-span of cutter can be more and more less, will cause like this causing frequent tool changing in process.The tool change time that exchanging knives process brings and extra tool setting time can affect the average processing time of part, therefore, also cutter life are taken into account when carrying out modeling to process time.The model of process time can be obtained thus as follows

$t_w=t_m+t_{ct}\frac{t_m}{T}+t_{ot}---(9-23)$

In formula: t _m---the cutting time of operation;

T _ct---the time that tool changing is once consumed;

T---tool life;

T _m/ T---number of changing knife;

T _ot---the non-cutting time outside exchanging knives process.

The cutting time of operation is

$t_m=\frac{V}{MRR}=\frac{\mathrm{\π}dV}{1000v_c{f}_t{\mathrm{Za}}_p{a}_e}---(9-24)$

Tool life formula is

$T=\frac{C_T}{{V_c}^x{a_p}^y{f_z}^u{a_e}^w{Z}^q}---(9-25)$

In formula: C _t---coefficient is relevant to workpiece material, machining condition and cutter itself;

X, y, u, w, q---index, represents the influence degree of each milling usage to tool life respectively.

The function of time can being derived part process by above relational expression is

$t_w=\frac{\mathrm{\π}dV}{1000V_c{f}_t{\mathrm{Za}}_p{a}_e}+\frac{t_{ct}{{\mathrm{\πdVV}}_c}^{x-1}{a_p}^{y-1}{f}_t^{u-1}{a_e}^{w-1}{Z}^{m-1}}{1000C_T}+t_{ot}---(9-26)$

In formula: d is tool diameter, Z is number of teeth, t _ctthe time that tool changing is once consumed, C _tbe coefficient, relevant to workpiece material, machining condition and cutter itself, x, y, u, w, q are index, represent the influence degree of each milling usage to tool life respectively, t _otit is the non-cutting time outside exchanging knives process.

When solving multi-objective problem, several target realizes optimum being difficult to often simultaneously, and different optimization aim has different dimensions and meaning, the energy consumption of the machine tooling process that such as the present invention optimizes and time.In this case, changing that mostly to be a kind of skill conventional in optimizing less, is single-object problem by multi-objective optimization question by rational method migration.Tax weights method is this problem of solution is a kind of method often used.Optimization aim of the present invention is that the energy consumption of lathe process is the shortest for minimum and process time, and concrete grammar is weighted sum again after energy consumption and time being normalized, and value after another weighted sum is minimum.Corresponding single object optimization function is

In formula: w ₁and w ₂weight coefficient, and w ₁+ w ₂=1.V _ccutting speed, f _tfeed engagement, a _pcutting depth, a _ecutting depth, t _w ^*be through the function of time of normalized, be through the power dissipation obj ectives function of normalized.

Due to function of time t _wdifferent from power dissipation obj ectives function E dimension, both can not carry out summation operation, can to two model normalizeds, and concrete disposal route is as follows

${t_w}^{*}=\frac{t_w(V_c,f_t,a_p,a_e)-t_{w\min }}{t_{wmax}-t_{w\min }}---(9-28)$

In formula: t _w ^*for the function of time of the part process after normalized; t _wminand t _wmaxonly to the minimum and maximal value that process time is optimized. for the cutting total energy consumption after normalized; E _{total min}and E _{total max}only to the minimum and maximal value that power consumption of polymer processing is optimized.

Single object optimization function model after normalization is

4) with the function of time t of the part process after normalized _w ^*with cutting total energy consumption weighted sum be optimization aim, adopt Revised genetic algorithum to solve optimization aim, obtain optimum cutting parameter.

In optimizing process, the iteration optimizing carrying out repeatedly according to the operation steps of genetic algorithm until when meeting termination condition, as shown in Figure 4, is the tax right-value optimization step of genetic algorithm.

Random the selecting individuality of roulette selection strategy, is conducive to the randomness keeping population at individual, avoids genetic algorithm in optimizing process to occur locally optimal solution.Inferior solution in population at individual is eliminated by algorithm of tournament selection strategy, ensure that the optimization efficiency of genetic algorithm.For above two kinds of selection strategies, propose a kind of system of selection of championship-roulette, both can ensure the search efficiency of algorithm, the diversity of population can be improved again.First be by algorithm of tournament selection method, the inferior solution in population at individual is eliminated, the validity that the setting impact heredity of mortality is distributed in the process of eliminating.If mortality is too high, the major part individuality selected will be eliminated, and so genetic algorithm is just probably absorbed in local convergence; If the setting of mortality is too low, so the selection strategy of championship-roulette will be substantially identical with the selection strategy of roulette.

So adopt championship-roulette selection method to carry out population select operation, the concrete steps of championship-roulette selection method are as follows:

Step 4.1, set the quantity of population as h, calculate each kind of group unit C _jrelative adaptability degrees R _j, j={1,2 ..., h};

Step 4.2, relative adaptability degrees R according to each unit _jto C _jsort, setting mortality, eliminates the unit that relative adaptability degrees is low;

Step 4.3, [0, R] the interior random generation h group random number of scope, according to the method for roulette, be with the individuality of the corresponding relative adaptability degrees of random number produced the individuality chosen, wherein R be not eliminated in step 4.2 individual relative adaptability degrees and.

The scale supposing initial population is 7, obtains the relative adaptability degrees R of each kind of group unit _i, as shown in Fig. 5-(a); After sorting according to the size of relative adaptability degrees, certain mortality is set, as 25%, then can eliminates C ₂and C ₅these two individualities, as shown in Fig. 5-(b); Finally, adopt roulette selection method to select remaining individuality, as shown in Fig. 5-(c), new population at individual is C ₆, C ₆, C ₃, C ₃, C ₇, C ₄, C ₄and C ₁.

