WO2020090019A1

WO2020090019A1 - Odor-sensor-data correcting device, odor-sensor-data correcting method, and computer-readable recording medium

Info

Publication number: WO2020090019A1
Application number: PCT/JP2018/040418
Authority: WO
Inventors: 鈴木　亮太; 山田　聡; 江藤　力; ひろみ清水; 純子渡辺
Original assignee: 日本電気株式会社
Priority date: 2018-10-30
Filing date: 2018-10-30
Publication date: 2020-05-07
Also published as: JPWO2020090019A1; JP7205549B2

Abstract

An odor-sensor-data correcting device 1 for suppressing a measurement error due to the individual difference among odor sensors includes: a calculation unit 2 that calculates a correction coefficient on the basis of first odor data indicating a first odor and second odor data obtained as a result of an odor sensor measuring the first odor; and a correcting unit 3 that corrects, on the basis of the correction coefficient, third odor data obtained as a result of the odor sensor measuring a target odor.

Description

Odor sensor data correction device, odor sensor data correction method, and computer-readable recording medium

The present invention relates to an odor sensor data correction device and an odor sensor data correction method for correcting odor data output by an odor sensor, and further relates to a computer-readable recording medium recording a program for realizing these.

Non-Patent Document 1 discloses an odor sensor provided with a plurality of sensor elements. Specifically, these sensor elements are provided with a sensitive film having different characteristics for each sensor element. In addition, the sensor element is configured to have a unique reaction with respect to the molecule adsorbed on the sensitive film.

However, in the odor sensor having the above-mentioned sensitive film, there are individual differences between the odor sensors depending on the state after the production of the sensitive film, so when measuring odors using different odor sensors, a measurement error occurs between the odor sensors. Occurs.

One example of an object of the present invention is to provide an odor sensor data correction device, an odor sensor data correction method, and a computer-readable recording medium that suppress measurement errors due to individual differences between odor sensors.

In order to achieve the above object, an odor sensor data correction device according to one aspect of the present invention,
Based on the first odor data indicating the first odor and the second odor data obtained by measuring the first odor by the odor sensor, a calculation unit, a calculation unit,
A correction unit that corrects the third odor data obtained by measuring the target odor with the odor sensor based on the correction coefficient,
It is characterized by having.

Further, in order to achieve the above object, an odor sensor data correction method according to one aspect of the present invention,
(A) calculating a correction coefficient based on first odor data indicating a first odor and second odor data obtained by measuring the first odor with an odor sensor;
(B) correcting the third odor data obtained by the odor sensor measuring the target odor based on the correction coefficient,
It is characterized by having.

Further, in order to achieve the above object, a computer-readable recording medium recording the program according to one aspect of the present invention,
On the computer,
(A) calculating a correction coefficient based on first odor data indicating a first odor and second odor data obtained by measuring the first odor with an odor sensor;
(B) correcting the third odor data obtained by the odor sensor measuring the target odor based on the correction coefficient,
It is characterized in that a program including an instruction to execute is recorded.

As described above, according to the present invention, it is possible to suppress measurement errors due to individual differences between odor sensors.

FIG. 1 is a diagram showing an example of an odor sensor data correction device. FIG. 2 is a diagram illustrating an example of a system in a phase for calculating the correction coefficient. FIG. 3 is a diagram showing an example of the data structure of the standard odor data and the measured odor data. FIG. 4 is a diagram showing an example of a waveform of odor data. FIG. 5 is a diagram showing an example of the data structure of the correction coefficient data. FIG. 6 is a diagram showing an example of a system for correcting measured odor data. FIG. 7 is a diagram showing an example of the operation of calculating the correction coefficient of the odor sensor data correction device. FIG. 8 is a diagram showing an example of the operation of the odor sensor data correction device for correcting odor data. FIG. 9 is a diagram illustrating an example of a computer that realizes the odor sensor data correction device.

(Embodiment)
Hereinafter, embodiments of the present invention will be described with reference to FIGS. 1 to 9.

[Device configuration]
First, the configuration of the odor sensor data correction device 1 according to the present embodiment will be described with reference to FIG. FIG. 1 is a diagram showing an example of an odor sensor data correction device.

The odor sensor data correction device 1 shown in FIG. 1 is a device that suppresses measurement errors due to individual differences between odor sensors. Further, as shown in FIG. 1, the odor sensor data correction device 1 includes a calculation unit 2 and a correction unit 3.

Of these, the calculation unit 2 uses the reference odor data (first odor data) indicating the reference odor (first odor) and the measured odor data obtained by the odor sensor measuring the reference odor (second odor data). The correction coefficient is calculated based on the odor data). The correction unit 3 corrects the measured odor data (third odor data) obtained by the odor sensor measuring the target odor based on the correction coefficient.

As described above, in the present embodiment, the calculated correction coefficient is used to correct the measured odor data obtained by the odor sensor measuring the target odor, so that the measurement error due to the individual difference between the odor sensors is suppressed. it can.

Further, the odor sensor data correction device 1 described above may have a configuration in which the calculation unit 2 and the correction unit 3 are divided into separate devices as long as the calculation unit 2 and the correction unit 3 are included.

If the calculation unit 2 and the correction unit 3 are separate devices, for example, the calculation unit 2 may be provided in the inspection device and the correction unit 3 may be provided in the odor analysis device. The inspection device is, for example, a device that sets the above-described correction coefficient and the like at the time of factory shipment. The odor analysis device is, for example, a terminal device directly connected to the odor sensor or a server computer connected to the odor sensor via a network.

[System configuration]
Next, the calculation of the correction coefficient according to the present embodiment will be described more specifically with reference to FIG. FIG. 2 is a diagram illustrating an example of a system in a phase for calculating the correction coefficient.

As shown in FIG. 2, in the phase of calculating the correction coefficient in the present embodiment, the system calculates the correction coefficient using the reference odor sensor 21a, the odor sensor 21b, and the acquisition unit 23 in addition to the calculation unit 2. .. The reference odor sensor 21a has a sensitive film 22a and a sensitive film 22b. The odor sensor 21b has a sensitive film 22c and a sensitive film 22d. Furthermore, the calculation unit 2 includes a preprocessing unit 24, a correction coefficient calculation unit 25, and a correction coefficient acquisition unit 26.

Explain the calculation of the correction coefficient. When calculating the correction coefficient, first, the acquisition unit 23 acquires a measurement result (reference odor data) obtained by the reference odor sensor 21a, which is the reference odor sensor, measuring the reference odor (reference gas).

Note that the reference odor data may be data obtained from one reference odor sensor, or may be a plurality of measurement results obtained by measuring the reference odors from a plurality of reference odor sensors. For example, the standard odor data may be obtained by using statistical processing such as average and median.

Subsequently, the calculation unit 2 acquires the measurement result (measured odor data) obtained by measuring the reference odor by the odor sensor 21b for which the correction coefficient is set. After that, the calculation unit 2 uses the reference odor data and the measured odor data obtained by measuring the reference odor to calculate a correction coefficient for correcting the measured odor data obtained by the odor sensor 21b when measuring the target odor. calculate.

The reference odor sensor 21a is a reference odor sensor used when calculating the correction coefficient, and has one or more sensitive films 22. In the example of FIG. 2, the reference odor sensor 21a has

sensitive films

22a and 22b. Specifically, the reference odor sensor 21a measures the reference odor and outputs the reference odor data for each of the

sensitive films

22a and 22b to the acquisition unit 23.

The odor sensor 21b is an odor sensor for which a correction coefficient is set, and has one or more sensitive films 22. In the example of FIG. 2, the odor sensor 21b has

sensitive films

22c and 22d. Specifically, when setting the correction coefficient, the odor sensor 21b measures the reference odor and outputs the reference odor data to the calculation unit 2 for each of the

sensitive films

22c and 22d. Further, at the time of measurement, the odor of the target is measured, and the measured odor data obtained by measuring the odor of the target for each of the

sensitive films

22c and 22d is output to the calculation unit 2.

Explain the odor sensor 21 (21a, 21b). The odor sensor is a sensor that detects a chemical substance in the air by using an element that detects a chemical substance. Specifically, an odor sensor is a cantilever if it utilizes changes in physical quantities related to device viscoelasticity and dynamic characteristics (mass, moment of inertia, etc.) due to adsorption and desorption of molecules on a sensitive film. Formula, film type, optical type, piezo, vibration response may be used. For example, as the odor sensor, a membrane surface stress sensor (MSS: Membrane-type Surface Stress Sensor) or the like can be considered.

MSS is an odor sensor that has multiple MSS elements. The MSS element has a sensitive film, a support member that supports the sensitive film, a wiring substrate that surrounds the support member, and a plurality of bridges that connect the support member and the wiring substrate. The bridge also has a piezoresistive element.

ㆍ In the sensitive film, when a substance is adsorbed, distortion occurs in the sensitive film. The support member is a member that supports the sensitive film, and has a mechanism of distorting according to the strain of the sensitive film. The bridge is connected to the support member, and when stress is applied to the bridge due to the strain described above, the electrical resistance of the piezoresistive element embedded in the bridge changes. That is, the MSS measures this electrical resistance and outputs odor data representing the measurement result via the wiring board.

Also, the material of the sensitive film of the MSS element differs for each MSS element, but the type of substance detected by the MSS element is not fixed to one. Therefore, the material of the sensitive film differs depending on the set of substances that make up the odor. The sensitive film is formed by depositing a chemical substance on the substrate for constructing the sensitive film. As the attachment method, for example, a method such as liquid application by an inkjet method, a dip method, or vapor deposition is used.

The acquisition unit 23 acquires reference odor data representing the measurement result for each sensitive film 22 from the reference odor sensor 21a. Specifically, the acquisition unit 23 stores the reference odor data acquired from the reference odor sensor 21a in a storage unit (not shown) in advance. For example, the acquisition unit 23 acquires reference odor data in which the sensitized film identification information for identifying each of the sensitized films 22 and the odor data output by each of the responsive films 22 are associated and stored in the storage unit.

The storage unit described above may be provided inside the odor sensor data correction device 1 or may be provided outside the odor sensor data correction device 1. When provided externally, the odor sensor data correction apparatus 1 communicates with the externally provided storage unit to acquire the reference odor data.

FIG. 3 is a diagram showing an example of the data structure of the standard odor data and the measured odor data. The reference odor data 31 in FIG. 3 is “1” and “2” that represent the sensitive film identification information that identifies the

sensitive films

22a and 22b, and “data1” and “data2” that represent the odor data output by each of the

sensitive films

22a and 22b. And are associated and stored.

The calculation unit 2 calculates a correction coefficient for each sensitive film 22. The calculation unit 2 also includes a preprocessing unit 24, a correction coefficient calculation unit 25, and a correction coefficient acquisition unit 26. However, the preprocessing unit 24 may not be provided in the calculation unit 2.

Specifically, the calculation unit 2 first acquires the measured odor data obtained by measuring the reference odor from the target odor sensor 21b. The measured odor data 32 in FIG. 3 represents "3" and "4" representing the sensitive film identification information for identifying the

sensitive films

22c and 22d, and the odor data obtained by measuring the reference odors output by the

sensitive films

22c and 22d. "data3" and "data4" are associated with each other.

Subsequently, the calculation unit 2 calculates the correction coefficient for each sensitive film 22 using the reference odor data and the measured odor data obtained by measuring the reference odor. Specifically, when the sensitive film “1” (22a) and the sensitive film “3” (22c) are the sensitive films corresponding to each other, the calculation unit 2 responds to the odor data “data1” of the sensitive film “1” and the sensitive film. The correction coefficient is calculated using the odor data “data3” of the film “3”. Further, when the sensitive film “2” (22b) and the sensitive film “4” (22d) are corresponding sensitive films, the calculation unit 2 calculates the odor data “data2” and the sensitive film “4” of the sensitive film “2”. The correction coefficient is calculated using the odor data “data4” of “”.

The calculation unit will be described in detail.
The pre-processing unit 24 performs pre-processing on the standard odor data and the measured odor data obtained by measuring the standard odor. Specifically, as shown in Equation 1, the standard odor data and the measured odor data obtained by measuring the standard odor are preprocessed using a linear conversion matrix. However, the linear conversion matrix does not necessarily have to be used for the preprocessing.

The pre-processing includes, for example, (A) acquisition of statistics, (B) downsampling, (C) smoothing, (D) offset removal, (E) feature extraction, and (F) weighting. is there.

Preprocessing will be explained using FIG. FIG. 4 is a diagram showing an example of a waveform of odor data. In FIG. 4, the vertical axis shows the level L of the odor data, and the horizontal axis shows the time t.

(A) Acquisition of statistics will be described.
The acquisition of the statistical amount is, for example, a process of calculating an amplitude and an average. In the process of calculating the amplitude, in the example of FIG. 4, the level L0 at the time t0 and the level Lm at the time tm are acquired, and the difference (Lm−L0) between the level Lm and the level L0 is set as the amplitude. The amplitude is related to the magnitude of the reaction of the sensitive membrane 22. In the example of FIG. 4, the process of calculating the amplitude from the level of the odor data can be expressed using, for example, a linear conversion matrix shown in Formula 2.

In the process of calculating the average, for example, the average of some or all of the levels of odor data is calculated. The process of calculating the average is such that, when calculating the average of the levels from time t0 to time tn in FIG. 4, when the number of odor data sampled during the time from time t0 to time tn is N, the odor data N Calculate the average of the individual levels. By calculating the average in this way, it is expected to reduce noise based on the odor data.

Note that the process of calculating the average from the level of the odor data can be expressed by using a linear conversion matrix as shown in Formula 3, for example.

(B) Down sampling will be described.
The downsampling is, for example, a process (thinning process) of acquiring odor data at predetermined intervals. By using downsampling, the amount of data can be suppressed, so that the processing speed can be improved.

The downsampling process can be expressed using, for example, a linear transformation matrix as shown in Equation 4. The linear transformation matrix of the example of Expression 4 thins out the odor data to 1/3.

(C) Smoothing will be described.
The smoothing is a process of applying a moving average filter, a Gaussian filter, a median filter, or the like to the odor data. By using the smoothing, it is possible to reduce the noise on the level of the odor data. In the case of linear conversion, the smoothing process uses, for example, a moving average filter or a Gaussian filter. For example, the kernels of the equations 5 and 6 are used, respectively.

Explain how to use a specific kernel by using an example of moving average filter. The kernel of the moving average filter is expressed by Equation 5. Here, if the size of the kernel is (1 × 3), the kernel is expressed as (1/3/1/3 1/3). The odor data can be smoothed by the moving average filter by multiplying the kernel and the odor data while shifting them and further adding them.

That is, the processing of the moving average filter can be expressed by using, for example, a linear conversion matrix shown in Expression 7. Also in the Gaussian filter, the odor data can be smoothed by Gaussian by performing the same calculation after determining the kernel by setting the size of the kernel to (1 × 3) and the magnitude of the variance to 1, for example.

In the case of non-linear conversion, for example, in the case of the median filter, if the kernel size is (1 × 3), the data is selected while shifting the odor data by three, and if the median of each is taken, it is smoothed by the median filter. Can be converted. Note that the moving average filter, Gaussian filter, and median filter are examples, and smoothing may be performed using another filter such as a Gabor filter. The presented numerical value is an example, and another numerical value may be used to determine the kernel.

(D) The removal of offset will be described.
The removal of the offset is, for example, a process of subtracting an average value of a part or all of the levels from each level of the odor data, or a process of subtracting the level of the odor data at a predetermined time from each level of the odor data. By removing such an offset, the bias can be removed. Note that the process of subtracting the average value of the levels from the level of the odor data is expressed using, for example, a linear conversion matrix shown in Expression 8.

Further, the process of subtracting the level L0 of the odor data at time t0 in FIG. 4 from each level L of the odor data is expressed by, for example, a linear conversion matrix shown in Expression 9.

(E) The extraction of the feature amount of the rate constant contribution will be described.
The odor data may be converted such that the magnitude of the contribution of the speed of adsorption / desorption of molecules to the sensitive film (rate constant) is used as the characteristic amount. For example, a linear transformation matrix as shown in Expression 10 may be used.

(F) Weighting will be described.
Weighting is a process of assigning weight by designating the odor data of a place to be emphasized. By weighting, any important part of the odor data can be emphasized. For weighting, a linear transformation matrix as shown in Expression 11 may be used. The linear conversion matrix shown in Expression 11 is a linear conversion matrix for making the levels L0 and Lm notice in the odor data acquired from the time immediately before the rising time t0 of the waveform in FIG. 4 to the time immediately after the falling time tm. Is. That is, the values 0.1, 0.3, and 0.1 of Expression 11 make the levels L1 and Lm and the level L in the vicinity thereof large.

Further, when performing regression analysis or discriminant analysis using odor data, for example, a linear model such as Expression 12 is used as it is or as a part of an equation. When this linear model is used, its output is determined based on the total value obtained by adding the weights w _i and x _i together. That is, it is considered that the larger the absolute values of w _i and x _i are, the more they contribute to the output. Therefore, by weighting the coefficient of the created linear model, the odor data of the part that contributes to the output of regression analysis or discriminant analysis can be emphasized. When the coefficient w _{i of the} linear model is used as the weight, for example, a linear conversion matrix shown in Expression 12 is used. At this time, the odor data may or may not be standardized.

Note that two or more of the above-mentioned pretreatments may be combined and used. For example, (C) smoothing may be performed after (D) offset is removed.

The correction coefficient calculation unit 25 calculates the correction coefficient α by, for example, minimizing the formula shown in Expression 13. For the minimization, for example, the least squares method or the stochastic gradient descent method may be used.

Alternatively, the correction coefficient calculation unit 25 may calculate the correction coefficient α by applying the above-described linear conversion matrix (pre-processing) and then minimizing it, as shown in Expression 14, for example.

Further, the correction coefficient calculation unit 25 may calculate the correction coefficient α _k by minimizing each condition, as shown in Expression 15. Here, the condition is a measurement condition in which the temperature, the humidity, the type of the standard odor, and the like are combined.

Alternatively, for example, the correction coefficient calculation unit 25 calculates the correction coefficient α _k by applying and minimizing the above-described linear conversion matrix (preprocessing) for each condition as shown in Expression 16. Good.

The correction coefficient acquisition unit 26 acquires the correction coefficient calculated by the correction coefficient calculation unit 25. Specifically, the correction coefficient acquisition unit 26 acquires the correction coefficient from the correction coefficient calculation unit 25, then generates the correction coefficient data, and stores the correction coefficient data in a storage unit (not illustrated). The correction coefficient acquisition unit 26 generates correction coefficient data using the correction coefficient α or α _k , for example, and is provided in the storage unit provided inside the odor sensor data correction device 1 as described above, or provided outside. Stored in the storage unit.

FIG. 5 is a diagram showing an example of the data structure of the correction coefficient data. The correction coefficient data 51 shown in FIG. 5 includes “S1” that represents sensor identification information that identifies the odor sensor 21b, “3” and “4” that represent sensitive film identification information that identifies the

sensitive films

22c and 22d, and the sensitive film. "Cr3" and "cr4" representing the correction coefficient α for each of 22c and 22d are stored in association with each other.

In addition, the correction coefficient data 52 shown in FIG. 5 includes “S1” indicating sensor identification information for identifying the odor sensor 21b, “con1” and “con2” indicating conditions, and a sensitive film identification for identifying the sensitive films 22c and 22d. "3" and "4" representing information and "cr3_1""cr4_1""cr3_2""cr4_2" representing the correction coefficients α _k for the

sensitive films

22c and 22d for each condition are stored in association with each other.

Next, the correction of the odor sensor data according to the present embodiment will be described more specifically with reference to FIG. FIG. 6 is a diagram showing an example of a system for correcting measured odor data.

As shown in FIG. 6, in the phase of correcting the measured odor data in the present embodiment, the system has the odor sensor 21 and the correction unit 3. Further, the system performs odor analysis using the odor analysis unit 61 and the output unit 62, and outputs the analysis result.

The correction unit 3 corrects the measured odor data for each of the

sensitive films

22c and 22d measured by the odor sensor 21b at the time of measurement, using the correction coefficient α or α _k corresponding to the

sensitive films

22c and 22d. To do. Specifically, the correction unit 3 first acquires the correction coefficient data corresponding to the odor sensor 21b. Further, the correction unit 3 acquires the measured odor data measured by the odor sensor 21b.

Next, the correction unit 3 corrects the measured odor data for each of the

sensitive films

22c and 22d using the measured odor data obtained by measuring the target odor and the correction coefficient data. When there is no condition, the correction unit 3 acquires the correction coefficient data 51 and multiplies the level of the odor data for each of the

sensitive films

22c and 22d by the correction coefficient α.

Further, if there is a condition, the correction unit 3 refers to the correction coefficient data 52 using the condition, acquires the correction coefficient α _k for each of the

sensitive films

22c and 22d associated with the condition, and sets the odor data level to Multiply by the correction coefficient α _k . After that, the correction unit 3 transmits the corrected measured odor data to the odor analysis unit 61 that performs odor analysis.

The odor analysis unit 61 analyzes the corrected measured odor data corresponding to the sensitive film 22. Specifically, the odor analysis unit 61 is an analyzer that performs odor analysis, and may perform regression, discriminant analysis, or the like using a prediction model created in advance, or may perform visualization, or simply You only need to store the data.

The output unit 62 acquires output information representing an analysis result (an odor identification result) converted into a format that can be output from the odor analysis unit 61, and outputs an image and a sound generated based on the output information. .. The output unit 62 is, for example, an image display device using a liquid crystal, an organic EL (Electro Luminescence), or a CRT (Cathode Ray Tube). Further, the image display device may include a voice output device such as a speaker. The output unit 62 may be a printing device such as a printer.

[Device operation]
Next, the operation of the odor sensor data correction device 1 according to the embodiment of the present invention will be described with reference to FIGS. 7 and 8. FIG. 7 is a diagram showing an example of the operation of calculating the correction coefficient of the odor sensor data correction device. FIG. 8 is a diagram showing an example of the operation of the odor sensor data correction device for correcting odor data. In the following description, FIGS. 1 to 6 will be referred to as appropriate. Further, in the present embodiment, the odor sensor data correction method is implemented by operating the odor sensor data correction device 1. Therefore, the description of the odor sensor data correction method in the present embodiment will be replaced with the following description of the operation of the odor sensor data correction device 1.

Calculation of the correction coefficient according to the present embodiment will be described with reference to FIG. 7.
As shown in FIG. 7, first, the acquisition unit 23 acquires reference odor data for each sensitive film 22 from the reference odor sensor 21a (step A1). Specifically, in step A1, the acquisition unit 23 acquires reference odor data in which the sensitive film identification information for identifying each of the sensitive films 22 and the odor data output by each of the sensitive films 22 are associated with each other, and is stored in the storage unit. Remember. See the reference odor data 31 in FIG.

Subsequently, the calculation unit 2 acquires the measured odor data obtained by measuring the reference odor for each of the sensitive films 22 from the odor sensor 21b for which the correction coefficient is set (step A2). Specifically, in step A2, the calculation unit 2 associates the sensitive film identification information for identifying each of the sensitive films 22 with the odor data output by each of the sensitive films 22, and outputs the measured odor data obtained by measuring the reference odor. get.

Subsequently, the preprocessing unit 24 of the calculation unit 2 performs preprocessing on the standard odor data and the measured odor data obtained by measuring the standard odor (step A3). Specifically, in step A3, the preprocessing unit 24 preprocesses the reference odor data and the measured odor data by using a linear conversion matrix or the like, as shown in Formula 1. However, the pre-processing of step A3 may be omitted.

The preprocessing includes, for example, (A) acquisition of statistics, (B) downsampling, (C) smoothing, (D) offset removal, (E) feature extraction, and (F) weighting. Processing.

Subsequently, the correction coefficient calculation unit 25 of the calculation unit 2 calculates the correction coefficient for each sensitive film 22 using the reference odor data and the measured odor data (step A4). Specifically, in step A4, when the sensitive film “1” (22a) and the sensitive film “3” (22c) are the corresponding sensitive films, the correction coefficient calculation unit 25 determines that the odor of the sensitive film “1” is bad. The correction coefficient is calculated using the data “data1” and the odor data “data3” of the sensitive film “3”. Further, when the sensitive film “2” (22b) and the sensitive film “4” (22d) are corresponding sensitive films, the calculation unit 2 calculates the odor data “data2” and the sensitive film “4” of the sensitive film “2”. The correction coefficient is calculated using the odor data “data4” of “”.

The correction coefficient calculation unit 25 calculates the correction coefficient α by minimizing, for example, the formula shown in Expression 13. Alternatively, the correction coefficient calculation unit 25 may calculate the correction coefficient α by applying the above-described linear conversion matrix (preprocessing) and then minimizing it, as shown in Expression 14, for example.

Furthermore, the correction coefficient calculation unit 25 may calculate the correction coefficient α _k by minimizing it for each condition, as shown in Expression 15. Here, the condition is a measurement condition in which the temperature, the humidity, the type of the standard odor, and the like are combined. Alternatively, the correction coefficient calculation unit 25 may calculate the correction coefficient α _k by performing the above-described linear conversion matrix (preprocessing) for each condition and minimizing it, as shown in Expression 16.

Subsequently, the correction coefficient acquisition unit 26 of the calculation unit 2 acquires the correction coefficient calculated by the correction coefficient calculation unit 25 and stores it in the storage unit (step A5). Specifically, in step A5, the correction coefficient acquisition unit 26 acquires the correction coefficient α or α _k from the correction coefficient calculation unit 25 and then generates the correction coefficient data to be stored in the odor sensor data correction device 1. The data is stored in a storage unit provided or an external storage unit. See the

correction coefficient data

51 and 52 in FIG.

Correction of odor sensor data according to the present embodiment will be described with reference to FIG.
As shown in FIG. 8, first, the correction unit 3 acquires the measured odor data measured by the odor sensor 21b from the odor sensor 21b (step B1). The correction unit 3 acquires the correction coefficient data corresponding to the odor sensor 21b from the storage unit (step B2).

The correction unit 3 corrects the measured odor data obtained by measuring the target odor for each of the

sensitive films

22c and 22d using the measured odor data obtained by measuring the target odor and the correction coefficient data (step B3). Specifically, in step B3, the correction unit 3 corrects the measured odor data obtained by measuring the target odor by multiplying the level of the odor data for each of the

sensitive films

22c and 22d by the correction coefficient α or α _k .

When there is no condition, the correction unit 3 acquires the correction coefficient data 51 and multiplies the level of the odor data for each of the

sensitive films

22c and 22d by the correction coefficient α.

sensitive films

22c and 22d associated with the condition, and sets the odor data level to Multiply by the correction coefficient α _k .

The odor analysis unit 61 analyzes the corrected measured odor data corresponding to the sensitive film 22 (step B4). Specifically, in step B4, the odor analysis unit 61 acquires the odor data measured by the

sensitive films

22c and 22d of the odor sensor 21b, and analyzes the odor using an analyzer.

The output unit 62 acquires output information representing an analysis result (an odor identification result) converted into a format that can be output from the odor analysis unit 61, and outputs an image and a sound generated based on the output information. (Step B5).

[Effects of this Embodiment]
As described above, according to the present embodiment, the measured odor data is corrected using the calculated correction coefficient, so that the measurement error due to the individual difference between the odor sensors can be suppressed.

Also, since the measured odor data is corrected for each sensitive film using the correction coefficient calculated for each sensitive film, the measurement error due to the individual difference between the odor sensors can be further suppressed.

Further, if there is a manufacturing variation due to the shape, size, thickness, chemical properties, etc. of the sensitive film, the odor analysis cannot be performed accurately, so the accuracy of measuring the odor decreases, but according to the present embodiment, Since the measured odor data is corrected using the calculated correction coefficient, the accuracy of measuring the odor can be improved.

[program]
The program in the embodiment of the present invention may be a program which causes a computer to execute steps A1 to A5 shown in FIG. 7 or a program which causes steps B1 to B5 shown in FIG. 8 to be executed. The odor sensor data correction device and the odor sensor data correction method according to the present embodiment can be realized by installing and executing this program on a computer. In this case, the processor of the computer functions as the acquisition unit 23, the calculation unit 2 (the preprocessing unit 24, the correction coefficient calculation unit 25, the correction coefficient acquisition unit 26), the correction unit 3, the odor analysis unit 61, and the output unit 62, Perform processing.

Also, the program in the present embodiment may be executed by a computer system constructed by a plurality of computers. In this case, for example, each computer serves as any one of the calculation unit 2 (preprocessing unit 24, correction coefficient calculation unit 25, correction coefficient acquisition unit 26), correction unit 3, odor analysis unit 61, and output unit 62. May function.

[Physical configuration]
Here, a computer that realizes the odor sensor data correction device by executing the program according to the embodiment will be described with reference to FIG. 9. FIG. 9 is a block diagram showing an example of a computer that realizes the odor sensor data correction device.

As shown in FIG. 9, the computer 110 includes a CPU 111, a main memory 112, a storage device 113, an input interface 114, a display controller 115, a data reader / writer 116, and a communication interface 117. These units are connected to each other via a bus 121 so as to be able to perform data communication with each other. The computer 110 may include a GPU (Graphics Processing Unit) or an FPGA (Field-Programmable Gate Array) in addition to or instead of the CPU 111.

The CPU 111 expands the program (code) according to the present embodiment stored in the storage device 113 into the main memory 112 and executes these in a predetermined order to perform various calculations. The main memory 112 is typically a volatile storage device such as a DRAM (Dynamic Random Access Memory). Further, the program in the present embodiment is provided in a state of being stored in computer-readable recording medium 120. The program in the present embodiment may be distributed on the Internet connected via communication interface 117.

Further, as a specific example of the storage device 113, a semiconductor storage device such as a flash memory can be cited in addition to a hard disk drive. The input interface 114 mediates data transmission between the CPU 111 and an input device 118 such as a keyboard and a mouse. The display controller 115 is connected to the display device 119 and controls the display on the display device 119.

The data reader / writer 116 mediates data transmission between the CPU 111 and the recording medium 120, reads a program from the recording medium 120, and writes the processing result in the computer 110 to the recording medium 120. The communication interface 117 mediates data transmission between the CPU 111 and another computer.

Specific examples of the recording medium 120 include general-purpose semiconductor storage devices such as CF (Compact Flash (registered trademark)) and SD (Secure Digital), magnetic recording media such as a flexible disk, or CD- An optical recording medium such as a ROM (Compact Disk Read Only Memory) can be given.

Note that the odor sensor data correction device 1 in the present embodiment can be realized not by using a computer in which a program is installed but by using hardware corresponding to each unit. Further, part of the odor sensor data correction device 1 may be realized by a program, and the remaining part may be realized by hardware.

[Appendix]
Regarding the above-described embodiment, the following supplementary notes will be disclosed. The whole or part of the exemplary embodiments described above can be expressed by (Supplementary Note 1) to (Supplementary Note 15) described below, but the present invention is not limited to the following description.

(Appendix 1)
Based on the first odor data indicating the first odor and the second odor data obtained by measuring the first odor by the odor sensor, a calculation unit, a calculation unit,
A correction unit that corrects the third odor data obtained by measuring the target odor with the odor sensor based on the correction coefficient,
An odor sensor data correction device comprising:

(Appendix 2)
The odor sensor data correction device according to attachment 1,
The odor sensor has one or more sensitive films,
The odor sensor data correction device, wherein the calculation unit calculates the correction coefficient for each of the sensitive films.

(Appendix 3)
The odor sensor data correction device according to attachment 2,
The odor sensor data correction device, wherein the first odor data is generated for each of the sensitive films based on the first odor serving as a reference.

(Appendix 4)
The odor sensor data correction device according to attachment 3,
The odor sensor data correction device, wherein the correction unit corrects the third odor data for each of the sensitive films output by the odor sensor using the correction coefficient corresponding to the sensitive film.

(Appendix 5)
The odor sensor data correction device according to any one of appendices 1 to 4,
The odor sensor data correction device, wherein the calculation unit linearly converts the first odor data and the second odor data.

(Appendix 6)
(A) calculating a correction coefficient based on first odor data indicating a first odor and second odor data obtained by measuring the first odor with an odor sensor;
(B) correcting the third odor data obtained by the odor sensor measuring the target odor based on the correction coefficient,
A method for correcting odor sensor data, comprising:

(Appendix 7)
The odor sensor data correction method according to attachment 6,
The odor sensor has one or more sensitive films,
In the step (b), the correction coefficient is calculated for each of the sensitive films, and the odor sensor data correction method is characterized.

(Appendix 8)
The odor sensor data correction method according to attachment 7,
The odor sensor data correction method, wherein the first odor data is generated for each of the sensitive films based on the reference first odor.

(Appendix 9)
The odor sensor data correction method according to attachment 8,
In the step (b), the third odor data for each of the sensitive films output by the odor sensor is corrected using the correction coefficient corresponding to the sensitive film. Method.

(Appendix 10)
The odor sensor data correction method according to any one of appendices 6 to 9,
In the step (a), the odor sensor data correction method is characterized in that the first odor data and the second odor data are linearly converted.

(Appendix 11)
On the computer,
(A) calculating a correction coefficient based on first odor data indicating a first odor and second odor data obtained by measuring the first odor with an odor sensor;
(B) correcting the third odor data obtained by the odor sensor measuring the target odor based on the correction coefficient,
A computer-readable recording medium recording a program including an instruction to execute.

(Appendix 12)
The computer-readable recording medium according to attachment 11,
The odor sensor has a plurality of sensitive films,
A computer-readable recording medium characterized in that, in the step (b), the correction coefficient is calculated for each of the sensitive films.

(Appendix 13)
The computer-readable recording medium according to attachment 12,
The computer-readable recording medium, wherein the first odor data is generated for each of the sensitive films based on the reference first odor.

(Appendix 14)
The computer-readable recording medium according to attachment 13,
In the step (b), the third odor data for each of the sensitive films output by the odor sensor is corrected using the correction coefficient corresponding to the sensitive film, which is computer-readable. recoding media.

(Appendix 15)
The computer-readable recording medium according to any one of appendices 11 to 14,
A computer-readable recording medium, wherein in the step (a), the first odor data and the second odor data are linearly converted.

Although the present invention has been described with reference to the exemplary embodiments, the present invention is not limited to the above exemplary embodiments. Various modifications that can be understood by those skilled in the art can be made to the configuration and details of the present invention within the scope of the present invention.

As described above, according to the present invention, it is possible to suppress measurement errors due to individual differences between odor sensors. INDUSTRIAL APPLICABILITY The present invention is useful in fields where odor sensors require improved measurement accuracy.

1 Odor Sensor Data Correction Device 2 Calculation Unit 3

Correction Unit

21, 21a,

21b Odor Sensor

22, 22a, 22b, 22c, 22d Sensitive Membrane 23 Acquisition Unit 24 Pre-Processing Unit 25 Correction Coefficient Calculation Unit 26 Correction Coefficient Acquisition Unit 31 Standard Odor Data 32 Measured odor data 51 Correction coefficient data 61 Odor analysis unit 62 Output unit 110 Computer 111 CPU
112 Main Memory 113 Storage Device 114 Input Interface 115 Display Controller 116 Data Reader / Writer 117 Communication Interface 118 Input Equipment 119 Display Device 120 Recording Medium 121 Bus

Claims

Based on the first odor data indicating the first odor and the second odor data obtained by measuring the first odor by the odor sensor, a correction coefficient is calculated, and a calculating unit,
Based on the correction coefficient, a correction unit that corrects the third odor data obtained by measuring the target odor with the odor sensor,
An odor sensor data correction device comprising:
The odor sensor data correction device according to claim 1,
The odor sensor has one or more sensitive films,
The odor sensor data correction device, wherein the calculation means calculates the correction coefficient for each of the sensitive films.
The odor sensor data correction device according to claim 2,
The odor sensor data correction device, wherein the first odor data is generated for each of the sensitive films based on the first odor serving as a reference.
The odor sensor data correction device according to claim 3,
The odor sensor data correction device, wherein the correction means corrects the third odor data for each of the sensitive films output by the odor sensor using the correction coefficient corresponding to the sensitive film.
The odor sensor data correction device according to any one of claims 1 to 4,
The odor sensor data correction device, wherein the calculation means performs a linear conversion on the first odor data and the second odor data.
(A) calculating a correction coefficient based on first odor data indicating a first odor and second odor data obtained by measuring the first odor with an odor sensor;
(B) correcting the third odor data obtained by the odor sensor measuring the target odor based on the correction coefficient,
A method for correcting odor sensor data, comprising:
The odor sensor data correction method according to claim 6,
The odor sensor has a plurality of sensitive films,
In the step (b), the correction coefficient is calculated for each of the sensitive films, and the odor sensor data correction method is characterized.
The odor sensor data correction method according to claim 7,
The odor sensor data correction method, wherein the first odor data is generated for each of the sensitive films based on the reference first odor.
The odor sensor data correction method according to claim 8,
In the step (b), the third odor data for each of the sensitive films output by the odor sensor is corrected using the correction coefficient corresponding to the sensitive film. Method.
The odor sensor data correction method according to any one of claims 6 to 9,
The odor sensor data correction method, wherein in the step (a), the first odor data and the second odor data are linearly converted.
On the computer,
(A) calculating a correction coefficient based on first odor data indicating a first odor and second odor data obtained by measuring the first odor with an odor sensor;
(B) correcting the third odor data obtained by the odor sensor measuring the target odor based on the correction coefficient,
A computer-readable recording medium recording a program including an instruction to execute.
The computer-readable recording medium according to claim 11,
The odor sensor has a plurality of sensitive films,
A computer-readable recording medium characterized in that, in the step (b), the correction coefficient is calculated for each of the sensitive films.
The computer-readable recording medium according to claim 12,
The computer-readable recording medium, wherein the first odor data is generated for each of the sensitive films based on the reference first odor.
The computer-readable recording medium according to claim 13,
In the step (b), the third odor data for each of the sensitive films output by the odor sensor is corrected using the correction coefficient corresponding to the sensitive film, which is computer-readable. recoding media.
The computer-readable recording medium according to any one of claims 11 to 14,
A computer-readable recording medium, wherein in the step (a), the first odor data and the second odor data are linearly converted.