CN114417737B - Anomaly detection method and device for wafer etching process - Google Patents
Anomaly detection method and device for wafer etching process Download PDFInfo
- Publication number
- CN114417737B CN114417737B CN202210308558.0A CN202210308558A CN114417737B CN 114417737 B CN114417737 B CN 114417737B CN 202210308558 A CN202210308558 A CN 202210308558A CN 114417737 B CN114417737 B CN 114417737B
- Authority
- CN
- China
- Prior art keywords
- value
- module
- data
- time sequence
- wafer etching
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2413—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on distances to training or reference patterns
- G06F18/24133—Distances to prototypes
- G06F18/24143—Distances to neighbourhood prototypes, e.g. restricted Coulomb energy networks [RCEN]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/24323—Tree-organised classifiers
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Software Systems (AREA)
- Medical Informatics (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- General Factory Administration (AREA)
Abstract
The invention provides an anomaly detection method and device for a wafer etching process, wherein the method comprises the following steps: acquiring time sequence data and set parameter data corresponding to each module sensor in the production process of defective products of wafer etching equipment; calculating the data mean value and the overall mean value characteristics of different process stages according to the set parameter data, and acquiring the corresponding time sequence statistical characteristics of the different process stages according to the corresponding time sequence data; carrying out anomaly prediction on each module sensor according to the data mean value, the overall mean value characteristic and the corresponding time sequence statistical characteristic to obtain an anomaly prediction value; predicting by adopting a gradient lifting model according to the abnormal predicted value to obtain a characteristic importance value of each module sensor; respectively calculating the correlation coefficient of the abnormal predicted value of each module sensor and the product performance; carrying out normalization processing on the feature importance value and the correlation coefficient, and calculating a corresponding weighted average value; and performing exception sorting on each module sensor according to the weighted average value.
Description
Technical Field
The invention relates to the technical field of anomaly detection, in particular to an anomaly detection method and an anomaly detection device for a wafer etching process.
Background
Etching is a key process in the semiconductor manufacturing process, and unnecessary parts are selectively eroded from the surface of a wafer by a chemical or physical method in the etching process, so that the wafer mask pattern is carved. The semiconductor process is usually in nanometer level, has extremely harsh index requirements on the performance and stability of each module of equipment, an etching machine often comprises a complex etching control system, an electric device and a vacuum device, and the vulnerable parts of a subsystem need to be replaced and maintained regularly, performance drift of the etching equipment can be caused due to manufacturing, installation tolerance and small performance difference of parts in the process of replacing and maintaining the parts, so that the wafer etching yield is influenced, a certain number of samples need to be prepared for equipment warming up after the etching equipment is maintained in the actual production process, specific subsystem data are recorded and analyzed according to the experience of an equipment engineer in the warming up process, abnormal roots are checked out, and then relevant set parameters are adjusted to stabilize the performance of the etching machine.
In the related art, the online maintenance of the equipment is realized by selecting the spectral line intensity change in the emission spectral line spectrometer in the equipment for statistical analysis and drawing up a monitoring index according to the prior experience of an engineer, but the method depends heavily on the understanding of the engineer on the process and the equipment, and has poor applicability and low efficiency in different processes and equipment of manufacturers.
Disclosure of Invention
In order to solve the technical problems, the invention provides the abnormality detection method for the wafer etching process, when equipment is abnormal, the abnormality sensor can be automatically and quickly positioned, manual participation is not needed, and the detection efficiency and accuracy are greatly improved.
The technical scheme adopted by the invention is as follows:
an anomaly detection method for a wafer etching process comprises the following steps: acquiring time sequence data and set parameter data corresponding to each module sensor in the production process of defective products by wafer etching equipment; calculating the data mean value and the integral mean value characteristics of different process stages according to the set parameter data, and acquiring corresponding time sequence statistical characteristics of different process stages according to corresponding time sequence data; carrying out anomaly prediction on each module sensor by adopting a plurality of anomaly detection models according to the data mean value, the integral mean value characteristic and the corresponding time sequence statistical characteristic of different process stages so as to obtain an anomaly prediction value of each module sensor; predicting by adopting a gradient lifting model in machine learning according to the abnormal predicted value to obtain a characteristic importance value of each module sensor; respectively calculating correlation coefficients of the abnormal predicted values of the module sensors and product performance; carrying out normalization processing on the characteristic importance value and the correlation coefficient, and calculating a corresponding weighted average value; and performing exception sorting on each module sensor according to the weighted average value.
