CN1584433A - Noise source identifying method for air-conditioner based on nervous network - Google Patents
Noise source identifying method for air-conditioner based on nervous network Download PDFInfo
- Publication number
- CN1584433A CN1584433A CN 200410027476 CN200410027476A CN1584433A CN 1584433 A CN1584433 A CN 1584433A CN 200410027476 CN200410027476 CN 200410027476 CN 200410027476 A CN200410027476 A CN 200410027476A CN 1584433 A CN1584433 A CN 1584433A
- Authority
- CN
- China
- Prior art keywords
- noise
- neural network
- air conditioner
- source
- discrimination based
- 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
Images
Abstract
A neural network based detection of noise-source in air conditioner is carried out by: extracting dynamic characteristics; training neural network; and detecting noise source in air conditioner. It has distributive and parallel processing, non-linear mapping, self-adapting and robust error tolerance so as to be widely used in pattern identification, control optimization, intellectual data processing and fault diagnosis. Noises generated from different parts of air conditioner are detected in mode of neural network to determine effectively where a main noise exists with high robustness and intelligence. It can be used for noisy faults diagnosis of air conditioner or some electric appliance so as to qualify them.
Description
1, technical field:
The present invention relates to a kind of noise from air conditioner source discrimination, be used for the identification of air-conditioner noise sources and the noise failure diagnosis of air conditioner, belong to the innovative technology of noise from air conditioner source discrimination and noise from air conditioner method for diagnosing faults based on neural network (NN).
2, background technology:
Air conditioner as a kind of and people live its quality of closely bound up household electrical appliance especially noise qualities more and more obtain people's attention.The noise of air conditioner is very complicated, and parts such as compressor, pipeline, fan, air channel all may produce noise, and the noise qualities that improve air conditioner at first will identify the main source of noise.It is several that the recognition methods of noise source mainly contains subjective estimate method, branch's running method, surface strength method, superficial velocity mensuration, near field measurement method etc., but because the structural singularity of air conditioner, above-mentioned these traditional methods are difficult to identify well the overriding noise source of air conditioner.
3, summary of the invention:
The objective of the invention is to overcome above-mentioned shortcoming and propose a kind of new noise from air conditioner source discrimination based on neural network, make that the identification of air-conditioner noise sources is simple, for quality quality, the solution noise from air conditioner problem that improves air conditioner provides and instruct and help.
The present invention is based on the noise from air conditioner source discrimination of neural network, include following three basic steps:
1) status flag extracts: the running status of air conditioner being had only an overriding noise source is as a kind of pattern, noise under each pattern is done spectrum analysis, the spectrum value of getting on the different frequency range forms a vector, this vector can be used as the pattern feature amount under this state, and the spectrum value of noise in different frequency range got up just to constitute a status flag amount by series arrangement from low to high;
2) BP neural metwork training: design the BP neural network as requested, train the BP neural network with the pattern feature amount that step 1) obtains, and nerve network system is trained;
3) air-conditioner noise sources identification: noise from air conditioner data input step 2) inside the neural network that has trained, can identify the overriding noise source of air conditioner by neural network.
Above-mentioned steps 1) the spectrum analysis third-octave analysis of noise in, the centre frequency of 1/3 octave band is from 25Hz～4000Hz during analysis.
Above-mentioned steps 2) selects three layers BP neural network in, wherein contain the BP network of a hidden layer.
The input layer number of above-mentioned three layers of BP neural network can be 2～5, and the number of hidden nodes can be 5～12, and the output node number can be taken as 1.
The input number of nodes of above-mentioned three layers of BP neural network is 3, and the output node number is 1, and the number of hidden nodes is 7.
Above-mentioned BP neural network can design with mathematical tool software matlab.
