WO2023181354A1 - 情報処理装置、算出方法及び記憶媒体 - Google Patents
情報処理装置、算出方法及び記憶媒体 Download PDFInfo
- Publication number
- WO2023181354A1 WO2023181354A1 PCT/JP2022/014405 JP2022014405W WO2023181354A1 WO 2023181354 A1 WO2023181354 A1 WO 2023181354A1 JP 2022014405 W JP2022014405 W JP 2022014405W WO 2023181354 A1 WO2023181354 A1 WO 2023181354A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- abnormality
- data
- information processing
- degree
- time series
- Prior art date
Links
Images
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B64—AIRCRAFT; AVIATION; COSMONAUTICS
- B64G—COSMONAUTICS; VEHICLES OR EQUIPMENT THEREFOR
- B64G3/00—Observing or tracking cosmonautic vehicles
Definitions
- the present disclosure relates to the technical field of an information processing device, a calculation method, and a storage medium that perform processing related to detecting changes in space objects.
- Patent Document 1 the luminosity information of a target object, which is an artificial object orbiting the earth, is compared with the luminosity information of a reference star, and after correcting the luminosity information of the target object, the luminosity information of the target object is A monitoring system is disclosed that detects an abnormality in the posture or shape of a target object based on periodicity.
- one of the main objectives of the present disclosure is to provide an information processing device, a calculation method, and a storage medium that can perform processing to suitably detect changes in space objects. .
- One aspect of the information processing apparatus is an information processing apparatus, the information processing apparatus comprising: a data acquisition means for acquiring time series data representing the luminosity of a space object; and a degree of abnormality regarding the state of the space object based on the time series data.
- An information processing apparatus includes an abnormality degree calculation means.
- One aspect of the calculation method is a calculation method in which a computer acquires time series data representing the luminosity of a space object, and calculates the degree of abnormality regarding the state of the space object based on the time series data.
- One aspect of the storage medium is storage that stores a program that causes a computer to execute a process of acquiring time series data representing the luminosity of a space object and calculating an abnormality degree regarding the state of the space object based on the time series data. It is a medium.
- FIG. 1 shows a configuration of an observation system according to a first embodiment. This is an example of the data structure of observation data. An example of a block configuration of an information processing device is shown. This is an example of a functional block related to abnormality detection related processing. It is an example of the functional block of the abnormality degree calculation part when the abnormality degree calculation model is an autoencoder. It is an example of the table of the abnormality degree based on a 1st example of output.
- FIG. 3 is a diagram illustrating an overview of learning processing of an abnormality degree calculation model. It is an example of a flowchart of abnormality detection related processing. It is an example of the functional block regarding the abnormality detection related process based on a modification.
- FIG. 2 is a block diagram of an information processing device in a second embodiment. It is an example of the flowchart which shows the processing procedure in 2nd Embodiment.
- FIG. 1 shows the configuration of an observation system 100 according to the first embodiment.
- the observation system 100 mainly includes an optical observation device 1, an information processing device 2, and a storage device 4.
- the optical observation device 1 is installed on the ground and optically observes a space object 5 such as a satellite, which is an observation target located in the sky.
- the optical observation device 1 then supplies observation data “Da” indicating the observation results regarding the space object 5 to the information processing device 2.
- the space object 5 is not limited to a satellite, and may be any space object such as space debris.
- FIG. 2 is an example of the data structure of observation data Da.
- the observation data Da mainly includes information indicating observation time, luminous intensity, and sensor name.
- the "observation time” indicates the observation time (date and time) at which the corresponding luminosity was observed, and functions as a timestamp.
- Luminosity indicates the luminosity (brightness) of the observed space object 5.
- “Sensor name” indicates the name or identification information (ID) of the optical observation device 1 or a sensor included in the optical observation device 1 that observes the luminous intensity. Note that whether or not the space object 5 can be observed depends on the weather, etc., so there are times when the space object 5 cannot be observed. Therefore, the observation data Da generated by the optical observation device 1 becomes temporally discontinuous time-series data (that is, data in which observation intervals are not necessarily constant).
- the information processing device 2 performs processing related to abnormality detection of the space object 5 (also referred to as "anomaly detection related processing") based on the temporal change in luminosity (so-called light curve) indicated by the time-series observation data Da supplied from the optical observation device 1. call).
- the abnormality of the space object 5 corresponds to, for example, a change in attitude or shape of the space object 5.
- the information processing device 2 calculates the degree of abnormality indicating the degree of abnormality of the space object 5 (or the probability of occurrence of the abnormality).
- the storage device 4 is a memory that stores various information necessary for abnormality detection related processing by the information processing device 2.
- the storage device 4 stores observation data DB 41, parameter information 42, and training data 43.
- the observation data DB 41 is a database of observation data Da supplied from the optical observation device 1 to the information processing device 2.
- the information processing device 2 receives the observation data Da from the optical observation device 1, it adds a record corresponding to the received observation data Da to the observation data DB 41.
- the observation data DB 41 may further include information indicating a processing result of the information processing device 2, such as the degree of abnormality calculated by the information processing device 2.
- the parameter information 42 indicates the parameters of the model used to calculate the abnormality degree (also referred to as the "abnormality degree calculation model").
- the abnormality degree calculation model may be, for example, a learning model based on machine learning, a learning model based on a neural network, or another type of learning model such as a support vector machine, or a combination of these. There may be.
