CN103699945A - Method and device for extracting different-scale production performance signal of RMS (Reconfigurable Manufacturing System) - Google Patents

Abstract

The invention provides a method and a device for extracting different-scale production performance signals of an RMS (Reconfigurable Manufacturing System), belonging to the field of manufacturing techniques. The method comprises the steps of obtaining the production performance signals of the RMS; processing the production performance signals to obtain a function about scale factors and shift factors; obtaining boundary points on the scale factors in wavelet analysis of each scale in the RMS; respectively reconfiguring the scale factors in ranges corresponding to each scale to obtain the production performance change signals of the RMS on each scale. By adopting the technical scheme provided by the invention, the reconfigured production performance signals of the RMS on different scales can be extracted.

Description

Extract method and the device of Reconfigurable Manufacturing System different scale production performance signal

Technical field

The present invention relates to manufacturing technology field, refer to especially a kind of method and device that extracts Reconfigurable Manufacturing System different scale production performance signal.

Background technology

Manufacturing system has the architecture of level, Reconfigurable Manufacturing System (RMS) as Advanced Manufacturing System has same architectural feature, as shown in Figure 1, according to the object of analyzing and the difference of granularity, that RMS is divided into is system-level, cell level and lathe level, is illustrated in figure 2 the scale effect schematic diagram of reconstruct.

This architectural feature of RMS, has also determined that the reconstruct of RMS on space scale can be divided into the reconstruct of system-level reconstruct, cell level and the reconstruct of lathe level.Impact and effect that reconstruct on different scale produces RMS are not identical: the reconstruct of carrying out in large scale is larger on the impact of RMS, corresponding, the reconstruct of carrying out in small scale is less on the impact of RMS, the amplitude and the frequency that are reflected in production performance the fluctuation that is exactly production performance can exist obvious difference, but between the production performance signal of reconstruct on different scale, exist again complicated coupled relation, on different scale, the coupling of production performance signal has formed the final production performance curve of RMS.

At present the research of RMS is confined to more division of device layout, manufacturing cell etc., what the object of research tends to solve more is reconstruct and the problem of reconstruct how, seldom considers the Issues On Multi-scales of RMS.The Issues On Multi-scales of RMS is the importance of RMS theory development, for the essence that discloses RMS, the reconstruct efficiency and the quality that improve RMS have profound significance, and need to the production performance signal on each yardstick be extracted during performance on research RMS different scale, otherwise the multiple dimensioned research of RMS is also just had no way of doing it.

Summary of the invention

The technical problem to be solved in the present invention is to provide a kind of method and device that extracts Reconfigurable Manufacturing System different scale production performance signal, can extract the production performance signal of RMS reconstruct on different scale.

For solving the problems of the technologies described above, embodiments of the invention provide technical scheme as follows:

On the one hand, provide a kind of method of extracting Reconfigurable Manufacturing System different scale production performance signal, comprising:

Obtain the production performance signal of Reconfigurable Manufacturing System;

Described production performance signal is processed to obtain a function about scale factor and shift factor;

Obtain the separation of each yardstick in scale factor;

Scale factor is reconstructed respectively in scope corresponding to each yardstick, obtains the production performance variable signal on each yardstick.

Further, describedly described production performance signal processed to obtain described production performance signal about the function of time and yardstick, comprise:

Described production performance signal in Life cycle is carried out to continuous wavelet transform, obtain the function about scale factor and shift factor

$W_v(a,b)=\frac{1}{\sqrt{a}}{\∫}_{-\∞}^{\∞}v\left(t\right)\stackrel{\‾}{\mathrm{\ψ}\left(\frac{t-b}{a}\right)}\mathrm{dt},a>0$

Wherein, v(t) be described production performance signal, a is scale factor, and b is shift factor.

Further, described each yardstick comprises cell level reconstruct, the reconstruct of lathe level and system-level reconstruct, described in obtain the separation of each yardstick in scale factor and comprise:

Difference acquiring unit level reconstruct and the separation a of lathe level reconstruct on wavelet analysis yardstick _cand a _r.

Further, described acquiring unit level reconstruct and the separation a of lathe level reconstruct on wavelet analysis yardstick respectively _cand a _rcomprise:

Obtain respectively the longest time interval T that carries out adjacent twice cell level reconstruct and the reconstruct of lathe level _cand T _r, on frequency domain, the separation of cell level reconstruct and the reconstruct of lathe level is respectively ω _c=π/T _cand ω _r=π/T _r;

According to frequency and scale factor, be related to a=ω _w/ ω _aΔ, obtains cell level reconstruct and the separation of lathe level reconstruct on wavelet analysis yardstick is respectively a _c=ω _w/ ω _cΔ and a _r=ω _w/ ω _rΔ.

Further, described scale factor is reconstructed respectively in scope corresponding to each yardstick, the production performance variable signal obtaining on each yardstick comprises:

By scale factor a at (a _c,+∞) on be reconstructed, obtain the production performance variable signal of system-level reconstruct

$\mathrm{vs}\left(t\right)=C_{\mathrm{\ψ}}{\∫}_0^{\∞}{\∫}_{a_c}^{\∞}\frac{1}{a^2}{W}_v(a,b)\mathrm{\ψ}\left(\frac{t-b}{a}\right)\mathrm{dadb};$

By scale factor a at (a _r, a _c) on be reconstructed, obtain the production performance variable signal of cell level reconstruct

$\mathrm{vc}\left(t\right)=C_{\mathrm{\ψ}}{\∫}_0^{\∞}{\∫}_{a_r}^{a_c}\frac{1}{a^2}{W}_v(a,b)\mathrm{\ψ}\left(\frac{t-b}{a}\right)\mathrm{dadb};$

By scale factor a at (0, a _r) on be reconstructed, obtain the production performance variable signal of lathe level reconstruct

$\mathrm{vr}\left(t\right)=C_{\mathrm{\ψ}}{\∫}_0^{\∞}{\∫}_0^{a_r}\frac{1}{a^2}{W}_v(a,b)\mathrm{\ψ}\left(\frac{t-b}{a}\right)\mathrm{dadb}.$

The embodiment of the present invention also provides a kind of device that extracts Reconfigurable Manufacturing System different scale production performance signal, comprising:

Acquisition module, for obtaining the production performance signal of Reconfigurable Manufacturing System;

Processing module, for processing to obtain a function about scale factor and shift factor to described production performance signal;

Computing module, for obtaining the separation of each yardstick in scale factor;

Reconstructed module, for scale factor is reconstructed respectively in scope corresponding to each yardstick, obtains the production performance variable signal on each yardstick.