The present invention proposes a kind of new noise from air conditioner source discrimination based on neural network, neural network is as a kind of new method system, having distributed parallel handles, Nonlinear Mapping, characteristic such as adaptive learning and robust Fault-Tolerant makes it be widely used at aspects such as pattern-recognition, Control and Optimization, Intelligent Information Processing and fault diagnosises.The noise states that the different parts of air conditioner are produced can be used as different patterns is discerned air conditioner with the mode identificating ability of neural network overriding noise source.The present invention has utilized the characteristic information of dbjective state, can not only effective recognition go out the overriding noise source of air conditioner, and have very high robustness and intelligent.The present invention can be used for the noise failure diagnosis of noise from air conditioner fault diagnosis and other household electrical appliance.The present invention is a kind of simple noise from air conditioner source discrimination, and can be the quality quality, the solution noise from air conditioner problem that improve air conditioner provides and instruct and help.
4, description of drawings:
Describe concrete structure of the present invention in detail below in conjunction with accompanying drawing:
Fig. 1 is the process flow diagram of embodiment of the invention air-conditioner noise sources identification;
Fig. 2 is the structural representation of embodiment of the invention BP neural network.
5, embodiment:
Embodiment:
The present invention is based on the noise from air conditioner source discrimination of neural network, include following three basic steps:
1) status flag extracts
The first step of the present invention is exactly the status flag amount of extraction pattern.Extract the status flag amount from two air conditioners, the overriding noise source of two air-conditionings is respectively to allow compressor piping system and fan system.This two states of air conditioner can be considered as two kinds of patterns of air conditioner operation, gather the noise signal that air conditioner sends under two kinds of patterns, carry out spectrum analysis respectively, spectrum analysis is analyzed with third-octave, the centre frequency of 1/3 octave band is from 25Hz～4000Hz during analysis, gets spectrum value on the different frequency range and forms proper vector under this pattern.
Present embodiment is discerned two kinds of overriding noise sources of air conditioner.The definition fan is that the running status in overriding noise source is a pattern 0, and air conditioner compressor system is that the operational mode in overriding noise source is a pattern 1.At first gather its noise signal respectively under two kinds of patterns, do spectrum analysis with BSWA V302USB acoustic analysis instrument, the analysis result that draws under two kinds of patterns is as follows respectively, and each column data is the proper vector of an associative mode.
The proper vector of pattern 1 (fan noise):
-73.29 -73.81 -78.42 -72.64 -73.93 -76.01 -69.27 -71.54 -74.38
-79.13 -73.63 -81.13 -84.63 -76.10 -91.98 -74.84 -76.27 -77.96
-81.05 -80.51 -83.54 -85.76 -79.03 -82.89 -85.53 -85.19 -78.52
-81.12 -79.13 -92.08 -76.17 -80.39 -80.82 -84.84 -81.83 -81.12
-83.31 -85.84 -83.16 -83.28 -86.01 -82.56 -82.02 -80.45 -82.83
-85.19 -84.51 -84.96 -89.36 -88.07 -83.65 -85.55 -90.02 -90.96
-89.57 -86.92 -86.25 -86.99 -90.36 -90.73 -87.92 -88.97 -92.12
-94.39 -92.64 -86.50 -92.59 -91.81 -98.44 -89.25 -91.05 -88.94
-91.84 -90.74 -91.84 -91.12 -95.05 -90.90 -91.34 -90.96 -93.48
