CN103236383B - Dynamically determining and optimization method of plasma panel process parameters importance - Google Patents

Abstract

The invention discloses dynamically determining and optimization method of a kind of plasma panel process parameters importance, according to product hierarchy and the bad type of plasma panel production technology, adopt Feature Extraction Technology, after the result of given multiple feature extraction, the method that various features extracts organically is fused in the process of integration characteristic extraction, more stable feature extraction result is obtained from multiple angle, thus overcome the drawback and unsteadiness of applying single feature extracting method, final formation article on plasma display screen creation data more healthy and stronger, more stable, have more the displaying of practical significance.

Description

Dynamically determining and optimization method of plasma panel process parameters importance

Technical field

The present invention relates to a kind of data mining technology of plasma panel manufacture process characteristic, especially relate to dynamically determining and optimization method of a kind of plasma panel process parameters importance.

Background technology

The determination of plasma panel process parameters importance directly affects the final quality grade of plasma panel, be in the past by the EXPERIMENTAL DESIGN of construction period and repeatedly checking determine the importance of parameter in plasma panel manufacture process.Along with production programming extension and 7*24 hour uninterrupted volume production, need traditional test design method by a large amount of manpower of waste and kinetic energy consuming cost if continue, need production line to carry out stopping line to carry out the debugging design of test parameters importance simultaneously.This method reduces the production efficiency of volume production on the one hand, and test design method is by the manpower of at substantial and kinetic energy consuming cost simultaneously.Therefore a kind of new method is needed dynamically to carry out determination and the optimization of process parameters importance.

In plasma panel manufacture process volume production process, separate unit plasma panel process of producing product device parameter is included MES system storage in and is reached 1.2 ten thousand, every day more than data volume 10G, the parameter related to is more than 9000, in quantity, dimension and data generation speed has the large data characteristics of magnanimity.

The defining method that the plasma panel manufacturing process data of magnanimity like this further determines traditional parameter importance is inapplicable, and particularly " feature extraction " and " invariant feature extraction " technology must add plasma panel manufacturing process data characteristic.

1. the Feature Extraction Technology of plasma panel process parameters data: feature extraction is the characteristic distributions according to data, chooses a series of feature in order to build the process of model from feature database.An important hypothesis of feature extraction is containing a lot of redundancy or incoherent feature in data, and these redundant informations filter out by the object of feature extraction exactly, and finding out can the key character of characterization data.In this invention, we explore multiple basic Feature Extraction Technology according to plasma panel industrial nature, comprise information gain, information gain ratio, SVMs, relief etc.

2. the invariant feature extractive technique of plasma panel process parameters data: it is find out the feature with general character from the result of multiple feature extraction that integration characteristic extracts, can from the feature of multiple sides characterization data.By the method for each feature extraction, we can obtain the feature of a series of characterization data; But often kind of method can only remove decryption from a certain side.The object that the invariant feature of plasma panel process parameters data extracts effectively is merged the result of multiple feature extraction, therefrom extracts the consistency between multiple method gained feature, so that obtain can the invariant feature of characterization data.

Summary of the invention

In order to overcome the above-mentioned shortcoming of prior art, the invention provides dynamically determining and optimization method of a kind of plasma panel process parameters importance, according to product hierarchy and the bad type of plasma panel production technology, adopt Feature Extraction Technology, after the result of given multiple feature extraction, the method that various features extracts organically is fused in the process of integration characteristic extraction, more stable feature extraction result is obtained from multiple angle, thus overcome the drawback and unsteadiness of applying single feature extracting method, final formation article on plasma display screen creation data more healthy and stronger, more stable, have more the displaying of practical significance.

