CN102172849A - Cutter damage adaptive alarm method based on wavelet packet and probability neural network - Google Patents
Cutter damage adaptive alarm method based on wavelet packet and probability neural network Download PDFInfo
- Publication number
- CN102172849A CN102172849A CN 201010594595 CN201010594595A CN102172849A CN 102172849 A CN102172849 A CN 102172849A CN 201010594595 CN201010594595 CN 201010594595 CN 201010594595 A CN201010594595 A CN 201010594595A CN 102172849 A CN102172849 A CN 102172849A
- Authority
- CN
- China
- Prior art keywords
- value
- tool
- neural network
- wavelet packet
- alarm
- 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
Links
Images
Landscapes
- Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
Abstract
基于小波包和概率神经网络刀具破损自适应报警方法，先把声发射传感器固定到刀杆上，采集声发射信号，进行三层小波包分析，选取特征频带并取其均方根值，将均方根值进行归一化处理得到平滑因子和先验概率，利用概率神经网络建立刀具破损状态概率模型，根据模型和拉依达准则确定刀具磨损状态的报警值，形成一条动态报警线，依据该动态报警线，进行刀具运行状态的自适应报警监测，本发明能够找出与刀具磨损相关的均方根值的概率分布曲线，用数理统计的方法确定报警值，随着刀具磨损状态变化形成动态报警线，不会出现漏报和误报。
Based on wavelet packet and probabilistic neural network self-adaptive alarm method for tool damage, the acoustic emission sensor is fixed on the tool holder first, the acoustic emission signal is collected, and the three-layer wavelet packet analysis is carried out, the characteristic frequency band is selected and its root mean square value is taken, and the average The square root value is normalized to obtain the smoothing factor and prior probability, and the probability model of tool damage state is established by using the probability neural network, and the alarm value of the tool wear state is determined according to the model and Raida criterion, forming a dynamic alarm line. The dynamic alarm line is used for self-adaptive alarm monitoring of the tool running state. The present invention can find out the probability distribution curve of the root mean square value related to the tool wear, determine the alarm value by means of mathematical statistics, and form a dynamic Alarm line, there will be no false negatives and false negatives.
Description
Technical field
The present invention relates to machine tool status monitoring field, be specifically related to tool failure adaptive alarm method based on wavelet packet analysis and probabilistic neural network modeling.
Background technology
The adaptive alarm technology index that refers to report to the police should be that actual conditions such as condition of work, working time, power, speed along with equipment change and change, its target is to set up the dynamic judge rule of warning index and equipment ruuning situation, forms the dynamic warning curve of a variation.
Because the complexity and the diversity of manufacture process, the life-span of cutter is in a discrete distribution, make many cutters by in advance or postpone with changing, cause the unnecessary waste of cutter because process quality issue, make necessity that cutting tool state detects.Current equipment state alarm technique still is based upon on the static basis of reporting to the police, in case find that parameter has surmounted set-point in advance, just reports to the police immediately or takes appropriate measures.Easy like this appearance fails to report and reports by mistake.
Summary of the invention
In order to overcome the shortcoming of above-mentioned prior art, the object of the present invention is to provide tool failure adaptive alarm method based on WAVELET PACKET DECOMPOSITION and probabilistic neural network, can find out the probability distribution curve of the root-mean-square value relevant with tool wear, determine alarming value with mathematical statistic method, form dynamic alarming line along with cutting-tool wear state changes, can not occur failing to report and reporting by mistake.