- an autoencoder used for anomaly detection is used as an anomaly degree calculation model.
- the abnormality degree calculation model is a neural network that uses time-series data indicating the luminosity of the space object 5 in a normal state as input data and is trained to output data that reproduces the input data.
- the parameter information 42 is parameter information indicating the weight of the network of the autoencoder, etc., and is obtained by learning the autoencoder using the training data 43.
- the training data 43 is training data used for learning the abnormality degree calculation model.
- the abnormality degree calculation model is an autoencoder
- time-series data representing the luminosity of the space object 5 in a normal state is stored in the storage device 4 as the training data 43.
- the storage device 4 may be an external storage device such as a hard disk connected to or built in the information processing device 2, or may be a storage medium such as a flash memory that is detachable from the information processing device 2. . Furthermore, the storage device 4 may be composed of one or more server devices that perform data communication with the information processing device 2. Further, the database and the like stored in the storage device 4 may be distributed and stored in a plurality of devices or storage media.
- the configuration of the observation system 100 shown in FIG. 1 is an example, and various changes may be made to the configuration.
- the optical observation device 1 and the information processing device 2 may be configured as one unit.
- the information processing device 2 and the storage device 4 may be configured as one unit.
- the information processing device 2 may be composed of a plurality of devices. In this case, the plurality of devices constituting the information processing device 2 exchange information necessary for executing pre-assigned processing between these devices. In this case, the information processing device 2 functions as an information processing system.
- FIG. 3 shows an example of a block configuration of the information processing device 2.
- the information processing device 2 includes a processor 21, a memory 22, and an interface 23 as hardware.
- Processor 21, memory 22, and interface 23 are connected via data bus 29.
- the processor 21 executes a predetermined process by executing a program stored in the memory 22.
- the processor 21 is a processor such as a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), or a TPU (Tensor Processing Unit).
- Processor 21 may be composed of multiple processors.
- Processor 21 is an example of a computer.
- the memory 22 is composed of various types of volatile memory and nonvolatile memory such as RAM (Random Access Memory) and ROM (Read Only Memory).
- the memory 22 also stores programs for the information processing device 2 to execute various processes. Further, the memory 22 is used as a working memory and temporarily stores information etc. acquired from the storage device 4. Note that the memory 22 may function as the storage device 4. Similarly, the storage device 4 may function as the memory 22 of the information processing device 2. Note that the program executed by the information processing device 2 may be stored in a storage medium other than the memory 22.
- the interface 23 is an interface for electrically connecting the information processing device 2 and other devices by wire or wirelessly. These interfaces may be wireless interfaces such as network adapters for wirelessly transmitting and receiving data to and from other devices, or may be hardware interfaces for connecting to other devices via cables or the like. Furthermore, in this embodiment, the interface 23 performs interface operations for the input section 24, display section 25, and sound output section 26 included in the information processing device 2.
- the input unit 24 is a user interface for the user of the observation system 100 to input predetermined information, and includes, for example, a button, a switch, a touch panel, or a voice input device.
- the display unit 25 is, for example, a display or a projector, and displays predetermined information under the control of the processor 21.
- the sound output unit 26 is, for example, a speaker, and outputs sound (voice) under the control of the processor 21.
- the input section 24, the display section 25, and the sound output section 26 may be external devices that are electrically connected to the information processing device 2 via the interface 23 by wire or wirelessly. Further, the interface 23 may perform an interface operation for any device other than the input section 24, the display section 25, and the sound output section 26.
- FIG. 4 is an example of functional blocks related to abnormality detection related processing.
- the processor 21 of the information processing device 2 functionally includes an observation data acquisition section 31, a segment data generation section 32, an abnormality degree calculation section 33, and an output control section 34 regarding abnormality detection related processing. Note that in FIG. 4, blocks where data is exchanged are connected by solid lines, but the combination of blocks where data is exchanged is not limited to this. The same applies to other functional block diagrams to be described later.
- the observation data acquisition unit 31 acquires observation data Da indicating the observation results of the space object 5 from the optical observation device 1 via the interface 23. Then, the observation data acquisition unit 31 stores the acquired observation data Da in the observation data DB 41. In addition to or instead of storing the observation data Da in the observation data DB 41, the observation data acquisition unit 31 may supply the observation data Da to the segment data generation unit 32.
- the segment data generation unit 32 generates segment data based on the time-series observation data Da acquired by the observation data acquisition unit 31.
- the segment data indicates the time-series observed value and observation time of the luminosity of the space object 5 observed in a certain period, and is generated from observation data Da corresponding to the observation time of a predetermined number of objects. .
- the above-mentioned predetermined number may be a predetermined constant or a variable number.
- the segment data generation unit 32 supplies the generated segment data to the abnormality degree calculation unit 33.
- the segment data generation unit 32 divides the time-series observation data Da at the timing at which the observation of the space object 5 by the optical observation device 1 becomes discontinuous (for example, at the timing at which the observation interval is longer than a predetermined length of time). , divides the time-series observation data Da. Then, the segment data generation unit 32 generates segment data for each group of the divided observation data Da. According to this aspect, the segment data generation unit 32 can group pieces of observation data Da that have similar observation times as segment data. Thereby, the accuracy of the degree of abnormality can be improved. Note that in this case as well, an upper limit number of observation data Da to be included in the segment data may be determined.
