CN105707974A - Cigarette production process stability monitoring method - Google Patents

Abstract

The invention discloses a cigarette production process stability monitoring method. The monitoring method sequentially comprises the following steps: 1, classifying variables in a cigarette production process; 2, conducting quantitative characterization on the classified variables; 3, implementing single-variable monitoring, adopting different monitoring models in accordance with different variable types; and 4, implementing multi-variable monitoring, wherein the multi-variable monitoring includes two parts, namely process comprehensive monitoring and batch comprehensive monitoring, a process stability monitoring comprehensive index model is adopted and a batch stability comprehensive monitoring index model is adopted for the batch comprehensive monitoring. The cigarette production process stability monitoring method disclosed by the invention not only can detect the minimal stability variable in the production process but also can detect processing stability and batch comprehensive stability of the process, so that the method offers strong support for improving cigarette production and enhancing the stability of product quality.

Description

A kind of production of cigarettes process stability monitoring method

Technical field

The present invention relates to technical field of cigarette production, particularly relate to a kind of production of cigarettes process stability monitoring method.

Background technology

The purpose of production of cigarettes process stability monitoring is in process of production, when the input of production system, the resource relied on, control measure and manufacturing activities itself can continue when meeting technical standard, specification and requirement, it is ensured that the fluctuation of each quality management unit institute output products quality is in allowed band.

At present, production of cigarettes process monitoring Capability index C_PK, sigma level test etc. find broad application at tobacco business, achieve very good effect, but there is following defect:

(1) monitoring result is all based on normal distribution, and need to calculate in the steady state；

(2) qualification rate of discrete Monitoring Data, without distinguishing in interval of acceptance；

(3) quantization level is narrower.

Therefore, research and develop a kind of unification, have discrimination, the production of cigarettes process stability monitoring method that is prone to intuitive analysis, for triggering improving and optimizating of production of cigarettes process, there is important directive significance.

Summary of the invention

It is an object of the invention to provide a kind of production of cigarettes process stability monitoring method, for improving production of cigarettes, the offer that improves product quality stability provides powerful support for.

The technical solution used in the present invention is: a kind of production of cigarettes process stability monitoring method, comprises the following steps successively:

One, the variable in production of cigarettes process to be classified, classification type includes the inspection of the eyes control type of tolerance requirements, without the inspection of the eyes control type of tolerance requirements, limiting control type and scope control type；

Two, sorted variable is carried out quantization signifying, wherein, inspection of the eyes control type variable is characterized by irrelevance and dispersion, and limiting control type variable is characterized by the degrees of offset of measured value Yu setting value；

Three, carry out single argument monitoring, difference according to types of variables adopts different monitoring models, inspection of the eyes control type variable adopts the inspection of the eyes control variable model based on irrelevance and dispersion to be monitored, and limiting control type variable adopts the limiting control variate model based on measured value Yu the degrees of offset of setting value to be monitored；

Four, carry out multivariate monitoring, multivariate monitoring includes operation comprehensive monitoring and batch comprehensive monitoring two parts, operation comprehensive monitoring adopts based on the average weighted process stability monitoring Synthesized Index Model of single argument monitoring result, and batch comprehensive monitoring adopts the lot stability comprehensive monitoring exponential model based on operation comprehensive monitoring result geometric average.

The computing formula having the irrelevance of the inspection of the eyes control type variable of tolerance requirements in described step 2 isWherein, A represents irrelevance；Represent actual measurement meansigma methods；χ_pvExpression standard design load；δ_pvExpression standard design standard deviation, δ_pv=franchise/3；Subscript_pvRepresent standard value；

Being converted into limiting control type without the irrelevance of the inspection of the eyes control type variable of tolerance requirements by calculating relative variation in described step 2 to obtain, or control to limit the inspection of the eyes type being converted into franchise to obtain by increasing, the computing formula of relative variation isWherein, B represents relative variation；Represent actual measurement meansigma methods；χ_pvExpression standard design load；Subscript_pvRepresent standard value；

