JP7637116B2 - Measuring device, sensor output monitoring method, and sensor output monitoring program - Google Patents

Description

The present invention relates to a measuring device for measuring the properties of a sample, such as water quality.

For example, in a measuring device such as that described in Patent Document 1, which is equipped with a sensor for measuring water quality, conventionally, if an obvious abnormality occurred, such as the sensor's output value going out of range, the user could determine that maintenance was required for the sensor.

However, even if there is no obvious abnormality in the sensor as described above, the measurement value output by the sensor may deviate from the value that should be output due to sensor dirt or a malfunction of a part that affects the sensor output value.

In such cases, it is difficult for the user to determine whether the sensor output value has changed due to dirt or malfunction of the sensor or other components.

JP 2015-25794 A

Regarding the problems described above, the inventor realized that it is possible to determine whether each sensor is outputting the measurement value that it should be outputting by monitoring changes in the relationship between the measurement values of other correlated sensors.
Therefore, the main object of the present invention is to provide a measuring device that can detect when the measurement value output by a sensor deviates from the value that should be output, even when no obvious abnormality has occurred in the sensor, and further that can identify which sensor among multiple sensors has an output value that deviates from the original value.

In other words, the measuring device of the present invention is a measuring device comprising three or more sensors and an output monitoring unit that monitors the output values of these sensors, wherein the output monitoring unit calculates an estimated output value of the first sensor based on the output values of each sensor of a first sensor group including at least one sensor other than the first sensor among the sensors, calculates an estimated output value of the first sensor based on the output values of each sensor of a second sensor group different from the first sensor group, the second sensor group including at least one sensor other than the first sensor, and compares each of the calculated estimated output values with an actual measurement value actually measured by the first sensor, and determines whether the difference or ratio between each of the estimated output values and the actual measurement value is within a predetermined tolerance range.
Here, the first sensor group and the second sensor group may include only one sensor, or may include a plurality of sensors.

With the measuring device configured in this manner, it is possible to detect whether the correlation between the output values of each sensor has deviated from the previous correlation by determining whether the difference or ratio between each estimated output value and the actual measured value is within a predetermined tolerance range.
If the correlation maintains the previous correlation, it can be determined that there is no problem with the sensors or components that affect the sensor output values, and that the output values of each sensor are the values that should be output.

On the other hand, if the correlation deviates from the previous correlation, it can be determined that the output value of each sensor deviates from the value that should be output due to a malfunction of the sensor or part, or that the correlation between the output values of each sensor deviates from the previous correlation due to environmental changes such as sample processing conditions, measurement conditions, weather conditions, etc. If the correlation deviates from the previous correlation even though there is no environmental change as described above, it can be determined that there is a problem with the sensor or a part that affects the sensor output value, and that the output value of each sensor may deviate from the value that should be output.
Furthermore, by providing three or more sensors and examining the correlations between them, it is possible to identify which of the three or more sensors has an output value that deviates from the value that should originally be output.

If the system further includes a memory unit that stores a correlation equation calculated based on past output values of each sensor, and the output monitoring unit calculates an estimated output value using the correlation equation, a more accurate estimated output value can be calculated.
The correlation equation may be calculated, for example, by regression analysis.

Specific embodiments of the present invention include a sensor that measures one or more of a number of indicators, such as the sample's conductivity, resistivity, ORP, turbidity, color, temperature, pressure, oil film, and the concentrations of various substances contained in the sample.

For example, the device may be provided with multiple sensors that measure the same indicator among the multiple indicators mentioned above, or multiple types of sensors that measure different types of indicators among the multiple indicators mentioned above.

As mentioned above, if the system is equipped with sensors that measure different types of indicators, it is possible not only to identify the sensors that need maintenance, but also to more specifically narrow down the parts that are likely to be related to changes in the output value of the indicator.

A specific embodiment of the present invention is one in which the measuring device is used for water quality analysis.