- the abnormality degree calculation unit 33 calculates the abnormality degree of the space object 5 at the time when the luminous intensity indicated by the segment data is observed.
- the abnormality degree calculation unit 33 configures an abnormality degree calculation model based on the parameter information 42, and calculates the abnormality degree based on the information output by the abnormality degree calculation model by inputting segment data to the abnormality degree calculation model. do.
- the abnormality degree calculation unit 33 supplies information regarding the calculated abnormality degree to the output control unit 34.
- the abnormality degree calculation unit 33 may record information regarding the calculated abnormality degree in the observation data DB 41.
- the output control unit 34 controls the output related to the degree of abnormality calculated by the degree of abnormality calculation unit 33.
- the output control unit 34 performs control to display information regarding the abnormality degree calculated by the abnormality degree calculation unit 33 on the display unit 25.
- the output control unit 34 detects the presence or absence of an abnormality in the space object 5 based on the degree of abnormality, and outputs information based on the detection result via the display unit 25 or the sound output unit 26.
- the output control unit 34 supplies a display signal based on the detection result to the display unit 25 via the interface 23 to cause the display unit 25 to display predetermined information, or outputs a sound output signal based on the detection result.
- the output control unit 34 is an example of a "display means" and a "warning means.”
- each component of the observation data acquisition unit 31, segment data generation unit 32, abnormality degree calculation unit 33, and output control unit 34 described in FIG. 4 can be realized, for example, by the processor 21 executing a program. Further, each component may be realized by recording necessary programs in an arbitrary non-volatile storage medium and installing them as necessary. Note that at least a part of each of these components is not limited to being implemented by software based on a program, but may be implemented by a combination of hardware, firmware, and software. Furthermore, at least a portion of each of these components may be realized using a user-programmable integrated circuit, such as a field-programmable gate array (FPGA) or a microcontroller.
- FPGA field-programmable gate array
- this integrated circuit may be used to implement a program made up of the above-mentioned components.
- at least a part of each component is configured by an ASSP (Application Specific Standard Produce), an ASIC (Application Specific Integrated Circuit), or a quantum processor (Quantum Computer Control Chip). may be done.
- ASSP Application Specific Standard Produce
- ASIC Application Specific Integrated Circuit
- quantum processor Quantum Computer Control Chip
- FIG. 5 is an example of functional blocks of the abnormality degree calculation unit 33 when the abnormality degree calculation model is an autoencoder.
- the abnormality degree calculation section 33 includes a model execution section 37 and an error calculation section 47.
- the model execution unit 37 configures an anomaly degree calculation model that is a self-encoder based on the parameter information 42, and when the segment data supplied from the segment data generation unit 32 is input to the abnormality degree calculation model, the abnormality degree calculation model Get the data output by .
- the anomaly degree calculation model outputs data (also referred to as "restored data") obtained by restoring the input segment data when segment data indicating the luminosity of the space object 5 in a normal state is input. has been studied. Therefore, the model execution unit 37 supplies the restored data output by the abnormality degree calculation model to the error calculation unit 47.
- the error calculation unit 47 calculates the error (so-called restoration error) between the restored data supplied from the model execution unit 37 and the segment data that is the input data of the abnormality degree calculation model, as the degree of abnormality.
- the error calculation unit 47 may calculate the above-mentioned error using any loss function used in machine learning or the like. For example, the error calculation unit 47 calculates the L1 norm or L2 norm of the difference vector between the input data and the restored data as the degree of abnormality. Then, the error calculation unit 47 outputs the calculated degree of abnormality.
- the anomaly degree calculation unit 33 uses the anomaly degree calculation model, which is a trained self-encoder, to suitably determine the anomaly degree, which is an index indicating the probability of occurrence of an attitude change, etc. of the space object 5. It can be calculated as follows.
- the output control unit 34 performs control to display information regarding the degree of abnormality calculated by the degree of abnormality calculation unit 33 on the display unit 25.
- the output control unit 34 may display on the display unit 25 a graph or a table showing the transition of the degree of abnormality in time series calculated by the degree of abnormality calculation unit 33.
- FIG. 6 is an example of a table showing the transition of the degree of abnormality output by the output control unit 34 in the first output example.
- the table shown in FIG. 6 has items of "time” and "abnormality degree", and for example, a record is generated for each segment data generated by the segment data generation unit 32.
- Time indicates the representative time of the observation time of the luminosity included in the corresponding segment data.
- the output control unit 34 may determine the representative time based on an arbitrary rule from a plurality of observation times corresponding to a plurality of luminosities included in the segment data. For example, the output control unit 34 may set the earliest or latest time among the plurality of observation times as the representative time, or may set the median value of the plurality of observation times as the representative time. Note that, instead of "time”, the output control unit 34 provides "time zone" as an item in the table, which indicates a time zone specified by the earliest time and latest time among the plurality of observation times described above. Good too.
- Abnormality degree indicates the abnormality degree calculated from the corresponding segment data. Note that the output control unit 34 may highlight a record in which the degree of abnormality is equal to or higher than a predetermined threshold value as a record corresponding to the observation result of the space object 5 suspected of causing an abnormality.
- the user can suitably use the degree of abnormality presented by the information processing device 2 as a criterion for determining whether or not an abnormality has occurred in the space object 5.