Further, described processing module comprises:

Converter unit, carries out continuous wavelet transform for the described production performance signal in Life cycle, obtains the function about scale factor and shift factor;

$W_v(a,b)=\frac{1}{\sqrt{a}}{\∫}_{-\∞}^{\∞}v\left(t\right)\stackrel{\‾}{\mathrm{\ψ}\left(\frac{t-b}{a}\right)}\mathrm{dt},a>0$

Wherein, v(t) be described production performance signal, a is scale factor, and b is shift factor.

Further, described each yardstick comprises cell level reconstruct, the reconstruct of lathe level and system-level reconstruct,

Described computing module is specifically for difference acquiring unit level reconstruct and the separation a of lathe level reconstruct on wavelet analysis yardstick _cand a _r.

Further, described computing module comprises:

The first computing unit, for obtaining respectively the longest time interval T that carries out adjacent twice cell level reconstruct and the reconstruct of lathe level _cand T _r, on frequency domain, the separation of cell level reconstruct and the reconstruct of lathe level is respectively ω _c=π/T _cand ω _r=π/T _r;

The second computing unit, for being related to a=ω according to frequency and scale factor _w/ ω _aΔ, obtains cell level reconstruct and the separation of lathe level reconstruct on wavelet analysis yardstick is respectively a _c=ω _w/ ω _cΔ and a _r=ω _w/ ω _rΔ.

Further, described reconstructed module comprises:

The first reconfiguration unit, for by scale factor a at (a _c,+∞) on be reconstructed, obtain the production performance variable signal of system-level reconstruct

$\mathrm{vs}\left(t\right)=C_{\mathrm{\ψ}}{\∫}_0^{\∞}{\∫}_{a_c}^{\∞}\frac{1}{a^2}{W}_v(a,b)\mathrm{\ψ}\left(\frac{t-b}{a}\right)\mathrm{dadb};$

The second reconfiguration unit, for by scale factor a at (a _r, a _c) on be reconstructed, obtain the production performance variable signal of cell level reconstruct

$\mathrm{vc}\left(t\right)=C_{\mathrm{\ψ}}{\∫}_0^{\∞}{\∫}_{a_r}^{a_c}\frac{1}{a^2}{W}_v(a,b)\mathrm{\ψ}\left(\frac{t-b}{a}\right)\mathrm{dadb};$

Reconstructed unit, for by scale factor a at (0, a _r) on be reconstructed, obtain the production performance variable signal of lathe level reconstruct

$\mathrm{vr}\left(t\right)=C_{\mathrm{\ψ}}{\∫}_0^{\∞}{\∫}_0^{a_r}\frac{1}{a^2}{W}_v(a,b)\mathrm{\ψ}\left(\frac{t-b}{a}\right)\mathrm{dadb}.$

Embodiments of the invention have following beneficial effect:

In such scheme, production performance signal is processed to obtain a function about scale factor and shift factor, obtain the separation of each yardstick in scale factor, scale factor is reconstructed respectively in scope corresponding to each yardstick, obtains the production performance variable signal on each yardstick.By technical scheme of the present invention, can extract the production performance variable signal of RMS reconstruct on different scale, for the advantage of optimization RMS, performance RMS and the efficiency of lifting RMS, there is very important meaning.

Accompanying drawing explanation

Fig. 1 is the architectural schematic of Reconfigurable Manufacturing System;

Fig. 2 is the scale effect schematic diagram of reconstruct;

Fig. 3 is the method flow schematic diagram that the embodiment of the present invention is extracted Reconfigurable Manufacturing System different scale production performance signal;

Fig. 4 is the process effect schematic diagram of RMS;

Fig. 5 is the trend of production performance signals spectrogram and the distribution schematic diagram of each yardstick reconstruct;

Fig. 6 is the time domain waveform figure of Mexican Hat (mexh) small echo;

Fig. 7 is the frequency-domain waveform figure of Mexican Hat (mexh) small echo;

Fig. 8 is reconstruct on each yardstick scope schematic diagram in scale factor and frequency.

Embodiment

For technical matters, technical scheme and advantage that embodiments of the invention will be solved are clearer, be described in detail below in conjunction with the accompanying drawings and the specific embodiments.

Embodiments of the invention provide a kind of method and device that extracts Reconfigurable Manufacturing System different scale production performance signal, can extract the production performance signal of RMS reconstruct on different scale.

Fig. 3 is the method flow schematic diagram that the embodiment of the present invention is extracted Reconfigurable Manufacturing System different scale production performance signal, and as shown in Figure 3, the present embodiment comprises:

Step 101: the production performance signal that obtains Reconfigurable Manufacturing System;

Step 102: described production performance signal is processed to obtain a function about scale factor and shift factor;

Step 103: obtain the separation of each yardstick in scale factor;

Step 104: scale factor is reconstructed respectively in scope corresponding to each yardstick, obtains the production performance variable signal on each yardstick.

