CN115098824A - Ultrasonic sensor sensitivity compensation curve construction method based on BP neural network - Google Patents
Ultrasonic sensor sensitivity compensation curve construction method based on BP neural network Download PDFInfo
- Publication number
- CN115098824A CN115098824A CN202210755513.8A CN202210755513A CN115098824A CN 115098824 A CN115098824 A CN 115098824A CN 202210755513 A CN202210755513 A CN 202210755513A CN 115098824 A CN115098824 A CN 115098824A
- Authority
- CN
- China
- Prior art keywords
- neural network
- data
- error
- value
- training
- 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.)
- Granted
Links
- 238000013528 artificial neural network Methods 0.000 title claims abstract description 41
- 230000035945 sensitivity Effects 0.000 title claims abstract description 28
- 238000010276 construction Methods 0.000 title description 2
- 238000000034 method Methods 0.000 claims abstract description 20
- 238000012360 testing method Methods 0.000 claims abstract description 19
- 238000012937 correction Methods 0.000 claims abstract description 5
- 238000004088 simulation Methods 0.000 claims abstract description 4
- 238000012549 training Methods 0.000 claims description 18
- 230000004913 activation Effects 0.000 claims description 6
- 238000004458 analytical method Methods 0.000 claims description 6
- 230000006870 function Effects 0.000 claims description 6
- 238000013507 mapping Methods 0.000 claims description 3
- 238000010606 normalization Methods 0.000 claims description 3
- 238000001514 detection method Methods 0.000 description 15
- 229910000831 Steel Inorganic materials 0.000 description 7
- 239000010959 steel Substances 0.000 description 7
- 239000002131 composite material Substances 0.000 description 4
- 229920001940 conductive polymer Polymers 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- 238000007689 inspection Methods 0.000 description 2
- 230000007774 longterm Effects 0.000 description 2
- 230000009471 action Effects 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000001125 extrusion Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000012805 post-processing Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000008439 repair process Effects 0.000 description 1
Classifications
-
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Mathematical Physics (AREA)
- General Physics & Mathematics (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- Computing Systems (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Computation (AREA)
- Algebra (AREA)
- Computational Mathematics (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Databases & Information Systems (AREA)
- Investigating Or Analyzing Materials By The Use Of Ultrasonic Waves (AREA)
Abstract
The method is characterized in that ultrasonic sensing data under different scenes are collected, and the sensitivity information of the ultrasonic sensor is recorded. The method comprises the steps of using ultrasonic sensors with different sensitivities and the same model in a laboratory environment to obtain ultrasonic sensing data under different sensitivities, carrying out error correction on simulation data according to actual data, using a trained BP neural network to detect test data, carrying out parameter configuration according to an actual application scene by the trained BP neural network, and inputting collected data into the BP neural network to train to obtain an ultrasonic sensor sensitivity compensation curve under the scene.
Description
Technical Field
The invention belongs to the field of strain sensor post-processing, and particularly relates to a method for compensating a strain sensor based on a flexible conductive polymer composite material to generate a hysteresis phenomenon in a use process based on a BP (back propagation) neural network, wherein the method is simple in calculation and accurate in compensation.
Background
Railway operation safety is a strategic issue in the country. The improvement of the railway operation speed puts forward higher requirements on safety and reliability, and is particularly important for railway maintenance detection work; in order to ensure the safety and smoothness of railway transportation and improve the operation efficiency, the perfectness rate of equipment must be kept for the whole national economy service, and the state of key equipment, in particular to a bearing body of a train, namely a steel rail, is monitored in real time; in the long-term use process of the steel rail in the railway, stripping, spot stripping, cracks, fractures and other steel rail damages influencing the performance of the steel rail occur due to fatigue; in a high-speed railway system, the effects of long-term collision, extrusion and the like generated by a high-speed train are more prominent, the probability of crack occurrence and the crack propagation speed are improved, if the detection is not timely performed and safety measures are taken, the crack is easy to propagate under the external force of subsequent continuous action, so that rail breakage, even train derailment and other major accidents are caused; therefore, the detection of the rail damage is a key technology for mastering the safety state of the rail and is an essential condition for ensuring the safe operation of the high-speed rail.
The inventor finds that in the current stage of railway steel rail flaw detection work, most of the railway steel rail flaw detection work is to collect data on a line through an ultrasonic flaw detection vehicle, then the collected flaw information is directly displayed on a terminal in a B scanning mode, flaw detection personnel judge whether a steel rail has a flaw or not, if suspected flaw exists, the flaw detection personnel send a re-inspection notice to the field flaw detection personnel, and after re-inspection and checking, the flaw is determined to be subjected to subsequent repair treatment; the working mode depends on flaw detection experience and interpretation capability of operators, the detection rate of damaged rails has a great relation with the professional level of equipment operators, and the process of finding out the damaged position from flaw detection data consumes a lot of time, which forms an important contradiction with the rapid growth of railway lines in China at present.