In described step 2, the computing formula of the dispersion of inspection of the eyes control type variable isWherein P represents dispersion；S represents actual measurement standard deviation；δ_pvExpression standard design standard deviation, δ_pv=franchise/3；Subscript_pvRepresent standard value；

In described step 2, the measured value of limiting control type variable with the computing formula of the degrees of offset of setting value isWherein, C represents the degrees of offset of measured value and setting value；Represent actual measurement meansigma methods；χ_pvExpression standard design load；χ_bestRepresentation theory or actual optimum value；Subscript_pvRepresent standard value；Subscript_bestRepresent optimal value.

Described inspection of the eyes control variable model is:Wherein, I_cRepresent inspection of the eyes type control variable STABILITY MONITORING index；I_bestRepresent optimum Monitoring Index；I_baseRepresent baseline Monitoring Index；A represents irrelevance；P represents dispersion；Subscript_cRepresent inspection of the eyes type control variable；Subscript_bestRepresent optimal value；Subscript_baseRepresent baseline value；

Described limiting control variate model is: I_d=I_base+(I_best-I_base) × C, wherein, I_dRepresent extreme value type control variable STABILITY MONITORING index；I_bestRepresent optimum Monitoring Index；I_baseRepresent baseline Monitoring Index；C represents the degrees of offset of measured value and setting value；Subscript_dRepresent extreme value type control variable；Subscript_bestRepresent optimal value；Subscript_baseRepresent baseline value.

Described process stability monitoring Synthesized Index Model is:Wherein, G represents operation comprehensive monitoring index；I_iRepresent the STABILITY MONITORING index of i-th variable；W_iRepresent the weight of i-th variable；Subscript i=1,2,3...k；

Described lot stability comprehensive monitoring exponential model is:Wherein, L represents a batch comprehensive monitoring index；G_iRepresent the STABILITY MONITORING index of i-th operation；D_iRepresent the weight of i-th operation；Subscript i=1,2,3...n.

The present invention carries out different quantization signifyings according to the difference of the variable in production of cigarettes process, and then carries out single argument monitoring, carries out multivariate monitoring, namely monitor processing stability and batch comprehensive stability of operation on the basis of single argument monitoring.Production of cigarettes process stability monitoring method of the present invention is possible not only to the stability of the single variable of monitor production process, the processing stability of operation and the stability of batch comprehensive process quality can also be monitored, for improving production of cigarettes, the offer that improves product quality stability provides powerful support for.

Accompanying drawing explanation

Fig. 1 is the flow chart of the present invention；

Fig. 2 is inspection of the eyes control type variable monitoring schematic diagram in the present invention；

Fig. 3 hopes big limiting control type variable monitoring schematic diagram in the present invention；

Fig. 4 hopes little limiting control type variable monitoring schematic diagram in the present invention.

Detailed description of the invention

As it is shown in figure 1, production of cigarettes process stability monitoring method of the present invention, comprise the following steps successively:

One, the variable (include control parameter and quality index) in production of cigarettes process to be classified, classification type includes the inspection of the eyes control type of tolerance requirements, without the inspection of the eyes control type of tolerance requirements, limiting control type and scope control type；

Inspection of the eyes control type includes the inspection of the eyes control type of tolerance requirements and without inspection of the eyes control type two class of tolerance requirements: have the inspection of the eyes control type of tolerance requirements, such as barrel temperature, material moisture, cigarette resistance to suction etc.；Inspection of the eyes control type without tolerance requirements, such as mass flow, feed ratio etc., but it can be converted into undulate quantity and precision controls or increases control limit, also has a class such as drum rotation speed and valve opening etc., substantially do not change over time after setting in process of production, monitor.

Limiting control type, such as Filling power, filament broken rate, end tobacco-falling and Medicated cigarette presentation quality etc..

Scope control type, such as strip bulking time etc., it is desirable in certain scope, namely meet the requirements.

Therefore, the monitoring of production of cigarettes process stability needs mainly inspection of the eyes control type variable and limiting control type variable that emphasis is monitored.

Two, sorted variable is carried out quantization signifying, wherein, inspection of the eyes control type variable is characterized by irrelevance and dispersion, and limiting control type variable is characterized by the degrees of offset of measured value Yu setting value；Sorted variable carries out quantization signifying realize compared with requiring with technical standard.

Inspection of the eyes type control variable is generally the continuous data of on-line data acquisition system acquisition, can be characterized by irrelevance and dispersion.

What irrelevance characterized is the departure degree of actual controlling value and technological standards value, i.e. accuracy.

Having the inspection of the eyes control type variable of tolerance requirements, general requirement controls in franchise or 3 times of standard design Standard deviation-Range, i.e. χ=χ_pv±3δ_pv, the computing formula of irrelevance is:

$A=\frac{\left|\stackrel{\‾}{\mathrm{\χ}}-{\mathrm{\χ}}_{pv}\right|}{{\mathrm{\δ}}_{pv}}---\left(1\right)$

Wherein, A represents irrelevance；Represent actual measurement meansigma methods；χ_pvExpression standard design load；δ_pvExpression standard design standard deviation, δ_pv=franchise/3；Subscript_pvRepresent standard value.A value is more little, characterize reality control average and technical standard value closer to, namely control accuracy is more good.

Without the irrelevance of inspection of the eyes control type variable of tolerance requirements by calculating after relative variation, it is converted into limiting control type and obtains, or increase and control limit and be converted into the inspection of the eyes type of franchise.The computing formula of relative variation is:

$B=\frac{\left|\stackrel{\‾}{\mathrm{\χ}}-{\mathrm{\χ}}_{pv}\right|}{{\mathrm{\χ}}_{pv}}---\left(2\right)$

Wherein, B represents relative variation；Represent actual measurement meansigma methods；χ_pvExpression standard design load；Subscript_pvRepresent standard value.

What dispersion characterized is the satisfaction degree of actual control fluctuation situation (standard deviation) and technical standard requirement, i.e. degree of accuracy.The computing formula of the dispersion of inspection of the eyes control type variable is:

$P=\frac{s}{{\mathrm{\δ}}_{pv}}---\left(3\right)$

Wherein, P represents dispersion；S represents actual measurement standard deviation；δ_pvExpression standard design standard deviation, δ_pv=franchise/3；Subscript_pvRepresent standard value.P value is more little, and the fluctuation situation characterizing reality control more can meet technical standard requirement, and namely control accuracy is better.

Extreme value type variable is generally offline inspection data, is generally expected to big (>=χ_pv) and hope little (≤χ_pv) set, characterize mainly through the degrees of offset of measured value Yu setting value.Hoping big namely compared with typical set value, measured value is the bigger the better；Hoping little namely compared with typical set value, measured value is the smaller the better.The computing formula of the measured value of limiting control type variable and the degrees of offset of setting value is:

$C=\frac{\stackrel{\‾}{\mathrm{\χ}}-{\mathrm{\χ}}_{pv}}{{\mathrm{\χ}}_{best}-{\mathrm{\χ}}_{pv}}---\left(4\right)$

Wherein, C represents the degrees of offset of measured value and setting value；Represent actual measurement meansigma methods；χ_pvExpression standard design load；χ_bestRepresentation theory or actual optimum value；Subscript_pvRepresent standard value；Subscript_bestRepresent optimal value.

Three, carry out single argument monitoring, difference according to types of variables adopts different monitoring models, inspection of the eyes control type variable adopts the inspection of the eyes control variable model based on irrelevance and dispersion to be monitored, and limiting control type variable adopts the limiting control variate model based on measured value Yu the degrees of offset of setting value to be monitored；Single argument is the minimum opinion scale of production process, and its level of stability requires directly related with technical standard.

