CN102590335A - SAW (Surface Acoustic Wave) sensor based embedded electronic nose testing system and testing method - Google Patents
SAW (Surface Acoustic Wave) sensor based embedded electronic nose testing system and testing method Download PDFInfo
- Publication number
- CN102590335A CN102590335A CN2012100061735A CN201210006173A CN102590335A CN 102590335 A CN102590335 A CN 102590335A CN 2012100061735 A CN2012100061735 A CN 2012100061735A CN 201210006173 A CN201210006173 A CN 201210006173A CN 102590335 A CN102590335 A CN 102590335A
- Authority
- CN
- China
- Prior art keywords
- training
- data
- rule
- saw sensor
- testing
- 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
- 238000012360 testing method Methods 0.000 title claims abstract description 77
- 238000010897 surface acoustic wave method Methods 0.000 title abstract description 25
- 238000012549 training Methods 0.000 claims abstract description 66
- 238000013528 artificial neural network Methods 0.000 claims abstract description 23
- 238000000034 method Methods 0.000 claims abstract description 21
- 230000008569 process Effects 0.000 claims abstract description 9
- 239000007789 gas Substances 0.000 claims description 18
- 238000004891 communication Methods 0.000 claims description 9
- 239000011159 matrix material Substances 0.000 claims description 8
- 230000004044 response Effects 0.000 claims description 8
- 238000010606 normalization Methods 0.000 claims description 7
- 230000008859 change Effects 0.000 claims description 5
- 238000004458 analytical method Methods 0.000 claims description 4
- 238000003909 pattern recognition Methods 0.000 claims description 3
- 230000008676 import Effects 0.000 claims description 2
- 238000004321 preservation Methods 0.000 claims description 2
- 238000004422 calculation algorithm Methods 0.000 abstract description 9
- 238000012545 processing Methods 0.000 abstract description 3
- 238000001514 detection method Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 238000010998 test method Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000035945 sensitivity Effects 0.000 description 2
- 238000013459 approach Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 210000001072 colon Anatomy 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005265 energy consumption Methods 0.000 description 1
- 239000012467 final product Substances 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 230000005055 memory storage Effects 0.000 description 1
- 230000002093 peripheral effect Effects 0.000 description 1
- 239000000047 product Substances 0.000 description 1
- 238000012113 quantitative test Methods 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
Images
Landscapes
- Testing Or Calibration Of Command Recording Devices (AREA)
Abstract
The invention discloses an SAW (Surface Acoustic Wave) sensor based embedded electronic nose testing system and testing method. The testing method comprises the steps of: (1) obtaining relevant training data and test data by a microcontroller; (2) if the format of the obtained data does not conform to a specification, processing; otherwise, skipping this step; (2) if a parameter setting signal is received, skipping to a step (6); otherwise, setting relevant parameters automatically; (4) adjusting relevant parameters according rules of neural networks until the training is finished; (5) if the training result reaches to requirements, skipping to a step (8); otherwise, continuing; (6) setting relevant parameters according to the reference setting signal until the training is finished; (7) if the training result reaches to the requirements, skipping to the step (8); otherwise, skipping back to the step (6); and (8) using the training result to mode identification or software measurement and obtaining practical output according to test data. According to the SAW sensor based embedded electronic nose testing system and testing method, disclosed by the invention, neural network algorithm is simplified and optimized, and the training process of the neural network can be transplanted in an embedded platform.
Description
Technical field
The present invention relates to the test macro that the gas qualitative and quantitative is analyzed, relate in particular to a kind of embedded radio electronics nose test macro based on SAW (surface acoustic wave) sensor.