5) cutting parameter obtained after adopting optimization carries out cut to workpiece.

In tax right-value optimization process, the size choosing population is 30.Wherein, the binary-coded figure place of cutting speed, feed engagement, axial cutting depth and the radial cutting degree of depth is respectively 4,4,3 and 3,14 altogether.The evolutionary generation of genetic algorithm is chosen as 100, and the crossing-over rate of population is set to 0.8, and aberration rate is set to 0.05.In the selection strategy of championship-roulette, the superseded probability of championship is 25%, and the change of scale factor scale in change of scale function is 0.01.Under MATLAB environment, as shown in table 4 to the optimum results of machine tooling process, objective function converges curve is (time weighting 0.5) as shown in Figure 6.When time weighting is 0, the energy consumption namely in an optimizing machining technology process, wherein minimum process energy consumption is 600.58kJ.When time weighting is 1, be namely only optimized the time of process, wherein minimum time is 1154.11s, compares 1845.86s process time of minimum process energy consumption, decreases about 700s, significantly shorten the time of process.

Right-value optimization result composed by table 4

Claims (5)

1., based on a machining parameters optimization method for general-cutting energy consumption model, it is characterized in that it comprises the following steps:

Energy consumption characteristics in step 1, analysis milling, turning, drilling process, set up the general-cutting energy consumption model of above-mentioned processing mode:

E＝SEC·V+P _air△t _air(1)

Each feed is then had

E _i＝SEC _i·V _i+P _airi·△t _airi(2)

Wherein, E cuts energy consumption; V is the volume removing material; P _airit is sky cutting power; △ t _airthat process time is cut in cut-in without ball; Footmark i is the sequence number of feed, E _i, SEC _i, V _i, P _airi, △ t _airibe respectively E, SEC, V, P that i-th feed is corresponding _air, t _air;

$SEC=\frac{P_{normal}}{MRR}=k_1\frac{n}{MRR}+k_2{\mathrm{MRR}}^{k_3}+k_4\frac{1}{MRR}---\left(3\right)$

Wherein, P _normalcut stage power; MRR is material removing rate; k ₁it is the constant coefficient that experiment obtains; N is the speed of mainshaft; k ₂it is power constant coefficient relevant with machine tool type in working angles; k ₃it is constant relevant with machine tool type in working angles; k ₄=P _standby+ P _fluid+ a is the constant coefficient in working angles; P _standbyit is lathe standby power; P _fluidit is cutting fluid consumed power; A is the power constant that experiment obtains;

P _air＝P _standby+P _fluid+k ₁n+a+k ₅f+b

F is feed rate, k ₅, b is feeding power of motor constant coefficient;

Step 2, determine to cut total energy consumption in milling, turning, drilling process according to formula (2):

Wherein, E _alwaysfor cutting total energy consumption; M is the feed number of times in process;

Step 3, set up the function of time t of part process _w, respectively by the function of time t of part process _wwith cutting total energy consumption E _alwaysbe normalized, obtain the function of time t of the part process after normalized _w ^*with cutting total energy consumption

Step 4, with the function of time t of the part process after normalized _w ^*with cutting total energy consumption weighted sum be optimization aim, adopt Revised genetic algorithum to solve optimization aim, obtain optimum cutting parameter.

2. a kind of machining parameters optimization method based on general-cutting energy consumption model according to claim 1, is characterized in that the function of time t setting up part process described in step 3 _was follows:

$t_w=\frac{\mathrm{\π}dV}{1000V_c{f}_t{\mathrm{Za}}_p{a}_e}+\frac{t_{ct}{{\mathrm{\πdVV}}_c}^{x-1}{a_p}^{y-1}{f_t}^{u-1}{a_e}^{w-1}{Z}^{q-1}}{1000C_T}+t_{ot}---\left(5\right)$

In formula, d is tool diameter; Z is number of teeth; t _ctit is the time that tool changing is once consumed; C _tcoefficient, relevant to workpiece material, machining condition and cutter itself; X, y, u, w, q are index, represent the influence degree of each milling usage to tool life respectively; t _otit is the non-cutting time outside exchanging knives process; V _cit is cutting speed; f _tit is feed engagement; a _pit is axial cutting depth; a _eit is the radial cutting degree of depth.

3. a kind of machining parameters optimization method based on general-cutting energy consumption model according to claim 2, it is characterized in that described in step 3 by the function of time t of part process _wthe process be normalized is as follows:

${t_w}^{*}=\frac{t_w(V_c,f_t,a_p,a_e)-t_{wmin}}{t_{wmax}-t_{wmin}}---\left(6\right)$

In formula, t _w ^*for the function of time of the part process after normalized; t _wminand t _wmaxonly to the minimum and maximal value that process time is optimized.

4. a kind of machining parameters optimization method based on general-cutting energy consumption model according to claim 3, is characterized in that cutting total energy consumption E described in step 3 _alwaysthe process be normalized is as follows:

In formula, for the cutting total energy consumption after normalized; E _{total min}and E _{total max}only to the minimum and maximal value that power consumption of polymer processing is optimized.

5. a kind of machining parameters optimization method based on general-cutting energy consumption model according to claim 4, is characterized in that the optimization aim described in step 4 is as follows:

In formula, w ₁and w ₂weight coefficient, and w ₁+ w ₂=1.