The technical solution adopted for the present invention to solve the technical problems is: dynamically determining and optimization method of a kind of plasma panel process parameters importance, comprises the steps:

Step one, the date of full operation creation data provided according to user, build the database of the creation data of full operation production technology;

Step 2, to the data content of all process steps in full operation do series connection process, and divide according to the grade of PDP display or bad type, the PDP display belonging to specific grade or specific bad type is divided into one group, other PDP display is divided into one group, then the parameter in the production technology corresponding to the PDP display between each group is alignd, be formed into full operation data;

Step 3, missing values process is carried out to the full operation data that step 2 is formed, and the measured value in all parameters to be normalized;

Step 4, employing feature extracting method carry out feature extraction to the product of different grouping, filter out the parameter list that can maximize the three groups of key characters comprising information gain, the ratio of gains, minimizing redundant degree and maximize the degree of correlation;

Step 5, the parameter list of three groups of key characters step 4 obtained are integrated, and obtain one group of unified important parameter list, which includes the parameter appeared in three groups of key character lists at least two group key characters.

Compared with prior art, good effect of the present invention is: saved time and resources costs spent by the mode by repetition test design and verification, improves the accuracy of the process parameters importance determined concerning plasma panel quality.And by the new historical data accumulation after adjustment, carry out the process parameters importance that this analytical model method of screw type recycling finds out plasma panel quality, manufacturing technology personnel are according to parameter importance, carry out the management and control of emphasis parameter, not only reach the object of lasting spiral lifting plasma panel yields, but also indirectly reduce the percent defective of manufacture process.Meanwhile, in the use procedure of method model, the breaks in production of production line can not be related to, avoid unnecessary line loss of stopping and lose.

Embodiment

Dynamically determining and an optimization method of plasma panel process parameters importance, comprises the steps:

Step one, the date of full operation creation data provided according to user, build the database of the creation data of full operation production technology;

Step 2, to the data content of all process steps in full operation do series connection process, and divide according to the grade of PDP display or bad type, the PDP display belonging to specific grade or specific bad type is divided into one group, other PDP display is divided into one group, then the parameter in the production technology corresponding to the PDP display between each group is alignd, be formed into full operation data;

Step 3, to step 2 formed full operation data carry out missing values process and normalized:

Described missing values process refers to: for missing values, fills according to the mean value of the measured value of the identical parameters of other PDP display, and target is that the missing values of this filling is not had an impact to analytical framework afterwards;

After missing values process, measured value in all parameters is normalized, by the measured value unification of this parameter in different PDP display to 0 average and unit variance this interval in and ensure that initial different relative size of shielding this measured value is constant simultaneously, illustrate as follows:

If α _t=(α ₁₁, α ₁₂, α _in) representing a group observations of n PDP display i-th parameter, after normalization, the adjustment mode of new argument measured value is:

in formula, μ _ifor α _iaverage, σ _ifor α _ivariance.

Step 4, employing feature extracting method carry out feature extraction to the product of different grouping, filter out the important parameter list that can maximize the important evaluation index comprising information gain, the ratio of gains, minimizing redundant degree and maximize the degree of correlation, concrete grammar is as follows:

Definition of T is a creation data set, comprises (x, y)=(x ₁, x ₂..., x _n, y), wherein x _i∈ vals (i) represents the value of i-th parameter in production PDP display data, and y represents the grade of this PDP display current or bad code.

(1) computing information gain (Information Gain):

$(T,a)=H\left(T\right)-{\mathrm{\Σ}}_{v\∈\mathrm{vals}\left(i\right)}\frac{\left|\right\{x\∈T|x_i=v\}|}{\left|T\right|}\·\left(\right\{x\∈T|x_i=v\}),$ Wherein

for comentropy, p (x _i) representative observe x _iprobability, x can be there is by historical data _inumber of times calculate.