An obstacle detection system for a vehicle generally employs a sensor mounted on a vehicle body, such as an ultrasonic sensor, for detecting a surrounding obstacle using ultrasonic waves. Generally, an ultrasonic sensor emits a sound wave around a vehicle and senses an obstacle by the sound wave reflected by the obstacle. To this end, the sensor determines the time interval at which the reflected sound waves are received and uses the known speed of sound to identify the distance between the obstacle and the vehicle.
It is understood that in dry air at 20 deg.C (68 deg.F), the speed of sound is 343 m/s. However, weather conditions affect the behavior of sound waves, and the speed of sound varies with pressure, temperature, and humidity. Ultrasonic sensors are generally used in a variety of vehicles to provide parking assistance, collision detection, automated parking, or any other type of obstacle avoidance capability. Therefore, suitable obstacle detection is required to avoid damage to the vehicle.
Currently, the speed of sound is compensated for atmospheric temperature variations by employing an ambient temperature sensor mounted on the vehicle. However, this can be expensive. In addition, there is currently no solution to compensate ultrasonic sensors for atmospheric pressure changes.
Disclosure of Invention
A method for constructing a sensitivity compensation curve of an ultrasonic sensor based on a BP neural network comprises the following steps:
the method comprises the steps that ultrasonic sensing data under different scenes are collected, sensitivity information of ultrasonic sensors is recorded, the ultrasonic sensors with different sensitivities and the same model are used in a laboratory environment, the ultrasonic sensing data under different sensitivities are obtained, and error correction is carried out on simulation data according to actual data;
using a BP neural network with double-layer input, positioning the number of hidden layer units as a set constant N, using ranging as a hidden layer activation function, using ReLU as an activation function, obtaining a sensitivity clustering center K of ultrasonic sensing by using a K-means-based clustering model according to actually acquired data, and obtaining the sensitivity K 1 Satisfy | K-K 1 |＜ts 1 As test data, where ts 1 Setting a first judgment threshold value, and taking the rest data as training data;
training the training data by using a well-defined BP neural network, and combining a threshold ts set by an output layer according to the range of 2N weights between an input layer and a hidden layer and the range of N weights between the hidden layer and the output layer 2 Calculating the error of the BP neural network, and modifying the weight and the threshold value until the error meets the error standard when the error does not meet the error standard;
detecting the test data by using the trained BP neural network, firstly carrying out normalization operation on the test data to obtain the maximum value K in the test data max And a minimum value K min For any one original test numberAccording to K 2 Performing standard mapping to obtain corrected data
Setting the maximum iteration number and the learning rate of the BP neural network, obtaining test data, obtaining an analysis value through the BP neural network, comparing the analysis value with an actual value to obtain an error value, and calculating a prediction curve { Kp i And the actual curve { Ks } i Uniformity of
Wherein N is sp For the number of points of the predicted curve and the actual curve
When g is ps Is greater than a set third determination threshold ts 3 Judging whether the BP neural network obtained by training can be used or not, and if not, extracting the error between the actual value and the predicted value to exceed a fourth judgment threshold value ts 4 The points are trained again until the BP neural network is obtained through training and can be used;
and carrying out parameter configuration according to the trained BP neural network according to an actual application scene, and inputting acquired data into the BP neural network for training to obtain an ultrasonic sensor sensitivity compensation curve under the scene.
Compared with the prior art, the invention has the advantages that:
compared with the prior art, the invention provides the strain sensor hysteresis compensation method based on the flexible conductive polymer composite material, and the method can compensate the hysteresis generated in the use of the sensor only by combining a small amount of measured data through simple steps.
Compared with a function correction method and a piecewise linear interpolation method, the strain sensor based on the flexible conductive polymer composite material is compensated by the BP neural network compensation sensor, so that the accuracy is remarkably improved, and the data processing and application are facilitated.
The 3-layer BP neural network is adopted to realize the compensation of the hysteresis of the strain sensor based on the flexible conductive polymer composite material, the method is simple to realize, the accuracy is high, and the accuracy of the sensor in the using process is greatly improved.
Detailed Description
The present invention will be further illustrated by the following specific examples. The examples are intended to better enable those skilled in the art to better understand the present invention and are not intended to limit the present invention in any way.
A method for constructing a sensitivity compensation curve of an ultrasonic sensor based on a BP neural network comprises the following steps:
the method comprises the steps that ultrasonic sensing data under different scenes are collected, sensitivity information of ultrasonic sensors is recorded, the ultrasonic sensors with different sensitivities and the same model are used in a laboratory environment, the ultrasonic sensing data under different sensitivities are obtained, and error correction is carried out on simulation data according to actual data;
the method comprises the steps of using a double-layer input BP neural network, positioning the number of hidden layer units as a set constant N, using range as a hidden layer activation function, using ReLU as an activation function, obtaining a sensitivity clustering center K of ultrasonic sensing by using a K-means-based clustering model according to actually acquired data, and enabling the sensitivity K to be equal to the sensitivity K 1 Satisfy | K-K 1 |＜ts 1 As test data, where ts 1 Setting a first judgment threshold value, and taking the rest data as training data;
training the training data by using a well-defined BP neural network, and combining a threshold ts set by an output layer according to the range of 2N weights between an input layer and a hidden layer and the range of N weights between the hidden layer and the output layer 2 Calculating the error of the BP neural network, and modifying the weight and the threshold value until the error meets the error standard when the error does not meet the error standard;
detecting the test data by using the trained BP neural network, firstly carrying out normalization operation on the test data to obtain the maximum value K in the test data max And a minimum value K min For any one of the original test data K 2 Go on to markQuasi mapping to obtain corrected data
Setting the maximum iteration number and the learning rate of the BP neural network, obtaining test data, obtaining an analysis value through the BP neural network, comparing the analysis value with an actual value to obtain an error value, and calculating a prediction curve { Kp i And the actual curve { Ks } i Uniformity of
Wherein N is sp For the number of points of the predicted curve and the actual curve
When g is ps Is greater than a set third determination threshold ts 3 Judging whether the BP neural network obtained by training can be used or not, and if not, extracting the error between the actual value and the predicted value to exceed a fourth judgment threshold value ts 4 The points are trained again until the BP neural network is obtained through training and can be used;
and carrying out parameter configuration according to the trained BP neural network according to the practical application scene, and inputting the acquired data into the BP neural network for training to obtain the sensitivity compensation curve of the ultrasonic sensor under the scene.
Claims (1)
1. A method for constructing a sensitivity compensation curve of an ultrasonic sensor based on a BP neural network is characterized by comprising the following steps:
the method comprises the steps that ultrasonic sensing data under different scenes are collected, sensitivity information of ultrasonic sensors is recorded, the ultrasonic sensors with different sensitivities and the same model are used in a laboratory environment, the ultrasonic sensing data under different sensitivities are obtained, and error correction is carried out on simulation data according to actual data;
using a BP neural network with double-layer input, positioning the number of hidden layer units as a set constant N, using the distance as the hidden layer activation function, and using the activation functionReLU, acquiring sensitivity clustering center K of ultrasonic sensor by using K-means-based clustering model according to actual acquired data, and determining sensitivity K 1 Satisfy | K-K 1 |＜ts 1 As test data, where ts 1 Taking the rest data as training data for a set first judgment threshold;
training the training data by using a well-defined BP neural network, and combining a threshold ts set by an output layer according to the range of 2N weights between an input layer and a hidden layer and the range of N weights between the hidden layer and the output layer 2 Calculating the error of the BP neural network, and modifying the weight and the threshold value until the error meets the error standard when the error does not meet the error standard;
detecting the test data by using the trained BP neural network, firstly carrying out normalization operation on the test data to obtain the maximum value K in the test data max And minimum value K min For any one of the original test data K 2 Performing standard mapping to obtain corrected data
Setting the maximum iteration number and the learning rate of the BP neural network, obtaining test data, obtaining an analysis value through the BP neural network, comparing the analysis value with an actual value to obtain an error value, and calculating a prediction curve { Kp i And the actual curve { Ks } i Uniformity of
Wherein N is sp For the number of points of the predicted curve and the actual curve
When g is ps Is greater than a set third determination threshold ts 3 Judging whether the BP neural network obtained by training can be used or not, and if not, extracting the error between the actual value and the predicted value to exceed a fourth judgment threshold value ts 4 The points are trained again until the BP neural network is obtained through training and can be used;
