CN112633329B - Method for detecting mechanical property defects of intelligent equipment - Google Patents
Method for detecting mechanical property defects of intelligent equipment Download PDFInfo
- Publication number
- CN112633329B CN112633329B CN202011405221.9A CN202011405221A CN112633329B CN 112633329 B CN112633329 B CN 112633329B CN 202011405221 A CN202011405221 A CN 202011405221A CN 112633329 B CN112633329 B CN 112633329B
- Authority
- CN
- China
- Prior art keywords
- data
- test
- training
- degree
- nonlinear regression
- 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
- 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
- G01—MEASURING; TESTING
- G01L—MEASURING FORCE, STRESS, TORQUE, WORK, MECHANICAL POWER, MECHANICAL EFFICIENCY, OR FLUID PRESSURE
- G01L5/00—Apparatus for, or methods of, measuring force, work, mechanical power, or torque, specially adapted for specific purposes
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computational Mathematics (AREA)
- Pure & Applied Mathematics (AREA)
- Mathematical Physics (AREA)
- Algebra (AREA)
- Databases & Information Systems (AREA)
- Software Systems (AREA)
- Probability & Statistics with Applications (AREA)
- Operations Research (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Computation (AREA)
- Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
- Investigating Strength Of Materials By Application Of Mechanical Stress (AREA)
Abstract
The invention discloses a method for detecting mechanical property defects of intelligent equipment, which belongs to the technical field of intelligent equipment detection and comprises the following steps: the testing software acquires stress data of the intelligent equipment in the motion process as testing data; processing the test data according to the optimal nonlinear regression degree of freedom; obtaining the interference degree of the abnormal factors according to the comparison of the decision coefficients; according to the method, the mechanical characteristic defect states are compared and output according to the fitted curve, on one hand, a sufficient amount of actual mechanical data of the motion process are obtained quickly by using a mechanical sensor, on the other hand, a nonlinear regression method is originally introduced, model training is carried out by using the mechanical data of the motion process of a standard sample, the control, data acquisition and operation of the whole measurement process are automatically completed through codes, the efficiency is greatly improved, and the conclusion reliability is higher; the output mechanical property defect conclusion is helpful for guiding the design and the assembly process to carry out optimization and improvement, and the defects caused by abnormal factors are effectively marked.
Description
Technical Field
The invention belongs to the technical field of intelligent equipment detection, and particularly relates to a method for detecting mechanical property defects of intelligent equipment.
Background
The intelligence lock in intelligence house lets people's daily life more careful and intelligent, and high-end intelligence wrist-watch has become the indispensable smart machine of people daily life through health monitoring, sports training and conversation function especially. The power consumption evaluation of the whole system is influenced by the mechanical motion process of the intelligent lock, the experience of a user and the fineness of application control are influenced by the motion mechanics of the rotary crown of the high-end intelligent watch, and therefore stricter and accurate requirements are provided for mechanical property description and detection in the motion process of a product. The mechanical characteristics comprise friction force in the motion process, torsion force in the deformation process, damping force in the rotation process and the like, on one hand, the mechanical characteristic detection needs to cover the whole motion process as much as possible, on the other hand, the influence of abnormal factors in the motion process accounts for the validity of data, and certain challenges are brought to the description and detection of the whole mechanical characteristics.
The existing test scheme is that a mechanical sensor is used for collecting data of a fixed point position, and fixed threshold value judgment is carried out; the method is simple in system building, can only reflect the numerical value of the fixed azimuthal force, cannot well reflect the mechanical state of the whole motion process, and in addition, the influence of abnormal factors on the stress process lacks an effective calibration mode.
Disclosure of Invention
The technical problem to be solved by the invention is as follows: the method for detecting the mechanical property defect of the intelligent equipment is provided, the whole detection process is automatically completed through codes, the efficiency is greatly improved, the reliability of a conclusion is higher, and the defect caused by an abnormal factor can be effectively marked.
In order to solve the technical problems, the technical scheme of the invention is as follows:
a method for detecting mechanical property defects of intelligent equipment comprises the following steps:
the testing software acquires stress data of the intelligent equipment in the motion process as testing data; the test software performs data processing on the test data according to the optimal nonlinear regression degree of freedom N to obtain a test decision coefficient, and performs nonlinear fitting to obtain a test curve; comparing the test decision coefficient with the training decision coefficient, and considering that the test data is greatly interfered by abnormal factors when the interference degree is greater than an interference threshold value; and comparing the test curve with the training curve, calibrating the test data with the deviation degree larger than the deviation threshold value as defect data, and calibrating the corresponding mechanical property defect position according to the defect data position.
