Estimation device, estimation method, and recording medium

Abstract

To provide an estimation device and the like that reliably estimate a posture of a space object having a function which allows the ground to be imaged from the outer space.SOLUTION: In an estimation device, feature amount generation means generates a feature amount using observation data obtained by observing a space object having an imaging function which allows the ground to be imaged from the outer space. In response to an input of the feature amount, posture estimation means estimates a posture of the space object at the time the observation data is obtained, by applying an estimation model learned to output an estimated result that shows whether a posture of the space object corresponds to any of a plurality of postures.SELECTED DRAWING: Figure 3

Description

[Technical Field]

[0001] The present disclosure relates to techniques that can be used to estimate the attitude of a space object. [Background technology]

[0002] Techniques have been proposed that can be used to estimate the attitude of space objects such as satellites.

[0003] Specifically, for example, Patent Document 1 discloses a perspective for estimating the attitude stability of a space object resident in low Earth orbit using radar cross section statistics from various radars. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Special Publication No. 2022-522170 Summary of the Invention [Problem to be solved by the invention]

[0005] However, the technique disclosed in Patent Document 1 has a problem in that, for example, there are cases where the attitude of a space object capable of capturing images of the earth from outer space cannot be accurately estimated.

[0006] One object of the present disclosure is to provide an estimation device capable of reliably estimating the attitude of a space object that has the function of capturing images of the earth from outer space. [Means for solving the problem]

[0007] In one aspect of the present disclosure, an estimation device includes a feature generation means for generating features using observation data obtained by observing a space object having an imaging function capable of imaging the ground from outer space, and an attitude estimation means for estimating the attitude of the space object at the time the observation data was obtained using an estimation model trained to output an estimation result indicating which of a plurality of attitudes the attitude of the space object corresponds to in response to input of the features.

[0008] In another aspect of the present disclosure, an estimation method executed by a computer generates features using observation data obtained by observing a space object having an imaging function capable of imaging the Earth from outer space, and estimates the attitude of the space object at the time the observation data was obtained using an estimation model trained to output an estimation result indicating which of multiple attitudes the attitude of the space object corresponds to in response to the input of the features.

[0009] In yet another aspect of the present disclosure, a recording medium records a program that causes a computer to execute a process of generating features using observation data obtained by observing a space object having an imaging function capable of imaging the Earth from outer space, and estimating the attitude of the space object when the observation data was obtained using an estimation model that has been trained to generate an estimation result indicating which of multiple attitudes the attitude of the space object corresponds to in response to the input of the features. [Effects of the Invention]

[0010] According to the present disclosure, it is possible to reliably estimate the attitude of a space object capable of capturing images of the Earth from outer space. [Brief explanation of the drawings]

[0011] [Figure 1] FIG. 1 is a diagram showing a schematic configuration of an observation system including an information processing device according to the present disclosure. [Figure 2] FIG. 1 is a block diagram showing an example of a hardware configuration of an information processing device according to the present disclosure. [Figure 3]FIG. 1 is a block diagram showing an example of a functional configuration of an information processing device according to the present disclosure. [Figure 4] FIG. 2 is a diagram for explaining an example of the contents of observation data relating to a space object. [Figure 5] FIG. 10 is a diagram for explaining an example of observation data used in processing related to generation of feature amounts. [Figure 6] FIG. 10 is a diagram showing an example of processing related to generation of feature amounts. [Figure 7] 10 is a flowchart showing an example of processing performed in an information processing device according to the present disclosure. [Figure 8] FIG. 1 is a block diagram showing an example of a functional configuration of an estimation device according to the present disclosure. [Figure 9] 10 is a flowchart illustrating an example of processing performed by an estimation device according to the present disclosure. DETAILED DESCRIPTION OF THE INVENTION

[0012] Hereinafter, preferred embodiments of the present disclosure will be described with reference to the drawings.

[0013] First Embodiment [System Configuration] 1 is a diagram showing a schematic configuration of an observation system including an information processing device according to the present disclosure. As shown in FIG. 1, the observation system 1 includes a radar device 50 and an information processing device 100.