The embodiment of the present invention processes to obtain a function about scale factor and shift factor to production performance signal, obtain the separation of each yardstick in scale factor, scale factor is reconstructed respectively in scope corresponding to each yardstick, obtains the production performance variable signal on each yardstick.By technical scheme of the present invention, can extract the production performance variable signal of RMS reconstruct on different scale, for the advantage of optimization RMS, performance RMS and the efficiency of lifting RMS, there is very important meaning.

Further, another embodiment of the present invention, comprises on the basis of above-mentioned steps 101-104, describedly described production performance signal is processed to obtain a function about scale factor and shift factor comprises:

Described production performance signal in Life cycle is carried out to continuous wavelet transform, obtains the function about scale factor and shift factor:

$W_v(a,b)=\frac{1}{\sqrt{a}}{\∫}_{-\∞}^{\∞}v\left(t\right)\stackrel{\‾}{\mathrm{\ψ}\left(\frac{t-b}{a}\right)}\mathrm{dt},a>0$

Wherein, v(t) be described production performance signal, a is scale factor, and b is shift factor.

Further, another embodiment of the present invention, comprises on the basis of above-mentioned steps 101-104, described each yardstick comprises cell level reconstruct, the reconstruct of lathe level and system-level reconstruct, described in obtain the separation of each yardstick in scale factor and comprise:

Difference acquiring unit level reconstruct and the separation a of lathe level reconstruct on wavelet analysis yardstick _cand a _r.

Particularly, described acquiring unit level reconstruct and the separation a of lathe level reconstruct on wavelet analysis yardstick respectively _cand a _rcomprise:

Obtain respectively the longest time interval T that carries out adjacent twice cell level reconstruct and the reconstruct of lathe level _cand T _r, on frequency domain, the separation of cell level reconstruct and the reconstruct of lathe level is respectively ω _c=π/T _cand ω _r=π/T _r;

According to frequency and scale factor, be related to a=ω _w/ ω _aΔ, obtains cell level reconstruct and the separation of lathe level reconstruct on wavelet analysis yardstick is respectively a _c=ω _w/ ω _cΔ and a _r=ω _w/ ω _rΔ.

Further, another embodiment of the present invention, comprises on the basis of above-mentioned steps 101-104, described scale factor is reconstructed respectively in scope corresponding to each yardstick, and the production performance variable signal obtaining on each yardstick comprises:

By scale factor a at (a _c,+∞) on be reconstructed, obtain the production performance variable signal of system-level reconstruct

$\mathrm{vs}\left(t\right)=C_{\mathrm{\ψ}}{\∫}_0^{\∞}{\∫}_{a_c}^{\∞}\frac{1}{a^2}{W}_v(a,b)\mathrm{\ψ}\left(\frac{t-b}{a}\right)\mathrm{dadb};$

By scale factor a at (a _r, a _c) on be reconstructed, obtain the production performance variable signal of cell level reconstruct

$\mathrm{vc}\left(t\right)=C_{\mathrm{\ψ}}{\∫}_0^{\∞}{\∫}_{a_r}^{a_c}\frac{1}{a^2}{W}_v(a,b)\mathrm{\ψ}\left(\frac{t-b}{a}\right)\mathrm{dadb};$

By scale factor a at (0, a _r) on be reconstructed, obtain the production performance variable signal of lathe level reconstruct

$\mathrm{vr}\left(t\right)=C_{\mathrm{\ψ}}{\∫}_0^{\∞}{\∫}_0^{a_r}\frac{1}{a^2}{W}_v(a,b)\mathrm{\ψ}\left(\frac{t-b}{a}\right)\mathrm{dadb}.$

The embodiment of the present invention also provides a kind of device that extracts Reconfigurable Manufacturing System different scale production performance signal, comprising:

Acquisition module, for obtaining a production performance signals of Reconfigurable Manufacturing System;

Processing module, for processing to obtain a function about scale factor and shift factor to described production performance signal;

Computing module, for obtaining the separation of each yardstick in scale factor;

Reconstructed module, for scale factor is reconstructed respectively in scope corresponding to each yardstick, obtains the production performance variable signal on each yardstick.

Device of the present invention processes to obtain a function about scale factor and shift factor to production performance signal, obtain the separation of each yardstick in scale factor, scale factor is reconstructed respectively in scope corresponding to each yardstick, obtains the production performance variable signal on each yardstick.By technical scheme of the present invention, can extract the production performance variable signal of RMS reconstruct on different scale, for the advantage of optimization RMS, performance RMS and the efficiency of lifting RMS, there is very important meaning.

Further, described processing module comprises:

Converter unit, carries out continuous wavelet transform for the described production performance signal in Life cycle, obtains a function about scale factor and shift factor;

$W_v(a,b)=\frac{1}{\sqrt{a}}{\∫}_{-\∞}^{\∞}v\left(t\right)\stackrel{\‾}{\mathrm{\ψ}\left(\frac{t-b}{a}\right)}\mathrm{dt},a>0$

Wherein, v(t) be described production performance signal, a is scale factor, and b is shift factor.

Further, described each yardstick comprises cell level reconstruct, the reconstruct of lathe level and system-level reconstruct,

Described computing module is specifically for difference acquiring unit level reconstruct and the separation a of lathe level reconstruct on wavelet analysis yardstick _cand a _r.

Further, described computing module comprises:

The first computing unit, for obtaining respectively the longest time interval T that carries out adjacent twice cell level reconstruct and the reconstruct of lathe level _cand T _r, on frequency domain, the separation of cell level reconstruct and the reconstruct of lathe level is respectively ω _c=π/T _cand ω _r=π/T _r;

The second computing unit, for being related to a=ω according to frequency and scale factor _w/ ω _aΔ, obtains cell level reconstruct and the separation of lathe level reconstruct on wavelet analysis yardstick is respectively a _c=ω _w/ ω _cΔ and a _r=ω _w/ ω _rΔ.