The time sequence statistical characteristics comprise data length, maximum value, minimum value, 25 percentile, 75 percentile, variance, standard deviation, skewness, kurtosis, peak number, graph symmetry, specific percentile change rate, graph area, enthalpy value of split box, extreme value occurrence position and data change gradient.
The normalizing the feature importance value and the correlation coefficient includes: and normalizing the feature importance value and the correlation coefficient by adopting a StandardScale mode in a sklern library.
An abnormality detection apparatus for a wafer etching process, comprising: the first acquisition module is used for acquiring time sequence data and set parameter data corresponding to each module sensor in the production process of defective products of the wafer etching equipment; the second acquisition module is used for calculating the data mean value and the overall mean value characteristics of different process stages according to the set parameter data and acquiring the corresponding time sequence statistical characteristics of the different process stages according to the corresponding time sequence data; a third obtaining module, configured to perform anomaly prediction on each module sensor according to the data mean value, the overall mean value characteristic, and the corresponding time sequence statistical characteristic at different process stages by using multiple anomaly detection models, so as to obtain an anomaly prediction value of each module sensor; the fourth acquisition module is used for predicting according to the abnormal prediction value by adopting a gradient lifting model in machine learning so as to acquire the characteristic importance value of each module sensor; the first calculation module is used for calculating correlation coefficients of the abnormal predicted values of the module sensors and product performance respectively; the second calculation module is used for carrying out normalization processing on the feature importance value and the correlation coefficient and calculating a corresponding weighted average value; and the sorting module is used for sorting the abnormality of each module sensor according to the weighted average value.
The time sequence statistical characteristics comprise data length, maximum value, minimum value, 25 percentile, 75 percentile, variance, standard deviation, skewness, kurtosis, peak number, graph symmetry, specific percentile change rate, graph area, enthalpy value of split box, position of occurrence of extreme value and data change gradient.
The second calculation module is specifically configured to: and normalizing the feature importance value and the correlation coefficient by adopting a StandardScale mode in a sklern library.
A computer device comprises a memory, a processor and a computer program which is stored on the memory and can run on the processor, wherein when the processor executes the computer program, the abnormality detection method for the wafer etching process is realized.
A non-transitory computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements the above-described anomaly detection method for a wafer etching process.
The invention has the beneficial effects that:
when the equipment is abnormal, the abnormal sensor can be automatically and quickly positioned without manual participation, and the detection efficiency and accuracy are greatly improved.
Drawings
FIG. 1 is a flowchart of an anomaly detection method for a wafer etching process according to an embodiment of the present invention;
fig. 2 is a schematic block diagram of an abnormality detection apparatus for a wafer etching process according to an embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be obtained by a person skilled in the art without making any creative effort based on the embodiments in the present invention, belong to the protection scope of the present invention.
Fig. 1 is a flowchart of an anomaly detection method for a wafer etching process according to an embodiment of the present invention.
As shown in fig. 1, the anomaly detection method for a wafer etching process according to the embodiment of the present invention may include the following steps:
and S1, acquiring time sequence data and set parameter data corresponding to each module sensor in the production process of the defective products of the wafer etching equipment.
Specifically, when defective products appear on wafer etching equipment, time sequence data and set parameter data of each module sensor in the production process of the defective products are collected, marking and grouping are carried out on the set parameter data and the time sequence data of the same process according to subsystems and technological processes corresponding to the module sensors, and the grouping is named as V1, V2, … … and Vn. Wherein different technological processes correspond to different sensors. Wherein, the V1 may include the time sequence data and setting parameter data of the first sensor; timing data and setting parameter data of the 2 nd sensor can be included in the V2; … …, respectively; the time series data and setting parameter data of the nth sensor may be included in Vn.
And S2, calculating the data mean value and the overall mean value characteristics of different process stages according to the set parameter data, and acquiring the corresponding time sequence statistical characteristics of different process stages according to the corresponding time sequence data.
Specifically, the data mean and the overall mean characteristic of different process stages can be calculated for the set parameter data, and the corresponding time sequence statistical characteristics of different process stages can be extracted for the time sequence data according to the data distribution characteristics.