-92.37 -91.24 -97.76 -94.55 -91.71 -92.08 -95.43 -91.46 -92.22
-95.38 -96.73 -93.16 -98.01 -93.30 -96.22 -97.11 -94.78 -93.74
-94.65 -93.43 -99.09 -97.19 -95.41 -97.61 -97.46 -84.42 -95.72
-100.63?-98.95 -99.20 -99.20 -99.32 -99.72 -100.60?-101.58?-100.23
-101.40?-101.63?-100.45?-103.27?-100.67?-101.81?-100.95?-101.92?-102.70
-102.29?-100.72?-101.04?-102.16?-103.02?-102.09?-101.91?-101.52?-101.63
-100.36 -99.30?-101.67?-100.16-?100.64?-101.20?-101.34?-100.53?-101.25
The proper vector of pattern 2 (compressor assembly noise):
-70.37 -68.89 -73.32 -78.38 -70.45 -67.65 -77.12 -72.58 -76.71
-79.06 -76.80 -76.03 -79.58 -80.55 -78.33 -79.39 -78.68 -83.01
-88.01 -79.92 -83.80 -85.99 -80.48 -81.20 -78.88 -82.12 -86.17
-92.46 -84.13 -89.78 -78.94 -80.70 -83.77 -76.76 -82.55 -85.08
-81.32 -85.40 -86.25 -87.93 -88.54 -84.88 -84.93 -87.61 -87.96
-87.63 -94.45 -83.51 -92.69 -88.07 -88.34 -85.58 -88.99 -88.62
-92.20 -85.70 -88.94 -89.22 -89.35 -89.49 -91.30 -88.27 -89.91
-90.20 -86.16 -95.91 -93.88 -91.76 -92.34 -89.02 -93.50 -92.18
-93.64 -91.47 -94.77 -95.07 -90.47 -94.28 -92.73 -95.48 -92.44
-93.13 -97.30 -92.59 -91.65 -90.13 -96.47 -92.21 -96.94 -95.50
-95.61 -99.06 -96.04 -97.39 -92.28 -93.32 -96.59 -97.42 -94.18
-98.66 -96.47 -98.55 -98.60 -95.96 -96.67 -97.96 -98.67 -98.09
-100.31?-97.87 -97.89 -98.04 -99.39 -98.98 -99.09 -99.61 -96.66
-97.99 -97.05 -96.29 -94.78 -96.25 -94.50 -97.61 -96.56 -94.91
-97.00 -96.93 -97.21 -99.17 -96.77 -98.27 -98.06 -98.08 -99.47
-97.25 -94.33 -97.40 -96.31 -95.88 -96.15 -96.98 -96.17 -98.66
2) network training
Because of three layers of BP network have the learning ability of approaching any nonlinear function, so the present invention selects three layers BP nerve net, wherein contains the BP network of a hidden layer, the input layer number of above-mentioned three layers of BP neural network can be 2～5, the number of hidden nodes can be 5～12, and the output node number can be taken as 1.In the present embodiment, the input number of nodes of three layers of BP neural network is 3, and the number of hidden nodes is 7, and the output node number is 1.The BP neural network can design with mathematical tool software matlab.
Before carrying out air-conditioner noise sources identification, at first the BP neural network that designs to be trained with the pattern character vector that previous step obtains.The parameter of neural network is as follows during training: learning rate is 0.5, and the inertia scale factor is 0.9, and the network iteration can obtain satisfied convergence effect after 1500 steps.
3) air conditioner overriding noise source identification (pattern-recognition)
The BP neural network just can be carried out the identification of air-conditioner noise sources after step 2 trains.The data of voice signal after the acoustic analysis instrument is done spectrum analysis that record when the operation of noise failure air conditioner will be arranged are input to the BP neural network that trains, and the output result of neural network just can discern the noise source of air conditioner.
In the present embodiment, measure the voice signal of the big air conditioner of two stage noises respectively, after spectrum analysis, obtain two proper vectors
V1＝[-74.93-76.50-79.83-82.36-85.71-89.27-91.34-90.85-94.35-92.11-93.37-96.01 -99.34-100.55-101.86-100.97]’
v2＝[-66.75-77.63-81.76-83.07-83.95-88.04-89.92-91.75-95.67-97.19-93.02-96.17 -98.12-95.53-97.24-97.13]’
In the BP network that above-mentioned two proper vector input steps 2 are trained, the neural network of proper vector V1 correspondence is output as 0, and the neural network of proper vector V2 correspondence is output as 1, thereby air-conditioner noise sources is discerned.
Claims (6)
1, a kind of noise from air conditioner source discrimination based on neural network is characterized in that including following three basic steps:
1) status flag extracts: the running status of air conditioner being had only an overriding noise source is as a kind of pattern, noise under each pattern is done spectrum analysis, the spectrum value of getting on the different frequency range forms a vector, this vector can be used as the pattern feature amount under this state, and the spectrum value of noise in different frequency range got up just to constitute a status flag amount by series arrangement from low to high;
2) BP neural metwork training: design the BP neural network as requested, train the BP neural network with the pattern feature amount that step 1) obtains, and nerve network system is trained;
3) air-conditioner noise sources identification: noise from air conditioner data input step 2) inside the neural network that has trained, can identify the overriding noise source of air conditioner by neural network.
2, the noise from air conditioner source discrimination based on neural network according to claim 1 is characterized in that above-mentioned steps 1) in the spectrum analysis third-octave analysis of noise, the centre frequency of 1/3 octave band is from 25Hz～4000Hz during analysis.
3, the noise from air conditioner source discrimination based on neural network according to claim 1 is characterized in that above-mentioned steps 2) in select three layers BP neural network, wherein contain the BP network of a hidden layer.
4, the noise from air conditioner source discrimination based on neural network according to claim 3 is characterized in that the input layer number of above-mentioned three layers of BP neural network can be 2～5, and the number of hidden nodes can be 5～12, and the output node number can be taken as 1.
5, the noise from air conditioner source discrimination based on neural network according to claim 4, the input number of nodes that it is characterized in that above-mentioned three layers of BP neural network is 3, and the output node number is 1, and the number of hidden nodes is 7.
6, the noise from air conditioner source discrimination based on neural network according to claim 1 is characterized in that above-mentioned BP neural network can design with mathematical tool software matlab.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2004100274760A CN1301387C (en) | 2004-06-04 | 2004-06-04 | Noise source identifying method for air-conditioner based on nervous network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2004100274760A CN1301387C (en) | 2004-06-04 | 2004-06-04 | Noise source identifying method for air-conditioner based on nervous network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1584433A true CN1584433A (en) | 2005-02-23 |
| CN1301387C CN1301387C (en) | 2007-02-21 |
Family
ID=34601360
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNB2004100274760A Expired - Fee Related CN1301387C (en) | 2004-06-04 | 2004-06-04 | Noise source identifying method for air-conditioner based on nervous network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN1301387C (en) |
Cited By (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102563808A (en) * | 2012-01-11 | 2012-07-11 | 华南理工大学 | Automatic control method of indoor environment comfort level |
| CN103139008A (en) * | 2011-11-23 | 2013-06-05 | 中兴通讯股份有限公司 | Self-adaption method and device capable of detecting message heartbeat period |
| CN103347070A (en) * | 2013-06-28 | 2013-10-09 | 北京小米科技有限责任公司 | Method, terminal, server and system for voice data pushing |
| CN104471501A (en) * | 2012-06-12 | 2015-03-25 | 西门子公司 | Generalized pattern recognition for fault diagnosis in machine condition monitoring |
| CN104931284A (en) * | 2015-05-29 | 2015-09-23 | 广东美的制冷设备有限公司 | Fault detection method and device for air conditioner, and air conditioner |
| CN105115119A (en) * | 2015-09-18 | 2015-12-02 | 珠海格力电器股份有限公司 | Fault detection method, device and equipment for air conditioning unit |
| CN105424395A (en) * | 2015-12-15 | 2016-03-23 | 珠海格力电器股份有限公司 | Method and device for determining equipment fault |
| CN106323454A (en) * | 2016-08-04 | 2017-01-11 | 广东美的制冷设备有限公司 | Air-conditioner indoor machine abnormal sound identification method and device |
| CN106527169A (en) * | 2017-01-20 | 2017-03-22 | 深圳大图科创技术开发有限公司 | Intelligent home control system based on Bluetooth |
| CN107123427A (en) * | 2016-02-21 | 2017-09-01 | 珠海格力电器股份有限公司 | A kind of method and device for determining sound quality |
| CN107798283A (en) * | 2016-08-31 | 2018-03-13 | 西安英诺视通信息技术有限公司 | A kind of neural network failure multi classifier based on the acyclic figure of decision-directed |
| WO2019207254A1 (en) * | 2018-04-26 | 2019-10-31 | Orange | Detector for detecting a complex state of an object, electronic ear and detecting method |
| CN110826583A (en) * | 2018-08-14 | 2020-02-21 | 珠海格力电器股份有限公司 | Fault determination method and device, storage medium and electronic device |
| CN111578445A (en) * | 2020-04-27 | 2020-08-25 | 青岛海尔空调器有限总公司 | Control method and device for air conditioner and air conditioner |
| CN111721401A (en) * | 2020-06-17 | 2020-09-29 | 广州广电计量检测股份有限公司 | Low-frequency noise analysis system and method |
| CN113298134A (en) * | 2021-05-20 | 2021-08-24 | 华中科技大学 | BPNN-based remote non-contact health monitoring system and method for fan blade |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP0557071B1 (en) * | 1992-02-19 | 1999-05-12 | Hitachi, Ltd. | Active noise control apparatus for three-dimensional space |
| JPH08320147A (en) * | 1995-05-25 | 1996-12-03 | Matsushita Electric Ind Co Ltd | Control device for air conditioner |
| CN1120366C (en) * | 1999-03-22 | 2003-09-03 | 西安交通大学 | Fault detecting and diagnosing method based on non-linear spectral analysis |
| TW564314B (en) * | 2002-03-21 | 2003-12-01 | Mang-Hui Wang | CMAC_based fault diagnosis of air condition systems |
-
2004
- 2004-06-04 CN CNB2004100274760A patent/CN1301387C/en not_active Expired - Fee Related
Cited By (22)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103139008A (en) * | 2011-11-23 | 2013-06-05 | 中兴通讯股份有限公司 | Self-adaption method and device capable of detecting message heartbeat period |
| CN102563808A (en) * | 2012-01-11 | 2012-07-11 | 华南理工大学 | Automatic control method of indoor environment comfort level |
| CN104471501A (en) * | 2012-06-12 | 2015-03-25 | 西门子公司 | Generalized pattern recognition for fault diagnosis in machine condition monitoring |
| CN103347070A (en) * | 2013-06-28 | 2013-10-09 | 北京小米科技有限责任公司 | Method, terminal, server and system for voice data pushing |
| CN104931284A (en) * | 2015-05-29 | 2015-09-23 | 广东美的制冷设备有限公司 | Fault detection method and device for air conditioner, and air conditioner |
| CN104931284B (en) * | 2015-05-29 | 2018-03-27 | 广东美的制冷设备有限公司 | Fault detection method, device and the air conditioner of air conditioner |
| CN105115119A (en) * | 2015-09-18 | 2015-12-02 | 珠海格力电器股份有限公司 | Fault detection method, device and equipment for air conditioning unit |
| CN105424395A (en) * | 2015-12-15 | 2016-03-23 | 珠海格力电器股份有限公司 | Method and device for determining equipment fault |
| CN105424395B (en) * | 2015-12-15 | 2018-05-18 | 珠海格力电器股份有限公司 | The definite method and apparatus of equipment fault |
| CN107123427B (en) * | 2016-02-21 | 2020-04-28 | 珠海格力电器股份有限公司 | Method and device for determining noise sound quality |
| CN107123427A (en) * | 2016-02-21 | 2017-09-01 | 珠海格力电器股份有限公司 | A kind of method and device for determining sound quality |
| CN106323454A (en) * | 2016-08-04 | 2017-01-11 | 广东美的制冷设备有限公司 | Air-conditioner indoor machine abnormal sound identification method and device |
| CN106323454B (en) * | 2016-08-04 | 2019-08-16 | 广东美的制冷设备有限公司 | The recognition methods of air conditioner indoor unit abnormal sound and device |
| CN107798283A (en) * | 2016-08-31 | 2018-03-13 | 西安英诺视通信息技术有限公司 | A kind of neural network failure multi classifier based on the acyclic figure of decision-directed |