Further, described reconstructed module comprises:

The first reconfiguration unit, for by scale factor a at (a _c,+∞) on be reconstructed, obtain the production performance variable signal of system-level reconstruct

$\mathrm{vs}\left(t\right)=C_{\mathrm{\ψ}}{\∫}_0^{\∞}{\∫}_{a_c}^{\∞}\frac{1}{a^2}{W}_v(a,b)\mathrm{\ψ}\left(\frac{t-b}{a}\right)\mathrm{dadb};$

The second reconfiguration unit, for by scale factor a at (a _r, a _c) on be reconstructed, obtain the production performance variable signal of cell level reconstruct

$\mathrm{vc}\left(t\right)=C_{\mathrm{\ψ}}{\∫}_0^{\∞}{\∫}_{a_r}^{a_c}\frac{1}{a^2}{W}_v(a,b)\mathrm{\ψ}\left(\frac{t-b}{a}\right)\mathrm{dadb};$

Reconstructed unit, for by scale factor a at (0, a _r) on be reconstructed, obtain the production performance variable signal of lathe level reconstruct

$\mathrm{vr}\left(t\right)=C_{\mathrm{\ψ}}{\∫}_0^{\∞}{\∫}_0^{a_r}\frac{1}{a^2}{W}_v(a,b)\mathrm{\ψ}\left(\frac{t-b}{a}\right)\mathrm{dadb}.$

Below in conjunction with accompanying drawing and specific embodiment, the method for extraction Reconfigurable Manufacturing System different scale production performance signal of the present invention is described in detail:

A life cycle that defines RMS in prior art is: the product in Yi Ge product family is produced the period of experiencing from starting to produce stopping completely.For traditional manufacturing system, in a production cycle, there is the phenomenon of overcapacity or inefficiency of production.For Reconfigurable Manufacturing System, in its each life cycle, the variation of RMS production performance all can meet to economy external demand, this is that variation by manufacturing system configuration realizes, and this advantage of RMS has just been determined at the beginning of the design of RMS, this is also the embodiment of RMS productive capacity flexibility.Production performance refers to productive capacity, crudy, economy of RMS etc. and can characterize rmc system and show fine or not index, generally by productive capacity, represents production performance.

As shown in Figure 4, Reconfigurable Manufacturing System can produce several process effects in its operational process, and the characteristic of this and RMS is undivided.These process effects are respectively: bird tail effect, oblique ascension effect, stochastic effects, mutation effect and deteriorating effect.Oblique ascension effect refers to manufacturing system after newly-built and reconstruct, and system performance index reaches design planning target and is stabilized in rising and the transient process this horizontal line from starting to rise to; Deteriorating effect refers to the phenomenon declining because of inside and outside reason system performance appearance in the lifetime later stage of reconstruct, and the appearance that general deteriorating effect and performance can not be mated demand phenomenon is the signal that need to manufacturing system is reconstructed and be adjusted; Bird tail effect refers to the phenomenon of the of short duration performance off-design desired value occurring in the process putting into effect at manufacturing system complete design, now needs manufacturing system to meet and produce actual adjustment, so that oblique ascension effect appears in its performance; Stochastic effects and mutation effect are because the enchancement factor occurring in manufacturing system operational process produces, and the generation of stochastic effects is that equipment failure tends to produce the mutation effect of production performance because of the fluctuation of equipment, personnel's state.

Above-mentioned these effects and artificial design and reconstruct acting in conjunction add the driving of external demand simultaneously, have finally formed the variation figure of RMS production performance as shown in Figure 4.

Might as well carry out reasonable assumption to following two conditions: 1. the production performance of RMS does not exist stochastic effects and mutation effect; 2. in RMS, causing production performance is the reconstruct of lathe itself compared with the reason of minor swing, and the reason that production performance moderate changes is that the reason that production performance significantly changes is the reconstruct that whole RMS is carried out to the reconstruct in unit, workshop (workshop layout).Under such hypotheses, can think that production performance signal waveform amplitude part less, that frequency is higher is to be produced by the reconstruct of lathe level, wave-shape amplitude part large, that frequency is very little is that the reconstruct by whole rmc system produces.The decomposition that production performance signal in a life cycle is carried out on frequency domain obtains the amplitude frequency spectrum figure of signal, thereby can analyze more intuitively the frequency distribution of production performance signal.

If temporarily the stopping of enterprise's activity in production while not considering reconstruct, the production performance of manufacturing system is continually varying, can be considered as continuous time signal.In time domain, production performance can be expressed as the function v (t) of time t.If continuous time signal v (t) is the voltage at resistance R two ends, average power is:

$P=\underset{a\→\∞}{\lim }\frac{1}{2a}{\∫}_{-a}^a\frac{v^2\left(t\right)}{R}\mathrm{dt}$ Formula 1

When resistance value is 1 ohm, average power is now only relevant with signal, for:

$P=\underset{a\→\∞}{\lim }\frac{1}{2a}{\∫}_{-a}^a{v}^2\left(t\right)\mathrm{dt}$ Formula 2

The energy of signal is:

$E={\&Integral;}_{-\&infin;}^{\&infin;}{v}^2\left(t\right)\mathrm{<figure-callout\; id="3"\; label="dt\; Formula"\; filenames="HDA0000451625560000041.png"\; state="\{\{state\}\}">dt</figure-callout>}$ Formula 3

If the energy of a signal is limited, claim that this signal is energy signal (energy signal), if the power of a signal is limited, claim that this signal is power signal (power signal).Production cycle is finite duration signal, and therefore, in a life cycle of RMS, the production performance signal energy of manufacturing system is limited, is typical energy signal.If think that desirable RMS has desirable reusability, regard its Life cycle as unlimited and continue, in the Life cycle of desirable RMS, its production performance signal is unlimited duration signal so.Obviously, the limit

restrain, the Infinite Energy of RMS and power limited, therefore the production performance signal under Life cycle is a power signal.This character of production performance signal meets the adequate condition of Fourier transform and wavelet transformation.