Background technology
Electronic Nose is the electronic system that a kind of response pattern that utilizes gas sensor array is discerned smell, is made up of sensor array and appropriate mode recognition system, can discern simple or complicated gas.Single-sensor in the Electronic Nose is nonspecific in response, can produce broad spectrum response to multiple gases.Because sensor itself is shortcoming inevitably.The single parameter measured sensor can produce very perturbation when measuring mixed gas, measuring error is difficult to control.Improving the antijamming capability valid approach is to adopt combined type or intelligent algorithms such as array multisensor and neural network to reach ideal effect.At present, mostly Electronic Nose product on the market is external production, and the general volume of these instruments is bigger, and costs an arm and a leg, and is difficult to satisfy the demand of present home market to portable set.
The SAW sensor has selects susceptibility high, and advantages such as repeatability and reliability can be used as the sensor array of Electronic Nose, and particularly, the SAW sensor has the following advantages:
1. high sensitivity, high linearity: the energy density of SAW sensor is very big, and is very sensitive to the disturbance on surface, and the fundamental frequency of SAW sensor can be machined to several GHz, so detection sensitivity is higher.
2. repeatability and good reliability: the critical component of SAW sensor is SAW resonator or lag line, when making, adopts the semiconductor technology of plane assignment, good reproducibility.And integrated easily, integrated, mechanism is firm, thereby better reliability.
3. signal is gathered easily and is handled, and can realize wireless sensing: adopt accurate digital signal output, be prone to digitizing.Aspect remote sensing and the remote measurement remarkable advantages is being arranged.
4. volume is little, and is in light weight, low in energy consumption: this is the common feature of all SAW sensors.
Embedded platform and equipment are more and more universal, and have low-power consumption, characteristics such as volume is little, integrated level is high, cost is low, be widely used.If the neural network scheduling algorithm can be applied to embedded device, and have good runnability, so just can solve the complex engineering problem, and possess miniaturization and portable characteristics.The neural network algorithm operand is big, complex algorithm, in order to reach portable and characteristics miniaturization, need simplify and handles algorithm.
Though the Embedded Application of a lot of proposition neural network algorithms has been arranged,, it all is on PC, to carry out emulation that great majority are used, the result that will obtain then is transplanted to embedded platform, rather than training process is transplanted in the embedded platform.
Summary of the invention
To above-mentioned prior art, the technical matters that the present invention will solve provides embedded radio electronics nose test macro and the method for testing based on the SAW sensor that a kind of cost is low, be convenient for carrying.
In order to solve the problems of the technologies described above, the present invention adopts following technical scheme: a kind of method of testing of the embedded electronic nose test macro based on the SAW sensor comprises the steps:
(1) microcontroller obtains relevant training data and test data;
(2) if the data layout that is obtained does not meet standard, then handle, otherwise skip this step;
(3) as if receiving parameter signal is set, then skipped to for (6) step, otherwise correlation parameter is set automatically;
(4) according to the rule adjustment correlation parameter of neural network, accomplish up to training;
(5), then skipped to for (8) step, otherwise continue if training result reaches requirement;
(6) according to parameter signal correlation parameter is set, accomplishes up to training;
(7) if training result reaches requirement, then skip to (8), otherwise (6) step of rebound;
(8) be used for pattern-recognition or software measurement to training result,, obtain actual output according to test data.
Further, the data layout in the said step (2) comprises: training sample dimension n, desirable output dimension m, training sample number, desirable output number, test sample book number N, the data of training sample and the data of desirable output;
The disposal route that the data form is not met standard comprises:
(2-1) according to following rule " training sample data 1, training sample data 2 ... training sample data N; Desirable output 1, desirable output 2 ... Desirable output data P; Test data 1, test data 2 ... Test data M ", wherein the M representative needs the number of test; The number of samples that the N representative is used to import; P is identical with the gas number of analysis;
(2-2) data normalization of training sample or test sample book is handled, and scope is 0.0-1.0;
(2-3) data normalization of output is handled, and scope is 0.0-1.0.
Further, step (3), (4), and the correlation parameter in (6) comprises that the node of neural network counts K, learning efficiency, and error, maximum cycle, stipulated time t are ended in training.
Further, the parameter signal in said step (3) and (6) is revised if desired, then according to following form " Y: neural network node number: learning efficiency: error is ended in training: maximum cycle: stipulated time "; Otherwise do not send signal; (such as accepting parameter modification, then form is Y:1:0.01:5000:15, wherein; Colon is in order to distinguish parameter, can in program, to add the if statement and distinguish parameter).
Further, the rule of the neural network in the step (4) comprises: the rule of 3 layers of BP neural network (Back Propagation Neural Network), weights initialization rule, learning efficiency, the number of hidden nodes selection rule, training stopping rule.
Further, the weights of described weights initialization rule are initialized as initialization at random, and initialized weights interval is (1,1).
Further; It is initial value that the rule of said learning efficiency is chosen bigger learning rate; The learning efficiency computing formula is:
; Wherein,
is the initial value of learning efficiency; Learning efficiency in the time of
expression maximum cycle;
representes maximum cycle; If after training iteration once; Error increases; Take strategy so:
adjusts; Up to the termination that satisfies condition; If after training iteration once; Error reduces; Take strategy:
adjusts, up to the termination that satisfies condition.
Further, said the number of hidden nodes selection rule adopts formula:
, wherein; K representes the number of hidden nodes; N representes the training sample dimension, and m representes desirable output dimension, and a is the constant greater than 1; It is 50 to the maximum; If the time that consumes, is directly chosen the corresponding K value of a less than the time of expection as the number of hidden nodes, otherwise change the value of a; Make a=a * 0.7, and with the value of the K of correspondence as the number of hidden nodes.
Further, said training stopping rule comprises: rule 1: " error current is ended error less than training, and keeps steady change "; Rule 2: " current cycle time is less than maximum cycle (such as 5000); and the error of follow-on test changes less than 1% "; Rule 3: " (such as 15s) accomplishes test in official hour, and current cycle time and satisfied respectively rule 1 of error current and rule 2 ".
Further, the training result that is used to calculate in the step (8) comprises:
1. to the preservation of weight matrix, making whole process all is under the condition of same weight matrix;
2. to the number of hidden nodes, learning efficiency, maximum cycle, error is ended in training, and program runtime is stored, and will test afterwards that the result is sent to remote port.
Say that further (2) are mentioned in the step (8) is used for analyzing and processing, mainly be meant that special finger sends to remote control terminal through radio-frequency module with the data after the test, thereby make remote control terminal obtain corresponding result, be convenient to analyze through output interface.
A kind of embedded radio electronics nose test macro based on the SAW sensor comprises SAW sensor and testing system platform, and said testing system platform comprises processor; Storer, communication interface, IO interface; Man Machine Interface, wherein, microcontroller is connected with storer, Man Machine Interface, communication interface, IO interface respectively; Input interface is connected with the SAW sensor, and output interface is connected with wireless module.
Further, said SAW sensor comprises the SAW sensor of three response heterogeneity gases and the sensor of two response humiture gases.
Compared with prior art; The present invention has following beneficial effect: simplify and optimized neural network algorithm; The training process of neural network algorithm is transplanted in the embedded platform, a kind of embedded radio electronics nose test macro based on the SAW sensor is provided, for the pattern-recognition or the software measurement of processes such as detection by electronic nose gas provides a kind of low cost; Portable, the method for high real-time.
Description of drawings
Fig. 1 is that the test macro of the embodiment of the invention is formed structured flowchart;
Fig. 2 is a test procedure process flow diagram of the present invention;
Fig. 3 is test philosophy figure of the present invention;
Fig. 4 is the right planar alignment structural drawing of the input and output interdigital transducer of SAW gas sensor;
Fig. 5 is sensor array and test macro syndeton synoptic diagram.
Embodiment
To combine accompanying drawing and embodiment that the present invention is done further description below.
As shown in Figure 1, be the composition frame chart of native system, wherein, microcontroller is connected with storer, Man Machine Interface, communication interface, IO interface respectively.
Man Machine Interface mainly refers to display screen, and major function is a display result.Storer comprises SDRAM, NANDFLASH, SD card.And storer directly links to each other with microcontroller.Communication interface comprises Ethernet interface, RS232 interface, USB interface, and the WIFI interface, described communication interface is connected with microcontroller.By the data that sensor obtains, the process circuit conversion is as the input of system.Raw data and treated data can be preserved in the SD card, are convenient to data are carried, and output interface is a radio-frequency communication module, and convenient data to acquisition transmit at a distance.
Microcontroller is the core of total system, and from the input or the communication interface reception data of signal, through data processing and analysis, the result shows on touch-screen, and sends corresponding data through radio-frequency module.Simultaneously, microcontroller also need provide communicating by letter of external units such as corresponding interface support and PC, and will obtain data and preserve.
The S3C6410 (ARM1176JZF-S) that microcontroller adopts SAMSUNG to provide; Its dominant frequency is 667MHz, supports Mobile DDR and multiple NAND FLASH, and it has optimized the interface of external memory storage simultaneously; Support numerous peripheral interface functions, make interface satisfy the demand of Industry Control.
Need obtain data to Electronic Nose and test, so at first analyze, guarantee the correctness of the design's input for the sensor of Electronic Nose.Such as, for gas detection,, also need consider the influence of environment except considering the response of the corresponding heterogeneity of sensor.Suppose that the SAW gas sensor has 3, arranges as shown in Figure 4.Sensor array one has five, and wherein two is to consider that humiture can influence the performance of gas sensor, as auxiliary; And be connected with test platform; As shown in Figure 5, so, for the single sample input of neural network; Be the data of five dimensions, the sample dimension of desirable output then is a three-dimensional data.With 50 groups of samples is training sample, and 30 groups is test sample book, and saves as " training sample 1 (sensor 1 data, sensor 2 data, sensor 3 data, sensor 4 data, sensor 5 data); Training sample 50 (sensor 1 data, sensor 2 data, sensor 3 data, sensor 4 data, sensor 5 data); Desirable output 1 (desirable sensing output 1, desirable sensing output 2, desirable sensing output 3); Desirable output 50 (desirable sensing output 1, desirable sensing output 2, desirable sensing output 3); Test sample book 1 (sensor 1 data, sensor 2 data, sensor 3 data, sensor 4 data; Sensor 5 data) ... Test sample book 30 (sensor 1 data, sensor 2 data, sensor 3 data; Sensor 4 data, sensor 5 data) " information that obtains through data normalization so then is: training sample dimension n=5, desirable output dimension m=3; training sample number 50, desirable output number 50, test sample book number N=30.
Be assumed to and do not receive parameter signal is set; Promptly select to be provided with automatically pattern; Learning efficiency
=10 can be set so; Maximum cycle
=5000; The number of hidden nodes K=50; Ending error is a random number, and the termination time is 15s.If receive parameter signal is set, system can be provided with learning rate according to semaphore request so, maximum cycle and the number of hidden nodes, and stand-by time.
At first, with 50 groups of data training sample neural network trainings, each training all can obtain the numerical value of error current, working time and current cycle time.System is through error current, and regulation rule changes learning rate; Through working time, declare the number of hidden nodes, and simultaneously these three data are judged, whether inquiry arrives training termination condition.If do not arrive the termination condition, continue training so when satisfying condition, preserve the relevant weights and the parameter of neural network, be used for test data.
After training finishes, during test data, for the gas observational measurement; The data of screen display are matrixes, and the numerical value of matrix has only 0 and 1 to constitute, and this gas of 0 expression is not major component; And 1 be expressed as major component, and through output interface, this matrix is sent to long-range receiving end; For gasometric analysis, after data are carried out normalization and handle, need to preserve normalization matrix; The data of output are handled with this matrix doing mathematics, accomplished quantitative test, and pass through output interface to the data that obtain; Send to receiving end, be convenient to User Recognition.
Through computing after a while,, need carry out correction again to data owing to reasons such as model drifts.Again timing repeats said process and gets final product.
Above-mentioned instance explanation is used for explaining the present invention, rather than the restriction to inventing, and in the protection domain of the present invention's spirit and claim, any modification and change to the present invention makes all fall into protection scope of the present invention.
Claims (12)
1. the method for testing based on the embedded electronic nose test macro of SAW sensor is characterized in that, comprises the steps:
(1) microcontroller obtains relevant training data and test data;
(2) if the data layout that is obtained does not meet standard, then handle, otherwise skip this step;
(3) as if receiving parameter signal is set, then skipped to for (6) step, otherwise correlation parameter is set automatically;
(4) according to the rule adjustment correlation parameter of neural network, accomplish up to training;
(5), then skipped to for (8) step, otherwise continue if training result reaches requirement;
(6) according to parameter signal correlation parameter is set, accomplishes up to training;
(7) if training result reaches requirement, then skip to (8), otherwise (6) step of rebound;
(8) be used for pattern-recognition or software measurement to training result,, obtain actual output according to test data.
2. the method for testing of the embedded electronic nose test macro based on the SAW sensor according to claim 1; It is characterized in that the data layout in the said step (2) comprises: training sample dimension n, desirable output dimension m, training sample number, desirable output number, test sample book number N, the data of training sample and the data of desirable output;
The disposal route that the data form is not met standard comprises:
(2-1) according to following rule " training sample data 1, training sample data 2 ... training sample data N; Desirable output 1, desirable output 2 ... Desirable output data P; Test data 1, test data 2 ... Test data M ", wherein the M representative needs the number of test; The number of samples that the N representative is used to import; P is identical with the gas number of analysis;
(2-2) data normalization of training sample or test sample book is handled, and scope is 0.0-1.0;
(2-3) data normalization of output is handled, and scope is 0.0-1.0.
3. the method for testing of the embedded electronic nose test macro based on the SAW sensor according to claim 1 is characterized in that step (3); (4); (6) correlation parameter in comprises that the node of neural network counts K, learning efficiency, and error is ended in training; Maximum cycle, stipulated time t.
4. the method for testing of the embedded electronic nose test macro based on the SAW sensor according to claim 1; It is characterized in that; Parameter signal in said step (3) and (6) is revised if desired; Then according to following form " Y: neural network node number: learning efficiency: error is ended in training: maximum cycle: stipulated time ", otherwise do not send signal.
5. the method for testing of the embedded electronic nose test macro based on the SAW sensor according to claim 1; It is characterized in that the rule of the neural network in the step (4) comprises: the rule of 3 layers of BP neural network, weights initialization rule, learning efficiency, the number of hidden nodes selection rule, training stopping rule.
6. the method for testing of the embedded electronic nose test macro based on the SAW sensor according to claim 5 is characterized in that, the weights of described weights initialization rule are initialized as initialization at random, and initialized weights interval is (1,1).
7. the method for testing of the embedded electronic nose test macro based on the SAW sensor according to claim 5; It is characterized in that; It is initial value that the rule of said learning efficiency is chosen bigger learning rate; The learning efficiency computing formula is:
; Wherein,
is the initial value of learning efficiency; Learning efficiency in the time of
expression maximum cycle;
representes maximum cycle; If after training iteration once; Error increases; Take strategy so:
adjusts; Up to the termination that satisfies condition; If after training iteration once; Error reduces; Take strategy:
adjusts, up to the termination that satisfies condition.
8. the method for testing of the embedded electronic nose test macro based on the SAW sensor according to claim 5 is characterized in that, said the number of hidden nodes selection rule adopts formula:
; Wherein, K representes the number of hidden nodes, and n representes the training sample dimension; M representes desirable output dimension; A is the constant greater than 1, and it is 50 to the maximum, if the time that consumes is less than the time of expection; Directly choose the corresponding K value of a as the number of hidden nodes; Otherwise change the value of a, make a=a * 0.7, and with the value of the K of correspondence as the number of hidden nodes.
9. the method for testing of the embedded electronic nose test macro based on the SAW sensor according to claim 5 is characterized in that said training stopping rule comprises: rule 1: " error current is ended error less than training, and keeps steady change "; Regular 2: " current cycle time is less than maximum cycle, and the error of follow-on test changes less than 1% ", rule 3: " in official hour, accomplish test, and current cycle time and satisfied respectively rule 1 of error current and rule 2 ".
10. the method for testing of the embedded electronic nose test macro based on the SAW sensor according to claim 1 is characterized in that the training result that is used to calculate in the step (8) comprises
1. to the preservation of weight matrix, making whole process all is under the condition of same weight matrix;
2. to the number of hidden nodes, learning efficiency, maximum cycle, error is ended in training, and program runtime is stored, and will test afterwards that the result is sent to remote port.
11. the embedded radio electronics nose test macro based on the SAW sensor, it is characterized in that: comprise SAW sensor and testing system platform, said testing system platform comprises processor; Storer, communication interface, IO interface; Man Machine Interface, wherein, microcontroller is connected with storer, Man Machine Interface, communication interface, IO interface respectively; Input interface is connected with the SAW sensor, and output interface is connected with wireless module.
12. the embedded radio electronics nose test macro based on the SAW sensor according to claim 11 is characterized in that: said SAW sensor comprises the SAW sensor of three response heterogeneity gases and the sensor of two response humiture gases.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210006173.5A CN102590335B (en) | 2012-01-06 | 2012-01-06 | Based on embedded electronic nose test macro and the method for testing of SAW sensor |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210006173.5A CN102590335B (en) | 2012-01-06 | 2012-01-06 | Based on embedded electronic nose test macro and the method for testing of SAW sensor |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102590335A true CN102590335A (en) | 2012-07-18 |
| CN102590335B CN102590335B (en) | 2015-09-02 |
Family
ID=46479268
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201210006173.5A Expired - Fee Related CN102590335B (en) | 2012-01-06 | 2012-01-06 | Based on embedded electronic nose test macro and the method for testing of SAW sensor |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102590335B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103346799A (en) * | 2013-06-05 | 2013-10-09 | 中国科学院微电子研究所 | Method for identifying gas based on compressed sensing theory |
| CN106295801A (en) * | 2016-08-01 | 2017-01-04 | 安徽农业大学 | A kind of Folium Camelliae sinensis based on fruit bat algorithm optimization generalized regression nerve networks algorithm stores method chronological classification |
| CN107121482A (en) * | 2017-04-13 | 2017-09-01 | 中国农业大学 | A kind of embedded volt-ampere electronic tongues detecting system and method |
| RU2699366C1 (en) * | 2018-12-07 | 2019-09-05 | Общество с ограниченной ответственностью "Научно-производственное объединение "АМБ" | Compact device of "electronic nose" type |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1381721A (en) * | 2002-06-04 | 2002-11-27 | 复旦大学 | Portable intelligent electronic nose and its preparing process |
| US20060161814A1 (en) * | 2003-07-09 | 2006-07-20 | Carl Wocke | Method and system of data analysis using neural networks |
| JP2007287153A (en) * | 2006-04-13 | 2007-11-01 | Fisher Rosemount Syst Inc | Robust process model identification method and system in model based control technique |
| CN101382439A (en) * | 2008-10-27 | 2009-03-11 | 哈尔滨工业大学 | Multi-parameter self-confirming sensor and state self-confirming method thereof |
| CN101706885A (en) * | 2009-11-23 | 2010-05-12 | 浙江大学 | System and method for online training support vector machine soft-sensing model |
| CN101706882A (en) * | 2009-11-23 | 2010-05-12 | 浙江大学 | Embedded platform based neural network model online training method |
-
2012
- 2012-01-06 CN CN201210006173.5A patent/CN102590335B/en not_active Expired - Fee Related
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1381721A (en) * | 2002-06-04 | 2002-11-27 | 复旦大学 | Portable intelligent electronic nose and its preparing process |
| US20060161814A1 (en) * | 2003-07-09 | 2006-07-20 | Carl Wocke | Method and system of data analysis using neural networks |
| JP2007287153A (en) * | 2006-04-13 | 2007-11-01 | Fisher Rosemount Syst Inc | Robust process model identification method and system in model based control technique |
| CN101382439A (en) * | 2008-10-27 | 2009-03-11 | 哈尔滨工业大学 | Multi-parameter self-confirming sensor and state self-confirming method thereof |
| CN101706885A (en) * | 2009-11-23 | 2010-05-12 | 浙江大学 | System and method for online training support vector machine soft-sensing model |
| CN101706882A (en) * | 2009-11-23 | 2010-05-12 | 浙江大学 | Embedded platform based neural network model online training method |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103346799A (en) * | 2013-06-05 | 2013-10-09 | 中国科学院微电子研究所 | Method for identifying gas based on compressed sensing theory |
| CN106295801A (en) * | 2016-08-01 | 2017-01-04 | 安徽农业大学 | A kind of Folium Camelliae sinensis based on fruit bat algorithm optimization generalized regression nerve networks algorithm stores method chronological classification |
| CN106295801B (en) * | 2016-08-01 | 2019-01-29 | 安徽农业大学 | A kind of tealeaves storage time classification method based on drosophila algorithm optimization generalized regression nerve networks algorithm |
| CN107121482A (en) * | 2017-04-13 | 2017-09-01 | 中国农业大学 | A kind of embedded volt-ampere electronic tongues detecting system and method |
| RU2699366C1 (en) * | 2018-12-07 | 2019-09-05 | Общество с ограниченной ответственностью "Научно-производственное объединение "АМБ" | Compact device of "electronic nose" type |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102590335B (en) | 2015-09-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Jaladi et al. | Environmental monitoring using wireless sensor networks (WSN) based on IOT | |
| US8717146B2 (en) | Methods and systems for integrated interrogation of RFID sensors | |
| CN110115071B (en) | Method and terminal for controlling transmitting power | |
| CN101917736B (en) | Sensor network test bed system and test method thereof | |
| Shah et al. | IoT-enabled low power environment monitoring system for prediction of PM2. 5 | |
| CN103743453B (en) | A kind of control method of radar level gauge system | |
| CN102590335B (en) | Based on embedded electronic nose test macro and the method for testing of SAW sensor | |
| CN104914069A (en) | Transferable calculation meat quality near infrared detection method and device | |
| CN103000996B (en) | Uniform circular array direction-finder antenna receiving mutual impedance test and mutual coupling compensation system | |
| KR101376016B1 (en) | A Measuring Device for Food Freshness and Maleficence Using Visible and Infrared Light Spectrometric Sensor | |
| CN104461534A (en) | Android mobile terminal-based portable intelligent ultrasonic flaw detection system | |
| CN101799321A (en) | Intelligent vibration monitor system | |
| CN205898894U (en) | Digital storage sweep generator | |
| CN103685655A (en) | Air quality testing device based on mobile phone system | |
| CN106211298A (en) | A kind of information processing method and mobile terminal | |
| CN106162559A (en) | For estimating equipment and the method for position in a wireless communication system | |
| CN110784837B (en) | Indoor positioning method, device, medium and electronic equipment | |
| CN1889486B (en) | Dynamic energy managing method and apparatus in wireless sensor network | |
| CN110146855A (en) | Radar Intermittent AF panel thresholding calculation method and device | |
| CN105404529A (en) | Popup method, device and system of operation system popup for reinstallation | |
| CN110049442A (en) | Indoor WiFi fingerprint location automatic calibration method and system based on smart phone | |
| CN202931564U (en) | Mobile internet technology based terminal of Internet of Things | |
| CN202210001U (en) | Wireless multi-path multi-monitoring-mode temperature measuring system | |
| Krupanek et al. | Investigations of transmission delays in ZigBee networks | |
| CN201928268U (en) | Multi-frequency point receiver |
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 | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20150902 |