[0014] The radar device 50 is an observation facility installed on the ground and is configured to obtain observation data KD by observing a space object SB, such as an artificial satellite, in the sky. The radar device 50 acquires observation data KD including a plurality of pieces of observation information KJ indicating observation results related to the space object SB, and transmits the acquired observation data KD to the information processing device 100. Hereinafter, unless otherwise specified, the space object SB will be described as a satellite orbiting in low Earth orbit. Hereinafter, unless otherwise specified, the space object SB will be described as a satellite having an imaging function capable of capturing images of the Earth from space and a charging function capable of storing and utilizing power obtained from sunlight. The imaging function of the space object SB can be realized, for example, by a synthetic aperture radar. The charging function of the space object SB can be realized, for example, by a solar cell.

[0015] The information processing device 100 performs processing related to estimation of the attitude of the space object SB using the observation data KD received from the radar device 50. The information processing device 100 also has a function as an estimation device.

[0016] [Hardware configuration] 2 is a block diagram showing an example of a hardware configuration of an information processing device according to the present disclosure. As shown in FIG. 2, the information processing device 100 includes an interface (IF) 111, a processor 112, a memory 113, a recording medium 114, a database (DB) 115, a display device 116, and an input device 117.

[0017] The IF 111 inputs and outputs data to and from an external device, and receives, for example, observation data KD obtained by the radar device 50.

[0018] The processor 112 is a computer such as a CPU (Central Processing Unit), and executes a program prepared in advance to control the entire information processing device 100. The processor 112 performs, for example, processing related to estimating the attitude of the space object SB.

[0019] The memory 113 is configured by a ROM (Read Only Memory), a RAM (Random Access Memory), etc. The memory 113 is also used as a working memory while the processor 112 is executing various processes.

[0020] The recording medium 114 is a non-volatile, non-transitory recording medium such as a disk-shaped recording medium or a semiconductor memory, and is configured to be detachable from the information processing device 100. The recording medium 114 records various programs to be executed by the processor 112. When the information processing device 100 executes various processes, the programs recorded on the recording medium 114 are loaded into the memory 113 and executed by the processor 112.

[0021] The DB 115 stores, for example, data received by the IF 111, data obtained by processing by the processor 112, and the like.

[0022] The display device 116 has, for example, a liquid crystal display, etc. Furthermore, the display device 116 displays the estimated results of the attitude of the space object SB, etc., as necessary.

[0023] The input device 117 includes, for example, at least one of a keyboard, a mouse, a touch panel, etc. The input device 117 also issues instructions to the processor 112 in response to user operations.

[0024] [Function Configuration] 3 is a block diagram showing an example of the functional configuration of an information processing device according to the present disclosure. As shown in FIG. 3, the information processing device 100 includes a feature generating unit 11 and a posture estimating unit 12.

[0025] The feature generator 11 functions as a feature generator. The feature generator 11 generates feature values ​​FV using observation information KJ included in observation data KD obtained from the radar device 50, and outputs the generated feature values ​​FV to the attitude estimation unit 12. A specific example of the process of generating feature values ​​FV performed by the feature generator 11 will be described later.

[0026] The attitude estimation unit 12 functions as an attitude estimation means. Furthermore, the attitude estimation unit 12 uses an estimation model 12A that has been trained to output an estimation result ER indicating which of a plurality of attitudes the attitude of the space object SB corresponds to in response to the input of the feature FV, and estimates the attitude of the space object SB when the observation data KD is obtained. Furthermore, the attitude estimation unit 12 outputs the estimation result ER obtained by processing using the estimation model 12A to the outside. The estimation result ER may be stored in DB 115, for example, or may be displayed on a display device 116.

[0027] [Specific example] Next, specific examples of processing etc. performed in the present disclosure will be described.

[0028] The radar device 50 acquires observation data KD for each segment and sequentially transmits the acquired observation data KD to the information processing device 100. A segment is defined in advance as a period during which the radar device 50 acquires one piece of observation data KD. Specifically, a segment is defined as, for example, a period from when a space object SB enters the observation range of the radar device 50 to when the space object SB leaves the observation range of the radar device 50. Furthermore, the period corresponding to one segment can be set to a period depending on the observation performance of the radar device 50, such as 10 minutes or 1 hour.