Further, the method for obtaining the optimal nonlinear regression degree of freedom N includes the following steps: acquiring stress data of standard intelligent equipment in a motion process as training data through the test software; establishing a model for the training data by using a Linear regression ion method; creating a non-linear regression training model from the model using the PolynomialFeatures method; normalizing and fitting the nonlinear regression training model using a Fit _ transform method; calculating a standard decision coefficient, comparing and evaluating the standard decision coefficient obtained when N =1, 2, 3 \8230 \ 8230;/and marking the standard decision coefficient more than or equal to 0.95 as the training decision coefficient, wherein the degree of freedom under the training decision coefficient is the optimal nonlinear regression degree of freedom N; and storing the optimal nonlinear regression degree of freedom N.
Further, the processing of the test data comprises the following specific steps: establishing a model for the test data by using a Linear regression method; bringing the optimal nonlinear regression degree of freedom N into a PolynomialFeatures method, and creating a nonlinear regression test model according to the model created by the test data; normalizing and fitting the nonlinear regression test model using the Fit _ transform method; and calculating the test decision coefficient under the optimal nonlinear regression degree of freedom N.
Further, the interference degree is obtained by performing a difference operation on the test decision coefficient and the training decision coefficient.
Further, the interference threshold is 0.05.
Further, the training curve generated by the data after fitting the nonlinear regression training model is stored.
Further, the data after fitting the nonlinear regression test model is used for generating the test curve.
Further, the training curve and the testing curve are placed in the same coordinate system, data are selected according to a sampling period of 10ms, and the deviation degree is obtained by performing difference operation on mechanical data of the training data and the testing data.
Further, the deviation threshold is 0.15.
After the technical scheme is adopted, the invention has the beneficial effects that:
the method for detecting the mechanical property defects of the intelligent equipment comprises the following steps: the testing software acquires stress data of the intelligent equipment in the motion process as testing data; the test software performs data processing on the test data according to the optimal nonlinear regression degree of freedom N to obtain a test decision coefficient, and performs nonlinear fitting to obtain a test curve; comparing the test decision coefficient with the training decision coefficient, and considering that the test data is greatly interfered by abnormal factors when the interference degree is greater than an interference threshold; comparing a test curve with a training curve, calibrating test data with a deviation degree larger than a deviation threshold value as defect data, calibrating corresponding mechanical characteristic defect positions according to the defect data positions, on one hand, rapidly and sufficiently acquiring actual mechanical data in a motion process by using a mechanical sensor, on the other hand, innovatively introducing a nonlinear regression method, performing model training by using mechanical data in a standard sample motion process, calculating and marking the mechanical characteristic defects in the actual motion process by using a decision coefficient and a fitting curve, and obtaining an abnormal factor interference degree and a mechanical characteristic defect conclusion, wherein the control, data acquisition and nonlinear regression of data in the whole measurement process are automatically completed by codes, the efficiency is greatly improved, and the conclusion credibility is higher; the output mechanical property defect conclusion is helpful for guiding the design and the assembly process to carry out optimization and improvement, and the defects caused by abnormal factors are effectively marked in the whole motion process.
Drawings
FIG. 1 is a logic processing diagram of a method for mechanical property defect detection of a smart device of the present invention;
FIG. 2 is a training data logic processing diagram of the method for detecting defects in mechanical properties of smart devices of the present invention;
FIG. 3 is a test data logic processing diagram of the method for detecting defects in mechanical properties of smart devices according to the present invention;
FIG. 4 is a schematic diagram of a fitted curve of an embodiment.
Detailed Description
The invention is further illustrated with reference to the following figures and examples.
All directions referred to in the present specification are based on the drawings, and represent relative positional relationships only, and do not represent absolute positional relationships.