| CN106527169A (en) * | 2017-01-20 | 2017-03-22 | 深圳大图科创技术开发有限公司 | Intelligent home control system based on Bluetooth |
| WO2019207254A1 (en) * | 2018-04-26 | 2019-10-31 | Orange | Detector for detecting a complex state of an object, electronic ear and detecting method |
| FR3080726A1 (en) * | 2018-04-26 | 2019-11-01 | Orange | COMPLEX STATE DETECTOR OF OBJECT, ELECTRONIC EAR, AND DETECTION METHOD |
| CN110826583A (en) * | 2018-08-14 | 2020-02-21 | 珠海格力电器股份有限公司 | Fault determination method and device, storage medium and electronic device |
| CN111578445A (en) * | 2020-04-27 | 2020-08-25 | 青岛海尔空调器有限总公司 | Control method and device for air conditioner and air conditioner |
| CN111721401A (en) * | 2020-06-17 | 2020-09-29 | 广州广电计量检测股份有限公司 | Low-frequency noise analysis system and method |
| CN111721401B (en) * | 2020-06-17 | 2022-03-08 | 广州广电计量检测股份有限公司 | Low-frequency noise analysis system and method |
| CN113298134A (en) * | 2021-05-20 | 2021-08-24 | 华中科技大学 | BPNN-based remote non-contact health monitoring system and method for fan blade |
Also Published As
| Publication number | Publication date |
|---|---|
| CN1301387C (en) | 2007-02-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN1301387C (en) | Noise source identifying method for air-conditioner based on nervous network | |
| CN112818892B (en) | Multi-modal depression detection method and system based on time convolution neural network | |
| EP3198247B1 (en) | Device for capturing vibrations produced by an object and system for capturing vibrations produced by a drum. | |
| CN110472587B (en) | Micro vibration motor defect identification method and device based on CNN and sound time-frequency characteristic diagram | |
| CN1302456C (en) | Sound veins identifying method | |
| CN1930575A (en) | Techniques for separating and evaluating audio and video source data | |
| CN110120230B (en) | Acoustic event detection method and device | |
| CN1949364A (en) | System and method for testing identification degree of input speech signal | |
| CN101080028A (en) | An advertisement video detection method | |
| CN110956613A (en) | Image quality-based target detection algorithm performance normalization evaluation method and system | |
| CN113763986A (en) | Air conditioner indoor unit abnormal sound detection method based on sound classification model | |
| CN109584888A (en) | Whistle recognition methods based on machine learning | |
| Whitehill et al. | Whosecough: In-the-wild cougher verification using multitask learning | |
| CA2999839C (en) | Systems and methods for capturing and interpreting audio | |
| CN1174374C (en) | Method and device for parallelly having speech recognition, classification and segmentation of speaker | |
| CN1628337A (en) | Speech recognizing method and device thereof | |
| Umri et al. | Comparative Analysis of CLAHE and AHE on Application of CNN Algorithm in the Detection of Covid-19 Patients | |
| CN115424638A (en) | Disease detection method based on limited cough sound data | |
| CN114781600A (en) | Generation method of countermeasure sample and defense method of countermeasure sample | |
| CN114358321A (en) | Machine learning detection method for abnormal sound of motor | |
| CN1343967A (en) | Speech identification system | |
| CN110322894B (en) | Sound-based oscillogram generation and panda detection method | |
| CN1302454C (en) | Method for rebuilding probability weighted average deletion characteristic data of speech recognition | |
| CN1819019A (en) | Phonetic identifier based on matrix characteristic vector function and identification thereof | |
| Jhansi et al. | Detection of COVID-19 Patients using Speech Recognition with Support Vector Machine” and Comparing with “K Nearest Neighbour Algorithm” |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| C19 | Lapse of patent right due to non-payment of the annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |