Observation control device and observation control method

Abstract

An observation control device 4 comprises: a contribution degree calculation unit 41 that acquires prediction information and observation data from a simulation model 3 that predicts a prescribed phenomenon by assimilating the observation data observed by a plurality of observation apparatuses 2 included in an observation network 1, and calculates, on the basis of the prediction information and the observation data, a degree of contribution of each piece of the observation data to the prediction information for each of the observation apparatuses; a status calculation unit 42 that calculates, for each of the observation apparatuses, a status value indicating the state of the observation apparatus on the basis of apparatus information acquired from the plurality of observation apparatuses 2; and a control unit that controls the plurality of observation apparatuses on the basis of the contribution degree and the status value.

Description

Observation control device and observation control method

[0001] The present disclosure relates to an observation control device and an observation control method.

[0002] When predicting a phenomenon through simulation, it is common in various fields, such as meteorology and engineering, to calculate values ​​closer to reality by adding information from observation data to the initial or predicted values ​​in order to improve the accuracy of the prediction (Non-Patent Documents 1 and 2).

[0003] Obtaining observation data is costly, so research is being conducted to identify observation locations and items that will be effective in improving prediction accuracy. Methods are also being considered for actively using observation equipment to observe locations where it will be most effective.

[0004] Yamazaki et al., 2021, “EFSO at different geographical locations verified with observing-system experiments”Yamazaki et al., 2023, “Estimation of AMSU-A Radiance Observation Impacts in an LETKF-Based Atmospheric Global Data Assimilation System: Comparison with EFSO and Observing System Experiments”

[0005] Conventional observations estimate the effects of observations over a period of a few months at most, and do not assume that observation equipment will be operated over the long term. As a result, in actual operation, if a situation arises in which observations are not possible due to wear and tear on the observation equipment, the effect of improving the prediction accuracy of simulations through observations cannot be maintained.

[0006] The present disclosure has been made in consideration of the above circumstances, and an object of the present disclosure is to provide a technology for efficiently acquiring observation data while taking into account the state of observation equipment.

[0007] In order to achieve the above-mentioned object, one aspect of the present disclosure is an observation control device that acquires prediction information and the observation data from a simulation model that assimilates observation data observed by multiple observation devices included in an observation network to predict a predetermined phenomenon, and includes: a contribution calculation unit that calculates, for each observation device, the contribution of each observation data to the prediction information based on the prediction information and the observation data; a status calculation unit that calculates, for each observation device, a status value that indicates the status of the observation device based on device information acquired from the multiple observation devices; and a control unit that controls the observation devices based on the contribution and the status value.

[0008] One aspect of the present disclosure is an observation control method performed by an observation control device, which obtains prediction information and the observation data from a simulation model that assimilates observation data observed by multiple observation devices included in an observation network to predict a specified phenomenon, calculates the contribution of each observation data to the prediction information for each observation device based on the prediction information and the observation data, calculates a status value indicating the state of each observation device for each observation device based on device information obtained from the multiple observation devices, and controls the observation devices based on the contribution and the status value.

[0009] According to the present disclosure, it is possible to provide a technology for efficiently acquiring observation data while taking into account the state of observation equipment.

[0010] Fig. 1 is a diagram showing the overall configuration of an example of an observation system according to this embodiment. Fig. 2 is a diagram showing the distribution of multiple observation devices before control. Fig. 3 is a diagram showing the distribution of multiple observation devices after control. Fig. 4 is an example of a hardware configuration.

[0011] Hereinafter, embodiments of the present disclosure will be described with reference to the drawings.

[0012] 1 is a diagram showing the overall configuration of an example of an observation system according to this embodiment. The observation system shown in the figure includes an observation network 1, a simulation model 3, and an observation control device 4.

[0013] The simulation model 3 is a model that simulates and predicts some phenomenon. Here, a simulation model 3 corresponding to the phenomenon to be predicted is prepared, and phenomenon information 5 related to the phenomenon to be predicted is input to the simulation model 3 (S1). For example, when predicting a typhoon, a simulation model SCALE is prepared. A user inputs phenomenon information related to the phenomenon to be predicted into the simulation model 3 using an input device such as a PC.

[0014] The simulation model 3 of this embodiment predicts a predetermined phenomenon by assimilating observation data observed by a plurality of observation devices 2 included in the observation network 1. Specifically, the simulation model 3 corrects the simulation prediction value using the observation data input from the observation devices 2 through data assimilation. This allows the simulation model 3 to make highly accurate predictions of phenomena. For example, SCALE-LETKF may be used for data assimilation. The simulation model 3 then outputs the prediction information and the observation data transmitted from the observation devices 2 to the observation control device 4 (S4).

[0015] The observation network 1 includes multiple observation devices 2. The observation devices 2 observe various meteorological elements, such as temperature and wind. The observation devices 2 may also observe phenomena and conditions other than meteorological elements. The observation devices 2 of this embodiment may be mobile devices equipped with transportation means. For example, the observation devices 2 may be unmanned aerial vehicles (UAVs) equipped with at least one observation device (sensor, detector, etc.).

[0016] Each observation device 2 inputs the observed observation data into the simulation model 3 (S2). The observation data may be, for example, meteorological data such as atmospheric pressure, temperature, wind direction, and wind speed. The observation data may be, for example, data in which atmospheric observation data (air pressure, temperature, wind direction, wind speed, etc.) observed by the observation devices 2 (UAV1, UAV2, UAV3, etc.) is associated with a device ID (identification information) for identifying the observation device 2.

[0017] Each observation device 2 acquires device information indicating its own status and transmits the acquired device information together with its own device ID (identification information) to the observation control device 4 (S3). The device information includes, for example, the remaining battery level of the observation device 2 and the responsiveness of the sensors. The device information will be described later.

[0018] The observation control device 4 evaluates and controls each observation device 2 based on the device information acquired from each observation device 2 and the prediction information and observation data acquired from the simulation model 3. Specifically, the observation control device 4 generates control information for controlling each observation device 2 and transmits it to each corresponding observation device 2 (S5).

[0019] The illustrated observation control device 4 includes a score calculation unit 41 (contribution calculation unit), a status calculation unit 42, and a control unit 43.

[0020] The score calculation unit 41 acquires the prediction information and the observation data from the simulation model 3 (S4), and calculates the contribution of each observation data to the prediction information for each observation instrument based on the prediction information and the observation data. The prediction information includes a predicted value predicted by data assimilation using all the observation data. The prediction information also includes a correction amount obtained by correcting a predicted value predicted without data assimilation to a predicted value obtained by data assimilation using all the observation data.

[0021] Specifically, the score calculation unit 41 breaks down the amount of correction brought about by the entire observation data from the predicted value before data assimilation to the predicted value after data assimilation into the contribution (amount of correction) of each observation data, and aggregates them for each observation device 2. Then, the score calculation unit 41 calculates a score for each observation device 2 by normalizing the contribution of each observation data.

[0022] The score calculation unit 41 can calculate the contribution of each piece of observation data to the forecast using, for example, EFSO (Ensemble-based Forecast Sensitivity to Observations). EFSO is a method for diagnosing the impact (contribution) of each piece of observation data on the forecast. Using EFSO, the score calculation unit 41 normalizes the amount of error reduction of each piece of observation data as its respective contribution and generates a score. This allows the score calculation unit 41 to quantify and express, as a score (contribution), how much each piece of observation data contributed to improving the forecast accuracy of the simulation model 3. Note that in this embodiment, a normalized score is used as the contribution, but the contribution does not have to be normalized. Any index may be used as the contribution as long as it indicates how much each piece of observation data contributed to improving the forecast accuracy.

[0023] The status calculation unit 42 calculates a status value indicating the state of each observation device based on device information acquired from multiple observation devices 2. Specifically, the status calculation unit 42 calculates a status value that quantifies the continuity (continuous operability) of each observation device 2.

[0024] The equipment information may be, for example, data in which various status information (remaining battery level, sensor responsiveness, etc.) indicating the status of each observation equipment 2 (UAV1, UAV2, UAV3, etc.) is associated with an equipment ID.

[0025] UAV1_remaining battery capacity, UAV1_sensor responsiveness UAV2_remaining battery capacity, UAV2_sensor responsiveness UAV3_remaining battery capacity, UAV3_sensor responsiveness The status calculation unit 42 may calculate the status values, for example, using the following method. That is, the status calculation unit 42 defines the remaining battery capacity and sensor responsiveness as follows. Then, the status calculation unit 42 converts the acquired remaining battery capacity into a value between "0" and "4" and converts the sensor responsiveness into a value between "0" and "3."

[0026] [Battery Remaining] 0-20%: "0", 20-40%: "1", 40-60%: "2", 60-80%: "3", 80-100%: "4" [Sensor Responsiveness] No response: "0", Unstable response: "1", Temporarily unstable response: "2", Stable response: "3" To obtain the sensor responsiveness, the status calculation unit 42 may query the sensor of the observation device 2 a predetermined number of times and determine the response status as "No response," "Unstable response," "Temporarily unstable response," or "Stable response." The status calculation unit 42 may then multiply the two item values ​​for each observation device 2 and divide the result by the maximum of the two item values, "12" (4 x 3), to obtain a normalized status value. Here, the device information includes two items: battery remaining and sensor responsiveness. However, this is not limiting. Other items may also be used as the device information, and the number of items may be one, three, or more.

[0027] The control unit 43 controls the observation device 2 based on the score (degree of contribution) and the status value. Specifically, the control unit 43 may control the observation device 2 so that the observation device 2 whose score is smaller than a first predetermined value and whose status value is equal to or greater than a second predetermined value observes the observation device 2 whose score is equal to or greater than the second predetermined value and whose status value is smaller than the second predetermined value. The control unit 43 may remove the observation device whose status value is lower than the second predetermined value from the observation network 1. The first predetermined value and the second predetermined value may be the same value or different values.

[0028] 2 is a diagram showing the distribution of observation devices 2 in the observation network 1 before control. In the graph shown, the vertical axis represents the score and the horizontal axis represents the status value. Each observation device 2 is mapped to the graph shown using the normalized score and status value. Before control by the control unit 43, multiple observation devices 2 are distributed throughout the entire area 6A. Area 6A is divided into areas 61 to 64.

[0029] The observation devices 2 located in the area 61 are observation devices 2 whose scores and statuses are smaller than a predetermined value (for example, 0.2). In other words, the observation devices 2 in the area 61 are observation devices 2 whose contribution to prediction is small and whose equipment condition is poor.

[0030] The observation devices 2 located in the area 62 are observation devices 2 whose scores are equal to or greater than a predetermined value but whose statuses are smaller than the predetermined value. In other words, the observation devices 2 in the area 62 are observation devices 2 whose contribution to prediction is relatively large but whose equipment condition is poor.

[0031] The observation devices 2 located in the area 63 have scores smaller than a predetermined value but have statuses equal to or greater than a predetermined value. In other words, the observation devices 2 in the area 63 have a small contribution to prediction, but are in relatively good condition.

[0032] The observation devices 2 located in the area 64 have scores and statuses equal to or greater than predetermined values. In other words, the observation devices 2 in the area 64 have a relatively large contribution to prediction and are in relatively good condition.

[0033] Since the status value of the observation equipment 2 located in areas 61 and 62 is small (the equipment is in poor condition), the control unit 43 transmits control information (control signal) to the observation equipment 2 to cause it to leave the observation network 1 and return to a specific location (for example, a base where maintenance is possible). Upon receiving this control information, the observation equipment 2 stops observation and moves to the specified location in accordance with the control information.

[0034] The observation device 2 located in area 63 has a small contribution but a relatively large status value and is not in bad condition, so the control unit 43 sends a control signal to that observation device 2 to control it to perform observation in place of the observation device 2 located in area 62. In other words, the control unit 43 causes the observation device 2 located in area 62 to leave the observation network 1, and instead moves an observation device 2 located in area 63 that is in good condition to the observation location of the left observation device 2, replacing it and continuing the observation of the left observation device 2. As a result, the observation device 2 in area 63 takes over the observation location and observation items that were previously handled by the left observation device 2 in area 62.

[0035] The control unit 43 may prioritize the observation device 2 in the area 62 with the higher score and replace it with the observation device 2 in the area 63. The control unit 43 may transmit control information to each observation device 2 using satellite communication.

[0036] 3 is a diagram showing the distribution of observation devices 2 after control in the observation network 1. In the graph shown, the vertical axis represents the score and the horizontal axis represents the status value, similar to FIG. 2. Each observation device 2 is mapped to the graph shown using the normalized score and status value.

[0037] By controlling (removing or replacing) the observation devices 2 by the control unit 43, the observation devices 2 are distributed only in the area 6B. That is, an observation device 2 whose status value is less than a predetermined value (poor condition) will no longer exist by leaving the observation network 1. Also, an observation device 2 whose status value is equal to or greater than a predetermined value (relatively good condition) but whose score is less than a predetermined value (low contribution) will no longer exist by being replaced with an observation device 2 whose status value is less than the predetermined value (poor condition) and whose score is equal to or greater than a predetermined value (relatively high contribution).

[0038] The observation control device 4 of this embodiment described above is equipped with a score calculation unit 41 that acquires prediction information and observation data from a simulation model 3 that assimilates observation data observed by multiple observation devices 2 included in an observation network 1 to predict a specified phenomenon, and calculates the contribution of each observation data to the prediction information for each observation device based on the prediction information and the observation data, a status calculation unit 42 that calculates a status value indicating the state of each observation device 2 for each observation device based on device information acquired from the multiple observation devices 2, and a control unit 43 that controls the observation devices 2 based on the contribution and the status value.

[0039] The observation control method performed by the observation control device 4 of this embodiment acquires prediction information and observation data from a simulation model 3 that assimilates observation data observed by multiple observation devices 2 included in an observation network 1 to predict a specified phenomenon, calculates the contribution of each observation data to the prediction information for each observation device based on the prediction information and the observation data, calculates a status value indicating the state of each observation device for each observation device based on device information acquired from the multiple observation devices 2, and controls the observation devices based on the contribution and the status value.

[0040] As a result, in this embodiment, it is possible to acquire efficient observation data while taking into account the state of the observation device 2. Specifically, it is possible to take into account wear and tear on the observation device 2 by introducing an index called a status value that takes into account the possibility of continuing observation by the observation device 2. Then, it is possible to maintain the effect of improving prediction accuracy through observation by taking into account the mobility and wear and tear of the observation device 2, such as by replacing worn observation device 2 with another observation device.

[0041] By introducing the status value, it is possible to replace "an observation device 2 in good condition but with little contribution to the prediction" with "an observation device 2 in poor condition that contributes greatly to the prediction." In other words, by replacing observation devices 2 in poor condition with observation devices 2 in good condition for observations that contribute greatly to the prediction, it is possible to maintain the effect of improving the prediction accuracy of the entire observation network 1.

[0042] By considering the contribution of the observation equipment 2 to the prediction and the state of the equipment in real time, it is possible to efficiently maintain and power the observation equipment, and to deal with wear and tear on the observation equipment 2.

[0043] The observation control device 4 described above can use, for example, a general-purpose computer system such as that shown in Fig. 4. The computer system shown in the figure includes a CPU (Central Processing Unit, processor) 901, memory 902, storage 903 (HDD: Hard Disk Drive, SSD: Solid State Drive), a communication device 904, an input device 905, and an output device 906. The memory 902 and storage 903 are storage devices. In this computer system, the CPU 901 executes a predetermined program loaded into the memory 902, thereby realizing each function of the observation control device 4.

[0044] The observation control device 4 may be implemented on a single computer or multiple computers. The observation control device 4 may also be a virtual machine implemented on a computer. The program for the observation control device 4 can be stored on a computer-readable recording medium such as a HDD, SSD, USB (Universal Serial Bus) memory, CD (Compact Disc), or DVD (Digital Versatile Disc), or can be distributed via a network. The computer-readable recording medium is, for example, a non-transitory recording medium.

[0045] The present disclosure is not limited to the above-described embodiments, and various modifications are possible within the scope of the present disclosure.

[0046] 1: Observation network 2: Observation equipment 3: Simulation model 4: Observation control device 41: Score calculation unit (contribution calculation unit) 42: Status calculation unit 43: Control unit 5: Phenomenon information

Claims

1. An observation control device comprising: a contribution calculation unit that acquires prediction information and observation data from a simulation model that assimilates observation data observed by multiple observation devices included in an observation network to predict a specified phenomenon, and calculates the contribution of each observation data to the prediction information for each observation device based on the prediction information and the observation data; a status calculation unit that calculates a status value indicating the status of each observation device for each observation device based on device information acquired from the multiple observation devices; and a control unit that controls the observation devices based on the contribution and the status value.

2. The observation control device described in claim 1, wherein the control unit controls the observation equipment whose contribution rate is less than a first predetermined value and whose status value is greater than or equal to a second predetermined value to observe the observation equipment whose contribution rate is greater than or equal to the first predetermined value and whose status value is less than the second predetermined value.

3. The observation control device according to claim 1, wherein the control unit causes observation equipment whose status value is lower than a predetermined value to be removed from the observation network.

4. An observation control method performed by an observation control device, comprising: acquiring prediction information and the observation data from a simulation model that assimilates observation data observed by multiple observation instruments included in an observation network to predict a specified phenomenon; calculating, for each observation instrument, the contribution of each observation data to the prediction information based on the prediction information and the observation data; calculating, for each observation instrument, a status value indicating the state of the observation instrument based on instrument information acquired from the multiple observation instruments; and controlling the observation instruments based on the contribution and the status value.

Images