CN114859821A - Self-detection, self-analysis and self-adaptive numerical control machine tool fuzzy control system - Google Patents

Self-detection, self-analysis and self-adaptive numerical control machine tool fuzzy control system Download PDF

Info

Publication number
CN114859821A
CN114859821A CN202210439003.XA CN202210439003A CN114859821A CN 114859821 A CN114859821 A CN 114859821A CN 202210439003 A CN202210439003 A CN 202210439003A CN 114859821 A CN114859821 A CN 114859821A
Authority
CN
China
Prior art keywords
data
self
fuzzy
real
time
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.)
Pending
Application number
CN202210439003.XA
Other languages
Chinese (zh)
Inventor
彭望
陈刚
杨鑫
徐一栋
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Zhejiang Sci Tech University ZSTU
Original Assignee
Zhejiang Sci Tech University ZSTU
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Zhejiang Sci Tech University ZSTU filed Critical Zhejiang Sci Tech University ZSTU
Priority to CN202210439003.XA priority Critical patent/CN114859821A/en
Publication of CN114859821A publication Critical patent/CN114859821A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B19/00—Programme-control systems
    • G05B19/02—Programme-control systems electric
    • G05B19/18—Numerical control [NC], i.e. automatically operating machines, in particular machine tools, e.g. in a manufacturing environment, so as to execute positioning, movement or co-ordinated operations by means of programme data in numerical form
    • G05B19/408—Numerical control [NC], i.e. automatically operating machines, in particular machine tools, e.g. in a manufacturing environment, so as to execute positioning, movement or co-ordinated operations by means of programme data in numerical form characterised by data handling or data format, e.g. reading, buffering or conversion of data
    • G05B19/4086—Coordinate conversions; Other special calculations
    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B2219/00—Program-control systems
    • G05B2219/30—Nc systems
    • G05B2219/35—Nc in input of data, input till input file format
    • G05B2219/35356—Data handling
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
    • Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
    • Y02P90/02—Total factory control, e.g. smart factories, flexible manufacturing systems [FMS] or integrated manufacturing systems [IMS]

Landscapes

  • Engineering & Computer Science (AREA)
  • Human Computer Interaction (AREA)
  • Manufacturing & Machinery (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Automation & Control Theory (AREA)
  • Numerical Control (AREA)

Abstract

The invention belongs to the field of numerical control machine tool data acquisition and artificial intelligence. The technical scheme is as follows: a self-detecting, self-analyzing and self-adapting fuzzy control system of a numerical control machine tool comprises the following components: the self-detection module: the industrial personal computer acquires real-time data of the numerical control machine tool collected by various sensors through network communication, acquires real-time information from a numerical control machine tool control system through the network communication, and then displays and preprocesses the acquired real-time data and real-time information in real time and stores the real-time data and the real-time information in a database; a self-analysis module: and reading the data of the database through SAS analysis software, and performing visualization and direct display on the read data by using a big data analysis algorithm of the SAS software so as to clearly and visually show the association between the data, and further performing data processing and analysis on the processing data to obtain key parameters influencing the processing precision. The system can automatically realize data acquisition and storage, rapid analysis and adjustment of numerical control machine tool processing parameters, and achieves higher processing precision.