[0029] FIG. 4 is a diagram illustrating an example of the contents of observation data related to a space object. The observation data KD includes multiple pieces of observation information KJ acquired in one segment. Each piece of observation information KJ included in the observation data KD includes, for example, five types of observation values ​​linked to an observation time KTM, as shown in FIG. 4. Specifically, the observation information KJ includes the five types of observation values, for example, a distance DST, a range rate DVR, an azimuth angle AZA, an elevation angle ELA, and a radar cross section RCS. The observation time KTM represents the time at which the five types of observation values ​​included in the observation data KD were obtained. The distance DST represents the distance from the radar device 50 to the space object SB. The range rate DVR represents the Doppler shift frequency value obtained in response to changes in the distance DST. The azimuth angle AZA represents the angle corresponding to the horizontal position of the space object SB, calculated clockwise from true north as 0 degrees. The elevation angle ELA represents the angle corresponding to the upward position of the space object SB, calculated vertically from the horizon as 0 degrees. The radar cross section RCS represents a value corresponding to the intensity of electromagnetic waves when the electromagnetic waves emitted from the radar device 50 are reflected by the space object SB. That is, each piece of observation information KJ included in the observation data KD includes multiple observation values ​​obtained during the period from when the space object SB enters the observation range of the radar device 50 to when the space object SB leaves the observation range.

[0030] The feature generator 11 uses the observation values ​​included in the observation information KJ of the observation data KD received from the radar device 50 to generate a predetermined number of types of feature values ​​FV for each observation time KTM included in the observation information KJ.

[0031] Here, a specific example of the process of generating feature quantities by the feature quantity generating unit 11 will be described. In this specific example, five types of feature quantities are generated as feature quantities for each observation time.

[0032] 5 is a diagram illustrating an example of observation data used in processing related to the generation of feature quantities. The feature quantity generator 11 acquires, for example, observation data KD1 as shown in FIG. 5 from the radar device 50 as observation data KD corresponding to one segment. The observation data KD1 includes observation information KJ1, observation information KJ2, and observation information KJ3 as information indicating time-series observation results related to the space object SB. The observation information KJ1 includes five types of observation values ​​(hereinafter also referred to as an observation value group KVG1) linked to an observation time KTM1: a distance DST1, a range rate DVR1, an azimuth angle AZA1, an elevation angle ELA1, and a radar cross section RCS1. Observation information KJ2 includes five types of observation values ​​(hereinafter also referred to as observation value group KVG2) linked to observation time KTM2, which is later than observation time KTM1: distance DST2, range rate DVR2, azimuth angle AZA2, elevation angle ELA2, and radar cross section RCS2. Observation information KJ3 includes five types of observation values ​​(hereinafter also referred to as observation value group KVG3) linked to observation time KTM3, which is later than observation time KTM2: distance DST3, range rate DVR3, azimuth angle AZA3, elevation angle ELA3, and radar cross section RCS3.

[0033] 6 is a diagram illustrating an example of a process for generating features. First, the feature generator 11 uses an observation value set KVG1 included in observation information KJ1 to generate five types of features (hereinafter also referred to as feature set FVG1) corresponding to an observation time KTM1 included in the observation information KJ1. Next, the feature generator 11 uses the observation value set KVG1 included in observation information KJ1 and an observation value set KVG2 included in observation information KJ2 to generate five types of features (hereinafter also referred to as feature set FVG2) corresponding to an observation time KTM2 included in the observation information KJ2 and which are the same type as the features included in the feature set FVG1. Next, the feature generation unit 11 uses the observation value group KVG1 contained in the observation information KJ1, the observation value group KVG2 contained in the observation information KJ2, and the observation value group KVG3 contained in the observation information KJ3 to generate five types of features FVA3, FVB3, FVC3, FVD3, and FVE3 (hereinafter also referred to as feature group FVG3) that correspond to the observation time KTM3 contained in the observation information KJ3 and are of the same type as each feature contained in the feature group FVG1.

[0034] According to the above-described process, the feature generator 11 can generate the 15 feature values ​​included in the feature groups FVG1, FVG2, and FVG3 as the feature FV1 corresponding to the observation data KD1 (see FIG. 6). According to the above-described process, the feature generator 11 can generate feature values ​​using a plurality of observation values ​​associated with a plurality of observation times of the observation data KD1. According to the above-described process, the feature generator 11 can generate a feature corresponding to a given observation time included in a piece of observation information in the observation data KD1 using the observation values ​​obtained before the given observation time in the observation data KD1.