Another aspect of the present invention is a sensor output monitoring method or a sensor output monitoring program, which comprises: calculating an estimated output value of a first sensor based on the output value of each sensor of a first sensor group including at least one sensor other than the first sensor among three or more sensors; calculating an estimated output value of the first sensor based on the output value of each sensor of a second sensor group including at least one sensor other than the first sensor and different from the first sensor group; comparing each of the calculated estimated output values with an actual measurement value actually measured by the first sensor; and determining whether or not a difference or ratio between each of the estimated output values and the actual measurement value is within a predetermined tolerance range.

According to the present invention, it is possible to detect whether the correlation between the output values of each sensor has deviated from the previous correlation. As a result, it is possible to easily determine whether the output value of each sensor has changed due to deterioration of the sensor or a part that affects the output value of the sensor.

In this way, if there is a possibility that the output value of each sensor is caused by deterioration of the sensor or a part that affects the sensor output value, the user can be advised to perform maintenance on a certain sensor and parts related to that sensor even if the next maintenance is scheduled for a long time away.
In addition, since it is possible to identify which of the three or more sensors has an output value that deviates from the value that should be output, it is possible to inform the user of which sensor needs maintenance.

1 is a schematic diagram showing the configuration of a measurement device according to an embodiment of the present invention. FIG. 2 is a functional block diagram showing functions of the information processing apparatus according to the embodiment. FIG. 4 is a schematic diagram showing an example of sensor output estimation performed using the measurement device of the embodiment. 6 is a graph showing the results of sensor output estimation performed using the measurement device of the embodiment. 4 is an example of an image displayed on a display unit in the embodiment. FIG. 13 is a schematic diagram showing an example of estimating an output of a sensor according to another embodiment of the present invention. 13 is an example of an image on a display unit according to another embodiment of the present invention. 13 is an example of an image on a display unit according to another embodiment of the present invention.

Reference Signs List 1 Measuring device 2 Sensor unit 3 Information processing device 31 Calculation unit 32 Display unit 33 Storage unit 34 Output monitoring unit

Below, one embodiment of the measuring device according to the present invention is described with reference to the drawings.

The measuring device 1 according to this embodiment measures indicators of water quality, such as the concentration of specific components, such as hydrogen ions, NH4-N, residual chlorine, dissolved oxygen, salt, fluoride ions, nitrate ions, HF, KOH, TMAH, dissolved ozone, phosphoric acid, citric acid, nitrogen, phosphorus, COD or silica, contained in a liquid sample (sample) such as clean water or sewage, for example in a water treatment facility, etc., and the sample's conductivity, resistivity, oxidation-reduction potential (ORP), turbidity, chromaticity, temperature, pressure, etc. As shown in FIG. 1, the measuring device 1 is equipped with a sensor unit 2 having a plurality of sensors for measuring the various indicators as described above, and an information processing device 3 for processing information output from the sensor unit 2.

The sensor unit 2 measures the various indices of a sample as described above.
Specifically, the sensor unit 2 includes three or more sensors of different types in order to measure a plurality of types of indices among the various indices described above.

In this embodiment, the sensor unit 2 is placed, for example, in each of different aeration tanks of a water treatment facility to measure a number of different indicators of the sample treated water. This can be modified as appropriate depending on the purpose of the measurement, but specifically, it is equipped with various sensors 21 such as an oxidation-reduction potential sensor 21a, a pH sensor 21b, a DO sensor 21c, a sludge turbidity sensor (MLSS measuring device 21d), an ammonia concentration sensor 21e, and a temperature sensor 21f.

The various output values output from the sensor unit 2 configured in this manner are processed by the information processing device 3.

The information processing device 3 is a dedicated or general-purpose computer having a CPU, memory, AD converter, etc., and is configured to operate according to a program stored in a specified area of the memory, and thereby perform the functions of a calculation unit 31 that calculates measurement values such as various component concentrations based on the output values of each sensor 21 for at least each indicator, as shown in Figure 2, a display unit 32 that displays the output values of each sensor, and a memory unit 33 that stores the output values of each sensor and the measurement values calculated by the calculation unit 31, etc.
The display unit 32 displays, for example, a graph on a screen in which one axis represents time and the other axis represents output value, and displays on this graph the change over time in the output value from each sensor 21. This display unit 32 may be a display provided in the information processing device, or may be a separately provided display, etc.

