JP2017156164A - Sensor system, sensor correction device, sensor correction method, and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a sensor system, a sensor correction device, a sensor correction method, and a program which can increase the accuracy of the sensor while suppressing increase in the cost related to the sensor.SOLUTION: A sensor system 100 includes a first sensor 10 as a target of a correction, a second sensor 20 as a reference of a correction, and a sensor correction device 30. The sensor correction device 30 includes a data acquisition part 31 for acquiring first sensor data output from the first sensor 10 and second sensor data output from the second sensor 20, an error learning unit 32 for learning an error of the first sensor by comparing the first sensor data and the second sensor data, and a correction processing unit 33 for correcting the first sensor data based on the result of learning by the error learning unit 32.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a sensor system using a sensor for detecting an environmental situation, a sensor correction apparatus and a sensor correction method for correcting an output value of the sensor, and further relates to a program for realizing them.

  In general, the production of agricultural products is performed based on the experience of the producer and the intuition backed by the experience, and it is difficult for an inexperienced producer to stably produce the agricultural product. On the other hand, in Japan, experienced producers are aging, and it is required that producers with little experience can stably produce agricultural products in a short period of time.

  Therefore, in recent years, in the agricultural field, attempts have been made to support producers by utilizing IT technology. Specifically, a system for monitoring the growth state of agricultural products based on sensor data from various sensors such as a humidity sensor, a temperature sensor, and an illuminance sensor installed in a farm (see Non-Patent Documents 1 and 2, for example). ) Has been proposed. In addition, a system (for example, selecting the optimum farm work at that time by collating current sensor data with a knowledge base that associates and accumulates past sensor data and history data of farm work executed in the past (for example, Non-patent document 3) has also been proposed.

NTT DoCoMo, “Aiming to solve social issues + d's challenge”, [online], DOCOMO's corporate website DOCOMO Business Online, [Search January 12, 2016], Internet <URL: http: // www .docomo.biz / plus_d_agriculture / # chart_area> NEC Solution Innovator Co., Ltd., “NEC Agricultural Guidance Support System”, [online], Realization of unified management of garden information and growth information, etc. [Search January 12, 2016], Internet <URL: http: // www .nec-solutioninnovators.co.jp / sl / einou / feature / feature_1.html> Kyowa Exeo Co., Ltd., “Agriculture ICT Solution”, [online], Kyowa Exeo Network Integration, [Search January 12, 2016], Internet <URL: https://www.exeo.co.jp/jigyou/ns /pdf/nougyou_sh.pdf>

  By the way, the systems disclosed in Non-Patent Document 1, Non-Patent Document 2 and Non-Patent Document 3 are characterized by using sensor data, and therefore the accuracy of the sensor is important. However, the higher the accuracy of the sensor, the higher the cost. Therefore, if the accuracy of the sensor is increased, the cost of the system is increased.

  An example of an object of the present invention is to provide a sensor system, a sensor correction device, a sensor correction method, and a program that can improve the accuracy of the sensor while solving the above-described problems and suppressing an increase in cost of the sensor. There is.

In order to achieve the above object, a sensor system according to an aspect of the present invention includes:
A first sensor to be corrected, a second sensor to be a correction reference, and a sensor correction device,
The sensor correction device includes:
A data acquisition unit that acquires first sensor data output from the first sensor and second sensor data output from the second sensor;
An error learning unit that compares the first sensor data with the second sensor data and learns an error in the first sensor;
A correction processing unit that corrects the first sensor data based on a result of learning by the error learning unit.
It is characterized by that.

In order to achieve the above object, a sensor correction apparatus according to one aspect of the present invention includes:
A data acquisition unit that acquires first sensor data output from the first sensor to be corrected and second sensor data output from the second sensor that is a reference for correction;
An error learning unit that compares the first sensor data with the second sensor data and learns an error in the first sensor;
A correction processing unit that corrects the first sensor data based on a result of learning by the error learning unit.
It is characterized by that.

Furthermore, in order to achieve the above object, a sensor correction method according to one aspect of the present invention includes:
(A) acquiring first sensor data output from a first sensor to be corrected and second sensor data output from a second sensor serving as a correction reference;
(B) comparing the first sensor data and the second sensor data to learn an error in the first sensor;
(C) correcting the first sensor data based on the result of learning in the step (b),
It is characterized by that.

In order to achieve the above object, a program according to one aspect of the present invention includes:
On the computer,
(A) acquiring first sensor data output from a first sensor to be corrected and second sensor data output from a second sensor serving as a correction reference;
(B) comparing the first sensor data and the second sensor data to learn an error in the first sensor;
(C) correcting the first sensor data based on the learning result in the step (b);
Is executed.

  As described above, according to the present invention, it is possible to improve the accuracy of the sensor while suppressing an increase in cost of the sensor.

FIG. 1 is a block diagram showing a configuration of a sensor system and a sensor correction apparatus according to Embodiment 1 of the present invention. 2A and 2B are diagrams for explaining sensor data interpolation performed in the first embodiment of the present invention. FIG. 2A shows an example of sensor data before interpolation, and FIG. 2B shows interpolation. An example of the later sensor data is shown. FIG. 3 is a diagram for explaining error learning performed in the first embodiment of the present invention. FIG. 3A shows an example of an error between correction target sensor data and reference sensor data. (B) shows an example of a function obtained by learning. FIG. 4 is a flowchart showing the operation during the learning process of the sensor correction apparatus according to Embodiment 1 of the present invention. FIG. 5 is a flowchart showing the operation during the correction process of the sensor correction apparatus according to the first embodiment of the present invention. FIG. 6 is a block diagram showing the configuration of the sensor system and the sensor correction apparatus according to Embodiment 2 of the present invention. FIG. 7 is a diagram showing deterioration determination in Embodiment 2 of the present invention. FIG. 8 is a block diagram illustrating an example of a computer that implements the sensor correction apparatus according to Embodiments 1 and 2 of the present invention.

(Embodiment 1)
Hereinafter, a sensor system, a sensor correction device, a sensor correction method, and a program according to Embodiment 1 of the present invention will be described with reference to FIGS.

[System configuration]
First, the configuration of the sensor system and the sensor correction apparatus according to the first embodiment will be described with reference to FIG. FIG. 1 is a block diagram showing a configuration of a sensor system and a sensor correction apparatus according to Embodiment 1 of the present invention.

  As shown in FIG. 1, the sensor system 100 according to the first embodiment includes a first sensor 10 to be corrected (hereinafter referred to as “correction target sensor”) 10 and a second sensor to be a correction reference. (Referred to as “reference sensor”) 20 and a sensor correction device 30. Further, the correction sensor 10 and the reference sensor 20 are installed in the same field and detect the same target.

  As illustrated in FIG. 1, the sensor correction device 30 includes a data acquisition unit 31, an error learning unit 32, and a correction processing unit 33. The data acquisition unit 31 includes first sensor data output from the correction target sensor 10 (hereinafter referred to as “correction target sensor data”) and reference sensor data output from the reference sensor 20 (hereinafter referred to as “reference sensor data”). ) And get. The error learning unit 32 compares the correction target sensor data and the reference sensor data to learn an error in the correction target sensor 10. The correction processing unit 33 corrects the correction target sensor data based on the learning result by the error learning unit 32.

  As described above, in the first embodiment, since the error of the correction target sensor 10 is learned and correction is performed based on the learning result, even if the correction target sensor 10 is not a very expensive sensor, the correction target is corrected. The accuracy of the sensor 10 is improved. In addition, since the reference sensor 20 is not required after the learning is completed, it is not necessary to use an expensive sensor as the correction target sensor 10, and according to the present embodiment, the cost of the sensor increases. It is also possible to suppress this.

  Next, the sensor system 100 and the sensor correction device 30 according to the first embodiment will be described more specifically. As shown in FIG. 1, in the first embodiment, the sensor correction device 30 is connected to the farming support system 200 via a network (not shown in FIG. 1). The sensor correction device 30 improves the accuracy of sensor data required by the farming support system 200.

The farming support system 200 is a system that monitors the growth state of crops based on sensor data from various sensors and accumulates sensor data and farm work history data. Specific examples of the farming system 200 include the systems disclosed in Non-Patent Document 1 to Non-Patent Document 3.