[0035] According to this specific example, the feature generator 11 applies the above-described process for generating the feature FV1 to generate a predetermined number of types of feature for each type of observation value included in the observation data KD. In this case, the feature generator 11 can generate, for example, five types of feature corresponding to each of the five types of observation values: distance DST, range rate DVR, azimuth angle AZA, elevation angle ELA, and radar cross section RCS.

[0036] The attitude estimation unit 12 inputs the feature FV generated in a manner similar to the feature FV1 into the estimation model 12A, and obtains an estimation result ER indicating which of multiple attitudes the attitude of the space object SB corresponds to.

[0037] The multiple attitudes may include at least two attitudes: an "imaging state" and a "charging state." That is, the estimation model 12A may be trained to output an estimation result ER indicating which of the multiple attitudes, including the "imaging state" and the "charging state," the attitude of the space object SB corresponds to, in response to the input of the feature FV. The "imaging state" in this specific example may be rephrased as, for example, an "attitude in which the space object is imaging the ground" or a "state in which the space object is in an attitude in which it can image the ground." Furthermore, the "charging state" in this specific example may be rephrased as, for example, an "attitude in which the space object is not imaging the ground" or a "state in which the space object is in a charging attitude." Furthermore, the multiple attitudes may include attitudes different from both the "imaging state" and the "charging state," such as a "fuel injection state" and / or a "collision avoidance state." Furthermore, when training the estimation model 12A, for example, training data can be used in which a label indicating one of the multiple postures mentioned above is assigned to a feature of the same type as the feature FV obtained from an observation value of the same type as the observation value contained in the observation data KD.

[0038] According to the processing described above, the attitude estimation unit 12 can obtain an estimation result ER that can identify which of multiple attitudes, including the ``imaging state'' and the ``charging state,'' the attitude of the space object SB in the segment in which the observation data KD was obtained corresponds to.

[0039] According to the present disclosure, the space object SB may be a satellite orbiting the Earth in a geostationary orbit, as long as it has an imaging function. In such a case, for example, the attitude estimation unit 12 inputs a feature quantity generated from observation data for a predetermined time corresponding to one segment into the estimation model 12A, thereby obtaining an estimation result of the attitude of the space object SB in that one segment.

[0040] According to the present disclosure, it is sufficient that the observation data KD includes at least the radar cross section RCS as an observation value associated with each of the multiple observation times KTM. In such a case, the feature generator 11 can generate the feature FV using the radar cross section RCS included in the observation data KD.

[0041] According to the present disclosure, the observation data KD can further include, in addition to the radar cross section RCS, at least one of the range DST, the range rate DVR, the azimuth angle AZA, and the elevation angle ELA as observation values ​​associated with each of the multiple observation times KTM. In such a case, the feature generator 11 can generate the feature FV using the radar cross section RCS included in the observation data KD and at least one of the range DST, the range rate DVR, the azimuth angle AZA, and the elevation angle ELA included in the observation data KD.

[0042] According to the present disclosure, the feature generator 11 may generate multiple types of feature for each observation time KTM included in the observation data KD, or may generate one type of feature.

[0043] According to the present disclosure, the attitude estimation unit 12 may acquire an estimation result ER relating to the attitude of the space object SB by inputting, into the estimation model 12A, five types of observation values ​​used to generate the feature FV in addition to the feature FV generated by the feature generation unit 11. In other words, the estimation model 12A may be trained to output an estimation result ER indicating which of a plurality of attitudes the attitude of the space object SB corresponds to in response to input of the feature FV generated using the observation data KD and the observation values ​​included in the observation data KD that were used to generate the feature FV.

[0044] [Processing flow] Next, a description will be given of the flow of processing performed in the information processing device 100. Fig. 7 is a flowchart showing an example of processing performed in the information processing device according to the present disclosure.

[0045] First, the information processing device 100 receives observation data KD including a plurality of pieces of observation information KJ indicating observation results relating to the space object SB from the radar device 50 (step S11).

[0046] Next, the information processing device 100 uses the observation values ​​included in the observation information KJ of the observation data KD obtained in step S11 to generate a predetermined number of types of feature values ​​FV for each observation time KTM included in the observation information KJ (step S12).

[0047] Next, the information processing device 100 estimates the attitude of the space object SB when the observation data KD was obtained in step S11 by inputting the feature value FV obtained in step S12 into the estimation model 12A (step S13). The estimation result obtained in step S13 only needs to include information that can identify which of multiple attitudes, including the "imaging state" and the "charging state," the attitude of the space object SB in the segment corresponding to the observation data KD obtained in step S11 corresponds to.

[0048] As described above, according to this embodiment, by generating feature values ​​FV from observation values ​​included in the observation data KD of the space object SB and inputting the generated feature values ​​FV into the trained estimation model 12A, it is possible to obtain an estimation result ER indicating which of a plurality of attitudes, including an "imaging state" and a "charging state," the attitude of the space object SB when the observation data KD was obtained corresponds to. Therefore, according to this embodiment, it is possible to reliably estimate the attitude of a space object capable of imaging the ground from outer space.

[0049] Second Embodiment FIG. 8 is a block diagram illustrating an example of a functional configuration of an estimation device according to the present disclosure.

[0050] The estimation device 500 has the same hardware configuration as the information processing device 100. The estimation device 500 also has a feature generating unit 511 and a posture estimating unit 512.

[0051] The feature generating means 511 can be realized, for example, by using the function of the feature generating unit 11. Furthermore, the posture estimating means 512 can be realized, for example, by using the function of the posture estimating unit 12.

[0052] FIG. 9 is a flowchart illustrating an example of processing performed by the estimation device according to the present disclosure.

[0053] The feature generating means 511 generates feature data using observation data obtained by observing a space object having an imaging function capable of imaging the earth from space (step S51).

[0054] The attitude estimation means 512 uses an estimation model that has been trained to output an estimation result indicating which of multiple attitudes the attitude of the space object corresponds to in response to the input of features, and estimates the attitude of the space object when the observation data was obtained (step S52).

[0055] According to this embodiment, it is possible to reliably estimate the attitude of a space object that has the function of capturing images of the earth from outer space.

[0056] A part or all of the above-described embodiments can be described as, but not limited to, the following supplementary notes.

[0057] (Appendix 1) a feature generating means for generating feature values ​​using observation data obtained by observing a space object having an imaging function capable of imaging the ground from outer space; an attitude estimation means for estimating the attitude of the space object when the observation data was obtained, using an estimation model trained to output an estimation result indicating which of a plurality of attitudes the attitude of the space object corresponds to in response to input of the feature quantity; and An estimation device having:

[0058] (Appendix 2) the space object further has a charging function capable of storing and utilizing power obtained from sunlight; The estimation device of Appendix 1, wherein the estimation model is trained to output an estimation result indicating which of the multiple attitudes, including an imaging state and a charging state, the attitude of the space object corresponds to in response to the input of the feature.

[0059] (Appendix 3) the space object is a satellite orbiting in low Earth orbit; 2. The estimation device of claim 1, wherein the observation data includes at least a value corresponding to the intensity of electromagnetic waves when the electromagnetic waves irradiated from the observation facility are reflected by the space object, as an observation value obtained during the period from the time the space object enters the observation range of an observation facility located on the ground to the time the space object leaves the observation range.

[0060] (Appendix 4) 4. The estimation device of claim 3, wherein the observation values ​​further include at least one of a distance from the observation facility to the space object, a distance change rate obtained in response to a change in the distance, an angle corresponding to the horizontal position of the space object, and an angle corresponding to the upward position of the space object.

[0061] (Appendix 5) 2. The estimation device according to claim 1, wherein the feature generating means generates the feature using a plurality of observation values ​​associated with a plurality of observation times of the observation data.

[0062] (Appendix 6) 6. The estimation device of claim 5, wherein the feature generation means generates the feature corresponding to a given observation time included in the observation data by using each observation value obtained before the given observation time in the observation data.

[0063] (Appendix 7) 6. The estimation device of claim 5, wherein the plurality of observation values ​​include at least a value corresponding to the intensity of electromagnetic waves when the electromagnetic waves are irradiated from an observation facility on the ground and reflected by the space object.

[0064] (Appendix 8) the estimation model is trained to output the estimation result in response to inputs of the feature generated using the observation data and the observation value included in the observation data and used to generate the feature.

[0065] (Appendix 9) 1. A computer-implemented estimation method comprising: Generate features using observation data obtained by observing a space object having an imaging function capable of imaging the ground from space; An estimation method for estimating the attitude of a space object when the observation data was obtained, using an estimation model trained to output an estimation result indicating which of multiple attitudes the attitude of the space object corresponds to in response to the input of the feature.

[0066] (Appendix 10) Generate features using observation data obtained by observing a space object having an imaging function capable of imaging the ground from space; A recording medium having a program recorded thereon that causes a computer to execute a process of estimating the attitude of a space object when the observation data is obtained, using an estimation model that has been trained to output an estimation result indicating which of multiple attitudes the attitude of the space object corresponds to in response to the input of the feature.

[0067] Although the present disclosure has been described above with reference to the embodiments, the present disclosure is not limited to the above-described embodiments and examples. Various modifications that can be understood by those skilled in the art can be made to the configuration and details of the present disclosure within the scope of the present disclosure. Furthermore, each embodiment can be combined with other embodiments as appropriate. [Explanation of symbols]

[0068] 11 Feature generation unit 12 Posture estimation section 12A Estimation Model 50 Radar equipment 100 Information processing device

Claims

1. a feature generating means for generating feature values ​​using observation data obtained by observing a space object having an imaging function capable of imaging the ground from outer space; an attitude estimation means for estimating the attitude of the space object when the observation data was obtained, using an estimation model trained to output an estimation result indicating which of a plurality of attitudes the attitude of the space object corresponds to in response to input of the feature quantity; and An estimation device having:

2. the space object further has a charging function capable of storing and utilizing power obtained from sunlight; The estimation device described in claim 1, wherein the estimation model is trained to output an estimation result indicating which of the multiple attitudes, including an imaging state and a charging state, the attitude of the space object corresponds to in response to the input of the feature.

3. the space object is a satellite orbiting in low Earth orbit; 2. The estimation device of claim 1, wherein the observation data includes at least a value corresponding to the intensity of electromagnetic waves when electromagnetic waves irradiated from the observation facility are reflected by the space object, as an observation value obtained during the period from the time the space object enters the observation range of an observation facility located on the ground to the time the space object leaves the observation range.

4. 4. The estimation device of claim 3, wherein the observation values ​​further include at least one of the distance from the observation facility to the space object, the distance change rate obtained in response to a change in the distance, an angle corresponding to the horizontal position of the space object, and an angle corresponding to the upward position of the space object.

5. The estimation device according to claim 1 , wherein the feature generating means generates the feature using a plurality of observation values ​​associated with a plurality of observation times of the observation data.

6. 6. The estimation device according to claim 5, wherein the feature generating means generates the feature corresponding to a given observation time included in the observation data by using each observation value obtained before the given observation time in the observation data.

7. The estimation device according to claim 5, wherein the plurality of observation values ​​include at least a value corresponding to the intensity of electromagnetic waves when the electromagnetic waves are irradiated from an observation facility on the ground and reflected by the space object.

8. 2. The estimation device according to claim 1, wherein the estimation model is trained to output the estimation result in response to inputs of the feature generated using the observation data and the observation value included in the observation data and used to generate the feature.

9. 1. A computer-implemented estimation method comprising: Generate features using observation data obtained by observing a space object having an imaging function capable of imaging the ground from space; An estimation method for estimating the attitude of a space object when the observation data was obtained, using an estimation model trained to output an estimation result indicating which of multiple attitudes the attitude of the space object corresponds to in response to the input of the feature.

10. Generate features using observation data obtained by observing a space object having an imaging function capable of imaging the ground from space; A recording medium having a program recorded thereon that causes a computer to execute a process of estimating the attitude of a space object when the observation data is obtained, using an estimation model that has been trained to output an estimation result indicating which of multiple attitudes the attitude of the space object corresponds to in response to the input of the feature.

Images