(2) calculated gains ratio (Gain Ratio):

IGR (T, a)=IG (T, a)/IV (T, a) wherein, IG (T, a) is information gain, IV (T, a) account form is as follows:

$\mathrm{IV}(T,a)=-{\mathrm{\Σ}}_{v\∈\mathrm{vals}\left(i\right)}\frac{\left|\right\{x\∈T|x_i=v\}|}{\left|T\right|}\·{\log }_2\left(\frac{\left|\right\{x\∈T|x_i=v\}|}{\left|T\right|}\right);$

(3) maximum relation degree and minimum redundancy (minimum Redundancy and Maximum Relevancy, mRMR) is calculated

1) computational minimization redundancy (mR):

minW _I， $W_I=\frac{1}{{\left|T\right|}^2}\underset{a,b\∈T}{\mathrm{\Σ}}I,(a,b)$

Wherein, $I(a,b)={\mathrm{\Σ}}_{i,j}p(a_i,b_j)\·\log \frac{p(a_i,b_j)}{p\left(a_i\right)p\left(b_j\right)}$ , be defined as the interactive information between Two Variables a and b;

2) the maximization degree of correlation (MR) is calculated:

maxV _I, $V_I=\frac{1}{\left|T\right|}\underset{a\∈T}{\mathrm{\Σ}}I(y,a)$

3) calculating the mode that can meet mR and MR is to the full extent found the feature (parameter) of satisfied following condition by the method for searching of incremental:

maxΦ(W _I，V _I)， Φ＝V _I-W _I

Suppose that existing key character (parameter) set is for S _m-1, wherein contain m-1 key character (parameter), target is to find m key character (parameter) to make it to maximize Φ from residue character (parameter) set.

The computational methods of each feature selecting can export one group of key character (parameter) above, namely by IG, three stack features (parameter) that IGR, mRMR export respectively characterize has the feature (parameter) of discrimination to combine most in current PDP display grouping.

Step 5, three groups of important parameter lists step 4 obtained are integrated, and obtain out one group of unified important parameter list, which includes the parameter appeared in three groups of key character lists at least two groups:

The target of calculating can be divided into two classes for different comparison task by the inventive method:

1) important parameter for the PDP display of specific grade judges;

2) important parameter for the PDP display of specific bad type judges.

Claims (2)

1. dynamically the determining and an optimization method of plasma panel process parameters importance, is characterized in that: comprise the steps:

Step one, the date of full operation creation data provided according to user, build the database of the creation data of full operation production technology;

Step 2, to the data content of all process steps in full operation do series connection process, and divide according to the grade of PDP display or bad type, the PDP display belonging to specific grade or specific bad type is divided into one group, other PDP display is divided into one group, then the parameter in the production technology corresponding to the PDP display between each group is alignd, be formed into full operation data;

Step 3, missing values process is carried out to the full operation data that step 2 is formed, and the measured value in all parameters to be normalized;

Step 4, filter out the parameter list of three groups of key characters that can maximize and comprise information gain, the ratio of gains, minimizing redundant degree and maximize the degree of correlation; Wherein, the computational methods of information gain, the ratio of gains, minimizing redundant degree and the maximization degree of correlation are as follows:

Definition of T is a creation data set, comprises (x, y)=(x ₁, x ₂..., x _n, y), wherein x _i∈ vals (i) represents the value of i-th parameter in production PDP display data, and y represents the grade of this PDP display current or bad code;

(1) computing information gain IG (T, a):

wherein:

for comentropy, p (x _i) representative observe x _iprobability;

(2) calculated gains ratio:

IGR (T, a)=IG (T, a)/IV (T, a), wherein:

(3) minimum redundancy and maximum relation degree is calculated:

1) computational minimization redundancy mR:

Wherein, be defined as the interactive information between Two Variables a and b;

2) maximization degree of correlation MR is calculated:

Step 5, the parameter list of three groups of key characters step 4 obtained are integrated, and obtain one group of unified important parameter list, which includes the parameter appeared in three groups of key character lists at least two group key characters.

2. dynamically the determining and optimization method of plasma panel process parameters importance according to claim 1, it is characterized in that: described missing values process refers to: for missing values, fills according to the mean value of the measured value of the identical parameters of other PDP display.