The time sequence statistical characteristics can comprise data length, maximum value, minimum value, 25 percentile, 75 percentile, variance, standard deviation, skewness, kurtosis, peak number, figure symmetry, specific percentile change rate, figure area, binned enthalpy value, extreme value occurrence position and data change gradient.
And S3, performing anomaly prediction on each module sensor by adopting a plurality of anomaly detection models according to the data mean value, the overall mean value characteristic and the corresponding time sequence statistical characteristic of different process stages to obtain an anomaly prediction value of each module sensor.
In one embodiment of the present invention, the module sensors may be grouped first, for example, the module sensors may be divided into a temperature control module, a radio frequency module, and the like. And then respectively adopting a plurality of anomaly detection models to carry out anomaly prediction on each module sensor in different groups according to the data mean value, the integral mean value characteristic and the corresponding time sequence statistical characteristic of different process stages. For example, an ocsv (class support vector machine) model, an LOF (Local Outlier Factor) model, a CBLOF (Cluster-based Local Outlier Factor) model, and a KNN (K-nearest neighbor classification algorithm) model may be used to perform anomaly prediction on each module sensor in the temperature control module, and a KNN model, an Isolation Forest model, and an HBOS (Histogram-based Outlier Score) model may be used to perform anomaly prediction on each module sensor in the radio frequency module.
Further, a weighting algorithm is adopted for each module sensor to perform weighting processing on the prediction result of each abnormality detection model, so as to obtain an abnormality prediction value of each module sensor. Specifically, for a certain module sensor, the corresponding anomaly detection models can be randomly divided into n groups, the average fraction of the prediction results of each anomaly detection model in the n groups is calculated, n average fractions are obtained, n is a positive integer, and the highest fraction is selected from the n average fractions to serve as the anomaly prediction value of the module sensor. For example, for a certain module sensor in the temperature control module, the OCSVM model, the LOF model, the CBLOF model and the KNN model are randomly divided into 2 groups, i.e., the OCSVM model and the LOF model are one group, and the CBLOF model and the KNN model are the other group. Firstly, average scores of prediction results of an OCSVM model and average scores of prediction results of a CBLOF model and a KNN model are respectively calculated, and then the highest score of the two average scores is taken as an abnormal prediction value of the module sensor.
And S4, predicting according to the abnormal predicted value by adopting a gradient lifting model in machine learning to obtain the characteristic importance value of each module sensor.
Specifically, a gradient boost model (e.g., a LightGBM model) in machine learning may be adopted, the abnormal predicted values of the module sensors are used as input, whether the product performance meets the requirement is used as a target for classification training and prediction, and the feature importance values of the module sensors are output, wherein the division basis of the feature importance may be the Gain "of the gradient boost model.
And S5, calculating the correlation coefficient between the abnormal prediction value of each module sensor and the product performance.
Specifically, the speerman correlation coefficient of the abnormal prediction value of each module sensor and the product performance can be respectively calculated, wherein binarization processing can be performed according to whether the product performance test requirement is met, the value is assigned to 1 if the product performance test requirement is met, and the value is assigned to 0 if the product performance test requirement is not met.
And S6, normalizing the characteristic importance value and the correlation coefficient, and calculating a corresponding weighted average value.
In an embodiment of the present invention, the feature importance value and the correlation coefficient may be normalized by using a StandardScaler method in a sklern library, and then a weighted average value is calculated.
And S7, sorting the abnormality of each module sensor according to the weighted average value.
Specifically, the module sensors may be sorted according to the weighted average, that is, the probability of abnormality occurrence of each module sensor is sorted.
Specifically, the anomaly detection method comprises two stages, wherein the first stage is used for carrying out anomaly prediction on time sequence data and set parameter data of each sensor group and weighting prediction results of a plurality of anomaly detection models corresponding to each sensor, the second stage is used for taking the weighted results as input, carrying out product performance prediction by using a gradient lifting model and outputting a characteristic importance value, further calculating a correlation coefficient of the weighted results and product performance, and finally outputting the device anomaly sensor sequence through the characteristic importance data and the correlation coefficient.