As shown in fig. 1, a method for detecting defects of mechanical characteristics of intelligent devices includes the following steps:
s1, acquiring stress data of intelligent equipment in a motion process as test data by test software, and acquiring the stress condition of the intelligent equipment in the motion process in real time through a mechanical sensor;s2, the test software carries out data processing on the test data according to the optimal nonlinear regression degree of freedom N to obtain a test decision coefficient
Carrying out nonlinear fitting to obtain a test curve; s3, determining coefficients by testing
And training decision coefficient
Comparing, and determining that the test data is greatly interfered by abnormal factors when the interference degree is greater than the interference threshold, wherein the interference degree is determined by a test decision coefficient
And training decision coefficient
Performing difference operation to obtain; s4, comparing the test curve with the training curve, placing the training curve and the test curve in the same coordinate system, selecting data according to a sampling period of 10ms, and determining the degree of deviation
From training data (t) i ，F train ) And test data (t) i ，F T ) Is obtained by performing a difference operation on the mechanical data, i.e.
Degree of deviation
And calibrating the test data larger than the deviation threshold value as defect data, and calibrating the corresponding mechanical property defect position according to the defect data position.
As shown in fig. 2, the method for obtaining the optimal nonlinear regression degree of freedom N includes the following steps: s211, standard intelligent equipment is acquired through test softwareTaking stress data in the exercise process as training data; s212, establishing a model for the training data by using a Linear regression method; s213, creating a nonlinear regression training model F = W from the model by using a PolynomialFeatures method 0 +W 1 t+w 2 t 2 +…+w i t i Wherein, F represents the stress condition of the ith time node; s214, standardizing and fitting the nonlinear regression training model by using a Fit _ transform method; s215, calculating the standard determination coefficient
And the standard determination coefficient obtained for N =1, 2, 3 \8230; N
Performing comparison evaluation when
Then stopping the cyclic acquisition process of the steps S213, S214 and S215, and taking
Of the hour
Marking as training decision coefficients
Determining coefficients in training
The lower degree of freedom N is the optimal nonlinear regression degree of freedom N; s216, storing the optimal nonlinear regression degree of freedom in N steps, and storing a training curve generated by data after fitting a nonlinear regression training model. Wherein, linear regression, polymonomial features and Fit _ transform are all system self-contained functions, and the specific operation method is not described herein.
As shown in fig. 3, the processing of the test data in step S2 includes the following specific steps: s221, establishing a model for the test data by using a Linear regression method; s222, substituting the optimal nonlinear regression degree of freedom N into a PolynomialFeatures method, and creating a nonlinear regression test model F = w according to the model created by the test data 0 +w 1 t+w 2 t 2 +…+w i t i (ii) a S223, standardizing and fitting the nonlinear regression test model by using a Fit _ transform method; s224, calculating a test decision coefficient under the optimal nonlinear regression degree of freedom N
And generating a test curve by the data fitted by the nonlinear regression test model.
The following are examples:
(1) According to the process shown in fig. 2, after the standard sample calibrated in the laboratory is used to obtain the training data, the nonlinear regression training model between the time t and the stress F is established as follows:
F train ＝-1.1804+1.3175t-0.2495t 2 +0.0176t 3 -0.0004t 4 ，
(2) According to the flow shown in fig. 3, a non-linear regression test model between time t and force F using product 1 was established using the test data as follows:
F T1 ＝-0.7785+0.6218t+0.0187t 2 -0.0135t 3 +0.0007t 4 ，
will be provided with
And
comparing the difference value of the two, namely the interference degree is more than 0.05, the product 1 has larger unacceptable factors, and the t is obtained by comparing a test curve L2 with a training curve L1 according to the fitting curve shown in figure 4 4 -t 10 And if the deviation degree in the time period is more than or equal to 0.15 and the data is abnormal, outputting a mechanical property defect conclusion: t is t 4 -t 10 There are unacceptable drawbacks to the time period.
(3) The non-linear regression test model between time t and force F using product 2, established with the test data, is as follows:
F T2 ＝-1.1432+1.3333t-0.2866t 2 +0.0232t 3 +0.0006t 4 ，
will be provided with
And
comparing, the interference degree is more than 0.05, part of unacceptable factors exist in the product 2, and t is obtained by comparing a test curve L3 with a training curve L1 according to the fitted curve shown in FIG. 4 5 -t 7 And if the deviation degree in the time period is more than or equal to 0.15 and the data is abnormal, outputting a mechanical property defect conclusion: t is t 5 -t 7 There are unacceptable drawbacks to the time period.