  In the first embodiment, the sensor correction device 30 improves the accuracy of sensor data required for the farming support system 200. In the first embodiment, examples of the correction target sensor and the reference sensor include a humidity sensor, a temperature sensor, an illuminance sensor, and the type of sensor (detection target) is not particularly limited.

  Further, the detection targets of the correction target sensor 10 and the reference sensor 20 are the same as described above. For example, the correction target sensor 10 and the reference sensor 20 both use the humidity of the field as a detection target. However, in the present embodiment, the reference sensor has a higher detection accuracy than the correction target sensor, and an expensive sensor is used.

  Furthermore, although not shown in FIG. 1, a data transmission device for transmitting sensor data to the sensor correction device 30 may be attached to each of the correction target sensor 10 and the reference sensor 20. For example, the data transmission device converts sensor data output from the correction target sensor 10 or the reference sensor 20 into a digital signal, and transmits the obtained digital signal to the sensor correction device 30 by wire or wirelessly.

  In the first embodiment, when the data acquisition unit 31 acquires sensor data from each of the correction target sensor 10 and the reference sensor 20, there is correction target sensor data and reference sensor data whose output timing does not overlap with the other sensor data. Determine whether you are doing. Then, when such sensor data exists, the data acquisition unit 31 overlaps the output timing with the other sensor data based on the time-series change in either or both of the correction target sensor data and the reference sensor data. Interpolate sensor data. This is because humidity, temperature, illuminance, and the like, which are correction targets, change with time. This will be specifically described below.

  2A and 2B are diagrams for explaining sensor data interpolation performed in the first embodiment of the present invention. FIG. 2A shows an example of sensor data before interpolation, and FIG. 2B shows interpolation. An example of the later sensor data is shown. In the graphs shown in FIGS. 2A and 2B, the vertical axis indicates a value (measurement value) specified by sensor data, and the horizontal axis indicates time.

  As shown in FIG. 2A, it is assumed that there is a difference between the time when the correction target sensor data is acquired and the time when the reference sensor data is acquired. In this case, in the example of FIG. 2B, since the data acquisition unit 31 interpolates the correction target data, the data acquisition unit 31 performs linear interpolation using two correction target data at times before and after the time to be added. The correction target data being calculated is calculated. Then, the data acquisition unit 31 adds the calculated correction target data. Note that, unlike the example of FIG. 2B, the data acquisition unit 31 can also perform interpolation on the reference sensor data.

  In the first embodiment, in order to learn the error, the error learning unit 31 first calculates the difference between the values of the correction target sensor data and the reference sensor data whose output times overlap each other. Asking. Subsequently, the error learning unit 31 performs a statistical process on the obtained set of errors for each set to obtain a function that defines the relationship between the value of the correction target sensor data and the error. This will be specifically described below.

FIG. 3 is a diagram for explaining error learning performed in the first embodiment of the present invention. FIG. 3A shows an example of an error between correction target sensor data and reference sensor data. (B) shows an example of a function obtained by learning. In the graph shown in FIG. 3A, the vertical axis indicates a value (measurement value) specified by sensor data, and the horizontal axis indicates time. In the graph shown in FIG. 3B, the vertical axis indicates the error value, and the horizontal axis indicates the value (measurement value) of the correction target sensor data.

  As shown in FIG. 3 (a), the error learning unit 31 first identifies a set of correction target sensor data and reference sensor data whose time coincides, and sets error values d1 to d5 for each set. calculate. Subsequently, the error learning unit 31 creates a two-dimensional data string of the correction target sensor data value (measured value) and the corresponding error. Further, in the example of FIG. 3B, the error learning unit 31 obtains a linear regression line (y = ax + b) from the created two-dimensional data string, and obtains the obtained straight line “the value of the correction target sensor data and the error. A function that defines the relationship. In this function, y is an error and x is a value of sensor data to be corrected.

  In the first embodiment, the correction processing unit 33 corrects the correction target sensor data based on the function obtained by the error learning unit 31. Specifically, when the correction target sensor data is acquired by the data acquisition unit 31 after the above-described function is obtained, the correction processing unit 33 sets the value of the acquired correction target sensor data to the above-described function (y = Ax + b) to calculate the error. And the correction | amendment process part 33 subtracts an error from correction object sensor data, and transmits the obtained value (value after correction | amendment) to the farming support system 200 as correction object data.