Therefore, the information processing device 3 of this embodiment is configured to further function as an output monitoring unit 34 that monitors output values output from three or more sensors, as shown in Figure 2, by operating according to a program stored in a specified area of the memory.

The output monitoring section 34 monitors the correlation between the output values from the above-mentioned sensors 21 .
The procedure and method by which the output monitoring unit 34 monitors the correlation between the output values from the sensors 21 will be described below.

The output monitoring unit 34 monitors the correlation between output values from three or more sensors 21 as described above that are correlated with one another.
For ease of understanding, an example will be described in which the correlation between three sensors 21 as shown in FIG. 3 is monitored.

First, the output monitoring unit 34 estimates the current output value of one of these three sensors 21 (corresponding to the first sensor, which will be referred to as sensor A) using a correlation calculated using past output values of the other sensor (which will be referred to as sensor B, which corresponds to the first sensor group. In this embodiment, the first sensor group includes only one sensor, as shown in Figure 3), and the current output value of the other sensor B. Hereinafter, the output value estimated in this way will be referred to as the first estimated output value. Note that in this embodiment, regression analysis is used to estimate the output value.

The method of calculating the first estimated output value will be described below with reference to an example.
Based on the output values of past measurements, a regression equation such as that shown in the following formula 1 is obtained, and stored in the memory unit 33. Then, for example, the current output value of sensor B is substituted into this regression equation to obtain a first estimated output value of sensor A. Note that constants such as a and b in formula 1 can be obtained by, for example, calculating a straight line that minimizes the difference from each output value by the least squares method. The past output value may be an output value in a certain past period, or may be an output value in the entire past period. In this embodiment, as an example, output values in all periods after the most recent change in the measurement environment are used to obtain the regression equation. In addition, this regression equation may be calculated and input manually by a person, or may be calculated automatically by, for example, the calculation unit 31 or the output monitoring unit 34.

(Equation 1)
(Estimated output value of sensor A) = a × (output value from sensor B) + b

Furthermore, the output value of sensor A at the same time point as above is similarly estimated using past output values of a sensor (hereinafter referred to as sensor C; sensor C corresponds to the second sensor group; in this embodiment, the second sensor group includes only one sensor, as shown in FIG. 3) other than sensors A and B. The output value estimated in this way is hereinafter referred to as the second estimated output value.
Next, the first estimated output value and the second estimated output value calculated as described above are compared with the actual measured value, which is the output value actually output from sensor A at the same point in time, and, for example, the difference between them is calculated.
As shown in FIG. 3, when sensor B or sensor C is used as the first sensor, the estimated output value of sensor B or sensor C is calculated, and the difference between this estimated output value and the actual measured value of sensor B or sensor C is calculated.

It is checked whether the difference between the first estimated output value and the actual measurement value or the difference between the second estimated output value and the actual measurement value is within a predetermined allowable range, for example.
The tolerance may be preset, for example, as a range of ±10% of the estimated output value.

Until data on the output values of each sensor is accumulated in the memory unit 33, the tolerance range may be set uniformly to a range of ±10% of the estimated output value, and after the data has been accumulated, the width of the tolerance range may be re-set to an appropriate range depending on the index measured by each sensor, for example, by estimating the range of values that will be output with a 50% probability using Bayesian estimation, etc.
If the difference is within the allowable range, the output monitoring unit 34 determines that the correlation between the sensors is maintained.
On the other hand, if the difference described above exceeds the allowable range, for example, as in the circled portion in Fig. 4, it is determined that the correlation between the sensors has deviated from the previous correlation. Note that the allowable range in Fig. 4 is indicated by shading.

Next, it is determined whether this correlation occurs in conjunction with environmental changes, such as sample processing conditions, measurement conditions, or weather conditions, and if the correlation changes regardless of environmental changes, it is determined that there is a problem with the sensor or a component that affects the sensor output value.
Regarding whether or not there has been an environmental change, for example, information on the environmental change may be stored in the storage unit 33. This information on the environmental change may be input by the user, or, for example, when the user resets the measurement conditions, the input date and time, changes, and the like may be automatically stored in the storage unit 33.