- the output control unit 34 detects the presence or absence of an abnormality in the space object 5 based on the degree of abnormality, and outputs the abnormality detection result through the display unit 25 or the sound output unit 26. For example, the output control unit 34 determines that an abnormality has occurred in the space object 5 when the abnormality degree calculated by the abnormality degree calculation unit 33 based on the latest segment data generated by the segment data generation unit 32 is equal to or higher than a predetermined threshold. A warning to that effect is output from the display section 25 or the sound output section 26.
- the above threshold value is stored in the memory 22 or the storage device 4 in advance, for example.
- the output control unit 34 recognizes an abnormal period in which an abnormality has occurred in the space object 5 and another normal period based on the degree of abnormality in time series, and displays them on a graph or table representing the degree of abnormality.
- the abnormal period and normal period are clearly identified by color coding, etc.
- the information processing device 2 can autonomously detect an abnormality in the space object 5 and suitably notify the user of the abnormality detection result.
- the output control unit 34 may store the abnormality detection result in the storage device 4 instead of outputting the abnormality detection result on the display unit 25 or the sound output unit 26, and the output control unit 34 may store the abnormality detection result in the storage device 4.
- the information may be sent to another device (which may be a terminal used by an administrator, etc.) that manages the status of the system.
- FIG. 7 is a diagram illustrating an overview of the learning process of the abnormality degree calculation model by the information processing device 2.
- the processor 21 of the information processing device 2 functionally includes an input data generation section 38 and a parameter update section 39.
- the input data generation unit 38 generates data that matches the input format of the anomaly degree calculation model, based on the training data 43 that indicates the time-series luminosity of the space object 5 in a normal state. For example, the input data generation unit 38 generates segment data from the training data 43 by performing the same processing as the segment data generation unit 32. As another example, if a plurality of segment data matching the input format of the abnormality degree calculation model is stored in advance as the training data 43, the input data generation unit 38 generates segment data to be input to the abnormality degree calculation model. The data are sequentially extracted from the training data 43.
- the parameter update unit 39 inputs the data supplied from the input data generation unit 38 as input data to the abnormality degree calculation model, and calculates the restoration error between the data output from the abnormality degree calculation model and the input data. Then, the parameter updating unit 39 determines the parameters of the anomaly degree calculation model using a gradient descent method, an error backpropagation method, or the like so that the restoration error is minimized. Then, the parameter update unit 39 updates the parameter information 42 with the determined parameters.
- the learning process of the abnormality degree calculation model may be executed by a device other than the information processing device 2.
- a device other than the information processing device 2 executes the above-described learning process, and the parameter information 42 obtained by the learning process is stored in the storage device 4. Ru.
- FIG. 8 is an example of a flowchart of abnormality detection related processing.
- the information processing device 2 repeatedly executes the process shown in the flowchart shown in FIG.
- the information processing device 2 acquires observation data Da from the optical observation device 1, and stores the acquired observation data Da in the observation data DB 41 (step S11).
- the information processing device 2 determines whether it is time to generate segment data (step S12). For example, when the information processing device 2 determines that a predetermined number of observation data Da necessary for generating segment data has been accumulated, it determines that it is time to generate segment data. In another example, when the information processing device 2 detects that the acquisition interval of observation data Da is equal to or greater than a predetermined interval, the information processing device 2 determines that it is time to generate segment data, and immediately before the acquisition interval becomes equal to or greater than the predetermined interval. Segment data is generated based on observation data Da.
- the information processing device 2 when the information processing device 2 detects an external input (including a user input by the input unit 24) requesting the output of information regarding the degree of abnormality, the information processing device 2 outputs the observation data Da stored in the observation data DB 41. Generate segment data as a target. In this case, if the external input includes information specifying a period, the information processing device 2 extracts observation data Da corresponding to the specified period from the observation data DB 41, and extracts the observation data Da from the extracted observation data Da. It is a good idea to generate segment data.
- step S12 determines that it is time to generate segment data
- step S13 determines that it is not the segment data generation timing
- step S12 determines that it is not the segment data generation timing
- the information processing device 2 calculates the degree of abnormality of the space object 5 based on the generated segment data (step S14). In this case, the information processing device 2 calculates the degree of abnormality based on the data output by the degree of abnormality calculation model when the segment data is input to the degree of abnormality calculation model configured using the parameter information 42.
- the information processing device 2 performs output control regarding the degree of abnormality calculated in step S14 (step S15).
- the information processing device 2 displays the transition of the degree of abnormality in time series, detects the presence or absence of an abnormality in the space object 5 based on the degree of abnormality, and displays or outputs the detection result as audio. or
- the information processing device 2 may convert the segment data into feature amount data representing the feature amount by performing feature extraction processing, adding a lag feature amount, or the like.
- the feature amount data is data in a predetermined tensor format that matches the input format of the abnormality degree calculation model.
- FIG. 9 is an example of functional blocks related to abnormality detection related processing of the processor 21 of the information processing device 2 according to this modification.
- the processor 21 includes a feature generation unit 35 in addition to the processing units 31 to 34 shown in FIG.
- the feature amount generation unit 35 converts the segment data generated by the segment data generation unit 32 into feature amount data that matches the input format of the abnormality degree calculation model.
- the feature amount generation unit 35 may generate feature amount data from the segment data based on any feature extraction technique.