Description

Self-detection, self-analysis and self-adaptive numerical control machine tool fuzzy control system
Technical Field
The invention belongs to the field of numerical control machine tool data acquisition and artificial intelligence, and particularly relates to a self-detection, self-analysis and self-adaptive numerical control machine tool fuzzy control system.
Background
Along with the strategic support of intelligent manufacturing and the high-speed development of the numerical control machine tool industry in recent years, the precision, efficiency and speed of the numerical control machine tool manufacturing industry in China are greatly improved. However, the technical level of the current stage of China has a certain gap compared with the international leading level. The artificial intelligence technology in China has made a great breakthrough through the rapid development of the past few years, the artificial intelligence scene fusion capability is continuously improved along with the increasing maturity of the artificial intelligence theory and technology, at present, the artificial intelligence technology has been technically grounded in a plurality of fields of finance, medical treatment, security protection, education, traffic, manufacturing, retail and the like, and the application scenes are more and more abundant.
At present, the invention relates to a self-detection, self-analysis and self-adaptive numerical control machine tool fuzzy control system applied to intelligent manufacturing of numerical control machine tools, which can adjust precision influence parameters in the whole processing process in real time by the steps of collecting industrial real-time data, establishing a complete database, analyzing big data, autonomously making decision and adjusting and the like, and can achieve the aims of high precision, high efficiency and strong self-adaptive capacity. The existing data acquisition and monitoring system of the numerical control machine is mainly used for realizing data acquisition and remote monitoring and diagnosis, and does not relate to the aspects of autonomous analysis, inference, decision and the like in a human brain mode, so that the research of a highly intelligent numerical control machine is necessary.
Disclosure of Invention
The invention aims to overcome the defects of the background technology and provides a self-detection, self-analysis and self-adaptation fuzzy control system of a numerical control machine tool, which can automatically realize the collection and storage of data, the rapid analysis and the adjustment of the machining parameters of the numerical control machine tool so as to achieve higher machining precision.
The technical scheme provided by the invention is as follows:
a self-detection, self-analysis and self-adaptive numerical control machine tool fuzzy control system comprises:
the self-detection module: the industrial personal computer acquires real-time data of the numerical control machine tool collected by various sensors through network communication, acquires real-time information from a numerical control machine tool control system through the network communication, and then displays and preprocesses the acquired real-time data and real-time information in real time and stores the real-time data and the real-time information in a database;
a self-analysis module: reading the data of the database through SAS analysis software, and performing visualization and direct display on the read data by using a big data analysis algorithm of the SAS software so as to clearly and visually show the association between the data, and further performing data processing and analysis on the processing data to obtain key parameters influencing the processing precision; so as to facilitate the adjustment and correction of the following fuzzy control system;
the self-adaptive module is based on key parameters obtained by the self-analysis module, and adopts a Matlab tool kit to design a fuzzy controller module for factors influencing the machining precision, so that the control of the numerical control machine tool is realized; the fuzzy controller module comprises a fuzzification module, a fuzzy rule module library, a fuzzy reasoning module and a clarification module.
The fuzzification module comprises the determination of input and output variables and the selection of membership degrees; the fuzzy rule base module is used for making a corresponding fuzzy rule base according to expert experience; the fuzzy reasoning module carries out fuzzy reasoning according to rules in the fuzzy rule base, obtains fuzzy output quantity through fuzzy input quantity reasoning, and completes functions through the fuzzy reasoning and the fuzzy rule base; the clarification module comprises the determination of a clarification method and the conversion of a data type, the clarification (defuzzification) is carried out on the result obtained by the fuzzy inference, and then the control is realized by converting the data type into a language which can be recognized by a computer.
The real-time data comprises an actual current value of the main shaft, an actual voltage value of the main shaft, an actual rotating speed of the main shaft, a vibration value and a noise value generated during processing;
the real-time information comprises the actual feeding rate of the main shaft, the current tool number, tool compensation, the tool position, the system alarm state and specific alarm information;
and the real-time display is to display the real-time data and the real-time information on a human-computer interface of the industrial personal computer in real time.
The preprocessing is to calibrate the real-time information and the real-time data for time, and to perform data cleaning, data integration, data conversion and data protocol preprocessing.
And the storage is stored in a background real-time detection and monitoring system database.
The invention has the beneficial effects that:
the invention can realize the real-time acquisition of the processing data and the sensor data of the numerical control machine; meanwhile, the data are stored in a database after being preprocessed, and key parameters influencing the machining precision are found by reading the data in the database and then quickly and accurately analyzing the machining precision influencing factors by adopting big data analysis software. The fuzzy control algorithm is utilized to realize real-time adjustment on the numerical control machine tool processing parameters according to the results generated by analysis, so that higher processing precision is achieved; the method plays a positive role in the real realization of future intelligent numerical control machine workshops and unmanned factories.
Drawings
FIG. 1 is a schematic diagram of the system architecture of the present invention.
FIG. 2 is a schematic flow chart of the operation of the fuzzy controller of the present invention.
FIG. 3 is a schematic diagram of the fuzzy controller of the present invention.
FIG. 4 is a graph of the membership function of input error in the fuzzification step in example 1 of the present invention.
FIG. 5 is a graph of the membership function of rotational speed in the fuzzification step in example 1 of the present invention.
FIG. 6 is a graph of the feed rate membership function in the fuzzification step in example 1 of the present invention.
Fig. 7 is a rule display diagram of the fuzzy rule base in embodiment 1 of the present invention.
FIG. 8 is a fuzzy inference result graph of E and rotation speed in embodiment 1 of the present invention.
FIG. 9 is a diagram showing the fuzzy inference results of E and the feed rate in embodiment 1 of the present invention.
Detailed Description
The following further description is made with reference to the embodiments shown in the drawings.
The invention provides a self-detection, self-analysis and self-adaptive numerical control machine tool fuzzy control system, which is an intelligent system for data acquisition, analysis and decision autonomy designed on the basis of a new algebra control system.
The invention relates to a self-detection, self-analysis and self-adaptive numerical control machine tool fuzzy control system, which comprises a self-detection module, a self-analysis module and a self-adaptive module; the structure is shown in figure 1.
In the self-detection module, an industrial personal computer is in network communication with various sensors (including a three-axis vibration sensor, a noise sensor, an alternating current transmitter and an alternating voltage transmitter) for acquiring real-time data, and the data acquired by the various sensors in real time is acquired, wherein the data includes a main shaft actual current value, a main shaft actual voltage value, a main shaft actual rotating speed, a vibration value and a noise value generated during processing; meanwhile, the industrial personal computer is in network communication with the numerical control machine controller, and the numerical control machine controller transmits the actual feeding rate of the main shaft, the current tool number, the tool compensation, the tool position, the system alarm state and specific alarm information to the industrial personal computer in real time; the real-time data and the real-time information are displayed on a human-computer interface of the industrial personal computer in real time; meanwhile, the industrial personal computer calibrates the data information for time, and stores the data information into a background real-time detection and monitoring system database after preprocessing steps such as data cleaning, data integration, data conversion, data reduction and the like.
In the self-detection module, the adopted software is an open-source MYSQL database, visual studio 2012 and data acquisition software autonomously developed based on MySQL, visual studio 2012 and SYNTEC official API interface functions.
The numerical control system controller (preferably SYNTEC numerical control system controller) is connected with an industrial personal computer through a network cable to realize network communication, and the controller is adapted to a power supply to convert alternating current into 24V direct current for power supply; and sending instructions to read internal processing data by using data acquisition software and a human-computer interface of an industrial personal computer.
The data acquisition software is designed based on visual studio 2012 and MYSQL database, the visual studio 2012 and the MYSQL database are connected with the database through a visual studio for MYSQL plug-in, and main modules of the data acquisition software comprise a real-time display main interface, a user management module, a data processing module and a serial communication testing module.
The real-time display main interface mainly comprises a basic information and setting area, a data real-time display area, a curve real-time display area, a processing program display area, a processing track map area and the like of the acquisition equipment.
The basic information and setting area comprises an IP address of the acquisition equipment, an acquisition equipment port, Beijing time, a test button for judging whether the database connection is successful or not and a button for starting or ending acquisition.
The data real-time display area comprises information such as the starting time of the numerical control system, the cutting time, the processing time, the alarm state, specific alarm information, the actual current of the spindle, the actual voltage, the actual power, the actual feed rate, the actual spindle rotating speed, the coordinate position information of the cutter and the like.
The curve display area comprises real-time data curve drawing of the vibration data, the noise data and the actual power data in the three-axis direction (X, Y, Z) acquired.
The processing program display area is a program section for displaying the execution of the current numerical control system; the processing track map area is used for drawing a simulated tool route map by using the collected tool position information point data.
The user management module mainly comprises functions of adding, deleting and quitting the user information logged in by the software system.
The data processing module mainly comprises data query, data deletion and data export functions; the information in the database is inquired and deleted and the corresponding data is exported to a TXT text format or an Excel table by establishing the connection with the MySQL database and screening date and time conditions through visual studio for MySQL.
The serial communication test module is used for testing whether the data of each sensor can be normally received and transmitted and normally read and written and stored in the database, and comprises basic function settings of serial communication, such as serial number, baud rate, stop bit, check bit, data bit and the like, issuing of a read instruction, receiving of transmitted data and real-time drawing of the received data (the latest 10 data points are adopted).
The three-axis vibration sensor is arranged above the spindle motor in a magnetic suction mode and used for detecting three-axis direction vibration data of the spindle.
The noise sensor is arranged on a machine tool body in the numerical control machine tool in a threaded connection mode and used for detecting the noise generated during machining.
The alternating current transducer is arranged on the ground in a guide rail way, and a power output line of the spindle motor penetrates through a small hole in the alternating current transducer to detect the change of current in the alternating current transducer; the same applies to the alternating voltage transducer.
The three-axis vibration sensor, the noise sensor, the alternating current transducer and the alternating voltage transducer are all powered by a DC (24V) switching power supply; the sensors are connected to a 485 communication interface on the industrial personal computer through self-contained 485 connecting wires to realize network communication, and then corresponding data reading instructions are sent through the human-computer interface of the industrial personal computer to obtain data.
The man-machine interface of the industrial personal computer is a man-machine interface which is developed based on an open source MYSQL database and a visual studio 2012 platform and used for acquiring one key and automatically sending instructions; the actual current value of the main shaft, the actual voltage value of the main shaft, the actual rotating speed of the main shaft, the actual feeding rate of the main shaft, the current tool number and tool compensation, the position of the tool, the actual feeding rate, the vibration value and the noise value generated during processing, the system alarm state and specific alarm information can be acquired through the interface.