In order to achieve the above object, the technical solution used in the present invention is:
Based on wavelet packet and probabilistic neural network tool failure adaptive alarm method, may further comprise the steps:
The first step, with the knife bar position polishing of fixed sound emission sensor, coating butter is fixed to calibrate AE sensor on the knife bar then, adopts the acoustic emission signal data acquisition program based on Labview, gathers acoustic emission signal by pci card earlier;
Second step, the acoustic emission signal that collects is carried out three layers of wavelet packet analysis, each group signal decomposition is gone on eight frequency ranges, i.e. 0～124khz, 125～249khz, 250～499khz, 500～549khz, 550～599khz, 600～649khz, 650～699khz, 700～749khz, wherein two frequency band energy maximums of 125～249khz and 250～499khz are chosen it and are the feature band of signal, and get the root-mean-square value of feature band;
The 3rd step, root-mean-square value is carried out normalized earlier, carry out identical value again and handle, obtain smoothing factor with different value, obtain prior probability with identical value;
The 4th step, all equal based on the prior distribution of each sample of Bayes theory hypothesis, utilize probabilistic neural network to set up tool failure state probability model;
The 5th step, determine the alarming value of cutting-tool wear state according to tool failure state probability model and La Yida criterion, operation along with cutting, historical data increases, alarm threshold value constantly changes, form a dynamic alarming line,, carry out the adaptive alarm monitoring of cutter running status according to this dynamic alarming line.
Because the present invention has adopted calibrate AE sensor, by energy and the probabilistic neural network apparatus for establishing state probability model primary signal gathered, wavelet packet analysis extracts characteristic spectra, utilize historical data to form dynamic alarming line, set up the relation of alarming line and equipment actual motion state, so improved the precision of reporting to the police, had advantages such as real-time detection cutting tool state and the cutter of warning reminding replacing in time.
Description of drawings
Accompanying drawing is a framework flow chart of the present invention.
The specific embodiment
Below in conjunction with accompanying drawing the present invention is described in further detail.
With reference to accompanying drawing,, may further comprise the steps based on wavelet packet and probabilistic neural network tool failure adaptive alarm method:
The first step, with the knife bar position polishing of fixed sound emission sensor, coating butter is fixed to calibrate AE sensor on the knife bar then, adopts the acoustic emission signal data acquisition program based on Labview, gathers acoustic emission signal by pci card earlier;
Second step, the acoustic emission signal that collects is carried out three layers of wavelet packet analysis, each group signal decomposition is gone on eight frequency ranges, i.e. 0～124khz, 125～249khz, 250～499khz, 500～549khz, 550～599khz, 600～649khz, 650～699khz, 700～749khz, wherein two frequency band energy maximums of 125～249khz and 250～499khz are chosen it and are the feature band of signal, and get the root-mean-square value of feature band;
In the 3rd step, the root-mean-square value that calculates is carried out preliminary treatment earlier: normalized is handled with identical value, and the normalized process is:
Wherein: { x
iBe the equipment operation history data, x
iBe the data after the normalization, max (x
i) maximum in being, min (x
i) minimum of a value in being, in identical value was handled, different value was estimated smoothing factor, the method for Cain is adopted in the estimation of smoothing factor, wherein the sample point average minimum range of asking
d
Ij=| x
i-x
j|, in the formula: N is the data number of sample layer, and the smoothing factor in the probabilistic neural network can be expressed as by empirical equation
G=1.1～1.4;
The 4th step, all equal based on the prior distribution of each sample of Bayes theory hypothesis, utilize probabilistic neural network to set up tool failure state probability model, detailed process is: according to condition probability formula
N is the sum of training sample, x
iBe i sample value, δ is a smoothing factor, and x is certain index of state undetermined, calculates the conditional probability under each sample point, then the add up conditional probability of each sample point of summation layer, and the result according to summation constructs the equipment running status probabilistic model at last;
In the 5th step, determine the alarming value of cutting-tool wear state according to the La Yida criterion, along with the operation of cutting, historical data increases, and alarm threshold value constantly changes, and forms a dynamic alarming line, according to this dynamic alarming line, carry out the adaptive alarm monitoring of cutter running status.
Claims (1)
1. based on wavelet packet and probabilistic neural network tool failure adaptive alarm method, it is characterized in that: may further comprise the steps:
The first step, with the knife bar position polishing of fixed sound emission sensor, coating butter is fixed to calibrate AE sensor on the knife bar then, adopts the acoustic emission signal data acquisition program based on Labview, gathers acoustic emission signal by pci card earlier;
Second step, the acoustic emission signal that collects is carried out three layers of wavelet packet analysis, each group signal decomposition is gone on eight frequency ranges, i.e. 0～124khz, 125～249khz, 250～499khz, 500～549khz, 550～599khz, 600～649khz, 650～699khz, 700～749khz, wherein two frequency band energy maximums of 125～249khz and 250～499khz are chosen it and are the feature band of signal, and get the root-mean-square value of feature band;
The 3rd step, root-mean-square value is carried out normalized earlier, carry out identical value again and handle, obtain smoothing factor with different value, obtain prior probability with identical value;
The 4th step, all equal based on the prior distribution of each sample of Bayes theory hypothesis, utilize probabilistic neural network to set up tool failure state probability model;
The 5th step, determine the alarming value of cutting-tool wear state according to tool failure state probability model and La Yida criterion, operation along with cutting, historical data increases, alarm threshold value constantly changes, form a dynamic alarming line,, carry out the adaptive alarm monitoring of cutter running status according to this dynamic alarming line.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201010594595 CN102172849A (en) | 2010-12-17 | 2010-12-17 | Cutter damage adaptive alarm method based on wavelet packet and probability neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201010594595 CN102172849A (en) | 2010-12-17 | 2010-12-17 | Cutter damage adaptive alarm method based on wavelet packet and probability neural network |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN102172849A true CN102172849A (en) | 2011-09-07 |
Family
ID=44516192
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 201010594595 Pending CN102172849A (en) | 2010-12-17 | 2010-12-17 | Cutter damage adaptive alarm method based on wavelet packet and probability neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102172849A (en) |
Cited By (18)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102689230A (en) * | 2012-05-09 | 2012-09-26 | 天津大学 | Tool wear condition monitoring method based on conditional random field model |
| CN103439413A (en) * | 2013-08-07 | 2013-12-11 | 湘潭大学 | Acoustic emission signal analytical method for automatically identifying damage mode of thermal barrier coating |
| CN103465107A (en) * | 2013-09-24 | 2013-12-25 | 沈阳利笙电子科技有限公司 | Tool wear monitoring method |
| CN105033763A (en) * | 2015-09-02 | 2015-11-11 | 华中科技大学 | Method for predicting abrasion state of numerically-controlled machine tool ball screw |
| CN105196114A (en) * | 2015-11-05 | 2015-12-30 | 西安科技大学 | Real-time online tool wear monitoring method based on wavelet analysis and neural network |
| CN105225223A (en) * | 2015-08-27 | 2016-01-06 | 南京市计量监督检测院 | Based on the damage Detection of Smart Composite Structure method of wavelet analysis and BP neural network |
| CN106334969A (en) * | 2016-10-31 | 2017-01-18 | 南开大学 | Cutter life estimation method for cutting power tool |
| CN106446494A (en) * | 2016-05-11 | 2017-02-22 | 新疆大学 | Wavelet packet-neural network-based wind/photovoltaic power prediction method |
| CN107350900A (en) * | 2017-07-06 | 2017-11-17 | 西安交通大学 | A kind of tool condition monitoring method based on the extraction of chip breaking time |
| CN107690660A (en) * | 2016-12-21 | 2018-02-13 | 深圳前海达闼云端智能科技有限公司 | Image-recognizing method and device |
| CN108723895A (en) * | 2018-05-25 | 2018-11-02 | 湘潭大学 | A kind of signal dividing method monitored in real time for drilling machining state |
| TWI640390B (en) * | 2017-03-24 | 2018-11-11 | 國立成功大學 | Tool wear monitoring and predicting method |
| CN108942409A (en) * | 2018-08-26 | 2018-12-07 | 西北工业大学 | The modeling and monitoring method of tool abrasion based on residual error convolutional neural networks |
| CN111024820A (en) * | 2019-12-19 | 2020-04-17 | 南通大学 | An offshore wind power blade health monitoring system and its data processing method |
| CN111331429A (en) * | 2020-03-12 | 2020-06-26 | 中国民航大学 | Method and device for monitoring tool wear state based on wavelet packet energy analysis |
| CN111590390A (en) * | 2020-04-27 | 2020-08-28 | 黄河水利职业技术学院 | A method, system, storage medium and terminal for real-time evaluation of tool wear state |
| CN114833636A (en) * | 2022-04-12 | 2022-08-02 | 安徽大学 | Cutter wear monitoring method based on multi-feature space convolution neural network |
| CN115713027A (en) * | 2022-10-31 | 2023-02-24 | 国网江苏省电力有限公司泰州供电分公司 | Transformer state evaluation method, device and system |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20060006997A1 (en) * | 2000-06-16 | 2006-01-12 | U.S. Government In The Name Of The Secretary Of Navy | Probabilistic neural network for multi-criteria fire detector |
| CN101819253A (en) * | 2010-04-20 | 2010-09-01 | 湖南大学 | Probabilistic neural network-based tolerance-circuit fault diagnosis method |
-
2010
- 2010-12-17 CN CN 201010594595 patent/CN102172849A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20060006997A1 (en) * | 2000-06-16 | 2006-01-12 | U.S. Government In The Name Of The Secretary Of Navy | Probabilistic neural network for multi-criteria fire detector |
| CN101819253A (en) * | 2010-04-20 | 2010-09-01 | 湖南大学 | Probabilistic neural network-based tolerance-circuit fault diagnosis method |
Non-Patent Citations (2)
| Title |
|---|
| 《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》 20080815 魏春燕 小波分析与神经网络在刀具故障诊断中的应用 第12，25-27，45-46页 1 , 第8期 * |
| 《机床与液压》 20070531 李志农，等 基于小波包-概率神经网络的自适应报警技术的研究 第208-210页 1 第35卷, 第5期 * |
Cited By (26)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102689230A (en) * | 2012-05-09 | 2012-09-26 | 天津大学 | Tool wear condition monitoring method based on conditional random field model |
| CN103439413A (en) * | 2013-08-07 | 2013-12-11 | 湘潭大学 | Acoustic emission signal analytical method for automatically identifying damage mode of thermal barrier coating |
| CN103439413B (en) * | 2013-08-07 | 2015-11-18 | 湘潭大学 | A kind of Analysis of Acoustic Emission Signal method that heat barrier coatings damage pattern identifies automatically |
| CN103465107A (en) * | 2013-09-24 | 2013-12-25 | 沈阳利笙电子科技有限公司 | Tool wear monitoring method |
| CN105225223A (en) * | 2015-08-27 | 2016-01-06 | 南京市计量监督检测院 | Based on the damage Detection of Smart Composite Structure method of wavelet analysis and BP neural network |
| CN105033763A (en) * | 2015-09-02 | 2015-11-11 | 华中科技大学 | Method for predicting abrasion state of numerically-controlled machine tool ball screw |
| CN105196114A (en) * | 2015-11-05 | 2015-12-30 | 西安科技大学 | Real-time online tool wear monitoring method based on wavelet analysis and neural network |
| CN106446494A (en) * | 2016-05-11 | 2017-02-22 | 新疆大学 | Wavelet packet-neural network-based wind/photovoltaic power prediction method |
| CN106446494B (en) * | 2016-05-11 | 2019-01-11 | 新疆大学 | Honourable power forecasting method based on wavelet packet-neural network |
| CN106334969A (en) * | 2016-10-31 | 2017-01-18 | 南开大学 | Cutter life estimation method for cutting power tool |
| CN106334969B (en) * | 2016-10-31 | 2018-08-10 | 南开大学 | A kind of cutter life method of estimation for cutting power tool |
| CN107690660A (en) * | 2016-12-21 | 2018-02-13 | 深圳前海达闼云端智能科技有限公司 | Image-recognizing method and device |
| CN107690660B (en) * | 2016-12-21 | 2021-03-23 | 达闼机器人有限公司 | Image recognition method and device |
| TWI640390B (en) * | 2017-03-24 | 2018-11-11 | 國立成功大學 | Tool wear monitoring and predicting method |
| CN107350900A (en) * | 2017-07-06 | 2017-11-17 | 西安交通大学 | A kind of tool condition monitoring method based on the extraction of chip breaking time |
| CN108723895A (en) * | 2018-05-25 | 2018-11-02 | 湘潭大学 | A kind of signal dividing method monitored in real time for drilling machining state |
| CN108942409A (en) * | 2018-08-26 | 2018-12-07 | 西北工业大学 | The modeling and monitoring method of tool abrasion based on residual error convolutional neural networks |
| CN111024820A (en) * | 2019-12-19 | 2020-04-17 | 南通大学 | An offshore wind power blade health monitoring system and its data processing method |
| CN111024820B (en) * | 2019-12-19 | 2023-01-31 | 南通大学 | An offshore wind power blade health monitoring system and data processing method thereof |
| CN111331429A (en) * | 2020-03-12 | 2020-06-26 | 中国民航大学 | Method and device for monitoring tool wear state based on wavelet packet energy analysis |
| CN111590390A (en) * | 2020-04-27 | 2020-08-28 | 黄河水利职业技术学院 | A method, system, storage medium and terminal for real-time evaluation of tool wear state |
| CN111590390B (en) * | 2020-04-27 | 2021-08-27 | 黄河水利职业技术学院 | Cutter wear state real-time assessment method and system, storage medium and terminal |
| CN114833636A (en) * | 2022-04-12 | 2022-08-02 | 安徽大学 | Cutter wear monitoring method based on multi-feature space convolution neural network |
| CN114833636B (en) * | 2022-04-12 | 2023-02-28 | 安徽大学 | Cutter wear monitoring method based on multi-feature space convolution neural network |
| CN115713027A (en) * | 2022-10-31 | 2023-02-24 | 国网江苏省电力有限公司泰州供电分公司 | Transformer state evaluation method, device and system |
| CN115713027B (en) * | 2022-10-31 | 2024-12-24 | 国网江苏省电力有限公司泰州供电分公司 | A method, device and system for evaluating transformer status |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102172849A (en) | Cutter damage adaptive alarm method based on wavelet packet and probability neural network | |
| CN108356606A (en) | Tool wear on-line monitoring method based on wavelet packet analysis and RBF neural | |
| CN102073300B (en) | Tool wear monitoring system capable of realizing self-learning in numerical control machining state | |
| CN105469138B (en) | Control system actuator fault diagnosis method based on population and SVMs | |
| CN103412542A (en) | Data-driven abnormity early-warning technical method of integrated circuit technology device | |
| CN103439933B (en) | A kind of production run adaptive process monitoring system and method applying OCSVM | |
| CN110000610A (en) | A kind of Tool Wear Monitoring method based on multi-sensor information fusion and depth confidence network | |
| CN108763729B (en) | Coupling state evaluation method of electromechanical system in process industry based on network structure entropy | |
| CN109023429B (en) | Intelligent crust breaking and intelligent feeding system and method for aluminum electrolytic cell | |
| CN108334033A (en) | Punching machine group failure prediction method and its system based on Internet of Things and machine learning | |
| EP1927830A3 (en) | Device for overall machine tool monitoring | |
| CN109262368A (en) | A kind of tool failure determination method | |
| Ren et al. | Big data platform for intelligence industrial IoT sensor monitoring system based on edge computing and AI | |
| CN102176217A (en) | Method for estimating reliability of numerical control machine tool cutting tool based on logistic model | |
| CN115781697B (en) | Industrial Robot Control System | |
| CN111181971B (en) | System for automatically detecting industrial network attack | |
| CN116360331B (en) | Universal irrigation automation control system and control method | |
| CN103745279A (en) | Method and device for monitoring energy consumption abnormity | |
| CN118386024A (en) | Machine tool based on artificial intelligence and fault detection method thereof | |
| CN117291582A (en) | Industrial production interconnection monitoring system based on data analysis | |
| CN114749996A (en) | Tool residual life prediction method based on deep learning and time sequence regression model | |
| CN117311316A (en) | Machine fault diagnosis method and system for industrial Internet | |
| CN116594328A (en) | Intelligent industrial robot management and control system based on 5G network | |
| CN115994692B (en) | Intelligent river and lake management platform based on 5G and big data | |
| CN105867353B (en) | A kind of vibration monitoring method and system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20110907 |