In the above two embodiments, the interference threshold is set to 0.05, and the deviation threshold is set to 0.15. The interference threshold and the deviation threshold are set by workers according to experience, and specific numerical values can be flexibly set by the workers according to the type of the intelligent equipment and the magnitude of external force applied to the intelligent equipment in the movement process. The detection example used in the application is an intelligent lock, and the detection method is also suitable for other products such as intelligent wristbands and the like.
The method for detecting the mechanical property defect of the intelligent equipment acquires actual mechanical data of the intelligent equipment in the motion process through the mechanical sensor, introduces a nonlinear regression method, performs model training by using the mechanical data of a standard sample in the motion process, obtains the interference degree of an abnormal factor and the mechanical property defect conclusion through the comparison of a decision coefficient and a fitting curve, automatically completes the control, data acquisition and nonlinear regression of data in the whole measurement process through codes, greatly improves the efficiency, has higher conclusion credibility, and effectively marks the defect caused by the abnormal factor in the whole motion process.
While specific embodiments of the invention have been described above, it will be understood by those skilled in the art that the described embodiments are only some, and not all, of the present invention, which is presented by way of example only, and the scope of the invention is defined by the appended claims. Various changes or modifications to these embodiments can be made by those skilled in the art without departing from the principle and spirit of this invention, and these changes and modifications all fall within the scope of this invention.
Claims (8)
1. A method for detecting defects of mechanical characteristics of intelligent equipment is characterized by comprising the following steps:
the testing software acquires stress data of the intelligent equipment in the motion process as testing data;
the test software performs data processing on the test data according to the optimal nonlinear regression degree of freedom N to obtain a test decision coefficient, and performs nonlinear fitting to obtain a test curve;
performing difference operation on the test decision coefficient and the training decision coefficient to obtain an interference degree, and considering that the test data is greatly interfered by abnormal factors when the interference degree is greater than an interference threshold;
and comparing the test curve with the training curve, calibrating the test data with the deviation degree larger than the deviation threshold value as defect data, and calibrating the corresponding mechanical property defect position according to the defect data position.
2. The method for detecting the mechanical property defect of the intelligent device according to claim 1, wherein the method for obtaining the optimal nonlinear regression degree of freedom N comprises the following steps:
acquiring stress data of standard intelligent equipment in a motion process as training data through the test software;
establishing a model for the training data by using a Linear regression method;
creating a non-linear regression training model from the model using the PolynomialFeatures method;
normalizing and fitting the nonlinear regression training model using a Fit _ transform method;
calculating a standard decision coefficient, comparing and evaluating the standard decision coefficient obtained when N =1, 2, 3 \8230 \ 8230;/and marking the standard decision coefficient more than or equal to 0.95 as the training decision coefficient, wherein the degree of freedom under the training decision coefficient is the optimal nonlinear regression degree of freedom N;
and storing the optimal nonlinear regression degree of freedom N.
3. The method for detecting defects in mechanical characteristics of intelligent equipment according to claim 2, wherein the processing of the test data comprises the following specific steps:
establishing a model for the test data by using a Linear regression method;
bringing the optimal nonlinear regression degree of freedom N into a PolynomialFeatures method, and creating a nonlinear regression test model according to the model created by the test data;
normalizing and fitting the nonlinear regression test model using the Fit _ transform method;
and calculating the test decision coefficient under the optimal nonlinear regression degree of freedom N.
4. The method for defect detection of mechanical characteristics of intelligent devices according to claim 3, wherein the interference threshold is 0.05.
5. The method for detecting defects in mechanical characteristics of intelligent equipment according to any one of claims 3 to 4, wherein the training curve generated by fitting data of the nonlinear regression training model is stored.
6. The method for detecting defects in mechanical properties of intelligent equipment according to claim 5, wherein the test curve is generated from data fitted by the nonlinear regression test model.
7. The method for detecting defects of mechanical characteristics of intelligent equipment according to claim 6, wherein the training curve and the test curve are placed in the same coordinate system, data are selected according to a sampling period of 10ms, and the deviation degree is obtained by performing difference operation on mechanical data of the training data and the test data.
