CN101436036A - Method and device for estimating processing capacity of process - Google Patents

Abstract

The invention discloses a method for estimating process capability of a technical process, which comprises the following steps: according to a predetermined standard, determining an upper standard limit USL and a lower standard limit LSL of the technical process; according to actual application of the technical process, determining deviation sigma and a target value T of the technical process; detecting and obtaining an actual mean value mu of the technical process; according to control requirement, selecting one of a first model and a second model as an estimation model for estimating the process capability, wherein for T not positioned at a middle position of the LSL and the USL in the first model and the second model, obtained estimation values of the process capability are proportionately symmetrical at two sides of the T; then according to the selected model, using the obtained parameters USL, LSL, sigma, T and mu to calculate theoretic process capability ratio Cp and actual process capability ratio Cpk; and according to the obtained Cp and Cpk, controlling or regulating the process capability.

Description

The method of estimation of processing capacity of process and device

Technical field

The present invention relates to the control technology of technological process, more particularly, relate to the method for estimation and the device of processing capacity of process.

Background technology

For the enterprise of type of production,, need to pay close attention to the quality and the ruuning situation of its production process and final products such as manufacturing business.For a long time, theoretical working ability rate Cp and actual working ability rate Cpk are used to represent the working ability of technological process, therefore, theoretical working ability rate Cp and actual working ability rate Cpk are used to measure the index that a technological process or its product meet the ability of its preassigned.

For the calculating processing ability, promptly theoretical working ability rate Cp and actual working ability rate Cpk usually calculate by the formula shown in following.At present, the overwhelming majority is used to estimate the instrument of working ability, all is to utilize these formula to calculate such as commercial statistics software.

For theoretical working ability rate Cp:

$\mathrm{Cp}=\frac{\mathrm{USL}-\mathrm{LSL}}{6\mathrm{\σ}}$

Wherein, USL is the upper limit of technological process preassigned, and LSL is the lower limit of technological process preassigned, and σ is the technological process deviation.

For actual working ability rate Cpk, calculate by two following formula:

Cpk ₁＝Cp(1-k) (1)

Wherein, $k=\frac{|T-\mathrm{\μ}|}{(\mathrm{USL}-\mathrm{LSL})/2},$ T is the desired value of technological process, and μ is the actual mean value of technological process, thereby parameter k is also referred to as discrepancy factor (off-target factor).

$\mathrm{Cp}{k}_2=\min (\mathrm{CpU},\mathrm{CpL})=\min (\frac{\mathrm{USL}-\mathrm{\μ}}{3\mathrm{\σ}},\frac{\mathrm{\μ}-\mathrm{LSL}}{3\mathrm{\σ}}),---\left(2\right)$

When the desired value T of technological process is just in time between technological process preassigned upper limit USL and lower limit LSL, USL-T=T-LSL, so, the Cpk as a result that formula (1) obtains ₁And the Cpk as a result that obtains of formula (2) ₂Arithmetically equating.

For having bilateral technological process standard, and meet the technological process of normal distribution, after actual mean value that has obtained technological process and actual deviation, can be by top formula theory of computation working ability rate Cp and actual working ability rate Cpk.In application, also can calculate another one parameter DPPM usually, fall the value (defective parts per million) of the defect part outside the course of work preassigned in promptly per 1,000,000 units.

But in the actual technological process, the technological process that meets normal distribution fully is very rare, and most technological processs all can have bigger deflection (skewness) or long-tail (longtail) is arranged.So, may be insecure based on the resulting Cp of above-mentioned formula, Cpk and the DPPM of perfect condition.In addition, in many manufacture field, such as semiconductor manufacturing industry, all products all are that 100% process detects, and that is to say to have defective product to flow in client's the hand, so, obtain this function of DPPM parameter by Cp, Cpk and in these fields, can be used hardly.

Cp, Cpk also have been developed many new application functions, such as, industrial products above-mentioned (manufacturing industry) user can use the value of Cp, Cpk to estimate the efficient of its technological process and the stability of product.Cp, Cpk can also be used as lasting improvement tolerance (continues improvementmetrics).The purpose of these work all is in order to reduce the deviation in the technological process, and the desired value of Alignment Process process better.By observation, as seen on arithmetic mean, reduce the technological process deviations and can increase theoretical working ability rate Cp, and reduce discrepancy factor k simultaneously above-mentioned formula (1) and (2).

Consider the method for estimation of existing C p, Cpk based on above-mentioned new application point of view, will find that existing method of estimation can not satisfy these new requirements, because the angle difference during design, therefore Cp, the Cpk that uses existing method of estimation to obtain has very big deviation in these new applications.So, just need improve, to meet these new demands to the method for estimation of Cp, Cpk.

Summary of the invention

The present invention aims to provide a kind of improvement Cp, Cpk estimation approach.

According to an aspect of the present invention, disclose a kind of method of estimation of working ability, comprising:, determine standard max USL, the standard min LSL of technological process according to preassigned.According to the practical application of technological process, determine technological process deviations and desired value T.Detect and obtain the actual mean value mu of technological process.According to the control requirement, from first model and second model, select one as the estimation model of estimating working ability, wherein in this first model and this second model for the T that is not in LSL and USL centre position, resulting working ability estimated value is symmetry pro rata in the both sides of T.According to selected model, use parameter USL, LSL, σ, T and the μ theory of computation working ability rate Cp and the actual working ability rate Cpk that are obtained afterwards.According to Cp that is obtained and Cpk technological process is controlled or adjusted.

According to a further aspect in the invention, a kind of device that is used to realize said method also is provided, comprise: the preset parameter securing component, be used for obtaining predetermined parameters according to preassigned, comprise the upper limit USL, the lower limit LSL that determine according to preassigned, technological process deviations and the technological process desired value T definite according to the practical application of technological process; The operational factor securing component is used for obtaining the parameter of technological process operational process, mainly is actual mean value mu; The working ability estimation model, comprise first model and second model, the working ability estimation model requires to select one of them model according to concrete control, wherein in this first model and this second model for the T that is not in LSL and USL centre position, resulting working ability estimated value is symmetry pro rata in the both sides of T; The working ability estimation unit according to the model that the working ability estimation model provides, is estimated working ability Cp and Cpk on the basis of the parameter μ that parameter USL, LSL, σ, T and operational factor securing component that the preset parameter securing component obtains obtain; Process controller, the Cp and the Cpk that estimate according to the working ability estimation unit adjust and control technological process.

Description of drawings

The above and other features of the present invention, character and advantage will become more obvious by the description below in conjunction with drawings and Examples, in the accompanying drawings, identical Reference numeral is represented identical feature all the time, wherein,

Fig. 1 shows the process flow diagram according to the method for estimation of working ability of the present invention;

Fig. 2 shows the structural drawing according to the estimation unit of working ability of the present invention.

Embodiment

For the estimation mode of improving Cp, Cpk to satisfy the requirement of new application, at first need to analyze the attribute of Cp, Cpk.

By simple example, study the attribute of Cp, Cpk.

Example 1:LSL=0, USL=6, T=3, σ=0.5.

According to formula: $\mathrm{Cp}=\frac{\mathrm{USL}-\mathrm{LSL}}{6\mathrm{\σ}},$ Obtain theoretical working ability rate $\mathrm{Cp}=\frac{\mathrm{USL}-\mathrm{LSL}}{6\mathrm{\σ}}=\frac{6-0}{6\×0.5}=2.$

According to formula: Cpk ₁=Cp (1-k) and $\mathrm{Cp}{k}_2=\min (\mathrm{CpU},\mathrm{CpL})=\min (\frac{\mathrm{USL}-\mathrm{\μ}}{3\mathrm{\σ}},\frac{\mathrm{\μ}-\mathrm{LSL}}{3\mathrm{\σ}}),$ Can obtain actual working ability rate Cpk under different technology average value mu, shown in the reference table 1:

Table 1

	LSL						T						USL
														μ	0	0.5	1	1.5	2	2.5	3	3.5	4	4.5	5	5.5	6
Cpk	0	0.33	0.67	1.00	1.33	1.67	2.00	1.67	1.33	1.00	0.67	0.33	0

By above-mentioned example 1, can obtain the following attribute of relevant Cp, Cpk:

(A) Cp is independent of the desired value T of technological process.

(B) maximal value of Cpk equals Cp, and when the technological process deviations is a constant, and the technology actual mean value mu is when equaling desired value T, and Cpk obtains this maximal value that equates with Cp.This means that when the technology actual mean value mu equaled desired value T, whole abilities of technological process were utilized, without any loss, reached the highest efficient.

(C) when the technology actual mean value mu equals the upper limit USL of the lower limit LSL of preassigned or preassigned, Cpk=0.In this case, preassigned interval (specification width), promptly (USL-LSL) covered the scope in normal distribution zone 50%.For the actual process flow process, this is unacceptable, and therefore, the scope that covers normal state technological process distributed areas as preassigned interval (specification width) is less than 50% the time, the Cpk of this moment is defined as 0, shows that this technological process does not have technological ability (incapable).

(D) meet following condition when the technology actual mean value mu: when μ-LSL=3 σ or USL-μ=3 σ, Cpk=1.This means that in this case, preassigned interval (specificationwidth) covered the scope of technological process distributed areas 99.7% at least.

(E) when σ is constant, the value of Cpk is symmetrical about target T, and promptly (μ=T-a)=(μ=T+a), wherein a represents from the skew of desired value T Cpk Cpk.

(F) when σ is constant, in the both sides of desired value T, along with | T-μ | the perhaps increase of k, Cpk is reduction linearly.

Viewpoint from manufacturing and client's quality control, because the application for Cp and Cpk is in order to make whole technological process more efficient, final product is more stable, therefore, attribute (B) is valued most, it has reflected the key concept that Cpk uses, and attribute (C) to (F) has been explained the relation between the physical significance that some specific Cpk values and they are hinted, perhaps represents the function of Cpk variation.

Example 2:LSL=0, USL=6, T=2, σ=0.5.

Because T=2, so desired value T is not in the centre of preassigned upper limit USL and preassigned LSL.

According to formula: $\mathrm{Cp}=\frac{\mathrm{USL}-\mathrm{LSL}}{6\mathrm{\σ}},$ Obtain theoretical working ability rate $\mathrm{Cp}=\frac{\mathrm{USL}-\mathrm{LSL}}{6\mathrm{\σ}}=\frac{6-0}{6\×0.5}=2.$

According to formula: Cpk ₁=Cp (1-k) and $\mathrm{Cp}{k}_2=\min (\mathrm{CpU},\mathrm{CpL})=\min (\frac{\mathrm{USL}-\mathrm{\μ}}{3\mathrm{\σ}},\frac{\mathrm{\μ}-\mathrm{LSL}}{3\mathrm{\σ}}),$ Can obtain actual working ability rate Cpk under different technology average value mu, shown in the reference table 2:

Table 2

	LSL				T								USL
														μ	0	0.5	1	1.5	2	2.5	3	3.5	4	4.5	5	5.5	6
Cpk 1	0.67	1.00	1.33	1.67	2.00	1.67	1.33	1.00	0.67	0.33	0	0	0
														Cpk 2	0	0.33	0.67	1.00	1.33	1.67	2.00	1.67	1.33	1.00	0.67	0.33	0

By table 2 as seen, under the situation of example 2, when desired value T is not in preassigned upper limit USL and preassigned LSL middle, Cpk ₁And Cpk ₂Value no longer equate.Continue reference table 2, can find in table 2 Cpk ₂Value be consistent with Cpk in the table 1.This explanation, the Cpk that formula (2) calculates ₂Being incoherent with the desired value T of technological process, in other words, is the desired value T that is independent of technological process.When σ is constant, no matter the position of desired value T is where, the Cpk that formula (2) calculates ₂Be always can when the technology actual mean value mu equals the intermediate value of actual preassigned upper limit USL and preassigned LSL, obtain maximal value.Therefore, a problem has occurred, formula (2) does not have most important attribute (B) in the practical application, and unfortunately be that under most situations, desired value T is not in the centre of preassigned upper limit USL and preassigned LSL, therefore, if for the viewpoint of manufacturing and client's quality control, the conclusion that formula (2) obtains has very big misleading possibility, and practical value is not high.

Continue reference table 2, Cpk ₂(2＜μ＜4)〉Cpk ₂(μ=2) this means the Cpk that formula (2) calculates ₂Under the situation of the expression actual mean value mu value of departing from objectives T, the Cpk that departs from ₂Than the Cpk that reaches desired value ₂Better.Equally, the Cpk in the table 2 ₂=1.33 have two places, are respectively μ=2 and μ=4, and wherein, μ=2 expression actual mean value mu equal desired value T, and μ=4 expression actual mean value mu have departed from desired value T and reached 4 σ.But the Cpk that calculates by formula (2) ₂Expression, both working abilities are consistent.This obviously is the conclusion that does not meet logic.That is to say, for the viewpoint of manufacturing and client's quality control, the Cpk that formula (2) calculates ₂Value has more serious misleading situation to exist.

Refer now to the Cpk that calculates by formula (1) ₁Value, it does not possess above-mentioned attribute (C) and (D).Especially danger is, when actual mean value mu equaled LSL, in fact this technological process can not be implemented, but Cpk ₁It still has the ability of maximum working ability 1/3=0.67 expression.This obviously is unacceptable mistake.And when μ=5, Cpk ₁=0, indicate this technological process not possess working ability, in fact this moment, this technological process still possessed 5/6 working ability.

By example 1 and example 2 as seen, formula (1) and formula (2) all are based on T and are positioned at the hypothesis of having done on the position of intermediate point of USL and LSL.But in the actual conditions, T can not be positioned on the intermediate point of USL and LSL basically.At this moment, no matter be the Cpk that formula (1) obtains ₁Or the Cpk that formula (2) obtains ₂, all exist obvious defects, can't be applied to controlling the efficient of whole technological process and final product stability.

Therefore, must make correction, make that Cp and Cpk that revised estimation scheme obtains can be through the many attributes that comprises Cp and Cpk of possibility, especially of paramount importance attribute B above-mentioned evaluation method.

Victor.E.Kane has proposed a kind of improved Cp and Cpk estimation scheme in 1986, see " working ability index " (" Process Capa bility Indices ") for details, be published in Journalof Quality Technology, in January, 1986, No.1, Vol.18.

The scheme of Kane has been sacrificed attribute (A), and promptly theoretical working ability rate Cp be not independent of desired value T.

The evaluation method of Kane is as follows:

$\mathrm{Cp}=\min (\frac{T-\mathrm{LSL}}{3\mathrm{\σ}},\frac{\mathrm{USL}-T}{3\mathrm{\σ}}),$

Cpk ₃＝Cp(1-k)

Wherein, $k=\frac{|T-\mathrm{\μ}|}{\min \{T-\mathrm{LSL},\mathrm{USL}-T\}}---\left(3\right)$

Kane gives the alternative computing method of above-mentioned formula (2) simultaneously:

Cpk ₃′＝min(CPL，CPU)

$\mathrm{CPL}=\frac{T-\mathrm{LSL}}{3\mathrm{\σ}}(1-\frac{|T-\mathrm{\μ}|}{T-\mathrm{LSL}})$

(CPL＝0 if|T-μ|>T-LSL) (3)

$\mathrm{CPU}=\frac{\mathrm{USL}-T}{3\mathrm{\σ}}(1-\frac{|T-\mathrm{\μ}|}{\mathrm{USL}-T})$

(CPU＝0 if|T-μ|>USL-T)

Formula (3) is applied to example 2, can obtains following Cpk value, shown in the reference table 3:

Table 3

	LSL				T								USL
														μ	0	0.5	1	1.5	2	2.5	3	3.5	4	4.5	5	5.5	6
Cpk 3	0	0.33	0.67	1.00	1.33	1.00	0.67	0.33	0	0	0	0	0

In table 3, because in the left side of desired value T, T-LSL=2, therefore, by symmetrical manner, the scope of μ-T≤2 has also only been expanded on the right side, after exceeding this scope, promptly under the situation of μ 〉=4, Cpk ₃Value directly blocked.So, if the scope of calculating is narrowed down to situation about T symmetry (being that T is positioned at intermediate value), Cpk ₃Expressed situation is consistent (in the scope of dwindling) with the result that formula (1) and (2) are calculated.Cpk ₃Be not positioned under the situation in the middle of USL and the LSL (be the situation of most of actual process processes) Cpk though revised T effectively ₂The information that is worth expressed mistake, the value that promptly departs from Cpk under the T situation is better than the situation of aiming at T on the contrary.Also revised Cpk ₁The hazardous location, promptly when μ=LSL, Cpk still is not equal to 0.Should, Cpk ₃Be effective correction for formula (1) and (2), but Cpk ₃Also have defective, formula (3) is actually available scope has been dwindled, and directly one near scope of distance between T is mapped to opposite side to formula (3) symmetrically with LSL and USL, so in fact there are some available scopes to be blocked.In the superincumbent example (example 2), in the scope of μ=4-6, in fact technological process still can be carried out, such as, when μ=4, preassigned interval (specification width) covered the scope of technological process distributed areas 99.7% at least, but Cpk ₃Become 0.

So the present invention is intended to further to improve the indication ability for Cpk in above-mentioned μ=4-6 scope for example, more expansion ground is said, will be from the indication ability raising of the Cpk of the comparatively big side of desired value T.

In order to accomplish this point, the present invention need manage above-mentioned attribute (E) is modified to: when σ was constant, the value of Cpk was proportional symmetry about target T, promptly

((μ=T+ γ * a), wherein a represents from the skew of desired value T μ=T-a)=Cpk Cpk.

$\mathrm{\γ}=\frac{\mathrm{USL}-T}{T-\mathrm{LSL}}.$

Above-mentioned improved plan is used for example 2, owing to T in example 2 is offset to LSL one side, therefore obviously, responsive more and remarkable for side-play amount a than USL one side at LSL one side Cpk.In order to make Cpk, must make about desired value T " symmetry pro rata ":

Cpk (μ=0)=Cpk (μ=6), Cpk (μ=0.5)=Cpk (μ=5), Cpk (μ=1)=Cpk (μ=4),, γ=2 wherein.

By following formula, can calculate and obtain Cpk with " proportional symmetry ", and thereby expansion in whole LSL to USL scope for the deixis of working ability.

$\mathrm{Cp}=\frac{\mathrm{USL}-\mathrm{LSL}}{6\mathrm{\σ}}$

Cpk ₄＝Cp(1-k) (4)

$k=\left\{\begin{array}{cc}\frac{T-\mathrm{\μ}}{T-\mathrm{LSL}}& \mathrm{\μ}\≤T,\\ \frac{\mathrm{\μ}-T}{\mathrm{USL}-T}& \mathrm{\μ}>T.\end{array}\right.$

If the discrepancy factor k in the above-mentioned formula (4) is used for formula (3), can obtains following formula:

$\mathrm{Cp}=\min (\frac{T-\mathrm{LSL}}{3\mathrm{\σ}},\frac{\mathrm{USL}-T}{3\mathrm{\σ}})$

Cpk ₅＝Cp(1-k) (5)

$k=\left\{\begin{array}{cc}\frac{T-\mathrm{\μ}}{T-\mathrm{LSL}}& \mathrm{\μ}\≤T,\\ \frac{\mathrm{\μ}-T}{\mathrm{USL}-T}& \mathrm{\μ}>T.\end{array}\right.$

The formula (4) and (5) that obtain are applied to above-mentioned example 2, can obtain result as shown in table 4, in table 4, also show the data of parameter DPPM.

Table 4

	LSL				T								USL
														μ	0	0.5	1	1.5	2	2.5	3	3.5	4	4.5	5	5.5	6
Cpk4	0	0.50	1.00	1.50	2.00	1.75	1.50	1.25	1.00	0.75	0.50	0.25	0
														Cpk5	0	0.33	0.67	1.00	1.33	1.17	1.00	0.83	0.67	0.50	0.33	0.17	0
DPPM	500k	159k	22.8k	1350	31.7	0.29	.002	0.29	31.7	1350	22.8k	159k	500k

Understand the corresponding situation that Cpk value and related process process drop on DPPM if desired, the situation shown in can reference table 5:

Table 5

Cp	0.17	0.25	0.33	0.50	0.67	0.75	0.83	1.00	1.17	1.25	1.33	1.50	1.75	2.00
															DPPM	617k	453k	317k	134k	46k	24k	12k	2700	465	177	63	7	0.15	0.002

Consider Cpk by formula (4) acquisition ₄With the Cpk that obtains by formula (5) ₅, can find Cpk ₃The scope of value between LSL and T in (being the side that T is partial to) some is excessive, and some is too small in T and USL scope.And Cpk ₅Value all less than normal.Though Cpk ₄Value at some imbalance of the both sides of T, Cpk ₅Value underestimate a bit, but these two data have reflected the performance of technological process more truly, have avoided situation about misleading, and have also eliminated possible dangerous data.

Table 6 shows Cp, the Cpk attribute that Cpk that above-mentioned formula (1)-(5) are obtained can reflect.

Table 6

	(1)	(2)	(3)	(4)	(5)
						(A) Cp is independent of desired value T	Y	Y	N	Y	N
(B) maximal value of Cpk (=Cp) when μ=T, obtain	Y	N	Y	Y	Y
						(C) when μ=LSL or μ=USL, Cpk=0	N	Y	Y	Y	Y
(D) when μ-LSL=3 σ or USL-μ=3 σ, Cpk=1, expression USL-LSL cover at least 99.7% technological process and distribute	N	Y	Y	N	Y
						(E) about T symmetry or symmetrical pro rata	Y	N	Y	Y	Y
(F) Cpk is with | T-μ | increase reduce linearly	Y	N	Y	Y	Y
						Can be used for the non-central T of center T/	N	N	Y	Y	Y

As seen, formula (3), (4) and (5) meet 5 attributes respectively, can reflect the working ability of technological process preferably.Wherein, (4) and (5) have the effective range bigger than (3), therefore are the effective improvement for Cp, Cpk method of estimation.

For the best separately range of application in (4), (5), (5) are applicable in the application of paying close attention to DPPM, because the Cpk value of (5) is estimated that ground is smaller, under all the other situations, (4) are more better, because the Cpk value of (4) acquisition easy to understand more.

Based on above-mentioned improved formula (4), (5), the invention provides a kind of working ability and carry out estimation approach and device technological process.Wherein, the desired value of this technological process is not the centre position that is positioned at the upper and lower bound of preassigned, and this method is by the estimation of following process acquisition for working ability, with reference to shown in Figure 1:

According to preassigned, determine standard max USL, the standard min LSL (step 102) of technological process.According to the practical application of technological process, determine technological process deviations and desired value T (step 104).Detect and obtain the actual mean value mu (step 106) of technological process.According to the control requirement, wherein, this first model is meant formula (4) as the estimation model (step 108) of estimating working ability to select one from first model and second model, and this second model is meant formula (5).Choose formula (5) in the application of the standard of choosing for concern DPPM, and in other are used, choose formula (4).According to selected model, use parameter USL, the qLSL, σ, T and μ theory of computation working ability rate Cp and the actual working ability rate Cpk (step 110) that in step 102,104 and 106, obtain afterwards.Technological process is controlled or is adjusted (step 112) according to Cp that is obtained and Cpk.

The present invention also provides a kind of device that is used to realize said method, need to prove, said herein " device " should be understood to the instrument that makes one to be used to realize this scheme, this device can be that hardware is realized, also can is that software is realized, such as: carry out the general processor of special instruction, the special chip that is written into special instruction, programmable logic array or the like.This sentences this device of functional definition, it will be understood by those of skill in the art that its specific implementation can adopt multiple technique known to realize.

Fig. 2 has disclosed the structural drawing of said apparatus of the present invention, and it should comprise following functional module or perhaps assembly:

Preset parameter securing component 202 is used for obtaining predetermined parameters according to preassigned, comprises the upper limit USL, the lower limit LSL that determine according to preassigned, technological process deviations and the technological process desired value T definite according to the practical application of technological process;

Operational factor securing component 204 is used for obtaining the parameter of technological process operational process, mainly is actual mean value mu;

Working ability estimation model 206, this working ability estimation model 206 comprises two models, i.e. first model (referring to formula (4)) and second model (referring to formula (5)), working ability estimation model 206 requires to select one of them model according to concrete control;

Working ability estimation unit 208, according to the model that working ability estimation model 206 provides, on the basis of the parameter μ that parameter USL, LSL, σ, T and operational factor securing component 204 that preset parameter securing component 202 obtains obtain, working ability Cp and Cpk are estimated;

Process controller 210, the Cp and the Cpk that estimate according to working ability estimation unit 208 adjust and control technological process.

The foregoing description provides to being familiar with the person in the art and realizes or use of the present invention; those skilled in the art can be under the situation that does not break away from invention thought of the present invention; the foregoing description is made various modifications or variation; thereby protection scope of the present invention do not limit by the foregoing description, and should be the maximum magnitude that meets the inventive features that claims mention.

Claims (10)

1. the method for estimation of a processing capacity of process comprises:

According to preassigned, determine the preassigned upper limit USL and the preassigned lower limit LSL of technological process;

According to the practical application of technological process, determine technological process deviations and technological process desired value T;

Detect and obtain the actual mean value mu of technological process;

According to the control requirement, from first model and second model, select one as the estimation model of estimating working ability, wherein, in this first model and this second model, for the T that is not in LSL and USL centre position, resulting working ability estimated value is symmetrical pro rata in the both sides of T;

According to selected model, use at parameter USL, LSL, σ, T and μ theory of computation working ability rate Cp and actual working ability rate Cpk;

According to Cp that is obtained and Cpk technological process is controlled or adjusted.

2. the method for estimation of processing capacity of process as claimed in claim 1 is characterized in that,

Estimate that according to first model or second model Cpk that obtains meets

((μ=T+ γ * a), wherein a represents from the skew of desired value T μ=T-a)=Cpk Cpk; $\mathrm{\γ}=\frac{\mathrm{USL}-T}{T-\mathrm{LSL}}.$

3. the method for estimation of processing capacity of process as claimed in claim 1 is characterized in that, described first model definition is:

$\mathrm{Cp}=\frac{\mathrm{USL}-\mathrm{LSL}}{6\mathrm{\σ}}$

Cpk ₄＝Cp(1-k)

$k=\left\{\begin{array}{cc}\frac{T-\mathrm{\μ}}{T-\mathrm{LSL}}& \mathrm{\μ}\≤T,\\ \frac{\mathrm{\μ}-T}{\mathrm{USL}-T}& \mathrm{\μ}>T.\end{array}\right.$

4. the method for estimation of processing capacity of process as claimed in claim 1 is characterized in that, described second model definition is:

$\mathrm{Cp}=\min (\frac{T-\mathrm{LSL}}{3\mathrm{\σ}},\frac{\mathrm{USL}-T}{3\mathrm{\σ}})$

Cpk ₅＝Cp(1-k)

$k=\left\{\begin{array}{cc}\frac{T-\mathrm{\μ}}{T-\mathrm{LSL}}& \mathrm{\μ}\≤T,\\ \frac{\mathrm{\μ}-T}{\mathrm{USL}-T}& \mathrm{\μ}>T.\end{array}\right.$

。

5. the method for estimation of processing capacity of process as claimed in claim 1 is characterized in that,

In paying close attention to per 1,000,000 units, fall in the application of value DPPM of the defect part outside the course of work preassigned, select second model.

6. the estimation unit of a processing capacity of process comprises:

The preset parameter securing component is used for obtaining predetermined parameters according to preassigned, comprises the upper limit USL, the lower limit LSL that determine according to preassigned, and technological process deviations and the technological process desired value T definite according to the practical application of technological process;

The operational factor securing component is used for obtaining the parameter of technological process operational process, comprises actual mean value mu;

The working ability estimation model, comprise first model and second model, the working ability estimation model requires to select one of them model according to concrete control, wherein in this first model and this second model for the T that is not in LSL and USL centre position, resulting working ability estimated value is symmetry pro rata in the both sides of T;

The working ability estimation unit according to the model that the working ability estimation model provides, is estimated working ability Cp and Cpk on the basis of the parameter μ that parameter USL, LSL, σ, T and operational factor securing component that the preset parameter securing component obtains obtain;

Process controller, the Cp and the Cpk that estimate according to the working ability estimation unit adjust and control technological process.

7. the estimation unit of processing capacity of process as claimed in claim 6 is characterized in that,

Estimate that according to first model or second model Cpk that obtains meets

((μ=T+ γ * a), wherein a represents from the skew of desired value T μ=T-a)=Cpk Cpk; $\mathrm{\γ}=\frac{\mathrm{USL}-T}{T-\mathrm{LSL}}.$

8. the estimation unit of processing capacity of process as claimed in claim 6 is characterized in that, described first model definition is:

$\mathrm{Cp}=\frac{\mathrm{USL}-\mathrm{LSL}}{6\mathrm{\σ}}$

Cpk ₄＝Cp(1-k)

$k=\left\{\begin{array}{cc}\frac{T-\mathrm{\μ}}{T-\mathrm{LSL}}& \mathrm{\μ}\≤T,\\ \frac{\mathrm{\μ}-T}{\mathrm{USL}-T}& \mathrm{\μ}>T.\end{array}\right.$

9. the estimation unit of processing capacity of process as claimed in claim 6 is characterized in that, described second model definition is:

$\mathrm{Cp}=\min (\frac{T-\mathrm{LSL}}{3\mathrm{\σ}},\frac{\mathrm{YSL}-T}{3\mathrm{\σ}})$

Cpk ₅＝Cp(1-k)

$k=\left\{\begin{array}{cc}\frac{T-\mathrm{\μ}}{T-\mathrm{LSL}}& \mathrm{\μ}\≤T,\\ \frac{\mathrm{\μ}-T}{\mathrm{USL}-T}& \mathrm{\μ}>T.\end{array}\right.$

10. the estimation unit of processing capacity of process as claimed in claim 6 is characterized in that, falls in paying close attention to per 1,000,000 units in the application of value DPPM of the defect part outside the course of work preassigned, selects second model.

Images