In comparatively ideal situation, manufacturer generally should follow the principle of " selecting bad reconstruct ", when needing reconstruct reconstruct Object Selection be on production line, show poor, as bottleneck operation or the poor corresponding configuration of operation of crudy.When RMS is reconstructed, generally speaking, with regard to the frequency of reconstruct, the frequency of reconstruct constantly reduces along with the increase of reconstruct yardstick from the system-level reconstruct of being reconfigured to of lathe level; With regard to the effect of reconstruct, reconstruct also meets such rule to the amplitude of whole manufacturing system performance change.

Number of times to the adjustment of lathe and reconstruct is more, frequency is larger, be reflected in the effect of reconstruct is to adjust in thin sight aspect the production performance of RMS, to the influence time of production performance, also shorter (after reconstruct, the performance of lathe also meets the process effect of manufacturing system, later stage exists deteriorated, such as the wearing and tearing of cutter).With this, be reflected on production performance signal, can think that the production performance that the reconstruct of lathe level causes changes the back segment that is positioned at signal spectrum figure.

The reconstruct that unit is carried out, arranges and carries out re-configuration the lathe in workshop.The reconstruct of centering sight aspect is only carried out when being necessary, and such as adjusting greatly the performance of certain productive unit, or the technique of certain parts of production product has carried out significantly changing etc.The reconstruct of middle sight yardstick is longer to the influence time of production performance, affects amplitude larger, is reflected on production performance signal, and the production performance that the reconstruct of cell level causes changes the stage casing that is positioned at signal spectrum figure.

The system-level reconfiguration frequency that whole system is carried out is very low, a life cycle indegree seldom.General variation that can not meet external demand in the production of inter-product family or existing RMS configuration, the economy being reconstructed is higher---cost being reconstructed is less than and continues to produce produced opportunity cost under this configuration---, and in time, is reconstructed.The cost that whole system configuration is reconstructed is the highest, and frequency is extremely low, but reconstruct all can produce huge and long-term impact to the productive capacity of RMS and production function later.Be reflected on production performance signal, the production performance that system-level reconstruct causes changes the leading portion that is positioned at signal spectrum figure.

The above analysis, on the frequency spectrum of production performance curve, roughly has distribution as shown in Figure 5: what HFS was corresponding is the reconstruct of lathe level, and what intermediate-frequency section was corresponding is the reconstruct of cell level, and what low frequency part was corresponding is the reconstruct of rmc system level.Using frequency as benchmark is as separatrix, can determine with following method the separation of different scale reconstruct:

Be located in the reconstruct of a rmc system level, run through a repeatedly cell level reconstruct of life cycle needs, a cell level reconstruct comprises repeatedly lathe level reconstruct.For the product of a certain type, in a life cycle, be T the longest interval time of carrying out adjacent twice effective cell level reconstruct _c, be T the longest interval time of carrying out adjacent twice effective lathe level reconstruct _r.Analyze the figure of production performance signal, the fluctuation that a reconstruct of RMS produces is corresponding to the half period of sinusoidal signal, interval time T to correspond on frequency domain be ω=π/T.The reconstruct of each yardstick of guestimate corresponds on frequency domain, and its separation is respectively ω _c=π/T _cand ω _r=π/T _r.

When carrying out wavelet transformation, the selection of wavelet function is that wavelet analysis is applied to the difficulties in reality, is also a problem to be optimized.The general method adopting is to use different small echos by test, different results to be contrasted and made a choice, or carry out choosing of wavelet function according to the similarity of signal to be analyzed and wavelet function, consider the parameter such as vanishing moment, proper string, bearing length of small echo.In the embodiment of the present invention, according to the similarity of wavelet function and signal to be analyzed, use mexican hat wavelet (Mexican Hat) to carry out wavelet transformation to production performance signal.Mexican hat wavelet is the second derivative of Gaussian function, that is:

$\mathrm{\&psi;}\left(t\right)=(1-t^2)e^{-t^2/2}$ Formula 4

Mexican hat wavelet function has good localization in time domain and frequency domain, and meets:

${\&Integral;}_{-\&infin;}^{\&infin;}\mathrm{\&psi;}\left(t\right)\mathrm{dx}=0$ Formula 5

Its Fourier transform is:

$\widehat{\mathrm{\&psi;}}\left(\mathrm{\&omega;}\right)=\sqrt{2\mathrm{\&pi;}}{\mathrm{\&omega;}}^2{e}^{-{\mathrm{\&omega;}}^2/2}$ Formula 6

As Fig. 6 and Fig. 7 are respectively the time domain of Mexican Hat (mexh) small echo and the waveform schematic diagram of frequency domain.Mexican hat wavelet function finite energy, belongs to square-integrable real number space L ²(R).Obviously, Mexican hat wavelet function meets admissible condition (Admissible Condition):

$C_{\mathrm{\&psi;}}={\&Integral;}_R\frac{{\left|\widehat{\mathrm{\&psi;}}\left(\mathrm{\&omega;}\right)\right|}^2}{\mathrm{\&omega;}}\mathrm{d\&omega;}<\&infin;$ Formula 7

Be that sombrero function can be called a wavelet or female small echo (Mother Wavelet).