8. The method for defect detection of mechanical characteristics of intelligent devices according to claim 7, wherein the deviation threshold is 0.15.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011405221.9A CN112633329B (en) | 2020-12-05 | 2020-12-05 | Method for detecting mechanical property defects of intelligent equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011405221.9A CN112633329B (en) | 2020-12-05 | 2020-12-05 | Method for detecting mechanical property defects of intelligent equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112633329A CN112633329A (en) | 2021-04-09 |
| CN112633329B true CN112633329B (en) | 2023-01-31 |
Family
ID=75307875
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011405221.9A Active CN112633329B (en) | 2020-12-05 | 2020-12-05 | Method for detecting mechanical property defects of intelligent equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112633329B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105302848A (en) * | 2014-10-11 | 2016-02-03 | 山东鲁能软件技术有限公司 | Evaluation value calibration method of equipment intelligent early warning system |
| CN107025381A (en) * | 2017-04-18 | 2017-08-08 | 江苏省环境科学研究院 | Yangcheng Lake evaluation on Ecosystem Health method based on P IBI |
| CN108681633A (en) * | 2018-05-11 | 2018-10-19 | 上海电力学院 | A kind of condensate pump fault early warning method based on state parameter |
| CN111608899A (en) * | 2020-04-28 | 2020-09-01 | 茂盟（上海）工程技术股份有限公司 | Water pump running state abnormity discrimination method based on efficiency analysis |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US11488010B2 (en) * | 2018-12-29 | 2022-11-01 | Northeastern University | Intelligent analysis system using magnetic flux leakage data in pipeline inner inspection |
-
2020
- 2020-12-05 CN CN202011405221.9A patent/CN112633329B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105302848A (en) * | 2014-10-11 | 2016-02-03 | 山东鲁能软件技术有限公司 | Evaluation value calibration method of equipment intelligent early warning system |
| CN107025381A (en) * | 2017-04-18 | 2017-08-08 | 江苏省环境科学研究院 | Yangcheng Lake evaluation on Ecosystem Health method based on P IBI |
| CN108681633A (en) * | 2018-05-11 | 2018-10-19 | 上海电力学院 | A kind of condensate pump fault early warning method based on state parameter |
| CN111608899A (en) * | 2020-04-28 | 2020-09-01 | 茂盟（上海）工程技术股份有限公司 | Water pump running state abnormity discrimination method based on efficiency analysis |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112633329A (en) | 2021-04-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109781934A (en) | Processing unit, processing method, computer readable storage medium and the environment sensor system of environmental sensor detection data | |
| US11709474B2 (en) | Method and apparatus for detecting abnormality of manufacturing facility | |
| CN115982896B (en) | Bearing retainer service life detection method and device | |
| CN110060368A (en) | Mechanical method for detecting abnormality based on potential feature coding | |
| CN112633329B (en) | Method for detecting mechanical property defects of intelligent equipment | |
| CN116985183A (en) | Quality monitoring and management method and system for near infrared spectrum analyzer | |
| CN108645490A (en) | Cold-water meter vision detection system based on image processing techniques and detection method | |
| CN109470888B (en) | Calibration system and calibration method of high-g-value accelerometer based on deep learning | |
| CN117052970A (en) | Intelligent control system and method for pneumatic ball valve assembly | |
| CN113610851A (en) | Packaging decoration quality inspection method based on machine vision | |
| CN116520236A (en) | Abnormality detection method and system for intelligent ammeter | |
| CN116725546A (en) | Digital electrocardiograph verification method based on deep learning and corner detection | |
| Nelson | Analysis of residuals from censored data | |
| CN114236272B (en) | Intelligent detection system of electronic product | |
| CN113001546B (en) | Method and system for improving motion speed safety of industrial robot | |
| CN111931574B (en) | Robust fault diagnosis method for pneumatic regulating valve | |
| CN115219883A (en) | Automatic detection system and detection method for electric automation equipment | |
| CN115311447A (en) | Pointer instrument indicating number identification method based on deep convolutional neural network | |
| CN117224088B (en) | User sleep quality management system and method based on detection data analysis | |
| CN112798279B (en) | Novel detection method for diagnosing motor bearing fault | |
| CN114778114B (en) | Bearing health index construction method based on signal impact and periodicity | |
| CN113447045B (en) | Method and system for analyzing accuracy reliability of inertial system | |
| CN114841058A (en) | Neural network training method based on mechanism model, mobile equipment fault diagnosis method, device and storage medium | |
| CN116808535A (en) | Martial arts action recognition system and method based on sensor | |
| CN111710142A (en) | Remote data acquisition system for stroller and data analysis method thereof |
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 |