If it is determined that the correlation between the sensors has deviated from the previous correlation, the output monitoring unit 34 continues to determine which sensor 21 is causing the correlation to deviate.
For example, if the difference between the first estimated output value and the actual measured value exceeds the allowable range, while the difference between the second estimated output value and the actual measured value is within the allowable range, it is determined that the first estimated output value could not be estimated successfully because the output value from sensor B has deviated from the value that should have been output.
In this way, after determining which sensor 21's output value is the cause, the output monitoring unit outputs, for example, a signal to the effect that the value of sensor B is deviating.

When the display unit 32 receives a signal from the output monitoring unit 34 indicating that the correlation between the output values of the sensors 21 has deviated from the previous correlation due to the value of sensor B, the display unit 32 is configured to display, for example, an alert on the graph indicating that it is necessary to check for abnormalities in sensor B or in parts, etc. that may affect the output value of sensor B.
For example, as shown in FIG. 5, when the difference between the actual measured value and the estimated output value of each sensor exceeds the allowable range, a plot (circle) indicating an abnormality may be displayed on the graph.
In addition, by clicking on this graph, a new window may be displayed in which one or more pieces of information such as the measurement date and time, the actual measured value, the estimated output value, and the output value of the sensor used to calculate the estimated output value are displayed.

With the measuring device 1 configured in this manner, the correlation between the output values from each sensor 21 is monitored, making it possible to easily determine whether a change in the output value from each sensor 21 is due to a malfunction of each sensor 21 or a component that affects the output value of each sensor 21.
As a result, for example, even if the next maintenance is due in the future, a malfunction of each sensor 21 or a part that affects the output value of each sensor 21 can be detected early.

In particular, when performing water quality analysis as in this embodiment, it is expected that the output values output from each sensor 21 will differ significantly depending on the measurement sample and the measurement environment. On the other hand, the correlation between the output values from each sensor 21 is often unique to the measurement sample and the measurement environment. Therefore, by determining whether the output value from each sensor 21 is correct based on the correlation between the output values from each sensor 21, it is possible to more accurately detect malfunctions of each sensor 21 or parts that affect the output value of each sensor 21.

In addition, it is possible to identify which of the three or more sensors 21 has changed in output value, causing the correlation to be broken, so that the user can be more specifically informed of which sensor 21 or part needs maintenance, thereby reducing the effort required for maintenance.

The present invention is not limited to the above-described embodiment.
For example, in the above embodiment, the estimated output value of the first sensor is calculated from the output value of one other sensor (sensor B or sensor C), but for example, as shown in Fig. 6, the estimated output value of the first sensor may be calculated based on the output values of each sensor of a sensor group including two or more other sensors. When calculating the estimated output value based on the output values of each sensor of a sensor group including two or more sensors in this way, a formula for multiple regression analysis is used. A method for calculating the estimated output value in this case will be described below.

For example, when calculating an estimated output value of sensor A based on the output values of sensors B and C, the estimated output value of sensor A can be calculated using a multiple regression analysis formula such as the following equation 2. Note that constants c, d, e, etc. in equation 2 can be found, for example, by calculating a straight line using the least squares method that minimizes the difference from each output value.

(Equation 2)
(Estimated output value of sensor A) = c × (output value of sensor B) + d × (output value of sensor C) + e

In the above embodiment, it was determined whether the difference between the estimated output value and the actual measured value of the first sensor was within a predetermined allowable range, but it may also be possible to calculate the ratio between the estimated output value and the actual measured value and determine whether the ratio is within the allowable range.
For example, when the estimated output value of sensor A is calculated based on the output values of sensors B and C as described above, it may be checked whether the ratio of the actual measurement value of sensor A to the estimated output value of sensor A is within the allowable range. More specifically, when the allowable range is set to ±10%, for example, it may be determined whether the quotient obtained by dividing the estimated output value of sensor A by the actual measurement value of sensor A or the quotient obtained by dividing the actual measurement value of sensor A by the estimated output value of sensor A is within the range of 0.9 to 1.1.

For example, in a water treatment facility or the like where a single sample undergoes a series of treatment processes, the indicators of the sample at each treatment process may be measured sequentially over time using a sensor placed at a fixed location, or if each treatment process is carried out at a separate location, a sensor may be installed at each treatment process location to measure the indicators of the flowing sample at separate locations. Furthermore, the same sample may be measured at the same time in the same location.

The number of sensors must be three or more, and as mentioned above, it is possible to have three or more types of sensors, each of which measures a different indicator, or it is possible to have three or more sensors of the same type measuring the same indicator.
As described above, when three sensors are used, it is possible to identify which sensor's output value is deviating from the original value by calculating two estimated output values. In this way, when three or more sensors (n) are used, it is considered possible to more reliably identify which sensor's output value is deviating from the original value by calculating (n-1) estimated output values.

The sensor unit and information processing device do not need to be arranged in one location as a single device; multiple sensor units may be arranged in spatially separated locations, and the output values of multiple sensors provided in these multiple sensor units may be sent to the information processing device via wired or wireless communication.

Sensors that are correlated include not only those that measure multiple indicators as described above and that are known to be correlated in advance, but also those whose correlation has become empirically evident through continued actual measurements. For example, the correlation between the output values of each sensor may differ depending on the measurement environment, such as temperature or pressure.

The measuring device of the present invention is not limited to measuring liquid samples such as those used for water quality analysis as described above, but can also be used to measure gas samples such as NOx, SO2, CO, CO2, O2, and silica.

In the above embodiment, the output monitoring unit estimated the output value based on the output value of the sensor, but the measurement value may be estimated based on the measurement value calculated by the calculation unit. Similarly, the display unit may display an output value, such as the potential output from the sensor, instead of the measurement value calculated by the calculation unit.

In the above embodiment, the estimated output value is calculated by using past output values and regression analysis, but the output monitoring unit may also calculate the estimated output value by machine learning such as a neural network.
In addition, in the above embodiment, the sensor output value is estimated at the current time. However, it is also possible to calculate the output value of the sensor at, for example, a certain point in the past as an estimated output value, and compare the actual measured value of the sensor at that time with the estimated output value.

Some of the functions of the information processing device described in the above embodiment may be provided in a computer or the like other than the information processing device. As an example, a function as a data storage unit that stores data such as output values output from the sensor unit may be provided in, for example, a cloud server or the like other than the information processing device.

In the above-described embodiment, the sensor unit includes multiple sensors, but the sensor unit may include only one sensor. Also, for example, a single water treatment facility may be treated as a single sensor unit, and output values from multiple water treatment facilities may be processed by an information processing device on a cloud server.

The display may be provided for each sensor, each sensor unit, each measurement device, or each measurement facility.

The correlated sensors change depending on the measurement conditions, the measurement environment, the sample processing method, etc. Therefore, for example, the display unit may display a screen as shown in Figure 7, which allows the user to select a combination of correlated sensors.

By utilizing the regression equation stored in the memory unit, it is also possible to provide a simulation function that, for example, determines whether total phosphorus (TP) or total nitrogen (TN) should be changed in order to bring the chemical oxygen demand (COD) closer to the target value in the measurement environment.

In order to accurately determine the correlation between each sensor using the aforementioned regression analysis or machine learning, an input screen such as that shown in Figure 8 may be displayed on the display unit so that past data can be edited, such as by deleting measurement values that have caused the correlation to deviate from the previous correlation, or by inputting measurement values obtained by other methods such as manual analysis in place of the deleted measurement values.

In the above embodiment, whether or not there has been an environmental change is determined by user input, but for example, in the case of an apparatus that measures a specific component contained in a sample after subjecting the sample to a series of different processes, a threshold setting unit can be provided that acquires the output values over multiple stored past measurements and sets a threshold value for one or more of the analytical instruments for a new measurement based on these output values. For example, if the threshold value is set so that the output value of the analytical instrument in the event of a sudden abnormality exceeds the threshold value, the abnormality can be detected, and it can be determined that there is no malfunction in the sensor or parts unless this is detected, and that there has been an environmental change.

Needless to say, the present invention is not limited to the above-described embodiments, and various modifications are possible without departing from the spirit of the present invention.

Even if there is no obvious abnormality in the sensor, it is possible to detect that the measurement value output by the sensor deviates from the value that should be output, and further to identify which of multiple sensors the output value of which deviates from the correct value.

Claims (10)

A measurement device comprising three or more sensors and an output monitoring unit that monitors output values of the sensors,
the output monitoring unit calculates an estimated output value of the first sensor based on output values of each sensor of a first sensor group including at least one sensor other than the first sensor among the sensors;
calculating an estimated output value of the first sensor based on output values of each sensor of a second sensor group different from the first sensor group, the second sensor group including at least one sensor other than the first sensor;
comparing each of the estimated output values calculated for the first sensor with an actual measurement value actually measured by the first sensor, and determining whether or not a difference or ratio between each of the estimated output values for the first sensor and the actual measurement value is within a predetermined tolerance ;
calculating an estimated output value of a second sensor that is included in the first sensor group based on the output value of the first sensor;
calculating an estimated output value of the second sensor based on the output value of each sensor of the second sensor group;
A measuring device characterized in that it compares each of a plurality of estimated output values calculated for a second sensor with actual measured values actually measured by the second sensor, and determines whether or not a difference or ratio between each of the estimated output values and the actual measured values for the second sensor is within a predetermined tolerance range . 2. The measuring device according to claim 1, wherein the first sensor and the second sensor are different types of sensors. 3. The measuring device according to claim 1 or 2, which determines whether the current correlation between the output values of each sensor has deviated from the previous correlation by determining whether the difference or ratio between each estimated output value and the actual measured value is within a predetermined tolerance range. 4. The measuring device according to claim 3, which, when the current correlation between the output values of each sensor deviates from the previous correlation, determines whether or not this is due to a malfunction of the sensor or a part causing the output value of each sensor to deviate from the value that should be output. 5. The measuring device according to claim 1, wherein the sensor measures one or more of the conductivity, resistivity, ORP, turbidity , color, temperature, pressure, oil film, and concentration of various substances contained in the sample. The measuring device according to any one of claims 1 to 5, characterized in that the sensor measures one or more of the conductivity, resistivity, ORP, turbidity, color, temperature, pressure, oil film, and concentration of various substances contained in the sample, and the measuring device is provided with a plurality of different types of sensors. 7. The measuring apparatus according to claim 1, wherein the estimated output value is calculated by regression analysis. The measuring device according to any one of claims 1 to 7 , which is used for water quality analysis. calculating an estimated output value of the first sensor based on output values from a first sensor group including at least one sensor other than the first sensor among the three or more sensors;
calculating an estimated output value of the first sensor based on an output value from a second sensor group that is different from the first sensor group and includes at least one sensor other than the first sensor;
comparing each of the estimated output values calculated for the first sensor with an actual measurement value actually measured by the first sensor, and determining whether or not a difference or ratio between each of the estimated output values for the first sensor and the actual measurement value is within a predetermined tolerance ;
calculating an estimated output value of a second sensor that is included in the first sensor group based on the output value of the first sensor;
calculating an estimated output value of the second sensor based on the output value of each sensor of the second sensor group;
A sensor output monitoring method comprising: comparing each of a plurality of estimated output values calculated for a second sensor with an actual measurement value actually measured by the second sensor; and determining whether a difference or ratio between each of the estimated output values and the actual measurement value for the second sensor is within a predetermined tolerance range . calculating an estimated output value of the first sensor based on output values of each sensor of a first sensor group including at least one sensor other than the first sensor among the three or more sensors;
calculating an estimated output value of the first sensor based on output values of each sensor of a second sensor group different from the first sensor group, the second sensor group including at least one sensor other than the first sensor;
comparing each of the estimated output values calculated for the first sensor with an actual measurement value actually measured by the first sensor, and determining whether or not a difference or ratio between each of the estimated output values for the first sensor and the actual measurement value is within a predetermined tolerance ;
calculating an estimated output value of a second sensor that is included in the first sensor group based on the output value of the first sensor;
calculating an estimated output value of the second sensor based on the output value of each sensor of the second sensor group;
A sensor output monitoring program characterized by comparing each of a plurality of estimated output values calculated for a second sensor with an actual measurement value actually measured by the second sensor, and determining whether or not a difference or ratio between each of the estimated output values and the actual measurement value for the second sensor is within a predetermined tolerance range .

Images