- the feature amount generation unit 35 may generate segment data or data obtained by adding a lag feature amount to the feature amount as the feature amount data.
- the lag feature amount is generated based on, for example, a predetermined number of segment data generated immediately before the target segment data.
- the feature amount generation section 35 supplies the generated feature amount data to the abnormality degree calculation section 33. Thereafter, the abnormality degree calculation unit 33 calculates the abnormality degree based on the data output by the abnormality degree calculation model when the feature amount data is input to the abnormality degree calculation model.
- the information processing device 2 can calculate the degree of abnormality with higher accuracy.
- the abnormality degree calculation model may be a model other than an autoencoder.
- the abnormality degree calculation model may be a model that converts input data into data in a feature space with a predetermined number of dimensions.
- the data (normal data) on the above-mentioned feature space when the space object 5 is in a normal state is stored in advance in the storage device 4, etc., and the information processing device 2 outputs the abnormality degree calculation model.
- the error between the data and the normal data is calculated as the degree of abnormality.
- the information processing device 2 can suitably calculate the abnormality degree.
- FIG. 10 is a block diagram of the information processing device 2X in the second embodiment.
- the information processing device 2X includes a data acquisition means 32X and an abnormality degree calculation means 33X.
- the information processing device 2X may be composed of a plurality of devices.
- the data acquisition means 32X acquires time series data representing the luminosity of the space object.
- the data acquisition unit 32X can be, for example, the observation data acquisition unit 31 or the segment data generation unit 32 in the first embodiment (including modified examples, the same applies hereinafter).
- the abnormality degree calculation means 33X calculates the abnormality degree regarding the state of the space object based on the time series data.
- the abnormality degree calculation means 33X can be the abnormality degree calculation section 33 in the first embodiment.
- FIG. 11 is an example of a flowchart showing the processing procedure in the second embodiment.
- the data acquisition means 32X acquires time series data representing the luminosity of a space object (step S21).
- the abnormality degree calculating means 33X calculates the abnormality degree regarding the state of the space object based on the time series data (step S22).
- the information processing device 2X can suitably calculate the degree of abnormality regarding the state of the space object.
- Non-transitory computer-readable media include various types of tangible storage media.
- Examples of non-transitory computer-readable media include magnetic storage media (e.g., flexible disks, magnetic tapes, hard disk drives), magneto-optical storage media (e.g., magneto-optical disks), CD-ROMs (Read Only Memory), CD-Rs, CD-R/W, semiconductor memory (e.g., mask ROM, PROM (Programmable ROM), EPROM (Erasable PROM), flash ROM, RAM (Random Access Memory).
- Transitory computer readable media may be supplied to the computer by a transitory computer readable medium.
- Examples of transitory computer readable media include electrical signals, optical signals, and electromagnetic waves.
- Transitory computer readable media include electrical wires and optical
- the program can be supplied to the computer via a wired communication path such as a fiber or a wireless communication path.
- the information processing device according to appendix 1 or 2, further comprising an abnormality detection means for detecting an abnormality of the space object based on the degree of abnormality.
- the data acquisition means acquires, as the time series data, segment data in which the observed time series luminosity is divided into a predetermined number of data representing the luminosity;
- the information processing device according to any one of Supplementary Notes 1 to 3, wherein the abnormality degree calculation means calculates the abnormality degree based on the segment data.
- the data acquisition means acquires, as the time series data, segment data in which the observed time series luminosity is divided based on observation intervals;
- the information processing device according to any one of Supplementary Notes 1 to 3, wherein the abnormality degree calculation means calculates the abnormality degree based on the segment data.
- the information processing device according to any one of Supplementary Notes 1 to 5, further comprising a display unit that displays information regarding the degree of abnormality.
- the information processing device according to any one of Supplementary Notes 1 to 7, further comprising a warning unit that warns of the occurrence of an abnormality when an abnormality of the space object is detected based on the abnormality degree.
- the computer is Obtain time series data representing the luminosity of space objects, calculating an abnormality degree regarding the state of the space object based on the time series data; Calculation method.
- a data acquisition means for acquiring time series data representing the luminosity of a space object;
- An abnormality degree calculation means for calculating an abnormality degree regarding the state of the space object based on the time series data;
- the abnormality degree calculation means calculates the abnormality degree based on the time series data and the learning model, The information processing according to appendix 11, wherein the learning model is a model learned to output data obtained by restoring the input to the learning model based on training data representing the luminosity of the space object in a normal state. system.
- the information processing system according to appendix 11 or 12, further comprising an abnormality detection means for detecting an abnormality in the space object based on the degree of abnormality.
- the data acquisition means acquires, as the time series data, segment data in which the observed time series luminosity is divided into a predetermined number of data representing the luminosity;
- the information processing system according to any one of appendices 11 to 13, wherein the abnormality degree calculation means calculates the abnormality degree based on the segment data.
- the data acquisition means acquires, as the time series data, segment data in which the observed time series luminosity is divided based on observation intervals;
- the information processing system according to appendix 16 wherein the display means displays information indicating the transition of the degree of abnormality.
- 18 18.