[System operation]
Next, operations of the sensor system and the sensor correction apparatus according to the embodiment of the present invention will be described with reference to FIGS. In the first embodiment, the sensor correction method is performed by operating the sensor correction device. Therefore, the description of the sensor correction method in the present embodiment is replaced with the following description of the operation of the sensor system and sensor correction apparatus. In the following description, FIGS. 1 to 3 are referred to as appropriate.

  First, an error learning process performed by the sensor correction apparatus 10 will be described with reference to FIG. FIG. 4 is a flowchart showing the operation during the learning process of the sensor correction apparatus according to Embodiment 1 of the present invention.

  As shown in FIG. 4, first, the data acquisition unit 31 acquires correction target sensor data output from the correction target sensor 10 and reference sensor data output from the reference sensor 20 (step A1). .

  Next, the data acquisition unit 31 determines whether the set number of correction target sensor data and reference sensor data have been acquired (step A2). As a result of the determination, if the set number of sensor data cannot be acquired, the data acquisition unit 31 executes Step A1 again. On the other hand, if it is determined that the set number of sensor data has been acquired, the data acquisition unit 31 executes step A3.

  In the case of Yes in step A2, the data acquisition unit 31 determines whether or not sensor data interpolation is necessary (step A3). Specifically, the data acquisition unit 31 determines whether there is correction target sensor data and reference sensor data whose output timing does not overlap with the other sensor data.

  If the result of determination in step A3 is that interpolation is not necessary, step A5 is executed. On the other hand, if the interpolation is necessary as a result of the determination in step A3, the data acquisition unit 31 performs sensor processing on either the correction target sensor data or the reference sensor data as shown in FIG. Data interpolation is executed (step A4).

In the case of No in step A3, when step A4 is executed, the error learning unit 31 compares the correction target sensor data with the reference sensor data to learn an error in the correction target sensor data (step A5). Specifically, as shown in FIG. 3B, the error learning unit 31 calculates an error value for each set of correction target sensor data and reference sensor data having the same time, and corrects the correction target. A linear regression line (y = ax + b) is obtained from a two-dimensional data string of sensor data values (measured values) and corresponding errors.

  The execution of step A5 ends the learning process, but steps A1 to A5 may be executed again after the set period has elapsed. Further, when the learning process ends, the correction process by the correction processing unit 33 is enabled.

  Next, correction processing of correction target sensor data by the sensor correction device 10 will be described with reference to FIG. FIG. 5 is a flowchart showing the operation during the correction process of the sensor correction apparatus according to the first embodiment of the present invention.

  As shown in FIG. 5, first, the data acquisition unit 31 acquires correction target sensor data output from the correction target sensor 10 (step B1). Further, the data acquisition unit 31 passes the acquired correction target sensor data to the correction processing unit 33.

  When step B1 is executed, the correction processing unit 33 acquires the function obtained by the error learning unit, and corrects the correction target sensor data using the acquired function (step B2). Specifically, the correction processing unit 33 calculates the error by substituting the value of the correction target sensor data before correction into the linear regression line. The correction processing unit 33 calculates the correction target data after correction by subtracting the error from the value of the correction target sensor data before correction.

  Thereafter, the correction processing unit 33 transmits the correction target data after correction to the farming support system 200 (step B3).

  The program in the first embodiment may be a program that causes a computer to execute steps A1 to A5 shown in FIG. 4 and steps B1 to B3 shown in FIG. By installing and executing this program on a computer, the sensor correction apparatus 10 and the sensor correction method according to the first embodiment can be realized. In this case, a CPU (Central Processing Unit) of the computer functions as a data acquisition unit 31, an error learning unit 32, and a correction processing unit 33 to perform processing.

  In addition, the program in the first embodiment may be executed by a computer system constructed by a plurality of computers. In this case, for example, each computer may function as any of the data acquisition unit 31, the error learning unit 32, and the correction processing unit 33, respectively.

  As described above, according to the first embodiment, since correction is performed based on the sensor data of the reference sensor 20, the correction target sensor 10 is not expensive and the sensor cost is suppressed even if the accuracy is low. It becomes possible to improve the accuracy of the sensor. In particular, the accuracy of the humidity sensor is likely to decrease due to deterioration, but according to the present embodiment, correction according to the decrease in accuracy is possible, so the sensor correction device 10 is useful for the humidity sensor.

  Moreover, once Steps A1 to A5 are executed, even if the reference sensor 20 is removed, the correction target sensor 10 can be corrected. Therefore, the reference sensor 20 is installed in the same or different field. It can be used for error learning of another sensor. That is, in the example of FIGS. 1 to 5, the correction target sensor 10 is singular, but the present embodiment is not limited to this example. The present embodiment can also be applied when a plurality of correction target sensors 10 are installed.

(Embodiment 2)
Subsequently, a sensor system, a sensor correction device, a sensor correction method, and a program according to Embodiment 2 of the present invention will be described with reference to FIGS. FIG. 6 is a block diagram showing the configuration of the sensor system and the sensor correction apparatus according to Embodiment 2 of the present invention.

  As shown in FIG. 6, the sensor device 40 in the second embodiment is similar to the sensor device 30 in the first embodiment shown in FIG. 1, and includes a data acquisition unit 41, an error learning unit 42, and a correction processing unit 43. These function in the same manner as in the first embodiment. However, in the second embodiment, the sensor device 40 includes an abnormality determination unit 44 and a deterioration determination unit 45 in addition to these. The sensor device 40 according to the second embodiment is different from the sensor device 30 shown in FIG. 1 in this respect. Hereinafter, the difference will be mainly described.

  The abnormality determination unit 44 determines that an abnormality has occurred in the correction target sensor when the value of the correction target sensor data satisfies the setting condition. Further, when determining that an abnormality has occurred, the abnormality determination unit 44 notifies the farming support system 200 of the occurrence of the abnormality.

  The conditions for abnormality determination are set according to the type of sensor data to be corrected, the installation environment, the type of agricultural products growing in the field, and the like. For example, if the correction target sensor data is a temperature sensor, the conditions include “temperature measured in 10 minutes is 40 ° C. or higher”, “temperature is equivalent throughout the day”, and the like.

  Moreover, the condition for abnormality determination may be set by a value calculated using a statistical method. For example, first, spike-like mutation points are detected from sensor data to be corrected, then a histogram of mutation points is created based on past data, and mutation points are generated periodically based on the histogram. If it can be determined that it is not, this mutation point may be used as a criterion for determination.

  Further, it is possible to detect a portion where the same value continues from the correction target sensor data, and then determine that a portion where the same value continues based on past data does not occur periodically. In this case, the same value and the continuous time may be used as a criterion for determination.

  Furthermore, a sudden change in the correction target sensor data may be detected, and the detected change point may be used as a criterion for abnormality determination. In addition, a value obtained by a standard deviation test, a value obtained by a Smirnov-Grubbs test, or a value obtained by a Thompson test may be used as a criterion for abnormality determination.

  Furthermore, the condition for abnormality determination may be set by a value obtained by machine learning. Specifically, a value obtained using LOF (Local Outlier Factor), MSD (Modified Stahel-Donoho), One-Class SVM, and logistic regression may be used as a criterion for abnormality determination.

  Further, the deterioration determination unit 45 determines whether or not the correction target sensor is deteriorated based on the frequency at which the value of the correction target sensor data exceeds a threshold that varies with time. Furthermore, when determining that the deterioration has occurred, the deterioration determination unit 45 notifies the farming support system 200 of the occurrence of deterioration. A specific example will be described with reference to FIG. FIG. 7 is a diagram showing deterioration determination in Embodiment 2 of the present invention.

FIG. 7 shows a case where the inspection target sensor 10 is a temperature sensor. In the example of FIG. 7, a threshold value (abnormal value) for abnormality determination and a lower threshold value (quasi-abnormal value) are set in time series with respect to the measurement value of the temperature sensor. In this case, the deterioration determination unit 45 monitors the measurement value and measures the frequency at which the measurement value exceeds the quasi-abnormal value. Then, the deterioration determination unit 45 can determine that the correction target sensor data has deteriorated when the frequency per unit time increases beyond the set range.

  Also in the second embodiment, the sensor correction method is implemented by operating the sensor correction device 40. Further, also in the second embodiment, steps A1 to A5 shown in FIG. 4 and steps B1 to B3 shown in FIG. 5 are executed as in the first embodiment. However, in the second embodiment, an abnormality determination step by the abnormality determination unit 44 and a deterioration determination step by the deterioration determination unit 45 are further executed. The abnormality determination step and the deterioration determination step are executed separately from steps A1 to A5 shown in FIG. 4 and steps B1 to B3 shown in FIG.

  Moreover, the program in this Embodiment 2 should just be a program which performs step A1-A5 shown in FIG. 4, step B1-B3 shown in FIG. 5, an abnormality determination step, and a deterioration determination step. By installing and executing this program on a computer, the sensor correction device 40 and the sensor correction method according to the second embodiment can be realized. In this case, a CPU (Central Processing Unit) of the computer functions as a data acquisition unit 41, an error learning unit 42, a correction processing unit 43, an abnormality determination unit 44, and a deterioration determination unit 45 to perform processing.

  Further, the program according to the second embodiment may also be executed by a computer system constructed by a plurality of computers. In this case, for example, each computer may function as any of the data acquisition unit 41, the error learning unit 42, the correction processing unit 43, the abnormality determination unit 44, and the deterioration determination unit 45, respectively.

(Physical configuration)
Here, a computer that realizes the sensor correction apparatus by executing the program in the first and second embodiments will be described with reference to FIG. FIG. 8 is a block diagram illustrating an example of a computer that implements the sensor correction apparatus according to Embodiments 1 and 2 of the present invention.

  As shown in FIG. 8, 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 that data communication is possible.

  The CPU 111 performs various calculations by developing the program (code) in the present embodiment stored in the storage device 113 in the main memory 112 and executing them in a predetermined order. 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 a computer-readable recording medium 120. Note that the program in the present embodiment may be distributed on the Internet connected via the communication interface 117.

  Specific examples of the storage device 113 include a hard disk drive and a semiconductor storage device such as a flash memory. 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 display on the display device 119.

  The data reader / writer 116 mediates data transmission between the CPU 111 and the recording medium 120, and reads a program from the recording medium 120 and writes a 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 storage media such as a flexible disk, or CD- An optical storage medium such as ROM (Compact Disk Read Only Memory) can be used.

  As described above, according to the present invention, it is possible to improve the accuracy of the sensor while suppressing an increase in cost of the sensor. The present invention is useful in all fields where sensors are required.

10 Correction target sensor (first sensor)
20 Reference sensor (second sensor)
30 Sensor correction device (Embodiment 1)
31 Data Acquisition Unit 32 Error Learning Unit 33 Correction Processing Unit 40 Sensor Correction Device (Embodiment 2)
41 Data Acquisition Unit 42 Error Learning Unit 43 Correction Processing Unit 44 Abnormality Determination Unit 45 Degradation Determination Unit 100 Sensor System 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 Device 119 Display Device 120 Recording Medium 121 Bus 200 Farming System

Claims (20)

A first sensor to be corrected, a second sensor to be a correction reference, and a sensor correction device,
The sensor correction device includes:
A data acquisition unit that acquires first sensor data output from the first sensor and second sensor data output from the second sensor;
An error learning unit that compares the first sensor data with the second sensor data and learns an error in the first sensor;
A correction processing unit that corrects the first sensor data based on a result of learning by the error learning unit.
A sensor system characterized by that. In order to learn the error, the error learning unit first calculates a difference between the values of the first sensor data and the second sensor data whose output times overlap each other. Then, statistical processing is performed on the set of errors for each of the obtained sets to obtain a function that defines the relationship between the value of the first sensor data and the error,
The correction processing unit corrects the first sensor data based on the function;
The sensor system according to claim 1. When the data acquisition unit includes the first sensor data and the second sensor data whose output timing does not overlap with the other sensor data, the first sensor data and the second sensor data Either or both, based on the time-series change, interpolate the sensor data with the output time overlap with the other sensor data,
The sensor system according to claim 1. The sensor correction device further includes an abnormality determination unit that determines that an abnormality has occurred in the first sensor when a value of the first sensor data satisfies a setting condition.
The sensor system according to claim 1. A deterioration determination unit that determines whether or not the first sensor has deteriorated based on a frequency at which the value of the first sensor data exceeds a threshold that varies with time; In addition,
The sensor system according to claim 1. A data acquisition unit that acquires first sensor data output from the first sensor to be corrected and second sensor data output from the second sensor that is a reference for correction;
An error learning unit that compares the first sensor data with the second sensor data and learns an error in the first sensor;
A correction processing unit that corrects the first sensor data based on a result of learning by the error learning unit.
The sensor correction apparatus characterized by the above-mentioned. In order to learn the error, the error learning unit first calculates a difference between the values of the first sensor data and the second sensor data whose output times overlap each other. Then, statistical processing is performed on the set of errors for each of the obtained sets to obtain a function that defines the relationship between the value of the first sensor data and the error,
The correction processing unit corrects the first sensor data based on the function;
The sensor correction apparatus according to claim 6. When the data acquisition unit includes the first sensor data and the second sensor data whose output timing does not overlap with the other sensor data, the first sensor data and the second sensor data Either or both, based on the time-series change, interpolate the sensor data with the output time overlap with the other sensor data,
The sensor correction apparatus according to claim 6. The sensor correction device further includes an abnormality determination unit that determines that an abnormality has occurred in the first sensor when a value of the first sensor data satisfies a setting condition.
The sensor correction apparatus in any one of Claims 6-8. A deterioration determination unit that determines whether or not the first sensor has deteriorated based on a frequency at which the value of the first sensor data exceeds a threshold that varies with time; In addition,
The sensor correction apparatus in any one of Claims 6-9. (A) acquiring first sensor data output from a first sensor to be corrected and second sensor data output from a second sensor serving as a correction reference;
(B) comparing the first sensor data and the second sensor data to learn an error in the first sensor;
(C) correcting the first sensor data based on the result of learning in the step (b),
The sensor correction method characterized by the above-mentioned. In the step (b), in order to learn the error, first, for each set of the first sensor data and the second sensor data whose output times overlap each other, a difference between the two values is calculated. Obtained as the error, and then performing a statistical process on the obtained set of errors for each of the obtained sets to obtain a function that defines the relationship between the value of the first sensor data and the error,
In the step (c), the first sensor data is corrected based on the function.
The sensor correction method according to claim 11. In the step (a), when there is the first sensor data and the second sensor data whose output timing does not overlap with the other sensor data, the first sensor data and the second sensor Interpolate sensor data whose output timing overlaps with the other sensor data based on the time-series change in either or both of the data,
The sensor correction method according to claim 11. (D) further including a step of determining that an abnormality has occurred in the first sensor when a value of the first sensor data satisfies a setting condition;
The sensor correction method according to claim 11. (E) further comprising a step of determining whether or not the first sensor has deteriorated based on a frequency at which a value of the first sensor data exceeds a threshold that varies according to time.
The sensor correction method according to claim 11. On the computer,
(A) acquiring first sensor data output from a first sensor to be corrected and second sensor data output from a second sensor serving as a correction reference;
(B) comparing the first sensor data and the second sensor data to learn an error in the first sensor;
(C) correcting the first sensor data based on the learning result in the step (b);
A program that executes In the step (b), in order to learn the error, first, for each set of the first sensor data and the second sensor data whose output times overlap each other, a difference between the two values is calculated. Obtained as the error, and then performing a statistical process on the obtained set of errors for each of the obtained sets to obtain a function that defines the relationship between the value of the first sensor data and the error,
In the step (c), the first sensor data is corrected based on the function.
The program according to claim 16. In the step (a), when there is the first sensor data and the second sensor data whose output timing does not overlap with the other sensor data, the first sensor data and the second sensor Interpolate sensor data whose output timing overlaps with the other sensor data based on the time-series change in either or both of the data,
The program according to claim 16. In the computer,
(D) When the value of the first sensor data satisfies a setting condition, the step of determining that an abnormality has occurred in the first sensor is further executed.
The program according to any one of claims 16 to 18. In the computer,
(E) further executing a step of determining whether or not the first sensor has deteriorated based on a frequency at which a value of the first sensor data exceeds a threshold that varies according to time. Item 20. The program according to any one of Items 16 to 19.

Images