Right

, the continuous wavelet transform of f (t) (sometimes also referred to as integration wavelet transformation) is defined as:

$C{\mathrm{WT}}_f(a,b)=\frac{1}{\sqrt{a}}{\&Integral;}_{-\&infin;}^{\&infin;}f\left(t\right)\stackrel{\&OverBar;}{\mathrm{\&psi;}\left(\frac{t-b}{a}\right)}\mathrm{dt},a>0$ Formula 8

With frequency domain, be expressed as equivalently:

${\mathrm{CWT}}_f(a,b)=\frac{\sqrt{a}}{2\mathrm{\&pi;}}{\&Integral;}_{-\&infin;}^{\&infin;}F\left(\mathrm{\&omega;}\right)\stackrel{\&OverBar;}{\mathrm{\&psi;}\left(\mathrm{a\&omega;}\right)}{e}^{\mathrm{j\&omega;b}}\mathrm{d\&omega;}$ Formula 9

F (ω) wherein, ψ (ω) is respectively f (t), the Fourier transform of ψ (t).

By defining above, be not difficult to find out that wavelet transformation is the same with Fourier transform, be also a kind of integral transformation, WT _f(a, b) is wavelet conversion coefficient.The place that it is different from Fourier variation is that the function variable after wavelet transformation is scale factor a and shift factor b.So function through wavelet transformation, just means, a function of time is projected to when two-dimentional in m-yardstick phase plane, the third dimension represents the size of wavelet conversion coefficient, is conducive to like this extract some essential characteristic of signal function.

Production performance signal in Life cycle is carried out to continuous wavelet transform, obtains production performance signal about the function of time and yardstick:

$W_v(a,b)=\frac{1}{\sqrt{a}}{\&Integral;}_{-\&infin;}^{\&infin;}v\left(t\right)\stackrel{\&OverBar;}{\mathrm{\&psi;}\left(\frac{t-b}{a}\right)}\mathrm{dt},a>0$ Formula 10

If converted back time domain, be inversely transformed into:

$v\left(t\right)=C_{\mathrm{\&psi;}}{\&Integral;}_{R^{+}}{\&Integral;}_R\frac{1}{a^2}{W}_v(a,b)\mathrm{\&psi;}\left(\frac{t-b}{a}\right)\mathrm{dadb}$ Formula 11

In continuous wavelet transform, parameter scale factor a and shift factor b are continually varyings, and what shift factor b represented is the translation of wavelet function time, has the dimension of time; What scale factor a represented is the convergent-divergent of wavelet function.Obviously, scale factor a is less less wave period, and frequency is higher; A is larger less little wave period, and frequency is lower.Therefore scale factor a and frequency have a kind of fixing relation.

Scale factor and the shift factor of value are very inconvenient in wavelet analysis process continuously, have therefore just produced sampling thheorem, process, thereby with ordered series of numbers, signal is represented signal is carried out to certain discretize.In wavelet theory, sampling thheorem explanation is as long as sampling interval is enough little, and continuous time signal just can be by its sampling Exact recovery so.Along with sampling is more and more closeer, the error between sampled signal and original signal will be more and more less, and when sampling interval is tending towards infinitesimal, this error just trends towards zero.Sampling refers to multiplies each other production performance signal v (t) and an impulse function p (t), and sampled signal is exactly both products,

V _s(t)=v (t) p (t) formula 12

The product of time domain is to equaling the convolution of frequency domain:

V _s(ω)=V (ω) * P (ω) formula 13

Impulse function p (t) does Fourier transform and obtains the one-period impulse sequence on frequency domain, and the sampling rate of periodic sequence is ω _s, the maximal value of the frequency spectrum medium frequency of original signal is ω _m.From sampling thheorem, the necessary condition that original signal can be recovered completely by sampled signal is: sampling rate must not be less than 2 ω _m.2 ω _mthe minimum sample rate of sampled signal, the nyquist sampling rate that is otherwise known as (Nyquist rate).

For the purpose of convenient, when production performance signal is sampled, sampling rate gets the peaked twice of original signal frequency spectrum medium frequency, i.e. ω _s=2 ω _m.The general hits (hits/second) of rs representation unit in the time of using, Δ represents sampling time interval, Δ=1/r _s, ω again _s=2 π/T=2 π r _s, i.e. sampling time interval Δ=T=π/ω _m.

While mentioning, between m-yardstick, exist fixing relation above, this relation is relevant with the centre frequency of the sampling period of signal and the wavelet basis function of selection.Use ω _wthe centre frequency that represents Mexican hat wavelet function, the frequency that scale factor a is corresponding is so ω _a:

${\mathrm{\&omega;}}_a=\frac{{\mathrm{\&omega;}}_w}{\mathrm{a\&Delta;}}$ Formula 14

Equally, according to frequency, can obtain its corresponding scale factor, have:

$a=\frac{{\mathrm{\&omega;}}_w}{{\mathrm{\&omega;}}_a\mathrm{\&Delta;}}$ Formula 15

The frequency of the centre frequency of wavelet function and pure periodic signal (sinusoidal signal) is comparable, so the centre frequency ω of wavelet basis function _wcan be obtained by Fourier transform, the centre frequency of Mexican hat wavelet function approximates the frequency that reaches peak swing value on its Fourier transform frequency spectrum, can obtain thus the funtcional relationship of scale factor and respective frequencies.