Therefore, when performance drift occurs in the etching equipment, a field equipment engineer can be helped to quickly perform root cause analysis of equipment abnormity, the abnormal sensor is locked, a debugging strategy is formulated, and the reject ratio of the etching equipment is reduced.
To sum up, according to the anomaly detection method for the wafer etching process of the embodiment of the invention, the time sequence data and the set parameter data corresponding to each module sensor of the wafer etching equipment in the production process of the defective product are obtained, the data mean value and the overall mean value characteristics of different process stages are calculated according to the set parameter data, the time sequence statistical characteristics corresponding to different process stages are obtained according to the corresponding time sequence data, a plurality of anomaly detection models are adopted to carry out anomaly prediction on each module sensor according to the data mean value, the overall mean value characteristics and the corresponding time sequence statistical characteristics of different process stages so as to obtain the anomaly predicted value of each module sensor, a gradient promotion model in machine learning is adopted to carry out prediction according to the anomaly predicted value so as to obtain the characteristic importance value of each module sensor, and the correlation coefficient between the anomaly predicted value of each module sensor and the product performance is respectively calculated, and carrying out normalization processing on the characteristic importance value and the correlation coefficient, calculating a corresponding weighted average value, and carrying out exception sorting on each module sensor according to the weighted average value. Therefore, when the equipment is abnormal, the abnormal sensor can be automatically and quickly positioned, manual participation is not needed, and the detection efficiency and accuracy are greatly improved.
Corresponding to the above embodiment, the invention further provides an anomaly detection device for the wafer etching process.
As shown in fig. 2, the abnormality detecting apparatus for a wafer etching process according to an embodiment of the present invention may include: a first obtaining module 100, a second obtaining module 200, a third obtaining module 300, a fourth obtaining module 400, a first calculating module 500, a second calculating module 600 and a sorting module 700.
The first obtaining module 100 is configured to obtain time sequence data and set parameter data corresponding to each module sensor of the wafer etching apparatus in a defective product production process; the second obtaining module 200 is configured to calculate a data mean value and an overall mean value characteristic of different process stages according to the set parameter data, and obtain a timing statistic characteristic corresponding to the different process stages according to the corresponding timing data; the third obtaining module 300 is configured to perform anomaly prediction on each module sensor according to the data mean value, the overall mean value characteristic and the corresponding time sequence statistical characteristic of different process stages by using a plurality of anomaly detection models to obtain an anomaly prediction value of each module sensor; the fourth obtaining module 400 is configured to predict according to the abnormal prediction value by using a gradient lifting model in machine learning to obtain a feature importance value of each module sensor; the first calculation module 500 is used for calculating the correlation coefficient between the abnormal predicted value of each module sensor and the product performance; the second calculating module 600 is configured to perform normalization processing on the feature importance value and the correlation coefficient, and calculate a corresponding weighted average; the ranking module 700 is configured to rank the anomalies for each module sensor according to the weighted average.
According to one embodiment of the invention, the time sequence statistical characteristics comprise data length, maximum value, minimum value, 25 percentile, 75 percentile, variance, standard deviation, skewness, kurtosis, peak number, graph symmetry, specific percentile change rate, graph area, split-box enthalpy value, extreme value occurrence position and data change gradient.
The second computing module 600 is specifically configured to: and normalizing the feature importance value and the correlation coefficient by adopting a StandardScale mode in a sklern library.
It should be noted that, the anomaly detection apparatus for a wafer etching process according to the embodiment of the present invention may refer to the embodiment of the anomaly detection method for a wafer etching process, and details are not repeated here.
According to the abnormality detection device for the wafer etching process of the embodiment of the invention, the first acquisition module is used for acquiring time sequence data and set parameter data corresponding to each module sensor of the wafer etching equipment in the production process of defective products, the second acquisition module is used for calculating the data mean value and the overall mean value characteristic of different process stages according to the set parameter data and acquiring the time sequence statistical characteristic corresponding to different process stages according to the corresponding time sequence data, the third acquisition module is used for carrying out abnormality prediction on each module sensor according to the data mean value, the overall mean value characteristic and the corresponding time sequence statistical characteristic of different process stages by adopting a plurality of abnormality detection models to acquire the abnormality prediction value of each module sensor, and the fourth acquisition module is used for carrying out prediction according to the abnormality prediction value by adopting a gradient promotion model in machine learning to acquire the characteristic importance value of each module sensor, and the first calculation module is used for calculating the correlation coefficient between the abnormal predicted value and the product performance of each module sensor, the second calculation module is used for carrying out normalization processing on the characteristic importance value and the correlation coefficient, calculating a corresponding weighted average value, and carrying out abnormal sequencing on each module sensor according to the weighted average value through the sequencing module. Therefore, when the equipment is abnormal, the abnormal sensor can be automatically and quickly positioned, manual participation is not needed, and the detection efficiency and accuracy are greatly improved.