The man-machine interface calls the current system time to be synchronously stored when the collected information and data are displayed and stored in the MySQL database so as to realize the calibration time, and calls the MySQL insert statement instruction to be stored in the background real-time detection and monitoring system database after the preprocessing steps of data cleaning, data integration, data conversion, data reduction and the like.
The data cleaning uses a button (ETL cleaning tool) to simply process the collected information; types include cleanup of missing values and outliers. For missing or abnormal values, simple direct deletion can be adopted; if the missing value and the abnormal value are excessive, the missing value or the abnormal value is added to a set default value or an interpolation method (Lagrange interpolation and the like) is adopted to regularly fill the data after threshold judgment.
The data integration is a process of collecting and putting various data into a database. In the data collection, entity identification, redundant attribute identification and data value conflict are carried out when multiple data sources are integrated.
The entity identification refers to whether different data in one or more data sources are described as the same entity or not, and the same entity can be used for integration operations such as data deduplication in data integration. Briefly illustrated with a database example: if the A table has a field ID and the B table has a field NUM, then determine if the two fields are the same entity attribute? If the attribute is the same, the two fields can be used as a multi-table association condition during integration, and one value is reserved when a new table is generated.
The redundant attribute identification refers to whether there is a correlation between some attributes, or one attribute can be derived from another attribute. Correlation analysis detection may be used for redundant attributes, i.e., given two numeric attributes A, B, the degree of correlation of one attribute with the other is measured by the correlation number based on the attribute values. If the correlation number reaches a preset value, the two attributes are considered to be correlated, and one attribute is reserved; otherwise, both are retained in the new table. Common redundancy correlation analysis methods include pearson product distance coefficients, chi-square test, covariance of numerical attributes, and the like.
The data value conflict refers to that the attribute values of different data sources for the same entity are different, possibly due to unit inconsistency. Then, judging the proportional relation of the two data values on the value, judging whether the two data values are caused by the non-uniform unit through the proportional relation, and if so, retaining one data value; otherwise, both are retained in the new table.
The data conversion refers to the normalization, discretization and sparsification of data, so that the purpose of subsequent data analysis is achieved.
The data normalization processing means that dimensions of different features in the data may be inconsistent, and the difference between values may be large, so that the data analysis result is influenced without processing. Therefore, the data needs to be scaled to fall in a specific area for comprehensive analysis. The data normalization method mainly comprises the following three methods:
max-min normalization: mapping data to [0,1 ]]The interval of time is,
Figure BDA0003614286040000051
score normalization: the mean of the data after treatment was 0, the variance was 1, and the Z-Score was normalized: the mean value of the processed data was 0, the variance was 1,
Figure BDA0003614286040000052
log transformation: in time series data, for variables with large difference of data magnitude, Log function transformation is usually performed, x new ＝log 10 x。
The discretization processing means that continuous data is segmented into a segment discretization interval, and an equal frequency method and an equal width method can be adopted. The equal frequency method is to make the number of samples in each bin equal, for example, n is 50 for total samples, k is 5 for bins, and the binning principle is to ensure that the number of samples falling into each bin is 10. The equal width method comprises the following steps: the box widths of the attributes are made equal, for example, the length variable (between 0-1000), can be divided into five equal width boxes of [0,200], [200,400], [400,600], [600,800], [800,1000 ].
When the above-mentioned thinning processing is directed at discrete and nominal variables and cannot be performed in an ordered LabelEncoder, it is usually considered to perform thinning processing on the variables with 0,1 dummy variables, for example, a flower type variable contains four different values of trumpet flower, rose, peony and narcissus, and the variables are converted into four dummy variables of is _ trumpet flower, is _ rose, is _ peony and is _ narcissus. If the variable has more different values, the values with less occurrence times are uniformly classified as 'rare' according to the frequency. The sparse processing is beneficial to fast convergence of the model and can improve the anti-noise capability of the model.
The data specification refers to that a large amount of data is represented by a specification of a data set, the memory is reduced, and the integrity of the original data is still maintained, and the data specification comprises a dimension specification and a dimension transformation.
The data used by the dimension reduction test paper for data analysis may contain hundreds of attributes, and most of the attributes are irrelevant and redundant for our analysis. Then, irrelevant attributes are deleted through the dimension specification, the data volume is reduced, and the loss of information is guaranteed to be minimum. The function can be realized by selecting attribute subsets, namely finding out the minimum attribute set as a target, enabling the probability distribution of the data class to be as close to the original distribution using all attributes as possible, carrying out feature subset screening through a Recursive feature elimination algorithm (RFE) in python scinit-lean, and generally considering establishing an SVM or a regression model.
The dimension transformation refers to reducing existing data to a smaller dimension while ensuring the integrity of data information. The dimension transformation mainly comprises the following methods: 1. principal Component Analysis (PCA) and Factor Analysis (FA): PCA maps the current dimension to a lower dimension by means of spatial mapping, so that the variance of each variable in the new space is maximized. The FA is then a common factor (smaller dimension) to find the current feature vector, describing the current feature vector with a linear combination of common factors. 2. Singular Value Decomposition (SVD): SVD has lower dimensionality reduction interpretability and larger computation amount than PCA, and is generally used for dimensionality reduction on sparse matrices, such as picture compression, recommendation systems. 3. Clustering: features of a certain class that have similarities are grouped into a single variable, thereby greatly reducing dimensionality. 4. Linear combination: and performing linear regression on a plurality of variables, giving weight to the variables according to the voting coefficient of each variable, and combining the variables into one variable according to the weight.
The invention adopts MYSQL database to store real-time data in the processing process, and the storage format is as follows
User information table: ID (Primary Key), user name, password
Alarm information table ID (primary key), alarm information, alarm time, current date
Processing a data table: ID (master key), machining time, cutting time, working time, cycle time (work), number of workpieces, number of required workpieces, total number of workpieces, actual feed rate, actual spindle speed, current system time.
Sensor data sheet: ID (main key), vibration (X axis), vibration (Y axis), vibration (Z axis), noise, current value, voltage value, actual power.
The self-analysis module (prior art) reads a large amount of processing data from the database formed by self-detection through the connection of the SAS analysis software and the database, and processes and analyzes the processing data through the data processing and data analysis functions of the SAS software after reading the required data to obtain key parameters influencing the processing precision, so as to facilitate the adjustment and correction of a subsequent fuzzy control system.
The software part of the self-analysis module mainly comprises SAS software and a MySQL database. Analyzing data information exported from the MySQL database by using a big data analysis algorithm of SAS software, and realizing the visual processing of data by using the data presentation and graphic display functions of an industrial personal computer interface; and further obtaining factors causing machining precision fluctuation in the machining process, and outputting the influencing factors to an Excel table for storage so as to facilitate subsequent adaptive module control.
The self-adaptive module is mainly used for designing a fuzzy controller module for factors influencing the machining precision based on an analysis result in the self-analysis module; the fuzzy controller module comprises a fuzzification module, a fuzzy rule base module, a fuzzy reasoning module and a sharpening module. The fuzzification module comprises the determination of input and output variables and the selection of membership degrees; the fuzzy rule base module is mainly used for making a corresponding fuzzy rule base according to expert experience. The fuzzy reasoning module carries out fuzzy reasoning according to rules in the fuzzy rule base, obtains fuzzy output quantity through fuzzy input quantity reasoning, and completes functions through the fuzzy reasoning and the fuzzy rule base. The clarification module comprises the determination of a clarification method and the conversion of a data type, the clarification (defuzzification) is carried out on the result obtained by the fuzzy inference, and then the control is realized by converting the data type into a language which can be recognized by a computer.
The self-adaptive module reads an Excel table of an analysis result of factors influencing the machining precision, which is obtained from the analysis module, through an industrial personal computer, and designs the fuzzy controller aiming at the factor columns influencing the machining precision in the Excel.
The structure of the fuzzy controller is shown in FIG. 3; the CNC _ fuzzy in the figure is the saved file name and the Chinese name of mamdani is the traditional fuzzy reasoning method.
The fuzzy controller design is divided into four steps of fuzzification, fuzzy rule base, fuzzy reasoning and clarification.
The fuzzification process comprises fuzzy quantization of input and output variables and selection of membership.
The fuzzy quantization of the input and output variables comprises the determination of input and output quantities, the determination of basic discourse domain and the determination of fuzzy set. The input quantity is generally designed as an error E and an error change rate EC, and the output quantity u is designed as a control quantity. The basic universe of discourse refers to the actual range of variation of a variable, and the basic universe of discourse of the error is [ -x ] e ,x e ]The basic domain of error rate of discourse is [ -x [ ] c ,x c ]The fundamental universe of argument of the output variable of the fuzzy controller (the control variable of the system) is [ -y u ,y u ]。
There are two main methods for selecting the membership function: 1. a fuzzy statistical method; 2. an assignment method; to implement the obfuscation, a relationship is established between the precision quantity and the fuzzy quantity representing the fuzzy language, i.e. the degree of membership of each element in the domain to the respective fuzzy linguistic variable is determined. Membership describes the degree to which a certain amount is attributed to a certain fuzzy linguistic variable.
The fuzzy statistics method refers to determining the degree of membership of variables by making random surveys. Taking the determination of the age of the young as an example, taking the age of a person as a domain U, randomly surveying n random passers, proposing the most suitable section considered by the random passers after considering the age of the young, and dividing the number of people in each section by the total number of sample people to be taken as the membership frequency, namely the membership degree, of the age of the young.
The assignment method is to put the self instance into the proper function model to complete the determination of the membership degree through the existing function model. Common function models include rectangular, trapezoidal, parabolic, normal, and cauchy. For example, a membership function for "young" is determined. The more the age is, the less the younger population, so a mathematical model of a decreasing function is taken, such as selecting a Cauchy-type function model.
The fuzzy rule base comprises a database and a rule base, and the fuzzy rule base is made according to expert experience and daily experience of operators.
The database stores membership vector values (i.e. sets of corresponding values after domain level discretization) of all fuzzy subsets of all input and output variables, and if the domain is a continuous domain, the membership vector values are membership functions. And providing data for fuzzy inference in the process of solving the fuzzy relation equation of the rule inference.
The rules of the rule base are based on expert knowledge or long-term accumulated experience of a manual operator, and are a language representation of human intuitive reasoning. Fuzzy rules are usually formed by connecting a series of relation words, such as if-then, else, also, end, or, etc., and the relation words must be translated to digitize the fuzzy rules. The most commonly used relations are if-then, also, and for multivariable fuzzy control systems, and, or, etc.
The fuzzy inference refers to carrying out fuzzy inference through the synthesis of a fuzzy vector E of an error and a fuzzy vector EC of an error change rate and a fuzzy relation R to obtain a fuzzy vector of a controlled quantity. According to fuzzy set and fuzzy relation theory, different fuzzy reasoning methods can be used for different types of fuzzy rules, and the following takes the reasoning of the commonly used fuzzy rule of if A, then B is output if A is known as input, and B 'is output if A' is known as input, then B 'is obtained by using a synthesis rule to obtain A' R, wherein the fuzzy relation R is defined as mu R (x, y) min [ mu A (x), mu B (y)]. For example, the fuzzy set A of the known input and the fuzzy set B of the known output are respectively (wherein a is i ，b i Representing the discourse field elements corresponding to the fuzzy sets, and mu i Representing the corresponding degree of membership, "/" does not represent a notion of a score):
A＝1.0/a 1 +0.8/a 2 +0.5/a 3 +0.2/a 4 +0.0/a 5
B＝0.7/b 1 +1.0/b 2 +0.6/b 3 +0.0/b 4
Figure BDA0003614286040000081
Figure BDA0003614286040000082
note: in the above operation, "#" is the operator of fetch.
The above example only demonstrates a rule operation process, since the control rule base of the system is composed of several rules, a corresponding fuzzy relation can be obtained for each inference rule, n rules have n fuzzy relations, R 1 ,R 2 ,...,R n For all fuzzy relations corresponding to all control rules of the whole system and n fuzzy relations R i (i ═ 1,2,. n) and operated on to yield:
Figure BDA0003614286040000091
the above-mentioned clarification is that the control quantity obtained after fuzzy inference is a fuzzy vector, and can not be used as control quantity, and also it has need of making once data type conversion to obtain clear controllable quantity output. The current data type conversion methods (clarification) are as follows:
(1) maximum membership method
(2) Center of gravity method
(3) Weighted average method
The gravity center method is an example and is briefly described as follows:
Figure BDA0003614286040000092
e.g. u' ═ 0.1/2+0.8/3+1.0/4+0.1/6
Then u ═ 4 ═ 2 × 0.1+3 × 0.8+4 × 1+5 × 0.8+6 × 0.1)/0.1+0.8+1.0+0.8+0.1 ═ 0.8
The invention adopts fuzzy control to realize real-time adjustment of the parameters influencing the processing precision; because excessive parameters are involved in the fuzzy control, the fuzzy rule is very complicated, the calculation amount is very large, the calculation speed is very high, and the efficiency is greatly reduced, so that the key parameters influencing the machining precision are obtained by adopting a big data analysis method at first, and then the design of a fuzzy controller is carried out on the key parameters, so that the faster control effect is achieved, the machining precision and the production intelligence of parts are improved, and the production efficiency is improved.
Example 1
The whole controller design process is briefly described by a specific numerical control machine tool example: (the factors influencing the precision of the numerical control machine tool are many, only the rotating speed of the main shaft and the feed rate of the cutter are taken as examples)
1. Fuzzification:
(1) fuzzy quantization of input and output variables
The input variable is selected as the difference between the actual machining precision of the workpiece and the qualified value of the machining precision, namely the error E, and the output variable is selected as the variation of the feed rate of the tool and the variation of the rotating speed of the main shaft.
The fuzzy sets of input and output variables are all taken as NB, NS, ZE, PS, PB. Let the input domain be [ -33 ], and the domain of error E be X { -3, -2, -1, -0, +0, +1, +2, +3 }; the output domain is [ -33 ], and the domains of the output rotating speed variation and the feed rate variation are Y { -3, -2, -1, -0, +0, +1, +2, +3 }. The structure of the fuzzy controller is shown in fig. 3.
(2) Selection of degree of membership
The membership degree of each element in the discourse domain X and the discourse domain Y to the fuzzy set determined by the variable value of the error or the controlled variable language is given by the function type, and the trigonometric function is selected by the membership function type. The membership degree curve of the input and output linguistic variable values sequentially comprises an input error E membership degree function curve, a rotating speed membership degree function curve and a feeding rate membership degree function curve shown in figures 4 to 6. The fuzzy set of linguistic variable values E, rotational speed and feed rate over domains X and Y can be given by table 1.1.
TABLE 1.1 linguistic variable assignment Table
Figure BDA0003614286040000101
2. Fuzzy rule base
The language according to expert experience and the daily experience of the operator can be described as:
(1) if the actual machining precision grade is lower than the qualified machining precision grade, the feed rate is reduced, and the rotating speed of the main shaft is increased; the lower the phase difference accuracy level, the smaller the feed rate and the larger the rotation speed.
(2) If the actual machining precision grade meets the qualified machining precision grade, the parameters are kept unchanged.
(3) If the actual machining precision grade is higher than the qualified machining precision grade, considering the machining efficiency, increasing the feed rate and reducing the rotating speed of the main shaft; the higher the phase difference accuracy level, the larger the feed rate and the smaller the rotation speed.
The above experience is expressed as a fuzzy statement:
(1) if E, NB, then, NB and PB
(2) if E-NS-then speed NS and PS feed rate
(3) ZE the rotation speed and feed rate ZE
(4) if E is PS then speed PS and feed rate NS
(5) if E PB then PB and feed rate NB
Written as a control rule table as shown in Table 1.2
TABLE 1.2 control rules Table
Figure BDA0003614286040000102
Figure BDA0003614286040000111
The rules formed in Matlab are shown in fig. 7.
3. Fuzzy inference
In the embodiment, fuzzy reasoning is realized by means of a fuzzy control tool box in MATLAB, and an E and rotating speed reasoning result graph shown in FIG. 8 and an E and feed rate reasoning result graph shown in FIG. 9 are obtained.
4. Clarification
In the embodiment, the center of gravity method which is widely used is adopted for the clarification treatment. As shown in fig. 8 and 9, the center-of-gravity sharpening process is performed with E equal to 0.5.
And obtaining the rotating speed which is 1 and the feed rate which is 1 through MATLAB fuzzy reasoning calculation. And (3) taking the fuzzy set to which the maximum membership degree belongs from the table 1.1 to obtain the fuzzy set to which the rotating speed belongs at the moment as PS, and the fuzzy set to which the feed rate belongs as NS.
From table 1.1, the feed rate is 0.7/-2+0.7/-1 at 0.7/1+0.7/2
The gravity center method comprises the following steps:
Figure BDA0003614286040000112
the center of gravity method is clarified:
the rotation speed U1 is (0.7 × 1+0.7 × 2)/(1+2) is 0.7
The feed rate U2 ═ 0.7 × (-1) +0.7 × (-2))/(((-1) + (-2)) -0.7.

Claims (6)

1. A self-detecting, self-analyzing and self-adapting fuzzy control system of a numerical control machine tool comprises the following components:
the self-detection module: the industrial personal computer acquires real-time data of the numerical control machine tool collected by various sensors through network communication, acquires real-time information from a numerical control machine tool control system through the network communication, and then displays and preprocesses the acquired real-time data and real-time information in real time and stores the real-time data and the real-time information in a database;
a self-analysis module: reading the data of the database through SAS analysis software, and performing visualization and direct display on the read data information through a big data analysis algorithm of the SAS software so as to clearly and visually show the association between the data, and further performing data processing and analysis on the processing data information to obtain key parameters influencing the processing precision; so as to facilitate the adjustment and correction of the following fuzzy control system;
the self-adaptive module is based on key parameters obtained by the self-analysis module, and adopts a Matlab tool kit to design a fuzzy controller module for factors influencing the machining precision, so that the control of the numerical control machine tool is realized; the fuzzy controller module comprises a fuzzification module, a fuzzy rule module library, a fuzzy reasoning module and a clarification module.
2. The self-detecting, self-analyzing, and adaptive fuzzy control system of a numerically controlled machine tool according to claim 1, wherein: the fuzzification module comprises the steps of determining input and output variables and selecting membership degrees; the fuzzy rule base module is used for making a corresponding fuzzy rule base according to expert experience; the fuzzy reasoning module carries out fuzzy reasoning according to rules in the fuzzy rule base, obtains fuzzy output quantity by fuzzy input quantity reasoning, and completes functions by the fuzzy reasoning and the fuzzy rule base; the clarification module comprises the determination of a clarification method and the conversion of a data type, the clarification (defuzzification) is carried out on the result obtained by the fuzzy inference, and then the control is realized by converting the data type into a language which can be recognized by a computer.
3. The self-detecting, self-analyzing, and adaptive fuzzy control system of a CNC machine of claim 2, wherein: the real-time data comprises an actual current value of the main shaft, an actual voltage value of the main shaft, an actual rotating speed of the main shaft, a vibration value and a noise value generated during processing; the real-time information comprises the actual feeding rate of the main shaft, the current tool number, tool compensation, the tool position, the system alarm state and specific alarm information.
4. The self-detecting, self-analyzing, and adaptive fuzzy control system of a numerically controlled machine tool according to claim 3, wherein: and the real-time display is to display the real-time data and the real-time information on a human-computer interface of the industrial personal computer in real time.
5. The self-detecting, self-analyzing, and adaptive fuzzy control system of NC machine tool according to claim 4, wherein: the preprocessing is to calibrate the real-time information and the real-time data for time, and to perform data cleaning, data integration, data conversion and data protocol preprocessing.
6. The self-detecting, self-analyzing, and adaptive fuzzy control system of a CNC machine of claim 5, wherein: and the storage is stored in a background real-time detection and monitoring system database.
CN202210439003.XA 2022-04-25 2022-04-25 Self-detection, self-analysis and self-adaptive numerical control machine tool fuzzy control system Pending CN114859821A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202210439003.XA CN114859821A (en) 2022-04-25 2022-04-25 Self-detection, self-analysis and self-adaptive numerical control machine tool fuzzy control system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210439003.XA CN114859821A (en) 2022-04-25 2022-04-25 Self-detection, self-analysis and self-adaptive numerical control machine tool fuzzy control system

Publications (1)

Publication Number Publication Date
CN114859821A true CN114859821A (en) 2022-08-05

Family

ID=82632440

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210439003.XA Pending CN114859821A (en) 2022-04-25 2022-04-25 Self-detection, self-analysis and self-adaptive numerical control machine tool fuzzy control system

Country Status (1)

Country Link
CN (1) CN114859821A (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115963723A (en) * 2023-03-17 2023-04-14 深圳市鑫雅达机电工程有限公司 Method for automatically adjusting and controlling operation of intelligent electromechanical system equipment
CN117236663A (en) * 2023-11-14 2023-12-15 深圳前海橙色魔方信息技术有限公司 Computer data analysis method and system based on artificial intelligence

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115963723A (en) * 2023-03-17 2023-04-14 深圳市鑫雅达机电工程有限公司 Method for automatically adjusting and controlling operation of intelligent electromechanical system equipment
CN117236663A (en) * 2023-11-14 2023-12-15 深圳前海橙色魔方信息技术有限公司 Computer data analysis method and system based on artificial intelligence
CN117236663B (en) * 2023-11-14 2024-03-05 深圳前海橙色魔方信息技术有限公司 Computer data analysis method and system based on artificial intelligence

Similar Documents

Publication Publication Date Title
CN107584334B (en) A kind of complex structural member numerical control machining cutter status real time monitor method based on deep learning
CN114859821A (en) Self-detection, self-analysis and self-adaptive numerical control machine tool fuzzy control system
DE102016015017B4 (en) Control device with learning function for detecting a cause of noise
CN112085261B (en) Enterprise production status diagnosis method based on cloud fusion and digital twin technology
Rowlands et al. An approach of fuzzy logic evaluation and control in SPC
Zhu et al. A cyber-physical production system framework of smart CNC machining monitoring system
CN115630839B (en) Intelligent feedback production regulation and control system based on data mining
CN107263211B (en) A kind of tool condition monitoring method based on multi-sensor fusion
TW201615844A (en) Method and system of cause analysis and correction for manufacturing data
CN112487058A (en) Numerical control machine tool fault monitoring and diagnosing system based on data mining
CN101788811A (en) Data recorder for industrial automation systems
CN109333159B (en) Depth kernel extreme learning machine method and system for online monitoring of tool wear state
CN114273977A (en) MES-based cutter wear detection method and system
Oberle et al. A use case to implement machine learning for life time prediction of manufacturing tools
CN115438726A (en) Device life and fault type prediction method and system based on digital twin technology
CN113592314A (en) Silk making process quality evaluation method based on sigma level
CN114326593B (en) Cutter life prediction system and method
Antosz et al. Machining process time series data analysis with a decision support tool
Kaur et al. Machine Learning Approach to Recommender System for Web Mining
CN106054832B (en) Multivariable-based dynamic online monitoring method and device for intermittent chemical production process
Chaturvedi et al. Supporting complex real-time decision making through machine learning
Yan et al. A digital apprentice for chatter detection in machining via human–machine interaction
CN117113233A (en) Hierarchical energy structure scene construction method and energy consumption abnormal link tracing method
CN116475651A (en) Intelligent edge control method for welding overhaul and intelligent welding equipment
CN106779245A (en) Civil aviaton's needing forecasting method and device based on event

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