JP7345576B2 - Water level estimation system, water level estimation method, and information processing device - Google Patents

Description

The present invention relates to a water level estimation system, a water level estimation method, and an information processing device.

Conventionally, water level measurement methods using satellite positioning have been proposed.

Japanese Patent Application Publication No. 2008-58052

However, with the above-mentioned conventional techniques, it is not always possible to measure the water level with high accuracy.

For example, in the above-mentioned conventional technology, the water level is measured from altitude data determined by a float provided on the water surface based on satellite signals. However, in satellite positioning, it is known that the tropospheric effect on satellite signals causes errors, and the altitude data measured by the float may not be accurate.

In addition, it is thought that the float is constantly vibrating due to the influence of waves occurring on the water surface and other surrounding environments (e.g. wind), and this vibration is a cause of errors in satellite positioning. It is also possible.

From the above, the above-mentioned conventional technology has room for improvement in realizing highly accurate water level measurement.

The present application has been made in view of the above, and aims to provide a water level estimation system, a water level estimation method, and an information processing device that can measure water levels with high accuracy.

The water level measurement system according to the present application is a water level estimation system that includes a first receiver installed on the water surface to be measured, a second receiver installed at a predetermined location, and an information processing device. The information processing device includes first altitude data that is altitude data obtained from a satellite signal received by the first receiver, and altitude data obtained from a satellite signal received by the second receiver. an acquisition unit that acquires second altitude data, a calculation unit that calculates a correction coefficient based on predetermined information according to the situation in which the receiver is installed; and an estimation unit that estimates the water level of the water level measurement target based on corrected data obtained by correcting the first altitude data and the second altitude data.

Further, the calculation unit in the water level measurement system according to the present application calculates a correct coefficient determined according to altitude data acquired from a predetermined receiver and characteristic information according to a situation in which the predetermined receiver is installed. The correction coefficient is calculated based on the model that has learned the relationship between and the predetermined information.

According to one aspect of the embodiment, for example, the water level can be measured with high accuracy. According to another aspect of the embodiment, for example, a correction coefficient can be calculated using a learning model using so-called machine learning or supervised learning, and the water level can be measured with high accuracy.

FIG. 1 is an explanatory diagram illustrating the water level estimation logic according to the embodiment. FIG. 2 is a diagram illustrating an example of a system configuration according to the embodiment. FIG. 3 is a diagram illustrating an example of a situation in which water level estimation processing according to the embodiment is performed. FIG. 4 is a diagram illustrating an example of an overall image of the water level estimation process according to the embodiment. FIG. 5 is a diagram illustrating a configuration example of an information processing device according to an embodiment. FIG. 6 is a diagram illustrating an example of a learning data storage unit according to the embodiment. FIG. 7 is a flowchart showing the learning processing procedure according to the embodiment. FIG. 8 is a sequence diagram showing a water level estimation process executed by the water level estimation system according to the embodiment. FIG. 9 is a hardware configuration diagram showing an example of a computer that implements the functions of the information processing device.

EMBODIMENT OF THE INVENTION Below, the water level estimation system, the water level estimation method, and the form (henceforth called an "embodiment") for implementing an information processing apparatus which concern on this application are demonstrated in detail, referring drawings. Note that this embodiment does not limit the water level estimation system, water level estimation method, and information processing apparatus according to the present application. Further, in each of the embodiments below, the same parts are given the same reference numerals, and redundant explanations will be omitted.

[1. Introduction]
For example, in facilities such as dams that require disaster prevention operations, conditions such as water level positioning error within ±1 cm and real-time performance (for example, positioning within 1 minute) are required. However, conventional water level measurement methods using satellite positioning cannot always meet these conditions.

Furthermore, in satellite positioning, it is known that errors occur in positioning results due to various factors. For example, as a method of satellite positioning using signals from GNSS satellites (GNSS signals), RTK (Real Time Kinematic) positioning, which is more accurate than GPS positioning, is attracting attention, but the following various factors may affect the positioning results. It is known that errors can occur.

For example, error factors include ionospheric delay, tropospheric delay, and multipath error. In addition, in RTK positioning, by correcting the deviation by exchanging information between a dedicated receiver (mobile station) that receives GNSS signals and a base station (fixed station), independent positioning like GPS positioning is possible. Although higher accuracy is achieved, errors may occur based on the relative distance between the receiver and the reference station.

In addition, in conventional water level measurement methods, the receiver is installed so that it floats on the water surface, but the receiver vibrates on the water surface due to the influence of waves, wind, and other climatic conditions that occur on the water surface. There are cases where this happens. If the device vibrates in this way, stable positioning results cannot be obtained. Therefore, in water level measurement, when a receiver is installed on the water surface, vibration of the receiver also becomes a factor of error.

From the above, in water level measurement using a receiver, error factors originating from RTK positioning (or GPS positioning) (e.g., ionospheric delay, tropospheric delay, multipath error, relative distance error), error factors originating from the receiver, (For example, vibration of the receiver) may cause the altitude data obtained as a positioning result to deviate from the true altitude data. Furthermore, if an error occurs in the altitude data, the water level estimated based on the altitude data will also deviate from the true water level.

Therefore, in the present invention, we propose a correction logic that corrects to exclude error factors, specifically, error factors originating from RTK positioning and error factors originating from the receiver, in water level measurement using a receiver. do. As a result, in the present invention, it becomes possible to realize water level measurement with high precision, which has not been achieved with conventional water level measurement methods.

In addition, in conventional water level measurement methods, receivers were simply installed above the water surface, but in the present invention, multiple receivers are used, and the installation location of each receiver is varied. Ta. The present invention has focused on the fact that the above error factors can be corrected in real time by using machine learning to calculate correction coefficients when combining positioning results from each receiver. As a result, the present invention can also realize real-time performance. Note that the water level measurement method according to the present invention is not necessarily limited to machine learning, and can be realized using other statistical methods.

[2. Overview of embodiment]
Next, an overview of the water level estimation logic according to the embodiment will be explained using FIG. 1. FIG. 1 is an explanatory diagram illustrating the water level estimation logic according to the embodiment. In water level measurement using receivers, in this embodiment, as shown in FIG. A configuration was adopted in which a receiver was installed (floated) and one receiver was installed on the ground near the water level measurement target.

Regarding this point, FIG. 1 shows receivers D1 and D2 floatingly installed on the water level measurement target, and receiver D3 fixedly installed on the ground nearby.

First, for the purpose of reducing error factors derived from RTK positioning, one receiver D3 is installed on the ground among three devices, receivers D1, D2, and D3. For example, if the distance between receiver D1 (receiver D2) on the water surface and receiver D3 on the ground is relatively short, the fluctuation components due to error factors derived from RTK positioning are considered to be similar in both devices. . Therefore, in this embodiment, taking advantage of the fact that the receiver D3 is stably installed (fixed to the ground), the height of the ground is used as a reference position, and the altitude of the receiver D3 with respect to the reference position is changed. A process is performed to extract the component and correct the altitude of the receiver D1 (receiver D2) using the extracted fluctuation component.

For example, if the altitude of receiver D3 is ``+2m'' with respect to the altitude of the ground, subtracting the fluctuation component ``+2m'' from the altitude of receiver D1 (receiver D2) means that the receiver fixed on the ground This is an example of correction using D3.

In addition, in order to reduce error factors originating from RTK positioning, as described above, it is required that the fluctuation components be similar, so as shown in FIG. It is preferable that they be installed so that the distance relationship is within 500 m. Similarly, it is preferable that the receiver D2 and the receiver D3 are installed so that the distance relationship between them is within 500 m.

Furthermore, if the reference stations 30 used by the receivers D1, D2, and D3 are different, there is a possibility that the error cannot be appropriately reduced. and D3 are preferably installed.

Next, for the purpose of reducing error factors originating from the receivers, of the three devices D1, D2, and D3, receivers D1 and D2 are installed above the water surface. Here, the above correction using the receiver D3 fixed on the ground takes advantage of the fact that the fluctuation components are similar. On the other hand, receivers D1 and D2 are installed on the water surface, and it can be said that the fluctuation components are random. In such a case, it is considered that random fluctuation components can be suppressed by integrating the altitude of the receiver D1 and the altitude of the receiver D2. Therefore, in this embodiment, the altitudes of the receivers D1 and D2 are corrected by suppressing random fluctuation components between the two using a statistical method.

Regarding this point, in the present embodiment, specific information depending on the installation environment of the receiver D1 or D2 is used as feature information, and a correction coefficient is calculated based on this feature information and a machine learning model using teacher data. A correction process is performed in which both altitudes are weighted using the calculated correction coefficients. This correction process is a process of combining the altitudes of the receivers D1 and D2 by adjusting the balance of the altitudes of the receivers D1 and D2 by weighting with correction coefficients.

Note that each of the above-mentioned correction processes can also be interpreted as a type of leveling process for suppressing variations in altitude.

Also, when reducing error factors originating from the receiver, as shown in FIG. 1, it is preferable that the receiver D1 and the receiver D3 be installed so that the distance relationship between them is within 500 m. Similarly, it is preferable that the receiver D2 and the receiver D3 are installed so that the distance relationship between them is within 500 m. Furthermore, it is preferable that the receiver D1 and the receiver D2 be installed so that the distance relationship between them is within 50 meters.

In addition, in order to realize the water level estimation logic according to the embodiment, the conditions (installation conditions) required for the number of receivers to be used, variations in the location where the receivers are installed, and distance relationships between the receivers are as follows. , the one shown in FIG. 1 is considered to be the best. On the other hand, the conditions shown in FIG. 1 are not essential conditions. For example, the number of receivers may be two; in this case, one receiver D1 (or receiver D2) may be installed on the water surface, and the other receiver D3 may be installed on the ground. . Furthermore, as another example, the number of receivers installed on the ground may not be one, but may be multiple.

[3. System configuration〕
Next, a system configuration for realizing the water level estimation logic according to the embodiment will be described using FIG. 2. FIG. 2 is a diagram illustrating an example of a system configuration according to the embodiment. FIG. 2 shows a water level estimation system 1 as an example of the water level estimation system according to the embodiment.

In the example of FIG. 2, the water level estimation system 1 includes a receiver 10 and an information processing device 100. Further, the receiver 10 and the information processing device 100 are connected via a network N so that they can communicate by wire or wirelessly. Note that the number of information processing devices 100 may be plural.

Furthermore, in the following embodiments, as shown in FIG. 2, the receiver 10 may be explained separately as a first receiver 10-1. Further, the receiver 10 may be explained separately as second receivers 10-21 to 10-2N. In addition, when explaining the distinction in this way, it is assumed that the first receiver 10-1 is the receiver 10 installed on the water surface. Further, at least one of the second receivers 10-21 to 10-2N is the receiver 10 installed on the water surface, and at least another one is installed on the ground. Assume that the receiver is 10.

Furthermore, in the water level estimation system 1, the receiver 10 is located on the edge side. For this reason, users (for example, dam personnel) who wish to know the water level can install receivers 10 in various locations so as to satisfy the installation conditions shown in FIG. 1. That is, the receiver 10 according to the embodiment may be a portable information processing terminal that can be installed at any location. On the other hand, the information processing device 100 may be a server device existing on the cloud side.

Although not shown in FIG. 2, the water level estimation system 1 according to the embodiment may further include a user's terminal device T. The terminal device T may be, for example, a smartphone, a tablet terminal, a notebook PC (Personal Computer), a desktop PC, a mobile phone, a PDA (Personal Digital Assistant), or the like.

[4. On-site situation of water level measurement]
In the following description, among the receivers 10 included in the water level estimation system 1 shown in FIG. An example is shown in which water estimation processing is executed by the information processing device 100 in a state where one receiver 10 is installed. FIG. 3 shows one scene of such a situation. FIG. 3 is a diagram illustrating an example of a situation in which water level estimation processing according to the embodiment is performed.

According to the example of FIG. 3, the current water level of the dam lake DL1 is estimated using the first receiver 10-1, the second receiver 10-21, and the second receiver 10-22. Further, according to the example of FIG. 3, the first receiver 10-1 corresponds to the receiver D1 of FIG. 1, the second receiver 10-21 corresponds to the receiver D2 of FIG. The receiver 10-22 corresponds to the receiver D3 in FIG.

That is, in FIG. 3, the first receiver 10-1 and the second receiver 10-21 are installed on the water surface of the dam lake DL1, and the second receiver 10-22 is installed on the water surface of the dam lake DL1. An example is shown in which it is installed at a dam facility DAM1 (ground) located in . Further, the distance between the first receiver 10-1 and the second receiver 10-21 is "45 m", and the distance between the second receiver 10-21 and the second receiver 10-22 is "45 m". These three receivers 10 are installed in a state where the distance is "450 m" and the installation conditions explained in FIG. 1 are satisfied.

[5. Overall picture of water level measurement process]
Next, the overall picture of the water level measurement process according to the embodiment will be described using FIG. 4. FIG. 4 is a diagram illustrating an example of an overall image of the water level measurement process according to the embodiment. Further, the water level measurement process according to the embodiment is performed by the information processing device 100. For example, the information processing device 100 performs water level measurement processing according to a method determined by water level measurement logic. Further, the water level measurement logic may be realized by a water level measurement program installed in the information processing device 100.

Further, the example in FIG. 4 corresponds to FIG. 3. Specifically, in FIG. 4, the first receiver 10-1 and the second receiver 10-21 are installed on the water surface of the dam lake DL1, and the second receiver 10-22 is installed on the water surface of the dam facility DAM1. A scene is shown in which the current water level of dam lake DL1 is estimated in real time while the water is fixed on the ground.

In this state, the first receiver 10-1 calculates the altitude of its own device by positioning based on the GNSS signal, and transmits altitude data indicating the calculated altitude to the information processing device 100. FIG. 4 shows an example in which the first receiver 10-1 calculates the altitude X1 based on the GNSS signal and transmits first altitude data AD1 indicating this to the information processing device 100.

Further, the second receiver 10-21 calculates the altitude of its own device by positioning based on the GNSS signal, and transmits altitude data indicating the calculated altitude to the information processing device 100. FIG. 4 shows an example in which the second receiver 10-21 calculates the altitude N21 based on the GNSS signal and transmits second altitude data AD21 indicating this to the information processing device 100.

Further, the second receiver 10-22 calculates the altitude of its own device by positioning based on the GNSS signal, and transmits altitude data indicating the calculated altitude to the information processing device 100. FIG. 4 shows an example in which the second receiver 10-22 calculates the altitude N22 based on the GNSS signal and transmits second altitude data AD22 indicating this to the information processing device 100.

Among the altitude data received from each receiver 10, the information processing device 100 selects altitude data AD1 of the first receiver 10-1, which is the water surface receiver 10, and altitude data AD1 of the first receiver 10-1, which is the water surface receiver 10, and altitude data AD1 of the second receiver, which is the water surface receiver 10. Regarding the altitude data AD21 of No. 10-21, the distance from the water surface to the tip of the antenna may be excluded as an offset value.

From here, the information processing device 100 actually performs water level estimation processing. First, the information processing device 100 adjusts altitude data using the second receiver 10-22 fixed on the ground (dam facility DAM1) (step S1).

Specifically, the information processing device 100 converts the second altitude data AD22 obtained by positioning by the second receiver 10-22 into the non-uniformity of the first altitude data AD1 and the second altitude data AD21. Obtained as correction data for correcting. Then, the information processing device 100 extracts a variation component of the second altitude data AD22 with respect to the reference position using the altitude of the ground as a reference position, and based on the extracted variation component, the information processing device 100 extracts a variation component of the second altitude data AD22 with respect to the reference position. A correction process of adjusting the altitude data AD21 is performed.

According to the example in FIG. 4, the information processing device 100 extracts the difference in altitude N22 from the reference position as a variation component, and adjusts altitude X1 and altitude N21 based on the extracted variation component. For example, if the fluctuation component has a positive value, the information processing device 100 subtracts this positive value from the altitudes X1 and N21. Further, if the fluctuation component is a negative value, the information processing device 100 adds this negative value to the altitude X1 and the altitude N21.

FIG. 4 shows that as a result of the correction process in step S1, the altitude X1 is adjusted to the altitude X11, so that the information processing device 100 uses the altitude data indicating the altitude An example obtained is shown. Similarly, FIG. 4 shows that as a result of the leveling process, the altitude N21 is adjusted to the altitude N211, so that the information processing device 100 uses the altitude data indicating the altitude N211 as the adjusted second altitude data AD211. An example obtained is shown.

Next, the information processing device 100 performs a correction process of adjusting the altitude data using the first receiver 10-1 and the second receiver 10-21 installed above the water surface (step S2). . In the correction process here, the adjusted first altitude data AD11 obtained in step S1 and the adjusted second altitude data AD211 are balanced by weighting using correction coefficients.

As will be described later, at this point the information processing device 100 has a prediction model that calculates a correction coefficient according to the feature information. For this reason, the information processing device 100 inputs predetermined information according to the installation status of each of the first receiver 10-1 and the second receiver 10-21 as characteristic information into the model. , calculates a correction coefficient α used for correction processing from the output information.

Note that the predetermined information here includes the date and time when the receiver 10 observed the altitude data, position information indicating the position where the receiver 10 is installed, the positional relationship between the receiver 10 and the artificial satellite, and the date and time when the receiver 10 observed the altitude data. The information may be either sensor information detected by a sensor that the receiver 10 has, or the quality status of a communication line corresponding to the receiver 10. Of course, the information used as the characteristic information is not limited to this example, and may be any information as long as it is original information that can be obtained according to the installation situation of the receiver 10.

Returning to the explanation, the information processing device 100 corrects the altitude X11, which is the value of the first altitude data AD11, by weighting using the correction coefficient α. Furthermore, the information processing device 100 corrects the altitude N211, which is the value of the second altitude data AD211, by weighting using the correction coefficient α.

Finally, the information processing device 100 estimates the water level of the dam lake DL1 based on the corrected altitude data obtained in the correction process of step S2 (step S3). For example, the information processing device 100 performs a calculation to combine the corrected first altitude data AD11 and the corrected second altitude data AD211 by linear combination using the correction coefficient α as a weight value, and calculates the calculation result. Estimated as the water level of dam lake DL1.

[6. Configuration of information processing device]
From here, the information processing device 100 according to the embodiment will be described using FIG. 5. FIG. 5 is a diagram illustrating a configuration example of the information processing device 100 according to the embodiment. As shown in FIG. 5, the information processing device 100 includes a communication section 110, a storage section 120, and a control section 130.

(About communication department 110)
The communication unit 110 is realized by, for example, a NIC (Network Interface Card). The communication unit 110 is connected to the network N by wire or wirelessly, and transmits and receives information to and from the receiver 10, for example.

(About storage unit 120)
The storage unit 120 is realized by, for example, a semiconductor memory device such as a RAM (Random Access Memory) or a flash memory, or a storage device such as a hard disk or an optical disk. The storage unit 120 includes a learning data storage unit 121 , an altitude data storage unit 122 , and an estimation result storage unit 123 .

(About the learning data storage unit 121)
The learning data storage unit 121 stores information regarding learning data used for model learning. Here, FIG. 6 shows an example of the learning data storage unit 121 according to the embodiment. In the example of FIG. 6, the learning data storage unit 121 has items such as "date and time", "signal-1", "signal-2", "water level", "correct coefficient", and "feature information".

“Date and time” is information indicating the date and time when “signal-1” and “signal-2” were observed.

"Signal-1" is one of the elements of the teacher data, and indicates altitude data obtained by positioning by the receiver 10 used when collecting the teacher data. Note that such a receiver 10 may be, for example, one receiver 10 installed at an arbitrary location (for example, the water surface or the ground) in the dam lake DL1, and has been described so far as being used for this water level measurement. The receiver 10-1 may be regarded as the first receiver 10-1. For convenience, such a receiver 10 during learning will be referred to as receiver 10x1 below.

"Signal-2" is one of the elements of the teacher data, and indicates altitude data obtained by positioning by the receiver 10 used when collecting the teacher data. Note that such a receiver 10 may be, for example, one receiver 10 installed at an arbitrary location (for example, the water surface or the ground) in the dam lake DL1, and has been described so far as being used for this water level measurement. The second receiver 10-21 (or the second receiver 10-22) may be regarded as the second receiver 10-21 (or the second receiver 10-22) that has been used in the past. For convenience, such a receiver 10 during learning will be referred to as receiver 10x2 below.

“Water level” is information indicating the true water level of the dam lake DL1 at the “date and time” when “signal-1” and “signal-2” were observed. In addition, the value of the water level gauge provided in dam lake DL1 may be employ|adopted for the "water level" here.

The "correct coefficient" is a coefficient calculated by inputting "signal-1", "signal-2", and "water level" to equation (1) used for water level estimation.

"Feature information" is the date and time when receiver 10x1 observed "signal-1", location information indicating the location where receiver 10x1 is installed, and the source of the GNSS signal received by receiver 10x1 and receiver 10x1. The information may be any of the positional relationship with a GNSS satellite, sensor information detected by a sensor included in the receiver 10x1, or the quality status of a communication line corresponding to the receiver 10x1.

In addition, "characteristic information" includes the date and time when the receiver 10x2 observed "signal-2", location information indicating the position where the receiver 10x2 was installed, and the source of the GNSS signal received by the receiver 10x2 and the receiver 10x2. It may be any of the positional relationship with the GNSS satellite, sensor information detected by a sensor included in the receiver 10x2, or the quality status of the communication line corresponding to the receiver 10x2.

Here, in FIG. 6, the date and time "TM1", signal-1 "sg11", signal-2 "sg21", water level "WL1", correct answer coefficient "sα1", feature information "x11, x12, An example in which "x23" is associated is shown.

This example shows an example in which the receiver 10x1 installed at the dam lake DL1 calculates the altitude of signal-1 "sg11" at the date and time "TM1". Further, an example is shown in which the receiver 10x2 installed at the dam lake DL1 calculates the altitude of signal-1 "sg21" at the date and time "TM1".

In addition, in this example, at the date and time "TM1" when signal-1 "sg11" and signal-2 "sg21" were calculated by positioning, the water level gauge of dam lake DL1 was pointing to the water level "WL1". An example in which this value is adopted as the true water level at date and time "TM1" is shown.

Further, this example shows an example in which, at the date and time "TM1", the characteristic information corresponding to the receiver 10x1 (information according to the situation in which the receiver 10x1 is installed) is the characteristic x11 and the characteristic x12. Further, for the date and time "TM1", an example is shown in which the feature information corresponding to the receiver 10x2 (information according to the situation in which the receiver 10x2 is installed) is feature x23.

Further, in this example, a coefficient "sα1" is calculated by applying the water level "WL1" and the feature information "x11, x12, x23" to equation (1), and this is determined as the correct coefficient. show.

Further, according to the above example, a set of the correct coefficient "sα1" and the feature information "x11, x12, x23" becomes one teacher data.

Note that the "feature information" is not limited to the above example. For example, the "feature information" may be the S/N ratio of the GNSS signal received by the receiver 10x1 (receiver 10x2), or the number of satellites used for positioning by the receiver 10x1 (receiver 10x2). .

(About the altitude data storage unit 122)
The altitude data storage unit 122 stores information regarding altitude data indicating the altitude obtained by positioning by the receiver 10. FIG. 6 also shows an example of the altitude data storage unit 122 according to the embodiment. In the example of FIG. 6, the altitude data storage unit 122 has items such as "date and time", "signal-1", "signal-2", "water level", "correction coefficient", and "characteristic information".

"Date and time" is information indicating the current date and time at which the water level is to be estimated.

"Signal-1" indicates first altitude data AD1 obtained by positioning by the first receiver 10-1. More specifically, "signal-1" may be the first altitude data AD11 after the first altitude data AD1 has been adjusted through correction processing using the second receiver 10-22. . That is, according to the example of FIG. 4, "signal-1" corresponds to "altitude X11".

"Signal-2" indicates second altitude data AD21 obtained by positioning by the second receiver 10-21. More specifically, "signal-2" may be the second altitude data AD211 after the second altitude data AD21 has been adjusted through correction processing using the second receiver 10-22. . That is, according to the example of FIG. 4, "signal-2" corresponds to "altitude N211".

"Water level" indicates the water level that can be estimated in the current water level estimation process, and FIG. 6 shows the unknown state before the "water level" is estimated.

The "correction coefficient" is calculated based on the output information when predetermined information according to the installation status of the first receiver 10-1 and the second receiver 10-21 is input into the model as characteristic information. This coefficient corresponds to a weighting coefficient for the first altitude data AD11 and the second altitude data AD211. FIG. 6 shows an unknown state before the "correction coefficient" is calculated.

The "characteristic information" is predetermined information depending on the situation in which the first receiver 10-1 is installed, and is information acquired at the corresponding "date and time." Further, the "characteristic information" is predetermined information depending on the situation in which the second receiver 10-21 is installed, and may be information acquired at the corresponding "date and time." That is, the "feature information" may include information about both the first receiver 10-1 and the second receiver 10-21, or may include information about only one of them.

Moreover, the "characteristic information" here specifically indicates the date and time when the first receiver 10-1 observed "signal-1" and the position where the first receiver 10-1 was installed. Position information, the positional relationship between the first receiver 10-1 and the GNSS satellite that is the source of the GNSS signal received by the first receiver 10-1, and detected by the sensor of the first receiver 10-1. The information may be either the sensor information that has been received or the quality status of the communication line corresponding to the first receiver 10-1.

Further, the "characteristic information" specifically includes the date and time when the second receiver 10-21 observed "signal-2", position information indicating the position where the second receiver 10-21 is installed, The positional relationship between the second receiver 10-21 and the GNSS satellite that is the source of the GNSS signal received by the second receiver 10-21, and the sensor detected by the sensor included in the second receiver 10-21. The information may be either information or the quality status of the communication line corresponding to the second receiver 10-21.

(Regarding the estimation result storage unit 123)
The estimation result storage unit 123 may store information regarding the water level estimated by the water level estimation process.

(About the control unit 130)
Returning to FIG. 5, the control unit 130 executes various programs stored in the storage device inside the information processing device 100 using the RAM as a work area by the CPU (Central Processing Unit), MPU (Micro Processing Unit), etc. This is achieved by Further, the control unit 130 is realized by, for example, an integrated circuit such as an ASIC (Application Specific Integrated Circuit) or an FPGA (Field Programmable Gate Array).

As shown in FIG. 5, the control unit 130 includes a collection unit 131, a generation unit 132, an acquisition unit 133, an adjustment unit 134, a calculation unit 135, and an estimation unit 136. Achieve or execute a processing function or action. Note that the internal configuration of the control unit 130 is not limited to the configuration shown in FIG. 5, and may be any other configuration as long as it performs information processing to be described later. Further, the connection relationship between the respective processing units included in the control unit 130 is not limited to the connection relationship shown in FIG. 5, and may be other connection relationships.

(About the collection unit 131)
The collection unit 131 collects training data to be trained by the model in order to generate a prediction model that predicts correction coefficients.

(About the generation unit 132)
The generation unit 132 generates a prediction model used to calculate the correction coefficient. For example, when the feature information is input, the generation unit 132 generates a prediction model that outputs information used to calculate a correction coefficient.

For example, the generation unit 132 generates altitude data obtained from each of a plurality of receivers (receiver 10x1, receiver 10x2) corresponding to the set of first receiver 10-1 and second receiver 10-21. and the coefficient of the correct answer determined based on the water level (true water level) in the installation environment where multiple receivers are installed, and the characteristics according to the situation where at least one of the multiple receivers is installed. A predictive model is generated by making the model learn the relationship between the correct coefficient and feature information using the pair with the information as training data.

For example, the generation unit 132 generates a predictive model through supervised learning (also referred to as training) using teacher data. Here, if the learning model is realized as a regression model, the regression model can be regarded as a simple perceptron having an input layer and an output layer.

Further, the prediction model is generated by processing such as backpropagation (error backpropagation method) that corrects parameters (connection coefficients) so that the error between the output and input in the model is reduced. For example, the prediction model is generated by performing processing such as backpropagation to minimize a predetermined loss function such as an error function.

(About the acquisition unit 133)
The acquisition unit 133 receives first altitude data that is altitude data obtained from a satellite signal received by the first receiver 10 and altitude data obtained from a satellite signal received by the second receiver 10. and second altitude data.

In the example of FIG. 4, the acquisition unit 133 acquires first altitude data AD1 obtained by positioning by the first receiver 10-1. Furthermore, the acquisition unit 133 acquires second altitude data AD21 obtained by positioning by the second receiver 10-21. Further, the acquisition unit 133 acquires second altitude data AD22 obtained by positioning by the second receiver 10-22.

(About the adjustment section 134)
The adjustment unit 134 acquires the altitude data obtained from the satellite signal received by the second receiver 10 installed on the ground as correction data for correcting non-uniformity of the altitude data. The adjustment unit 134 then adjusts the first altitude data and the second altitude data using the acquired correction data. For example, the adjustment unit 134 uses the altitude of the ground as a reference position, extracts a variation component of the correction data with respect to the reference position, and adjusts the first altitude data and the second altitude data based on the extracted variation component. Further, the adjustment unit 134 may adjust the altitude data observed at the same timing as the correction data using the fluctuation component.

In the example of FIG. 4, the adjustment unit 134 adjusts the second altitude data AD22 obtained by positioning by the second receiver 10-22 to the nonuniformity of the first altitude data AD1 and the second altitude data AD21. It is acquired as correction data for correcting. Then, the adjustment unit 134 extracts a fluctuation component of the second altitude data AD22 with respect to the reference position using the ground altitude as a reference position, and adjusts the first altitude data AD1 and the second altitude data based on the extracted fluctuation component. A correction process is performed to adjust the data AD21.

(About the calculation unit 135)
The calculation unit 135 calculates a correction coefficient based on predetermined information depending on the situation in which the receiver 10 is installed. For example, the calculation unit 135 calculates the correct coefficient determined according to altitude data acquired from a predetermined receiver 10 (in the above example, receiver 10x1, receiver 10x2), and A correction coefficient is calculated based on a model that has learned the relationship with feature information according to the installation situation and predetermined information. As an example, the calculation unit 135 calculates at least one of predetermined information according to the situation in which the first receiver 10 is installed, or predetermined information according to the situation in which the second receiver 10 is installed. By inputting one of them to the model as feature information, a correction coefficient is calculated based on the output information.

In the example of FIG. 4, the calculation unit 135 inputs predetermined information according to the installation status of each of the first receiver 10-1 and the second receiver 10-21 as characteristic information into the model. A correction coefficient α used for correction processing is calculated from the output information.

(About the estimation unit 136)
The estimation unit 136 determines the water level measurement target (for example, dam lake DL1) based on the corrected data in which the first altitude data and the second altitude data are corrected using the correction coefficient calculated by the calculation unit 135. Estimate the water level. More specifically, the estimating unit 136 estimates the water level of the water level measurement target based on the corrected data obtained by correcting the altitude data adjusted by the adjusting unit 134 using a correction coefficient.

For example, the estimation unit 136 can estimate the water level by applying corrected data and a correction coefficient to Equation (1), which will be described later. Regarding this point, according to the example of FIG. 4, for example, the estimating unit 136 combines the corrected first altitude data AD11 and the corrected second altitude data by linear combination using the correction coefficient α as a weight value. A calculation is performed in combination with AD211, and the calculation result is estimated as the water level of the dam lake DL1.

[7. Learning processing procedure〕
From here on, a learning process procedure for generating a prediction model for predicting correction coefficients will be explained using FIG. 7. FIG. 7 is a flowchart showing the learning processing procedure according to the embodiment. The learning process is performed between the collection unit 131 and the generation unit 132.

Furthermore, the example of FIG. 6 will be used as appropriate to explain the learning processing procedure. Specifically, assume that a receiver 10x1 and a receiver 10x2 are installed on the water surface of dam lake DL1 to collect teacher data, and a situation where teacher data is collected and model generation is performed based on these multiple receivers 10 will be described. FIG. 7 will be described as an example.

In the example of FIG. 7, it is assumed that the receiver 10x1 and the receiver 10x2 constantly calculate the altitude of their own devices by continuously performing positioning. The receiver 10x1 and the receiver 10x2 may transmit altitude data indicating the calculated altitude to the information processing device 100 every time the altitude is calculated.

In such a case, the collection unit 131 may determine whether the timing has come to acquire the altitude data transmitted from the receiver 10x1 and the receiver 10x2 (step S701). Then, while the collection unit 131 determines that it is not the timing to acquire the altitude data (step S701; No), it waits until it can be determined that the timing to acquire the altitude data has arrived.

On the other hand, if the collection unit 131 determines that it is time to acquire altitude data (step S701; Yes), it acquires predetermined altitude data from among the altitude data transmitted from the receiver 10x1 and the receiver 10x2. (Step S702). For example, the collection unit 131 acquires the latest altitude data that has the same observation date and time (positioning date and time) between the altitude data transmitted from the receiver 10x1 and the altitude data transmitted from the receiver 10x2. be able to. Moreover, as a result, the collection unit 131 can acquire one set of altitude data corresponding to the set of receiver 10x1 and receiver 10x2.

In the example of FIG. 7, it is assumed that the collection unit 131 has acquired altitude data indicating an altitude "sg11" as signal-1, which is the positioning result of the positioning performed by the receiver 10x1 on the date and time "TM1". Further, it is assumed that the collection unit 131 has acquired altitude data indicating an altitude “sg21” as signal-2, which is the positioning result of the positioning performed by the receiver 10x2 on the date and time “TM1”.

In addition, the collection unit 131 acquires water level information indicating the water level of the dam lake DL1 at the current timing when the altitude data is acquired, as water level information indicating the true water level of the dam lake DL1 at this timing (step S703). . For example, the collection unit 131 may acquire water level information by accessing a system that controls a water level gauge provided in the dam lake DL1. Moreover, the water level information may be input manually.

In the example of FIG. 7, it is assumed that the collection unit 131 has acquired water level information indicating the water level "WL1" as the water level indicated by the water level gauge at the date and time "TM1".

The collection unit 131 also acquires characteristic information corresponding to each of the receiver 10x1 and the receiver 10x2 (step S704). Specifically, the collection unit 131 may acquire specific information according to the installation situation in which the receiver 10x1 is installed as the characteristic information. Similarly, the collection unit 131 may acquire specific information according to the installation situation in which the receiver 10x2 is installed as the characteristic information. Note that the feature information may be transmitted to the information processing device 100 together with the altitude data, and in such a case, the collection unit 131 selects the feature corresponding to the current timing from among the feature information that has been transmitted so far. You may obtain information.

In the example of FIG. 7, it is assumed that the collection unit 131 has acquired the feature information "x11, x12" that the receiver 10x1 transmitted at the date and time "TM1" as the feature information corresponding to the receiver 10x1. Further, it is assumed that the collection unit 131 has acquired characteristic information “x23” that the receiver 10x2 transmitted at the date and time “TM1” as the characteristic information corresponding to the receiver 10x2.

Next, the collection unit 131 calculates the correct coefficient corresponding to the altitude data set and the water level information based on the altitude data set acquired in step S702 and the water level information acquired in step S703. Determine (step S705). Specifically, the collection unit 131 calculates a coefficient α by inputting a set of altitude data and water level information into the following equation (1), and determines the calculated coefficient α as the correct coefficient.

Water level = {α×signal-1}+{(1-α)×signal-2}...(1)

According to the above example, the collection unit 131 obtains the coefficient α by calculating the water level “WL1”={αדsg11”}+{(1−α)דsg21”}.

In the example of FIG. 7, it is assumed that the collection unit 131 calculates "sα1" as the coefficient α. In this case, the collection unit 131 determines the coefficient sα1 as the correct coefficient corresponding to the altitude data set (sg11, sg21) and the water level information (WL1).

Further, the collection unit 131 may register the data obtained so far in the learning data storage unit 121 in association with each other (step S706). As a result, the first record is obtained in the learning data storage unit 121 shown in FIG. 6. Also, if we focus on the record in the first row, the set of the correct coefficient "sα1" and the feature information "x11, x12, x23" becomes one teacher data.

Here, the collection unit 131 determines whether or not such teacher data has been sufficiently collected (step S707). If the amount of training data is small, a highly accurate model cannot be obtained. For this reason, the collection unit may determine whether the learning data storage unit 121 has accumulated enough training data to be necessary and sufficient for machine learning to generate a predictive model.

If the collection unit 131 determines that a sufficient amount of teacher data has not been collected (step S707; No), it re-executes the process from step S701.

On the other hand, if the collection unit 131 determines that a sufficient amount of teacher data has been collected (step S707; Yes), the process shifts to the generation unit 132.

In response, the generation unit 132 generates a prediction model used for calculating the correction coefficient by machine learning using the teacher data (step S708). Specifically, the generation unit 132 creates a prediction model that outputs information used to calculate a correction coefficient when the feature information is input by having the model learn the relationship between the correct coefficient and the feature information. generate.

In addition, according to the example so far, the information input to the prediction model is specific information that depends on the installation situation of the first receiver 10-1 or the installation situation of the second receiver 10-21. It is context-specific information. Since the specific example of the feature information is as described above, the explanation here will be omitted.

Further, the correction coefficient calculated using the prediction model corresponds to the coefficient α shown in equation (1). For example, the information processing device 100 stores environment-dependent information (an example of predetermined information) corresponding to the first receiver 10-1 or environment-dependent information (an example of predetermined information) corresponding to the second receiver 10-21. The correction coefficient α can be obtained by inputting at least one of the following as feature information to the prediction model.

Furthermore, using the example of the altitude data storage unit 122 shown in FIG. 6, the information processing device 100 inputs “altitude X11” indicated by the first altitude data AD11 as “signal-1” into equation (1), "Altitude N211" indicated by the second altitude data AD211 is input into equation (1) as "signal-2". Furthermore, the information processing apparatus 100 also inputs the correction coefficient α calculated this time into equation (1). Then, the information processing device 100 can obtain information regarding the water level of the dam lake DL1 by solving Equation (1). This point will be explained again in FIG. 8 as well.

[8. Water level treatment procedure]
From here, the procedure of water level estimation processing using a receiver will be explained using FIG. 8. FIG. 8 is a sequence diagram showing a water level estimation process executed by the water level estimation system 1 according to the embodiment. In the example of FIG. 8, a processing procedure for estimating the water level of the dam lake DL1 will be described using the situation described in FIG. 3 as an example. Furthermore, the water level estimation process shown in FIG. 8 uses the prediction model generated in FIG. 7 . Therefore, the water level estimation process according to the embodiment may be performed in a separate phase from the learning process.

In the example of FIG. 8, the first receiver 10-1 calculates the altitude of its own device by positioning based on the GNSS signal, and transmits first altitude data indicating the calculated altitude to the information processing device 100. There is. Further, the first receiver 10-1 may also transmit specific information (information input to the model as characteristic information) depending on the installation situation together with the first altitude data. For example, each time the first receiver 10-1 calculates the first altitude data, the first receiver 10-1 associates the calculated first altitude data with specific information at this point and transmits it to the information processing device 100. You may do so.

Further, the specific information transmitted by the first receiver 10-1 includes the date and time when the first receiver 10-1 measured the position, and the location where the first receiver 10-1 is installed. information (position as of the positioning date and time), the positional relationship between the GNSS satellite that is the source of the GNSS signal and the first receiver 10-1 (positional relationship as of the positioning date and time), which the first receiver 10-1 has. It may be either sensor information detected by the sensor (sensor information at the positioning date and time) or quality status of the communication line corresponding to the first receiver 10-1 (quality status at the positioning date and time). . Note that the specific information is not limited to such examples; for example, the S/N ratio of the GNSS signal received by the first receiver 10-1, or the information used for positioning by the first receiver 10-1. It may also be the number of satellites.

In the example of FIG. 8, the second receiver 10-21 calculates the altitude of its own device by positioning based on the GNSS signal, and transmits second altitude data indicating the calculated altitude to the information processing device 100. are doing. Similarly, the second receiver 10-21 may transmit specific information (information input to the model as characteristic information) depending on the installation situation together with the second altitude data. For example, each time the second receiver 10-21 calculates the second altitude data, the second receiver 10-21 associates the calculated second altitude data with specific information at this point in time and transmits it to the information processing device 100. You may do so.

The specific information transmitted by the second receiver 10-21 includes the date and time when the second receiver 10-21 measured the position, and position information ( positional relationship between the GNSS satellite that is the source of the GNSS signal and the second receiver 10-21 (positional relationship as of the positioning date and time), and the sensor possessed by the second receiver 10-21. It may be either detected sensor information (sensor information at the positioning date and time) or quality status of the communication line corresponding to the second receiver 10-21 (quality status at the positioning date and time). Similarly, the specific information may be the S/N ratio of the GNSS signal received by the second receiver 10-21 or the number of satellites used for positioning by the second receiver 10-21. .

In the example of FIG. 8, the second receiver 10-22 calculates the altitude of its own device by positioning based on the GNSS signal, and transmits second altitude data indicating the calculated altitude to the information processing device 100. are doing. The second receiver 10-22 may similarly transmit specific information (information input to the model as characteristic information) depending on the installation situation together with the second altitude data. For example, every time the second receiver 10-22 calculates the second altitude data, the second receiver 10-22 associates the calculated second altitude data with specific information at this point in time and transmits it to the information processing device 100. You may do so.

The specific information transmitted by the second receiver 10-22 includes the date and time when the second receiver 10-22 measured the position, and location information ( (position at the positioning date and time), the positional relationship between the GNSS satellite that is the source of the GNSS signal and the second receiver 10-22 (positional relationship at the positioning date and time), and the sensor possessed by the second receiver 10-22. It may be either detected sensor information (sensor information at the positioning date and time) or quality status of the communication line corresponding to the second receiver 10-22 (quality status at the positioning date and time). Similarly, the specific information may be the S/N ratio of the GNSS signal received by the second receiver 10-22 or the number of satellites used for positioning by the second receiver 10-22. .

By transmitting altitude data and specific information from each receiver as described above, the information processing apparatus 100 receives altitude data and specific information from each receiver 10 at any time (step S801).

In such a state, the acquisition unit 133 is configured to detect altitude data transmitted from the first receiver 10-1, the second receiver 10-21, and the second receiver 10-22. In step S802, the latest altitude data having the same positioning date and time (observation date and time) may be acquired.

FIG. 8 shows an example in which the acquisition unit 133 acquires first altitude data AD1 indicating altitude X1 from among the first altitude data from the first receiver 10-1. Further, FIG. 8 shows an example in which the acquisition unit 133 acquires second altitude data AD21 indicating altitude N21 from among the second altitude data obtained by the second receiver 10-21. Further, FIG. 8 shows an example in which the acquisition unit 133 acquires second altitude data AD22 indicating altitude N22 from among the second altitude data obtained by the second receiver 10-22.

Note that, among the altitude data received from each receiver 10, the acquisition unit 133 acquires altitude data AD1 of the first receiver 10-1, which is the receiver 10 on the water surface, and altitude data AD1 of the second receiver 10, which is the receiver 10 on the water surface. Regarding the altitude data AD21 of aircraft 10-21, the distance from the water surface to the tip of the antenna may be excluded as an offset value.

In addition, the acquisition unit 133 selects the specific information transmitted from each of the first receiver 10-1, the second receiver 10-21, and the second receiver 10-22 in step S802. Information corresponding to the date and time when the altitude data obtained in step S803 was positioned is obtained as characteristic information (step S803).

FIG. 8 shows an example in which the acquisition unit 133 acquires characteristic information indicating a characteristic x41 according to the installation situation at the positioning date and time, from among the specific information corresponding to the first receiver 10-1. Further, FIG. 8 shows an example in which the acquisition unit 133 acquires characteristic information indicating the characteristic x42 according to the installation status at the positioning date and time, from among the specific information corresponding to the second receiver 10-21. It will be done. Further, FIG. 8 shows an example in which the acquisition unit 133 acquires characteristic information indicating the characteristic x43 according to the installation status at the positioning date and time, from among the specific information corresponding to the second receiver 10-22. It will be done.

Next, the adjustment unit 134 adjusts (corrects) the altitude data using the second receiver 10-22 fixed on the ground (dam facility DAM1). In the example of FIG. 8, the adjustment unit 134 corrects the second altitude data AD22 from the second receiver 10-22 to correct the non-uniformity of the first altitude data AD1 and the second altitude data AD21. It is obtained as data.

In such a state, the adjustment unit 134 uses the altitude of the ground as a reference position and extracts a fluctuation component of the second altitude data AD22 with respect to the reference position (step S804). For example, the adjustment unit 134 extracts the difference in altitude N22 (the altitude indicated by the second altitude data AD22) with respect to the reference position as a fluctuation component.

Then, the adjustment unit 134 performs a correction process to adjust the first altitude data AD1 and the second altitude data AD21 using the extracted fluctuation components (step S805). For example, if the extracted fluctuation component has a positive value, the adjustment unit 134 adjusts the altitude X1 (the altitude indicated by the first altitude data AD1) and the altitude N21 (the altitude indicated by the second altitude data AD21) by this positive value. (altitude indicated by). On the other hand, if the extracted fluctuation component has a negative value, the adjustment unit 134 adds this negative value to the altitudes X1 and N21.

The calculation unit 135 obtains a set of adjusted altitude data (step S806).

FIG. 8 shows an example in which the calculation unit 135 acquires altitude data indicating the altitude X11 as the adjusted first altitude data AD11 as the altitude X1 is adjusted to the altitude X11. Further, FIG. 8 shows an example in which the altitude N21 is adjusted to the altitude N211, and the calculation unit 135 acquires altitude data indicating the altitude N211 as the adjusted second altitude data AD211.

Further, in such an example, the information processing device 100 can obtain the altitude data storage unit 122 shown in FIG. 6. For example, as shown in FIG. 6, the calculation unit 135 sets "altitude 2", it can be registered in the altitude data storage unit 122. Further, as shown in FIG. 6, the calculation unit 135 includes feature information indicating a feature x41 corresponding to the first receiver 10-1, and feature information indicating a feature x42 corresponding to the second receiver 10-21. can be registered in the altitude data storage section 122.

In such an example, the calculation unit 135 calculates the correction coefficient α based on the feature information registered in the altitude data storage unit 122 and the prediction model generated by the procedure described in FIG. S807). Specifically, the calculation unit 135 calculates the correction coefficient α based on the information output by the prediction model by inputting the feature information indicating the feature x41 and the feature information indicating the feature x42 into the prediction model. do.

Then, the estimation unit 136 calculates the water level of the dam lake DL1 by applying the adjusted first altitude data AD11, the adjusted second altitude data AD211, and the correction coefficient α to equation (1). Estimate (step S808).

For example, the estimating unit 136 inputs "altitude X11" indicated by the first altitude data AD11 into equation (1) as "signal-1", and inputs "altitude 2'' into equation (1). Furthermore, the estimation unit 136 also inputs the correction coefficient α into equation (1). Then, the estimation unit 136 estimates the obtained value as the water level of the dam lake DL1 by solving Equation (1).

The estimating unit 136 may also perform a process of providing the estimation result to the user U1 who has requested to estimate the water level of the dam lake DL1. For example, the estimation unit 136 may transmit the estimation result to the terminal device T owned by the user U1 (step S809).

[9. Modified example]
The water level estimation system 1 according to the embodiment described above may be implemented in various different forms other than the embodiment described above. Therefore, other embodiments of the water level estimation system 1 will be described below.

[9-1. Modification example (1)]
In the above embodiment, the information processing device 100 uses the altitude data from the second receiver 10-22 fixed on the ground as correction data, and the altitude data from the first receiver 10-1 on the water surface and the altitude data from the first receiver 10-1 on the water surface. An example is shown in which the water level is estimated using the altitude data obtained by correcting the altitude data from the receiver 10-21 of No. 2 and adjusting through the correction process.

However, the information processing device 100 corrects the altitude data from the first receiver 10-1 on the water surface and the altitude data from the second receiver 10-22 fixed to the ground, and adjusts it through the correction process. Later altitude data may be used to estimate the water level. In such an example, at least one of the feature information corresponding to the first receiver 10-1 and the feature information corresponding to the second receiver 10-22 is input to the prediction model, A correction coefficient α is calculated. Then, the information processing device 100 estimates one water level in the dam lake DL1 by inputting the set of adjusted altitude data and the correction coefficient α into equation (1).

In addition, the information processing device 100 corrects the altitude data from the second receiver 10-21 on the water surface and the altitude data from the second receiver 10-22 fixed to the ground, and adjusts it by the correction process. Later altitude data may be used to estimate the water level. In such an example, at least one of the feature information corresponding to the second receiver 10-21 and the feature information corresponding to the second receiver 10-22 is input into the prediction model. A correction coefficient α is calculated. Then, the information processing device 100 estimates another water level in the dam lake DL1 by inputting the set of adjusted altitude data and the correction coefficient α into equation (1).

Further, as in the previous examples, the information processing device 100 corrects the altitude data from the first receiver 10-1 on the water surface and the altitude data from the second receiver 10-21 on the water surface, and The water level may be further estimated using the altitude data after being adjusted by the correction process. In such an example, at least one of the feature information corresponding to the first receiver 10-1 and the feature information corresponding to the second receiver 10-21 is input into the prediction model, A correction coefficient α is calculated. Then, the information processing device 100 estimates yet another water level in the dam lake DL1 by inputting the set of adjusted altitude data and the correction coefficient α into equation (1).

According to the above example, the information processing device 100 obtains three water levels in dam lake DL1 in real time. In such a case, the information processing device 100 may perform data cleaning using these three water levels to integrate the three water levels and obtain one final water level.

[9-2. Modification example (2)]
In the embodiment described above, an example has been shown in which the information processing device 100 includes the collection unit 131 and the generation unit 132 and may perform not only water level estimation but also a learning process to generate a predictive model used for water level estimation.

However, the learning process performed between the collection unit 131 and the generation unit 132 may be performed by an external device different from the information processing device 100. In this case, the external device may include a processing unit corresponding to the collection unit 131 and the generation unit 132. Furthermore, in such a case, the information processing device 100 only needs to acquire the prediction model generated by the external device without performing learning processing.

[10. Hardware configuration]
Further, the information processing apparatus 100 according to the embodiment described above is realized by, for example, a computer 1000 having a configuration as shown in FIG. FIG. 9 is a hardware configuration diagram showing an example of a computer 1000 that implements the functions of the information processing device 100. Computer 1000 has CPU 1100, RAM 1200, ROM 1300, HDD 1400, communication interface (I/F) 1500, input/output interface (I/F) 1600, and media interface (I/F) 1700.

CPU 1100 operates based on a program stored in ROM 1300 or HDD 1400, and controls each part. The ROM 1300 stores a boot program executed by the CPU 1100 when the computer 1000 is started, programs depending on the hardware of the computer 1000, and the like.

The HDD 1400 stores programs executed by the CPU 1100, data used by the programs, and the like. Communication interface 1500 receives data from other devices via communication network 50 and sends it to CPU 1100, and sends data generated by CPU 1100 to the other devices via communication network 50.

The CPU 1100 controls output devices such as a display and a printer, and input devices such as a keyboard and mouse via an input/output interface 1600. CPU 1100 obtains data from an input device via input/output interface 1600. Further, CPU 1100 outputs the generated data to an output device via input/output interface 1600.

Media interface 1700 reads programs or data stored in recording medium 1800 and provides them to CPU 1100 via RAM 1200. CPU 1100 loads the program from recording medium 1800 onto RAM 1200 via media interface 1700, and executes the loaded program. The recording medium 1800 is, for example, an optical recording medium such as a DVD (Digital Versatile Disc) or a PD (Phase change rewritable disk), a magneto-optical recording medium such as an MO (Magneto-Optical disk), a tape medium, a magnetic recording medium, or a semiconductor memory. etc.

For example, when the computer 1000 functions as the information processing device 100 according to the embodiment, the CPU 1100 of the computer 1000 realizes the functions of the control unit 130 by executing a program loaded onto the RAM 1200. Furthermore, data in the storage unit 120 is stored in the HDD 1400. The CPU 1100 of the computer 1000 reads these programs from the recording medium 1800 and executes them, but as another example, these programs may be acquired from another device via the communication network 50.

[11. others〕
Furthermore, each component of each device shown in the drawings is functionally conceptual, and does not necessarily need to be physically configured as shown in the drawings. In other words, the specific form of distributing and integrating each device is not limited to what is shown in the diagram, and all or part of the devices can be functionally or physically distributed or integrated in arbitrary units depending on various loads and usage conditions. Can be integrated and configured.

[12. summary〕
As mentioned above, the embodiments of the present application have been described in detail based on several drawings, but these are merely examples, and various modifications and variations can be made based on the knowledge of those skilled in the art, including the embodiments described in the disclosure section of the invention. It is possible to carry out the invention in other forms with modifications.

Further, the above-mentioned "section, module, unit" can be read as "means", "circuit", etc. For example, the acquisition unit can be read as an acquisition means or an acquisition circuit.

1 Water level estimation system 10 Receiver 100 Information processing device 120 Storage unit 121 Learning data storage unit 122 Altitude data storage unit 123 Estimation result storage unit 130 Control unit 131 Collection unit 132 Generation unit 133 Acquisition unit 134 Adjustment unit 135 Calculation unit 136 Estimation unit

Claims (11)

A first receiver installed on the water surface of the water level measurement target, and a second receiver installed on the water surface of the water level measurement target as a second receiver installed at a predetermined location, or A water level estimation system including a second receiver installed on land within a predetermined area including the water level measurement target , and an information processing device,
The information processing device includes:
First altitude data that is altitude data obtained from a satellite signal received by the first receiver; and second altitude data that is altitude data obtained from a satellite signal received by the second receiver. an acquisition unit that acquires and;
a calculation unit that calculates a correction coefficient based on predetermined information according to a situation in which the first receiver or the second receiver is installed;
and an estimation unit that estimates the water level of the water level measurement target based on corrected data in which the first altitude data and the second altitude data are corrected using the correction coefficient. Water level estimation system. The calculation unit includes a model that has learned a relationship between a correct coefficient determined according to altitude data obtained from a predetermined receiver and characteristic information according to a situation in which the predetermined receiver is installed; The water level estimation system according to claim 1, wherein the correction coefficient is calculated based on the predetermined information. The information processing device includes:
3. The model further includes a generation unit that generates a prediction model that outputs information used to calculate the correction coefficient when the predetermined information is input as feature information. water level estimation system. The generation unit includes, as the predetermined receiver, a set of altitude data obtained from each of a plurality of receivers corresponding to a set of the first receiver and the second receiver; The model uses a set of correct coefficients determined based on the water level in the installation environment where the receiver is installed and characteristic information according to the situation in which at least one of the plurality of receivers is installed as training data. The water level estimation system according to claim 3, wherein the prediction model is generated by learning the prediction model. The calculation unit calculates at least one of predetermined information according to the situation in which the first receiver is installed, or predetermined information according to the situation in which the second receiver is installed. The water level estimation system according to any one of claims 2 to 4, wherein the correction coefficient is calculated based on information outputted by inputting it into the model as characteristic information. The characteristic information includes the date and time when the predetermined receiver observed the altitude data, location information indicating the location where the predetermined receiver is installed, and the location of the artificial satellite that is the source of the satellite signal and the receiver. Any one of claims 2 to 5, characterized in that the information is information according to a relationship, sensor information detected by a sensor included in the receiver, or quality status of a communication line corresponding to the receiver. The water level estimation system described in one. The second receiver is installed on land in a predetermined area including the water level measurement target, and the second receiver is installed on land in an area where a distance relationship with the first receiver satisfies a predetermined condition.
7. The acquisition unit acquires, as the second altitude data, altitude data obtained from a satellite signal received by the second receiver installed on land. The water level estimation system described in (1) above. The second receiver installed on the water surface of the water level measurement target is installed on the water surface in an area where a distance relationship with the first receiver satisfies a predetermined condition,
The acquisition unit acquires, as the second altitude data, altitude data obtained from a satellite signal received by the second receiver installed above the water surface. Water level estimation system according to any one of the above. The information processing device includes:
The altitude data obtained from the satellite signal received by the second receiver installed on land is acquired as correction data for correcting unevenness of the altitude data, and using the acquired correction data, The water level estimation system according to claim 7 or 8, further comprising an adjustment unit that adjusts the first altitude data and the second altitude data. A first receiver installed on the water surface of the water level measurement target, and a second receiver installed on the water surface of the water level measurement target as a second receiver installed at a predetermined location, or A water level estimation method executed by a water level estimation system including a second receiver installed on land within a predetermined area including the water level measurement target and an information processing device,
The information processing device
First altitude data that is altitude data obtained from a satellite signal received by the first receiver; and second altitude data that is altitude data obtained from a satellite signal received by the second receiver. an acquisition step of acquiring the
a calculation step of calculating a correction coefficient based on predetermined information according to a situation in which the first receiver or the second receiver is installed;
and an estimation step of estimating the water level of the water level measurement target based on corrected data in which the first altitude data and the second altitude data are corrected using the correction coefficient. Water level estimation method. First altitude data, which is altitude data obtained from a satellite signal received by a first receiver installed above the water surface of the water level measurement target, and a second receiver installed at a predetermined location , Altitude data obtained from a satellite signal received by a second receiver installed on the water surface of the water level measurement target or a second receiver installed on land within a predetermined area including the water level measurement target an acquisition unit that acquires second altitude data,
a calculation unit that calculates a correction coefficient based on predetermined information according to a situation in which the first receiver or the second receiver is installed;
and an estimation unit that estimates the water level of the water level measurement target based on corrected data in which the first altitude data and the second altitude data are corrected using the correction coefficient. Information processing device.

Images