- Optical observation device 2 Information processing device 4 Storage device 21 Processor 22 Memory 23 Interface 24 Input section 25 Display section 26 Sound output section 41 Observation data DB 42 Parameter information 43 Training data 100 Observation system
Landscapes
- Engineering & Computer Science (AREA)
- Remote Sensing (AREA)
- Physics & Mathematics (AREA)
- Astronomy & Astrophysics (AREA)
- General Physics & Mathematics (AREA)
- Radar, Positioning & Navigation (AREA)
- Aviation & Aerospace Engineering (AREA)
- Testing Or Calibration Of Command Recording Devices (AREA)
Abstract
Description
(1)システム構成
図1は、第1実施形態に係る観測システム100の構成を示す。観測システム100は、主に、光学観測装置1と、情報処理装置2と、記憶装置4と、を有する。
図3は、情報処理装置2のブロック構成の一例を示す。情報処理装置2は、ハードウェアとして、プロセッサ21と、メモリ22と、インターフェース23とを含む。プロセッサ21、メモリ22及びインターフェース23は、データバス29を介して接続されている。
図4は、異常検知関連処理に関する機能ブロックの一例である。情報処理装置2のプロセッサ21は、異常検知関連処理に関し、機能的には、観測データ取得部31と、セグメントデータ生成部32と、異常度算出部33と、出力制御部34と、を有する。なお、図4では、データの授受が行われるブロック同士を実線により結んでいるが、データの授受が行われるブロックの組み合わせはこれに限定されない。後述する他の機能ブロックの図においても同様である。
次に、異常度算出部33の処理の具体例について説明する。
次に、出力制御部34の処理の具体例について説明する。
次に、異常度算出モデルの学習処理について補足説明する。図7は、情報処理装置2による異常度算出モデルの学習処理の概要を示す図である。学習処理において、情報処理装置2のプロセッサ21は、機能的には、入力データ生成部38と、パラメータ更新部39とを有する。
図8は、異常検知関連処理のフローチャートの一例である。情報処理装置2は、図8に示すフローチャートの処理を繰り返し実行する。
上述した実施形態に好適な変形例について説明する。以下の変形例は組み合わせて上述の実施形態に適用してもよい。
情報処理装置2は、セグメントデータに対し、特徴抽出処理やラグ特徴量の追加などを行うことで、特徴量を表す特徴量データに変換してもよい。この場合、特徴量データは、異常度算出モデルの入力形式に整合するような所定のテンソル形式のデータとなる。
異常度算出モデルは、自己符号化器以外のモデルであってもよい。例えば、異常度算出モデルは、入力データを所定次元数の特徴空間のデータに変換するモデルであってもよい。この場合、例えば、宇宙物体5が正常状態の場合における上述の特徴空間上でのデータ(正常データ)が予め記憶装置4等に記憶されており、情報処理装置2は、異常度算出モデルが出力するデータと、正常データとの誤差を、異常度として算出する。
図10は、第2実施形態における情報処理装置2Xのブロック図である。情報処理装置2Xは、データ取得手段32Xと、異常度算出手段33Xと、を備える。情報処理装置2Xは、複数の装置から構成されてもよい。
宇宙物体の光度を表す時系列データを取得するデータ取得手段と、
前記時系列データに基づき、前記宇宙物体の状態に関する異常度を算出する異常度算出手段と、
を有する情報処理装置。
[付記2]
前記異常度算出手段は、前記時系列データと、学習モデルとに基づき、前記異常度を算出し、
前記学習モデルは、正常状態での前記宇宙物体の光度を表す訓練データに基づき、前記学習モデルへの入力を復元したデータを出力するように学習されたモデルである、付記1に記載の情報処理装置。
[付記3]
前記異常度に基づき、前記宇宙物体の異常を検知する異常検知手段をさらに有する、付記1または2に記載の情報処理装置。
[付記4]
前記データ取得手段は、前記時系列データとして、観測された時系列の前記光度を所定個数分の前記光度を表すデータに区切ったセグメントデータを取得し、
前記異常度算出手段は、前記セグメントデータに基づき、前記異常度を算出する、付記1~3のいずれか一項に記載の情報処理装置。
[付記5]
前記データ取得手段は、前記時系列データとして、観測された時系列の前記光度を、観測間隔に基づき区切ったセグメントデータを取得し、
前記異常度算出手段は、前記セグメントデータに基づき、前記異常度を算出する、付記1~3のいずれか一項に記載の情報処理装置。
[付記6]
前記異常度に関する情報を表示する表示手段をさらに有する、付記1~5のいずれか一項に記載の情報処理装置。
[付記7]
前記表示手段は、前記異常度の遷移を示す情報を表示する、付記6に記載の情報処理装置。
[付記8]
前記異常度に基づき前記宇宙物体の異常を検知した場合、当該異常の発生を警告する警告手段をさらに有する、付記1~7のいずれか一項に記載の情報処理装置。
[付記9]
コンピュータが、
宇宙物体の光度を表す時系列データを取得し、
前記時系列データに基づき、前記宇宙物体の状態に関する異常度を算出する、
算出方法。
[付記10]
宇宙物体の光度を表す時系列データを取得し、
前記時系列データに基づき、前記宇宙物体の状態に関する異常度を算出する処理をコンピュータに実行させるプログラムを格納した記憶媒体。
[付記11]
宇宙物体の光度を表す時系列データを取得するデータ取得手段と、
前記時系列データに基づき、前記宇宙物体の状態に関する異常度を算出する異常度算出手段と、
を有する情報処理システム。
[付記12]
前記異常度算出手段は、前記時系列データと、学習モデルとに基づき、前記異常度を算出し、
前記学習モデルは、正常状態での前記宇宙物体の光度を表す訓練データに基づき、前記学習モデルへの入力を復元したデータを出力するように学習されたモデルである、付記11に記載の情報処理システム。
[付記13]
前記異常度に基づき、前記宇宙物体の異常を検知する異常検知手段をさらに有する、付記11または12に記載の情報処理システム。
[付記14]
前記データ取得手段は、前記時系列データとして、観測された時系列の前記光度を所定個数分の前記光度を表すデータに区切ったセグメントデータを取得し、
前記異常度算出手段は、前記セグメントデータに基づき、前記異常度を算出する、付記11~13のいずれか一項に記載の情報処理システム。
[付記15]
前記データ取得手段は、前記時系列データとして、観測された時系列の前記光度を、観測間隔に基づき区切ったセグメントデータを取得し、
前記異常度算出手段は、前記セグメントデータに基づき、前記異常度を算出する、付記11~13のいずれか一項に記載の情報処理システム。
[付記16]
前記異常度に関する情報を表示する表示手段をさらに有する、付記11~15のいずれか一項に記載の情報処理システム。
[付記17]
前記表示手段は、前記異常度の遷移を示す情報を表示する、付記16に記載の情報処理システム。
[付記18]
前記異常度に基づき前記宇宙物体の異常を検知した場合、当該異常の発生を警告する警告手段をさらに有する、付記11~17のいずれか一項に記載の情報処理システム。
2 情報処理装置
4 記憶装置
21 プロセッサ
22 メモリ
23 インターフェース
24 入力部
25 表示部
26 音出力部
41 観測データDB
42 パラメータ情報
43 訓練データ
100 観測システム
Claims (10)
- 宇宙物体の光度を表す時系列データを取得するデータ取得手段と、
前記時系列データに基づき、前記宇宙物体の状態に関する異常度を算出する異常度算出手段と、
を有する情報処理装置。 - 前記異常度算出手段は、前記時系列データと、学習モデルとに基づき、前記異常度を算出し、
前記学習モデルは、正常状態での前記宇宙物体の光度を表す訓練データに基づき、前記学習モデルへの入力を復元したデータを出力するように学習されたモデルである、請求項1に記載の情報処理装置。 - 前記異常度に基づき、前記宇宙物体の異常を検知する異常検知手段をさらに有する、請求項1または2に記載の情報処理装置。
- 前記データ取得手段は、前記時系列データとして、観測された時系列の前記光度を所定個数分の前記光度を表すデータに区切ったセグメントデータを取得し、
前記異常度算出手段は、前記セグメントデータに基づき、前記異常度を算出する、請求項1~3のいずれか一項に記載の情報処理装置。 - 前記データ取得手段は、前記時系列データとして、観測された時系列の前記光度を、観測間隔に基づき区切ったセグメントデータを取得し、
前記異常度算出手段は、前記セグメントデータに基づき、前記異常度を算出する、請求項1~3のいずれか一項に記載の情報処理装置。 - 前記異常度に関する情報を表示する表示手段をさらに有する、請求項1~5のいずれか一項に記載の情報処理装置。
- 前記表示手段は、前記異常度の遷移を示す情報を表示する、請求項6に記載の情報処理装置。
- 前記異常度に基づき前記宇宙物体の異常を検知した場合、当該異常の発生を警告する警告手段をさらに有する、請求項1~7のいずれか一項に記載の情報処理装置。
- コンピュータが、
宇宙物体の光度を表す時系列データを取得し、
前記時系列データに基づき、前記宇宙物体の状態に関する異常度を算出する、
算出方法。 - 宇宙物体の光度を表す時系列データを取得し、
前記時系列データに基づき、前記宇宙物体の状態に関する異常度を算出する処理をコンピュータに実行させるプログラムを格納した記憶媒体。
Priority Applications (3)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
PCT/JP2022/014405 WO2023181354A1 (ja) | 2022-03-25 | 2022-03-25 | 情報処理装置、算出方法及び記憶媒体 |
US18/729,954 US20250115377A1 (en) | 2022-03-25 | 2022-03-25 | Information processing device, calculation method, and storage medium |
JP2024509656A JPWO2023181354A5 (ja) | 2022-03-25 | 情報処理装置、算出方法及びプログラム |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
PCT/JP2022/014405 WO2023181354A1 (ja) | 2022-03-25 | 2022-03-25 | 情報処理装置、算出方法及び記憶媒体 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2023181354A1 true WO2023181354A1 (ja) | 2023-09-28 |
Family
ID=88100811
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2022/014405 WO2023181354A1 (ja) | 2022-03-25 | 2022-03-25 | 情報処理装置、算出方法及び記憶媒体 |
Country Status (2)
Country | Link |
---|---|
US (1) | US20250115377A1 (ja) |
WO (1) | WO2023181354A1 (ja) |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2011112284A (ja) * | 2009-11-26 | 2011-06-09 | Mitsubishi Electric Corp | 軌道推定システム |
JP5709651B2 (ja) * | 2011-06-03 | 2015-04-30 | 三菱電機株式会社 | 追尾装置 |
JP2015202809A (ja) * | 2014-04-15 | 2015-11-16 | 三菱重工業株式会社 | 監視システムおよび監視方法 |
JP2020504820A (ja) * | 2016-12-22 | 2020-02-13 | マイリオタ ピーティーワイ エルティーディーMyriota Pty Ltd | 拡張された衛星エフェメリス・データを生成するシステムおよび方法 |
WO2020085412A1 (ja) * | 2018-10-25 | 2020-04-30 | 国立研究開発法人宇宙航空研究開発機構 | 予測装置、予測方法、及び予測プログラム |
WO2021019672A1 (ja) * | 2019-07-30 | 2021-02-04 | 日本電信電話株式会社 | 異常度推定装置、異常度推定方法、およびプログラム |
CN112623283A (zh) * | 2020-12-30 | 2021-04-09 | 苏州三六零智能安全科技有限公司 | 太空物体异常检测方法、装置、设备及存储介质 |
-
2022
- 2022-03-25 WO PCT/JP2022/014405 patent/WO2023181354A1/ja active Application Filing
- 2022-03-25 US US18/729,954 patent/US20250115377A1/en active Pending
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2011112284A (ja) * | 2009-11-26 | 2011-06-09 | Mitsubishi Electric Corp | 軌道推定システム |
JP5709651B2 (ja) * | 2011-06-03 | 2015-04-30 | 三菱電機株式会社 | 追尾装置 |
JP2015202809A (ja) * | 2014-04-15 | 2015-11-16 | 三菱重工業株式会社 | 監視システムおよび監視方法 |