Inspection of the eyes control type variable requires control center value and standard deviation, namely irrelevance and dispersion is all required, and monitoring model is the function of A and P, and is all negative correlation with A and P.Inspection of the eyes control type Variable Control requires as in figure 2 it is shown, detection requirement is as follows:

(1) it is optimum when working as A=0, P=0, namely now controls level optimum；

(2) circumference or Dashed reference line are that most basic control requires or basic controlling line.

Described inspection of the eyes control variable model is:

$\begin{array}{c}{I}_c=I_{base}+(I_{best}-I_{base})\×\frac{(1-\sqrt{{\left(\frac{A}{3}\right)}^2+P^2})}{(r_{base}-r_{best})}\\ =I_{base}+(I_{best}-I_{base})\×\frac{(1-\sqrt{{\left(\frac{A}{3}\right)}^2+P^2})}{(1-0)}\\ =I_{best}-(I_{best}-I_{base})\×\sqrt{{\left(\frac{A}{3}\right)}^2+P^2}\end{array}---\left(5\right)$

Wherein, I_cRepresent inspection of the eyes type control variable STABILITY MONITORING index；I_bestRepresent optimum Monitoring Index；I_baseRepresent baseline Monitoring Index；A represents irrelevance；P represents dispersion；R_baseRepresentation theory or actual expectation baseline value, r_base=1；R_bestRepresentation theory or actual expectation optimal value, r_best=0；Subscript_cRepresent inspection of the eyes type control variable；Subscript_bestRepresent optimal value；Subscript_baseRepresent baseline value.

Make I_best=100, I_base=60, can obtain:

$I_c=100-40\×\sqrt{{\left(\frac{A}{3}\right)}^2+P^2}---\left(6\right)$

Wherein, I_cRepresent inspection of the eyes type control variable STABILITY MONITORING index；A represents irrelevance；P represents dispersion.

Limiting control type variable includes hoping big limiting control type and hoping little limiting control type two types, and that hopes big limiting control type variable controls requirement as it is shown on figure 3, hope that the control of little limiting control type variable requires as shown in Figure 4.Limiting control variate model is:

$\begin{array}{c}{I}_d=I_{base}+(I_{best}-I_{base})\×\left(\frac{\stackrel{\‾}{\mathrm{\χ}}-{\mathrm{\χ}}_{pv}}{{\mathrm{\χ}}_{best}-{\mathrm{\χ}}_{pv}}\right)\\ =I_{base}+(I_{best}-I_{base})\×C\end{array}---\left(7\right)$

Wherein, I_dRepresent extreme value type control variable STABILITY MONITORING index；I_bestRepresent optimum Monitoring Index；I_baseRepresent baseline Monitoring Index；Represent actual measurement meansigma methods；χ_pvExpression standard design load；χ_bestRepresentation theory or actual expectation optimal value；C represents the degrees of offset of measured value and setting value；Subscript_dRepresent extreme value type control variable；Subscript_bestRepresent optimal value；Subscript_baseRepresent baseline value；Subscript_pvRepresent standard value.

Make I_best=100, I_base=60, can obtain:

I_d=60+40 × C (8)

If cigarette finished product presentation quality is for hoping big limiting control type variable, it is desirable to >=90, namely it is divided into qualifying equal to 90, its I_d=60；Full marks 100, its I_d=100, formula (8) can obtain:

$\begin{array}{c}{I}_d=60+40\×\frac{\stackrel{\‾}{\mathrm{\χ}}-90}{10}\\ =4\stackrel{\‾}{\mathrm{\χ}}-300\end{array}$

Wherein, I_dRepresent extreme value type control variable STABILITY MONITORING index；Represent actual measurement meansigma methods.

Four, carry out multivariate monitoring, multivariate monitoring includes operation comprehensive monitoring and batch comprehensive monitoring two parts, operation comprehensive monitoring adopts based on the average weighted process stability monitoring Synthesized Index Model of single argument monitoring result, and batch comprehensive monitoring adopts the lot stability comprehensive monitoring exponential model based on operation comprehensive monitoring result geometric average.

Working procedure processing level of stability is evaluated mainly through parameter and the STABILITY MONITORING index of operation.The each variable of in-process is parallel, the method adopting statistical weight, builds the process stability monitoring Synthesized Index Model quantified:

$G=\underset{i=1}{\overset{k}{\Σ}}{w}_i{I}_i---\left(9\right)$

Wherein, G represents process stability comprehensive monitoring index；I_iRepresent the STABILITY MONITORING index of i-th variable；W_iRepresent the weight of i-th variable；Subscript i=1,2,3...k.

Batch comprehensive stability level is evaluated mainly through the stability comprehensive monitoring index of critical process.Batch each operation is front-to-back effect, the method adopting geometric average, builds the lot stability comprehensive monitoring exponential model quantified and is:

$L=\mathrm{\Σ}\sqrt[d_i]{\underset{i=1}{\overset{n}{\Π}}{G}_i^{d_i}}---\left(10\right)$

Wherein, L represents lot stability comprehensive monitoring index；G_iRepresent the STABILITY MONITORING index of i-th operation；D_iRepresent the weight of i-th operation；Subscript i=1,2,3...n.

Do auspicious further stating below with reference to embodiment, but be not limiting as the present invention.In production of cigarettes process stability Monitoring Index, if less than or equal to 100 be more than or equal in the scope of 80, then be excellent；If less than 80 be more than or equal in the scope of 60, then be good；If less than 60 be more than or equal in the scope of 0; for poor, production of cigarettes process need to be improved.

Utilize a certain class trade mark cigarette primary processing Central Control Room automatic data collection, respectively lot stability three kinds of situations of comprehensive monitoring index excellent, good, poor are illustrated.

(1) loosening and gaining moisture operation

Main investigation entrance mass flow, discharging moisture content and three variablees of hot blast temperature, entrance mass flow, discharging moisture content and hot blast temperature are the inspection of the eyes control type variable of tolerance requirements, calculate irrelevance and the dispersion of three variablees respectively, and then according to inspection of the eyes control type variate model computational stability Monitoring Index.Collection data and the STABILITY MONITORING Index for Calculation result of entrance mass flow are as shown in table 1；Collection data and the STABILITY MONITORING Index for Calculation result of discharging moisture content are as shown in table 2；Collection data and the STABILITY MONITORING Index for Calculation result of hot blast temperature are as shown in table 3.By table 1, table 2 and table 3 it can be seen that discharging moisture control is poor, carrying out quantitative watering with loosening and gaining moisture, the poor result of discharging moisture control is coincide.

The collection data of table 1 entrance mass flow and control ability Monitoring Index result of calculation

The collection data of table 2 discharging moisture content and control ability Monitoring Index result of calculation

Lot number	Standard value/%	Meansigma methods/%	Standard deviation/%	Franchise/%	A	P	Ic
								087	18.50	18.48	0.04	1.00	0.060	0.117	95.25
018	18.50	18.43	0.33	1.00	0.210	0.990	60.30
								075	18.50	18.47	0.36	1.00	0.090	1.080	56.78

The collection data of table 3 hot blast temperature and control ability Monitoring Index result of calculation

Lot number	Standard value/DEG C	Meansigma methods/DEG C	Standard deviation/DEG C	Franchise/DEG C	A	P	Ic
								087	60	60.03	0.40	2	0.045	0.596	76.14
018	60	59.98	0.56	2	0.030	0.844	66.25
								075	60	60.17	0.51	2	0.255	0.769	69.05

(2) Screening and casing operation

Main investigation entrance mass flow, charging precision, discharging moisture content and four variablees of hot blast temperature, entrance mass flow, discharging moisture content and hot blast temperature are the inspection of the eyes control type variable of tolerance requirements, calculate irrelevance and the dispersion of three variablees respectively, charging precision is limiting control type variable, the degrees of offset of measured value and setting value is calculated according to lining average in fact and standard design load, and then according to inspection of the eyes control variable model and limiting control variate model computational stability Monitoring Index respectively.The collection data of entrance mass flow and STABILITY MONITORING Index for Calculation result such as table 4；Collection data and the STABILITY MONITORING Index for Calculation result of charging precision are as shown in table 5；Collection data and the STABILITY MONITORING Index for Calculation result of discharging moisture content are as shown in table 6；Collection data and the STABILITY MONITORING Index for Calculation result of hot blast temperature are as shown in table 7.By table 4, table 5, table 6 and table 7 it can be seen that each index of Screening and casing operation all controls better.

The collection data of table 4 entrance mass flow and control ability Monitoring Index result of calculation

The collection data of table 5 charging precision and control ability Monitoring Index result of calculation

Lot number	Meansigma methods/%	Id
			087	0.014	98.88
018	0.019	98.48
			075	0.015	98.80

The collection data of table 6 discharging moisture content and control ability Monitoring Index result of calculation

Lot number	Standard value/%	Meansigma methods/%	Standard deviation/%	Franchise/%	A	P	Ic
								087	19.50	19.45	0.10	1.00	0.150	0.300	87.83
018	19.70	19.70	0.13	1.00	0.012	0.390	84.40
								075	19.50	19.50	0.10	1.00	0.003	0.290	88.40

The collection data of table 7 hot blast temperature and control ability Monitoring Index result of calculation

Lot number	Standard value/DEG C	Meansigma methods/DEG C	Standard deviation/DEG C	Franchise/DEG C	A	P	Ic
								087	95	95.00	0.47	3	0.000	0.472	81.13
018	95	94.99	0.60	3	0.010	0.605	75.81
								075	95	95.02	0.31	3	0.020	0.305	87.79

(3) thin plate drying process

Main investigation entrance mass flow, HT vapor flow rate, I district's barrel temperature, II district's barrel temperature, hot blast temperature and six variablees of discharging moisture content, entrance mass flow, HT vapor flow rate, I district's barrel temperature, II district's barrel temperature, hot blast temperature and discharging moisture content are the inspection of the eyes control type variable of tolerance requirements, calculate irrelevance and the dispersion of six variablees respectively, and then according to inspection of the eyes control variable model computational stability Monitoring Index.Collection data and the STABILITY MONITORING Index for Calculation result of entrance mass flow are as shown in table 8；Collection data and the STABILITY MONITORING Index for Calculation result of HT vapor flow rate are as shown in table 9；Collection data and the STABILITY MONITORING Index for Calculation result of I district's barrel temperature are as shown in table 10；Collection data and the STABILITY MONITORING Index for Calculation result of II district's barrel temperature are as shown in table 11；Collection data and the STABILITY MONITORING Index for Calculation result of hot blast temperature are as shown in table 12；Collection data and the STABILITY MONITORING Index for Calculation result of discharging moisture content are as shown in table 13.By table 8, table 9, table 10, table 11, table 12 and table 13 it can be seen that the dry inflow control of these two batches of cut tobaccos of lot number 075 and 018 is poor, cause that barrel temperature controls poor, consistent with practical situation.

The collection data of table 8 entrance mass flow and control ability Monitoring Index result of calculation

The collection data of table 9HT vapor flow rate and control ability Monitoring Index result of calculation

The collection data of table 10 I district barrel temperature and control ability Monitoring Index result of calculation

Lot number	Standard value/DEG C	Meansigma methods/DEG C	Standard deviation/DEG C	Franchise/DEG C	A	P	Ic
								087	128	128.00	0.07	2	0.000	0.105	95.79
018	127	128.09	0.38	2	1.635	0.573	68.36
								075	128	129.75	1.43	2	2.625	2.145	7.34

The collection data of table 11 II district barrel temperature and control ability Monitoring Index result of calculation

Lot number	Standard value/DEG C	Meansigma methods/DEG C	Standard deviation/DEG C	Franchise/DEG C	A	P	Ic
								087	128	128.00	0.09	2	0.000	0.136	94.58
018	127	128.10	0.47	2	1.650	0.702	64.31
								075	128	127.56	0.96	2	0.660	1.440	41.73

The collection data of table 12 hot blast temperature and control ability Monitoring Index result of calculation

Lot number	Standard value/DEG C	Meansigma methods/DEG C	Standard deviation/DEG C	Franchise/DEG C	A	P	Ic
								087	115	115	0.18	3	0.000	0.183	92.68
018	115	115	0.08	3	0.000	0.076	96.95
								075	115	115.7	0.42	3	0.700	0.420	80.78

The collection data of table 13 discharging moisture content and control ability Monitoring Index result of calculation

Lot number	Standard value/%	Meansigma methods/%	Standard deviation/%	Franchise/%	A	P	Ic
								087	13.00	12.97	0.056	0.50	0.180	0.336	86.35
018	13.00	13.01	0.044	0.50	0.060	0.264	89.41
								075	12.90	12.86	0.055	0.50	0.240	0.330	86.42

(4) mixed silk perfuming operation

Main investigation inlet flow rate undulating value, Flavouring Precision and two variablees of discharging moisture content, inlet flow rate undulating value and Flavouring Precision are limiting control type variable, the measured value of two variablees and the degrees of offset of setting value is calculated respectively according to lining average in fact and standard design load, discharging moisture content is the inspection of the eyes control type variable of tolerance requirements and calculates its irrelevance and dispersion, and then according to limiting control variate model and inspection of the eyes control variable model computational stability Monitoring Index respectively.Collection data and the STABILITY MONITORING Index for Calculation result of inlet flow rate undulating value are as shown in table 14；Collection data and the STABILITY MONITORING Index for Calculation result of Flavouring Precision are as shown in table 14；Collection data and the STABILITY MONITORING Index for Calculation result of discharging moisture content are as shown in Table 15.By table 14 and table 15 it can be seen that the mixed each index of silk perfuming all controls better.

The collection data of table 14 Flavouring Precision and the fluctuation of entrance mass flow and control ability Monitoring Index result of calculation

The collection data of table 15 discharging moisture content and control ability Monitoring Index result of calculation

Lot number	Standard value/%	Meansigma methods/%	Standard deviation/%	Franchise/%	A	P	Ic
								087	12.70	12.79	0.027	0.50	0.540	0.162	90.31
018	12.70	12.69	0.080	0.50	0.060	0.480	80.78
								075	12.70	12.74	0.041	0.50	0.240	0.246	89.65

(5) each operation and batch Comprehensive Control ability monitoring

Give weight to the parameter in each operation and quality index according to its significance level, be utilized respectively process stability comprehensive monitoring exponential model and calculate the process stability comprehensive monitoring index of loosening and gaining moisture operation, Screening and casing operation, thin plate drying process and mixed silk perfuming operation.And then, give weight according to its significance level to each operation, utilize lot stability comprehensive monitoring exponential model to calculate lot stability comprehensive monitoring index.Loosening and gaining moisture operation, Screening and casing operation, thin plate drying process and mixed silk perfuming operation weight arrange as shown in table 16.Loosening and gaining moisture operation, Screening and casing operation, thin plate drying process and the process stability comprehensive monitoring index of mixed silk perfuming operation and the result of calculation of lot stability comprehensive monitoring index are as shown in table 17.As shown in Table 17,087 total quality stability is better；018 loosening and gaining moisture and baking yarn quality have good stability, and total quality has good stability；075 loosening and gaining moisture quality stability is good, dries yarn quality poor stability, total quality less stable, need to control to improve to drying yarn quality.

Table 16 assessment item and weight are arranged

Table 17 stability comprehensive detection index assessment result

The invention is not restricted to above example, it is also possible to have many deformation.All within the spirit and principles in the present invention, all deformation that those of ordinary skill in the art can directly derive from present disclosure or associate, be all considered as protection scope of the present invention.

Claims (4)

1. a production of cigarettes process stability monitoring method, it is characterised in that: comprise the following steps successively:

One, the variable in production of cigarettes process to be classified, classification type includes the inspection of the eyes control type of tolerance requirements, without the inspection of the eyes control type of tolerance requirements, limiting control type and scope control type；

Two, sorted variable is carried out quantization signifying, wherein, inspection of the eyes control type variable is characterized by irrelevance and dispersion, and limiting control type variable is characterized by the degrees of offset of measured value Yu setting value；

Three, carry out single argument monitoring, difference according to types of variables adopts different monitoring models, inspection of the eyes control type variable adopts the inspection of the eyes control variable model based on irrelevance and dispersion to be monitored, and limiting control type variable adopts the limiting control variate model based on measured value Yu the degrees of offset of setting value to be monitored；

Four, carry out multivariate monitoring, multivariate monitoring includes operation comprehensive monitoring and batch comprehensive monitoring two parts, operation comprehensive monitoring adopts based on the average weighted process stability monitoring Synthesized Index Model of single argument monitoring result, and batch comprehensive monitoring adopts the lot stability comprehensive monitoring exponential model based on operation comprehensive monitoring result geometric average.

2. production of cigarettes process stability monitoring method according to claim 1, it is characterised in that:

The computing formula having the irrelevance of the inspection of the eyes control type variable of tolerance requirements in described step 2 isWherein, A represents irrelevance；Represent actual measurement meansigma methods；χ_pvExpression standard design load；δ_pvExpression standard design standard deviation, δ_pv=franchise/3；Subscript_pvRepresent standard value；

Being converted into limiting control type without the irrelevance of the inspection of the eyes control type variable of tolerance requirements by calculating relative variation in described step 2 to obtain, or control to limit the inspection of the eyes type being converted into franchise to obtain by increasing, the computing formula of relative variation isWherein, B represents relative variation；Represent actual measurement meansigma methods；χ_pvExpression standard design load；Subscript_pvRepresent standard value；

In described step 2, the computing formula of the dispersion of inspection of the eyes control type variable isWherein P represents dispersion；S represents actual measurement standard deviation；δ_pvExpression standard design standard deviation, δ_pv=franchise/3；Subscript_pvRepresent standard value；

In described step 2, the measured value of limiting control type variable with the computing formula of the degrees of offset of setting value isWherein, C represents the degrees of offset of measured value and setting value；Represent actual measurement meansigma methods；χ_pvExpression standard design load；χ_bestRepresentation theory or actual optimum value；Subscript_pvRepresent standard value；Subscript_bestRepresent optimal value.

3. production of cigarettes process stability monitoring method according to claim 1, it is characterised in that:

Described inspection of the eyes control variable model is: $I_c=I_{best}-(I_{best}-I_{base})\×\sqrt{{\left(\frac{A}{3}\right)}^2+P^2},$ Wherein, I_cRepresent inspection of the eyes type control variable STABILITY MONITORING index；I_bestRepresent optimum Monitoring Index；I_baseRepresent baseline Monitoring Index；A represents irrelevance；P represents dispersion；Subscript_cRepresent inspection of the eyes type control variable；Subscript_bestRepresent optimal value；Subscript_baseRepresent baseline value；

Described limiting control variate model is: I_d=I_base+(I_best-I_base) × C, wherein, I_dRepresent extreme value type control variable STABILITY MONITORING index；I_bestRepresent optimum Monitoring Index；I_baseRepresent baseline Monitoring Index；C represents the degrees of offset of measured value and setting value；Subscript_dRepresent extreme value type control variable；Subscript_bestRepresent optimal value；Subscript_baseRepresent baseline value.

4. production of cigarettes process stability monitoring method according to claim 1, it is characterised in that:

Described process stability monitoring Synthesized Index Model is:Wherein, G represents operation comprehensive monitoring index；I_iRepresent the STABILITY MONITORING index of i-th variable；W_iRepresent the weight of i-th variable；Subscript i=1,2,3...k；

Described lot stability comprehensive monitoring exponential model is:Wherein, L represents a batch comprehensive monitoring index；G_iRepresent the STABILITY MONITORING index of i-th operation；D_iRepresent the weight of i-th operation；Subscript i=1,2,3...n.