According to above-mentioned principle, in order to solve the problem of extracting production performance signal on Reconfigurable Manufacturing System different scale in restructuring procedure, the embodiment of the present invention proposes a kind of method of extracting production performance signal on Reconfigurable Manufacturing System different scale, specifically comprises the following steps:

Step 1, obtain the production performance signal of Reconfigurable Manufacturing System;

Step 2, this production performance signal is done to the frequency spectrum function F that Fourier transform obtains production performance signal _nor spectral density function F (ω), thereby obtain spectrogram;

Step 3, obtain the longest time interval T that carries out cell level reconstruct and lathe level reconstruct (must be effective) respectively _cand T _r, on frequency domain, the separation of each yardstick reconstruct is respectively ω _c=π/T _cand ω _r=π/T _r;

Step 4, production performance signal is carried out to wavelet transformation, obtain the function W about scale factor a and shift factor b _v(a, b);

Step 5, according to frequency and scale factor, be related to a=ω _w/ ω _aΔ, obtains cell level reconstruct and the separation of lathe level reconstruct on wavelet analysis yardstick is respectively a _c=ω _w/ ω _cΔ and a _r=ω _w/ ω _rΔ.

Step 6, to the waveform after the wavelet analysis in Life cycle, within the scope of different scale, scale factor a belongs to respectively (a _c,+∞), (a _r, a _c) and (0, a _r) time be reconstructed (integration), obtain the production performance variable signal of reconstruct on each yardstick.

Since then, just completed the process of extracting production performance signal on Reconfigurable Manufacturing System different scale.Production performance signal take below as v (t) is example, in conjunction with concrete formula explanation, extract the process of Reconfigurable Manufacturing System different scale production performance signal:

1, in Life cycle, production performance signal v (t) is carried out to continuous wavelet transform, obtains it about the function of time and yardstick:

$W_v(a,b)=\frac{1}{\sqrt{a}}{\&Integral;}_{-\&infin;}^{\&infin;}v\left(t\right)\stackrel{\&OverBar;}{\mathrm{\&psi;}\left(\frac{t-b}{a}\right)}\mathrm{dt},a>0$ Formula 16

2, for the product of a certain type, in a life cycle, be T the longest interval time of carrying out adjacent twice effective cell level reconstruct _c, be T the longest interval time of carrying out adjacent twice effective lathe level reconstruct _r.Analyze the figure of production performance signal, the fluctuation that a reconstruct of RMS produces is corresponding to the half period of sinusoidal signal, interval time T to correspond on frequency domain be ω=π/T.The reconstruct of each yardstick of guestimate corresponds on frequency domain, and its separation is respectively ω _c=π/T _cand ω _r=π/T _r.

3, due to time exist fixing relation between m-yardstick, this relation is relevant with the centre frequency of the sampling period of signal and the wavelet basis function of selection.Use ω _wthe centre frequency that represents Mexican hat wavelet function, the frequency that scale factor a is corresponding is so ω _a:

${\mathrm{\&omega;}}_a=\frac{{\mathrm{\&omega;}}_w}{\mathrm{a\&Delta;}}$ Formula 17

Equally, according to frequency, can obtain its corresponding scale factor, have:

$a=\frac{{\mathrm{\&omega;}}_w}{{\mathrm{\&omega;}}_a\mathrm{\&Delta;}}$ Formula 18

The frequency of the centre frequency of wavelet function and pure periodic signal (sinusoidal signal) is comparable, so the centre frequency ω of wavelet basis function _wcan be obtained by Fourier transform, the centre frequency of Mexican hat wavelet function approximates the frequency that reaches peak swing value on its Fourier transform frequency spectrum.Obtain thus cell level reconstruct and the separation of lathe level reconstruct on wavelet analysis yardstick is respectively a _c=ω _w/ ω _cΔ and a _r=ω _w/ ω _rΔ, is illustrated in figure 8 reconstruct on each yardstick scope schematic diagram in scale factor and frequency.

4, by scale factor a respectively at (a _c,+∞), (a _r, a _c) and (0, a _r) on be reconstructed (being integration), obtain the production performance variable signal of reconstruct on each yardstick,

System-level:

$\mathrm{vs}\left(t\right)=C_{\mathrm{\&psi;}}{\&Integral;}_0^{\&infin;}{\&Integral;}_{a_c}^{\&infin;}\frac{1}{a^2}{W}_v(a,b)\mathrm{\&psi;}\left(\frac{t-b}{a}\right)\mathrm{dadb}$ Formula 19

Cell level:

$\mathrm{vc}\left(t\right)=C_{\mathrm{\&psi;}}{\&Integral;}_0^{\&infin;}{\&Integral;}_{a_r}^{a_c}\frac{1}{a^2}{W}_v(a,b)\mathrm{\&psi;}\left(\frac{t-b}{a}\right)\mathrm{dadb}$ Formula 20

Lathe level:

$\mathrm{vr}\left(t\right)=C_{\mathrm{\&psi;}}{\&Integral;}_0^{\&infin;}{\&Integral;}_0^{a_r}\frac{1}{a^2}{W}_v(a,b)\mathrm{\&psi;}\left(\frac{t-b}{a}\right)\mathrm{dadb}$ Formula 21

Obtain thus the production performance signal of RMS reconstruct on different scale.

The present invention is directed to the multiple dimensioned characteristic of RMS, set up and described the multiple dimensioned mathematical model of RMS.With Fourier transform and wavelet transformation, the production problem in manufacturing is converted into mathematical problem, gather production performance signal as input signal, to input signal, utilize wavelet analysis to carry out wavelet transformation to it, according to (the frequency of reconstruct of reconstruct on different scale, the factors such as influence degree of reconstruct to production performance) difference, utilize scale factor in wavelet analysis and the relation of frequency to quantize the yardstick of RMS reconstruct, provided the yardstick of RMS on mathematical border.Utilize wavelet analysis to be reconstructed on different scale signal, obtain the production performance signal on Reconfigurable Manufacturing System different scale.

The production performance signal that extracts reconstruct on Reconfigurable Manufacturing System different scale has following meaning to RMS:

(1) can on the impact of production performance, independently analyze the reconstruct on each yardstick.On different scale, between production performance signal, exist complicated coupled relation, formed the final production performance curve of RMS.But the reconstruct on each yardstick is different on the impact of production performance, this coupled relation to be broken, the production performance signal extracting on different scale is particularly important to the further multiscale analysis of RMS.

(2) can obtain and analyze the frequency of reconstruct and the historical data of amplitude on each yardstick, for the reconstruct after Reconfigurable Manufacturing System provides reference and foundation.

(3) the production performance signal of analyzing reconstruct on different scale, for optimizing RMS, is brought into play the advantage of RMS, and the efficiency that promotes RMS has very important meaning.

The above is the preferred embodiment of the present invention; it should be pointed out that for those skilled in the art, do not departing under the prerequisite of principle of the present invention; can also make some improvements and modifications, these improvements and modifications also should be considered as protection scope of the present invention.

Claims (10)

1. a method of extracting Reconfigurable Manufacturing System different scale production performance signal, is characterized in that, comprising:

Obtain the production performance signal of Reconfigurable Manufacturing System;

Described production performance signal is processed to obtain described production performance signal about the function of time and yardstick:

Obtain the separation of each yardstick in scale factor;

Scale factor is reconstructed respectively in scope corresponding to each yardstick, obtains the production performance variable signal on each yardstick.

2. the method for extraction Reconfigurable Manufacturing System different scale production performance signal according to claim 1, is characterized in that, describedly described production performance signal is processed to obtain described production performance signal about the function of time and yardstick, comprises:

Described production performance signal in Life cycle is carried out to continuous wavelet transform, obtains the function about scale factor and shift factor:

Wherein, v(t) be described production performance signal, a is scale factor, and b is shift factor.

3. the method for extraction Reconfigurable Manufacturing System different scale production performance signal according to claim 2, it is characterized in that, described each yardstick comprises cell level reconstruct, the reconstruct of lathe level and system-level reconstruct, described in obtain the separation of each yardstick in scale factor and comprise:

Difference acquiring unit level reconstruct and separation ac and the ar of lathe level reconstruct on wavelet analysis yardstick.

4. the method for extraction Reconfigurable Manufacturing System different scale production performance signal according to claim 3, is characterized in that, described acquiring unit level reconstruct and the separation a of lathe level reconstruct on wavelet analysis yardstick respectively _cand a _rcomprise:

Obtain respectively the longest time interval T that carries out adjacent twice cell level reconstruct and the reconstruct of lathe level _cand T _r, on frequency domain, the separation of cell level reconstruct and the reconstruct of lathe level is respectively ω _c=π/T _cand ω _r=π/T _r;

According to frequency and scale factor, be related to a=ω _w/ ω _aΔ, obtains cell level reconstruct and the separation of lathe level reconstruct on wavelet analysis yardstick is respectively a _c=ω _w/ ω _cΔ and a _r=ω _w/ ω _rΔ.

5. the method for extraction Reconfigurable Manufacturing System different scale production performance signal according to claim 3, it is characterized in that, described scale factor is reconstructed respectively in scope corresponding to each yardstick, the production performance variable signal obtaining on each yardstick comprises:

By scale factor a at (a _c,+∞) on be reconstructed, obtain the production performance variable signal of system-level reconstruct

By scale factor a at (a _r, a _c) on be reconstructed, obtain the production performance variable signal of cell level reconstruct

By scale factor a at (0, a _r) on be reconstructed, obtain the production performance variable signal of lathe level reconstruct

6. a device that extracts Reconfigurable Manufacturing System different scale production performance signal, is characterized in that, comprising:

Acquisition module, for obtaining the production performance signal of Reconfigurable Manufacturing System;

Processing module, for described production performance signal is processed to obtain a function about scale factor and shift factor:

Computing module, for obtaining the separation of each yardstick in scale factor;

Reconstructed module, for scale factor is reconstructed respectively in scope corresponding to each yardstick, obtains the production performance variable signal on each yardstick.

7. the device of extraction Reconfigurable Manufacturing System different scale production performance signal according to claim 6, is characterized in that, described processing module comprises:

Converter unit, carries out continuous wavelet transform for the described production performance signal in Life cycle, obtains the function about scale factor and shift factor

Wherein, v(t) be described production performance signal, a is scale factor, and b is shift factor.

8. the device of extraction Reconfigurable Manufacturing System different scale production performance signal according to claim 7, is characterized in that, described each yardstick comprises cell level reconstruct, the reconstruct of lathe level and system-level reconstruct,

Described computing module is specifically for difference acquiring unit level reconstruct and the separation a of lathe level reconstruct on wavelet analysis yardstick _cand a _r.

9. the device of extraction Reconfigurable Manufacturing System different scale production performance signal according to claim 8, is characterized in that, described computing module comprises:

The first computing unit, for obtaining respectively the longest time interval T that carries out adjacent twice cell level reconstruct and the reconstruct of lathe level _cand T _r, on frequency domain, the separation of cell level reconstruct and the reconstruct of lathe level is respectively ω _c=π/T _cand ω _r=π/T _r;

The second computing unit, for being related to a=ω according to frequency and scale factor _w/ ω _aΔ, obtains cell level reconstruct and the separation of lathe level reconstruct on wavelet analysis yardstick is respectively a _c=ω _w/ ω _cΔ and a _r=ω _w/ ω _rΔ.

10. the device of extraction Reconfigurable Manufacturing System different scale production performance signal according to claim 8, is characterized in that, described reconstructed module comprises:

The first reconfiguration unit, for by scale factor a at (a _c,+∞) on be reconstructed, obtain the production performance variable signal of system-level reconstruct

The second reconfiguration unit, for by scale factor a at (a _r, a _c) on be reconstructed, obtain the production performance variable signal of cell level reconstruct

Reconstructed unit, for by scale factor a at (0, a _r) on be reconstructed, obtain the production performance variable signal of lathe level reconstruct

Images