JP2020504820A (ja) * | 2016-12-22 | 2020-02-13 | マイリオタ ピーティーワイ エルティーディーMyriota Pty Ltd | 拡張された衛星エフェメリス・データを生成するシステムおよび方法 |
WO2020085412A1 (ja) * | 2018-10-25 | 2020-04-30 | 国立研究開発法人宇宙航空研究開発機構 | 予測装置、予測方法、及び予測プログラム |
WO2021019672A1 (ja) * | 2019-07-30 | 2021-02-04 | 日本電信電話株式会社 | 異常度推定装置、異常度推定方法、およびプログラム |
CN112623283A (zh) * | 2020-12-30 | 2021-04-09 | 苏州三六零智能安全科技有限公司 | 太空物体异常检测方法、装置、设备及存储介质 |
Also Published As
Publication number | Publication date |
---|---|
JPWO2023181354A1 (ja) | 2023-09-28 |
US20250115377A1 (en) | 2025-04-10 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US10061989B2 (en) | Information transfer apparatus and method for image recognition | |
US20220277225A1 (en) | Method and device for detecting anomalies, corresponding computer program and non-transitory computer-readable medium | |
CN106471527B (zh) | 从全局模型进行本地化学习 | |
US10637751B2 (en) | Methods and systems for online monitoring using a variable data sampling rate | |
JP6551565B2 (ja) | 工程解析装置、工程解析方法、及び工程解析プログラム | |
JP6374466B2 (ja) | センサインタフェース装置、測定情報通信システム、測定情報通信方法、及び測定情報通信プログラム | |
TWI687711B (zh) | 震央距離推定裝置、震央距離推定方法及電腦可讀取之記錄媒體 | |
US20190266039A1 (en) | System analysis support device, system analysis support method, and program | |
US11275488B2 (en) | Augmenting user interface to convey a state of drilling operations | |
US10228994B2 (en) | Information processing system, information processing method, and program | |
US20240346112A1 (en) | Anomaly detection apparatus, anomaly detection method, and program | |
US20200342215A1 (en) | Model learning device, model learning method, and recording medium | |
CN112950687A (zh) | 一种确定跟踪状态的方法、装置、存储介质和电子设备 | |
EP3999983B1 (en) | Time-series data condensation and graphical signature analysis | |
WO2023181354A1 (ja) | 情報処理装置、算出方法及び記憶媒体 | |
JP7142530B2 (ja) | 情報処理システム、情報処理方法、およびプログラム | |
JP6273835B2 (ja) | 状態判定装置、状態判定方法および状態判定プログラム | |
JP7265127B2 (ja) | 機械学習プログラム、機械学習方法及び機械学習システム | |
US20200104644A1 (en) | Training apparatus, training method and recording medium | |
JP7397801B2 (ja) | 呼吸障害を有する被験者を検出する装置及び呼吸障害を有する被験者を検出する装置の作動方法 | |
US20250002182A1 (en) | Information processing device, detection method, and storage medium | |
US20240405866A1 (en) | Optical fiber sensing system, optical fiber sensing device, and optical fiber sensing method | |
JP2018097418A (ja) | 情報処理装置 | |
US20250148294A1 (en) | Conditional acoustic library generation for distributed fiber optic sensor | |
US20240377528A1 (en) | State determination device, state determination method, and recording medium having state determination program stored thereon |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 22933493 Country of ref document: EP Kind code of ref document: A1 |
|
WWE | Wipo information: entry into national phase |
Ref document number: 18729954 Country of ref document: US |
|
ENP | Entry into the national phase |
Ref document number: 2024509656 Country of ref document: JP Kind code of ref document: A |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 22933493 Country of ref document: EP Kind code of ref document: A1 |
|
WWP | Wipo information: published in national office |
Ref document number: 18729954 Country of ref document: US |