and carrying out parameter configuration according to the trained BP neural network according to the practical application scene, and inputting the acquired data into the BP neural network for training to obtain the sensitivity compensation curve of the ultrasonic sensor under the scene.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210755513.8A CN115098824B (en) | 2022-06-28 | 2022-06-28 | BP neural network-based ultrasonic sensor sensitivity compensation curve construction method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210755513.8A CN115098824B (en) | 2022-06-28 | 2022-06-28 | BP neural network-based ultrasonic sensor sensitivity compensation curve construction method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115098824A true CN115098824A (en) | 2022-09-23 |
| CN115098824B CN115098824B (en) | 2024-04-19 |
Family
ID=83295339
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210755513.8A Active CN115098824B (en) | 2022-06-28 | 2022-06-28 | BP neural network-based ultrasonic sensor sensitivity compensation curve construction method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115098824B (en) |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106679880A (en) * | 2016-12-21 | 2017-05-17 | 华南理工大学 | Pressure sensor temperature compensating method based on FOA-optimized SOM-RBF |
| CN111145042A (en) * | 2019-12-31 | 2020-05-12 | 国网北京市电力公司 | Power distribution network voltage abnormity diagnosis method adopting full-connection neural network |
| CN112287788A (en) * | 2020-10-20 | 2021-01-29 | 杭州电子科技大学 | Pedestrian detection method based on improved YOLOv3 and improved NMS |
| CN112528568A (en) * | 2020-12-26 | 2021-03-19 | 浙江天行健智能科技有限公司 | Road feel simulation method based on K-Means and BP neural network |
| WO2021082809A1 (en) * | 2019-10-29 | 2021-05-06 | 山东科技大学 | Training optimization method for foreign exchange time series prediction |
| US20210294479A1 (en) * | 2020-03-19 | 2021-09-23 | Sensel, Inc. | System and method for calibrating a touch sensor |
| JP2021149893A (en) * | 2020-03-24 | 2021-09-27 | 株式会社東芝 | Neural network analysis apparatus, neural network analysis method, and program |
-
2022
- 2022-06-28 CN CN202210755513.8A patent/CN115098824B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106679880A (en) * | 2016-12-21 | 2017-05-17 | 华南理工大学 | Pressure sensor temperature compensating method based on FOA-optimized SOM-RBF |
| WO2021082809A1 (en) * | 2019-10-29 | 2021-05-06 | 山东科技大学 | Training optimization method for foreign exchange time series prediction |
| CN111145042A (en) * | 2019-12-31 | 2020-05-12 | 国网北京市电力公司 | Power distribution network voltage abnormity diagnosis method adopting full-connection neural network |
| US20210294479A1 (en) * | 2020-03-19 | 2021-09-23 | Sensel, Inc. | System and method for calibrating a touch sensor |
| JP2021149893A (en) * | 2020-03-24 | 2021-09-27 | 株式会社東芝 | Neural network analysis apparatus, neural network analysis method, and program |
| CN112287788A (en) * | 2020-10-20 | 2021-01-29 | 杭州电子科技大学 | Pedestrian detection method based on improved YOLOv3 and improved NMS |
| CN112528568A (en) * | 2020-12-26 | 2021-03-19 | 浙江天行健智能科技有限公司 | Road feel simulation method based on K-Means and BP neural network |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115098824B (en) | 2024-04-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106950276B (en) | Pipeline defect depth inversion method based on convolutional neural network | |
| CN101561379B (en) | Tap-scanning method for detecting structural damages | |
| CN111322985A (en) | Tunnel clearance analysis method, device and system based on laser point cloud | |
| CN113878214B (en) | Welding quality real-time detection method and system based on LSTM and residual distribution | |
| WO2022179247A9 (en) | Damage determination method based on full cable system cable force measurement and error adaptive analysis | |
| CN110568082A (en) | cable wire breakage distinguishing method based on acoustic emission signals | |
| CN116776279A (en) | Multi-mode data collaborative power transmission line flash explosion early warning abnormal target detection method | |
| CN113805106B (en) | Rail transit train position and transformer direct current magnetic bias correlation analysis method | |
| Chang et al. | An efficient method for wheel-flattened defects detection based on acoustic emission technique | |
| Louhi Kasahara et al. | Unsupervised learning approach to automation of hammering test using topological information | |
| CN111855793A (en) | Seamless rail internal temperature stress early diagnosis method based on surface magnetic memory signal | |
| CN112836274B (en) | Data fusion method for tracing and auditing hidden engineering | |
| CN115098824A (en) | Ultrasonic sensor sensitivity compensation curve construction method based on BP neural network | |
| CN111855825B (en) | Rail head nuclear injury identification method and system based on BP neural network | |
| CN117607629A (en) | Cable terminal insulation defect detection method based on meteorological information | |
| CN110428408B (en) | Flaw detection method based on ELM-in-ELM | |
| CN114487103B (en) | Damage detection analysis method based on old part acoustic emission signal chaos characteristic value | |
| CN113687192B (en) | Method for collecting and positioning discharge signal of power transmission line | |
| CN114970610A (en) | Power transformer state identification method and device based on gram angular field enhancement | |
| CN113447570A (en) | Ballastless track disease detection method and system based on vehicle-mounted acoustic sensing | |
| CN113312801A (en) | Intelligent monitoring system and method for deepwater high-water-pressure steel wailing parameters of ultra-large bridge span | |
| CN111311591A (en) | Method for detecting lifting amount of high-speed railway contact net | |
| Zhou et al. | Detection of rail bottom damage based on BLS | |
| CN117579368A (en) | Method for detecting attack of exposed surface | |
| CN116626170B (en) | Fan blade damage two-step positioning method based on deep learning and sound emission |
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 |