Corresponding to the above embodiment, the present invention further provides a computer device.
The computer device comprises a memory, a processor and a computer program which is stored on the memory and can run on the processor, and when the processor executes the program, the abnormality detection method for the wafer etching process of the embodiment is realized.
According to the computer equipment provided by the embodiment of the invention, when the equipment is abnormal, the abnormal sensor can be automatically and quickly positioned, manual participation is not needed, and the detection efficiency and accuracy are greatly improved.
In response to the above embodiments, the present invention also provides a non-transitory computer-readable storage medium.
The non-transitory computer readable storage medium of the embodiment of the present invention stores a computer program, and when the program is executed by a processor, the method for detecting an abnormality of a wafer etching process according to the above embodiment is implemented.
According to the non-transitory computer readable storage medium provided by the embodiment of the invention, when the equipment is abnormal, the abnormal sensor can be automatically and quickly positioned without manual participation, so that the detection efficiency and accuracy are greatly improved.
In the description of the present invention, the terms "first" and "second" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implying any number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. The meaning of "plurality" is two or more unless specifically limited otherwise.
In the present invention, unless otherwise expressly stated or limited, the terms "mounted," "connected," "secured," and the like are to be construed broadly and can, for example, be fixedly connected, detachably connected, or integrally formed; can be mechanically or electrically connected; either directly or indirectly through intervening media, either internally or in any other relationship. The specific meanings of the above terms in the present invention can be understood by those skilled in the art according to specific situations.
In the present invention, unless otherwise expressly stated or limited, the first feature "on" or "under" the second feature may be directly contacting the first and second features or indirectly contacting the first and second features through an intermediate. Also, a first feature "on," "over," and "above" a second feature may be directly or diagonally above the second feature, or may simply indicate that the first feature is at a higher level than the second feature. A first feature being "under," "below," and "beneath" a second feature may be directly under or obliquely under the first feature, or may simply mean that the first feature is at a lesser elevation than the second feature.
In the description herein, references to the description of the term "one embodiment," "some embodiments," "an example," "a specific example," or "some examples," etc., mean that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the invention. In this specification, the schematic representations of the terms used above are not necessarily intended to refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples. Furthermore, various embodiments or examples and features of different embodiments or examples described in this specification can be combined and combined by one skilled in the art without contradiction.
Any process or method descriptions in flow charts or otherwise described herein may be understood as representing modules, segments, or portions of code which include one or more executable instructions for implementing specific logical functions or steps of the process, and alternate implementations are included within the scope of the preferred embodiment of the present invention in which functions may be executed out of order from that shown or discussed, including substantially concurrently or in reverse order, depending on the functionality involved, as would be understood by those reasonably skilled in the art of the present invention.
The logic and/or steps represented in the flowcharts or otherwise described herein, such as an ordered listing of executable instructions that can be considered to implement logical functions, can be embodied in any computer-readable medium for use by or in connection with an instruction execution system, apparatus, or device, such as a computer-based system, processor-containing system, or other system that can fetch the instructions from the instruction execution system, apparatus, or device and execute the instructions. For the purposes of this description, a "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transport the program for use by or in connection with the instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of the computer-readable medium would include the following: an electrical connection (electronic device) having one or more wires, a portable computer diskette (magnetic device), a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Further, the computer-readable medium could even be paper or another suitable medium upon which the program is printed, as the program can be electronically captured, via for instance optical scanning of the paper or other medium, then compiled, interpreted or otherwise processed in a suitable manner if necessary, and then stored in a computer memory.
It should be understood that portions of the present invention may be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, various steps or methods may be implemented in software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, any one or combination of the following technologies, which are well known in the art, may be used: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application specific integrated circuit having an appropriate combinational logic gate circuit, a Programmable Gate Array (PGA), a Field Programmable Gate Array (FPGA), or the like.
It will be understood by those skilled in the art that all or part of the steps carried by the method for implementing the above embodiments may be implemented by hardware related to instructions of a program, which may be stored in a computer readable storage medium, and when the program is executed, the program includes one or a combination of the steps of the method embodiments.
In addition, functional units in the embodiments of the present invention may be integrated into one processing module, or each unit may exist alone physically, or two or more units are integrated into one module. The integrated module can be realized in a hardware mode, and can also be realized in a software functional module mode. The integrated module, if implemented in the form of a software functional module and sold or used as a stand-alone product, may also be stored in a computer readable storage medium.
The storage medium mentioned above may be a read-only memory, a magnetic or optical disk, etc. Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention, and that variations, modifications, substitutions and alterations can be made to the above embodiments by those of ordinary skill in the art within the scope of the present invention.
Claims (8)
1. An anomaly detection method for a wafer etching process is characterized by comprising the following steps:
acquiring time sequence data and set parameter data corresponding to each module sensor in the production process of defective products of wafer etching equipment;
calculating the data mean value and the overall mean value characteristics of different process stages according to the set parameter data, and acquiring the corresponding time sequence statistical characteristics of the different process stages according to the corresponding time sequence data;
carrying out anomaly prediction on each module sensor by adopting a plurality of anomaly detection models according to the data mean value, the integral mean value characteristic and the corresponding time sequence statistical characteristic of different process stages so as to obtain an anomaly prediction value of each module sensor;
predicting by adopting a gradient lifting model in machine learning according to the abnormal predicted value to obtain the characteristic importance value of each module sensor;
Respectively calculating correlation coefficients of the abnormal predicted values of the module sensors and product performance;
carrying out normalization processing on the characteristic importance value and the correlation coefficient, and calculating a corresponding weighted average value;
and performing exception sorting on each module sensor according to the weighted average value.
2. The abnormality detection method for the wafer etching process according to claim 1,
the time sequence statistical characteristics comprise data length, maximum value, minimum value, 25 percentile, 75 percentile, variance, standard deviation, skewness, kurtosis, peak number, graph symmetry, specific percentile change rate, graph area, enthalpy value of split box, position of occurrence of extreme value and data change gradient.
3. The anomaly detection method for the wafer etching process as claimed in claim 1, wherein the normalizing the feature importance value and the correlation coefficient comprises:
and normalizing the feature importance value and the correlation coefficient by adopting a StandardScaler mode in a sklern library.
4. An abnormality detection device for a wafer etching process, comprising:
The first acquisition module is used for acquiring time sequence data and set parameter data corresponding to each module sensor in the production process of defective products of the wafer etching equipment;
the second acquisition module is used for calculating the data mean value and the overall mean value characteristics of different process stages according to the set parameter data and acquiring the corresponding time sequence statistical characteristics of the different process stages according to the corresponding time sequence data;
a third obtaining module, configured to perform anomaly prediction on each module sensor according to the data mean value, the overall mean value characteristic, and the corresponding time sequence statistical characteristic at different process stages by using multiple anomaly detection models, so as to obtain an anomaly prediction value of each module sensor;
the fourth acquisition module is used for predicting according to the abnormal prediction value by adopting a gradient lifting model in machine learning so as to acquire the characteristic importance value of each module sensor;
the first calculation module is used for calculating correlation coefficients of the abnormal predicted values of the module sensors and product performance respectively;
the second calculation module is used for carrying out normalization processing on the feature importance value and the correlation coefficient and calculating a corresponding weighted average value;
And the sorting module is used for sorting the abnormality of each module sensor according to the weighted average value.
5. The apparatus for detecting abnormality in a wafer etching process according to claim 4,
the time sequence statistical characteristics comprise data length, maximum value, minimum value, 25 percentile, 75 percentile, variance, standard deviation, skewness, kurtosis, peak number, graph symmetry, specific percentile change rate, graph area, enthalpy value of split box, position of occurrence of extreme value and data change gradient.
6. The abnormality detection apparatus for the wafer etching process according to claim 4, wherein the second calculation module is specifically configured to:
and normalizing the feature importance value and the correlation coefficient by adopting a StandardScaler mode in a sklern library.
7. Computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor, when executing the computer program, implements the method for anomaly detection for a wafer etching process according to any one of claims 1-3.
8. A non-transitory computer readable storage medium having stored thereon a computer program, wherein the program, when executed by a processor, implements the method for anomaly detection for a wafer etching process according to any one of claims 1-3.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210308558.0A CN114417737B (en) | 2022-03-28 | 2022-03-28 | Anomaly detection method and device for wafer etching process |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210308558.0A CN114417737B (en) | 2022-03-28 | 2022-03-28 | Anomaly detection method and device for wafer etching process |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114417737A CN114417737A (en) | 2022-04-29 |
| CN114417737B true CN114417737B (en) | 2022-06-28 |
Family
ID=81264071
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210308558.0A Active CN114417737B (en) | 2022-03-28 | 2022-03-28 | Anomaly detection method and device for wafer etching process |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114417737B (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112562866A (en) * | 2021-02-19 | 2021-03-26 | 北京声智科技有限公司 | Detection system, detection method, detection device, and storage medium |
| CN112801497A (en) * | 2021-01-26 | 2021-05-14 | 上海华力微电子有限公司 | Anomaly detection method and device |
-
2022
- 2022-03-28 CN CN202210308558.0A patent/CN114417737B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112801497A (en) * | 2021-01-26 | 2021-05-14 | 上海华力微电子有限公司 | Anomaly detection method and device |
| CN112562866A (en) * | 2021-02-19 | 2021-03-26 | 北京声智科技有限公司 | Detection system, detection method, detection device, and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114417737A (en) | 2022-04-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP5091604B2 (en) | Distribution evaluation method, product manufacturing method, distribution evaluation program, and distribution evaluation system | |
| US9110461B2 (en) | Semiconductor manufacturing equipment | |
| JP2002318617A (en) | Management method for manufacture process | |
| KR102384189B1 (en) | Method and apparatus for predicting semiconductor failure time based on machine learning | |
| JP4568786B2 (en) | Factor analysis apparatus and factor analysis method | |
| KR102142167B1 (en) | Metrology target characterization | |
| CN107851294A (en) | The preventive maintenance device and method based on state of large-scale runtime | |
| CN115698882A (en) | Abnormal modulation cause identification device, abnormal modulation cause identification method, and abnormal modulation cause identification program | |
| US20230138127A1 (en) | Information processing method and information processing apparatus including acquiring a time series data group measured duirng a processing cycle for a substrate | |
| CN115698881A (en) | Abnormal modulation cause identification device, abnormal modulation cause identification method, and abnormal modulation cause identification program | |
| JP4723544B2 (en) | Substrate classification method and apparatus, program for causing a computer to execute the substrate classification method, and a computer-readable recording medium storing the program | |
| CN110928265B (en) | Data processing method, device and system, and computer readable recording medium | |
| CN112000081A (en) | Fault monitoring method and system based on multi-block information extraction and Mahalanobis distance | |
| CN114417737B (en) | Anomaly detection method and device for wafer etching process | |
| KR20230102269A (en) | Abnormal condition check method in wafer fabrication equipment and apparatus therefor | |
| CN117110847B (en) | Internal resistance test method, device and equipment for printed circuit board and storage medium | |
| CN117193088A (en) | Industrial equipment monitoring method and device and server | |
| CN114556350A (en) | Grain-level product modeling without grain-level input data | |
| KR100827442B1 (en) | Managing method of processing and managing system of processing | |
| JP7127477B2 (en) | LEARNING METHOD, APPARATUS AND PROGRAM, AND EQUIPMENT FAILURE DIAGNOSIS METHOD | |
| WO2023073941A1 (en) | Error factor estimation device, error factor estimation method, and computer-readable medium | |
| JP2008258486A (en) | Distribution analysis method and system, abnormality facility estimation method and system, program for causing computer to execute its distribution analysis method or its abnormality facility estimation method, and recording medium readable by computer having its program recorded therein | |
| TWI841020B (en) | Error factor estimation device, error factor estimation method, and computer readable medium | |
| WO2022263713A1 (en) | Monitoring and control of a semiconductor manufacturing process | |
| JP5384091B2 (en) | Inspection data management system and inspection data management method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |