JP2021143499A - Rail condition prediction method, program, computer memory medium and rail condition prediction system - Google Patents

Abstract

To efficiently and properly predict rail conditions.SOLUTION: A method to predict rail conditions on which railroad vehicles travel has: a sway measurement step (S1) for measuring sway of railway vehicles using a measuring unit provided on the railroad vehicles; a sway prediction step (S4) for predicting future sway from the sway measurement data obtained in the sway measurement step using a prediction model; and a rail prediction step (S5) for predicting rail conditions from the sway prediction data obtained in the sway prediction step based on a predetermined sway threshold value.SELECTED DRAWING: Figure 4

Description

The present invention relates to orbital state prediction methods, programs, computer storage media and orbital state prediction systems.

As the railroad vehicle repeatedly travels on the track on which the railroad vehicle travels, the track displacement (track deviation) progresses and the state deteriorates. Therefore, it is important to grasp the state of the orbit and take appropriate care. In particular, in recent years, it has been required to evaluate the state of the orbit and perform maintenance (CBM: Condition Based Maintenance) based on the evaluated state of the orbit.

As a method of evaluating the state of the track, for example, a method of measuring the track displacement or the like by a dedicated track inspection vehicle which is a non-commercial vehicle is known. However, in such a case, the above-mentioned CBM cannot be carried out because the frequency of measurement by the track inspection vehicle is low.

Therefore, a method has been proposed in which a track measuring device is mounted on a commercial vehicle and the state of the track is evaluated during the commercial running of the commercial vehicle. For example, in the track evaluation method described in Patent Document 1, first, rail height information measured during traveling by a measuring device mounted on a railroad vehicle is accumulated. Next, based on the accumulated rail height information, track displacement information including the values of the rail height and low displacements of the rails in the evaluation target section at a plurality of time points is generated. Then, the evaluation value of the trajectory of the evaluation target section is calculated by using a predetermined displacement theoretical formula that fits the generated trajectory displacement information.

Japanese Unexamined Patent Publication No. 2019-19454

When the track evaluation method described in Patent Document 1 is implemented as described above, the railway vehicle needs to be equipped with a measuring device for measuring rail height information (track displacement). However, the existing railway vehicle (commercial vehicle) may not be equipped with the above-mentioned measuring device, and in such a case, it takes time and cost to provide the measuring device in the railway vehicle.

Further, in order to efficiently maintain the orbit, it is required to predict the state of the orbit in the future, but it has not reached that level at present.

The present invention has been made in view of this point, and an object of the present invention is to efficiently and appropriately predict the state of an orbit.

In order to achieve the above object, the present invention is a method of predicting the state of a track on which a railroad vehicle travels, and a sway measurement for measuring the sway of the railroad vehicle by using a measuring unit provided on the railroad vehicle. A sway prediction process that predicts future sway from the sway measurement data measured in the sway measurement step using a process and a prediction model, and a sway prediction predicted in the sway prediction process based on a predetermined threshold of sway. It is characterized by having an orbit prediction process for predicting the state of the orbit from data.

As described above, in the conventional track evaluation method described in Patent Document 1, it is necessary to newly provide a measuring device for measuring track displacement in a railway vehicle, whereas according to the present invention, it is provided in a normal railway vehicle. The sway measurement data is acquired using the existing measuring unit. Therefore, the orbital state can be predicted more easily and cheaply than before.

Further, according to the present invention, a prediction model is used to predict future sway (sway prediction data) from sway measurement data, and further, based on a predetermined threshold of sway, the orbital state is predicted from the sway prediction data. For example, when the sway prediction data is equal to or less than a predetermined threshold value, the orbit is predicted to be in a normal state, while when the sway prediction data is larger than the predetermined threshold value, the orbit is predicted to be in an abnormal state. Therefore, the state of the track can be appropriately predicted, and as a result, the maintenance of the track can be performed efficiently.

The orbital state method may further include a data processing step of processing the sway measurement data and a data visualization step of visualizing the processed sway measurement data for each position on the orbit.

In the orbital state method, the prediction model may be constructed by a regression model.

In the orbital state method, outliers may be excluded from the sway measurement data in the sway prediction step.

In the track state method, the outlier is the data measured by a railroad vehicle having a traveling speed of a predetermined speed or less among the shaking measurement data, and is measured by a railroad vehicle having a traveling speed different from that of normal traveling at a position on the same track. It may be either the data obtained or the data when the position on the orbit in the sway measurement data deviates from the actual position.

In the orbital state method, in the orbital prediction step, the orbital state may be predicted by further using the rate of increase of the sway prediction data.

In the orbital state method, the orbital state may be predicted by further using the error between the sway measurement data and the sway prediction data in the orbit prediction step.

According to another aspect of the present invention, there is provided a program that operates on a computer that controls the orbital state prediction system so that the orbital state prediction method is executed by the orbital state prediction system.

According to the present invention from another aspect, a readable computer storage medium in which the program is stored is provided.

From yet another point of view, the present invention is a system for predicting the state of the track on which the railroad vehicle travels, and stores the sway measurement data of the railroad vehicle measured by using the measuring unit provided in the railroad vehicle. Using a storage unit and a prediction model, the sway prediction unit that predicts future sway from the sway measurement data, and the orbital state from the sway prediction data predicted by the sway prediction unit based on a predetermined threshold of sway. It is characterized by having an orbit prediction unit for prediction.

The orbital state prediction system may further include a data processing unit that processes the sway measurement data and a data visualization unit that visualizes the processed sway measurement data for each position on the orbit.

In the orbital state prediction system, the prediction model may be constructed by a regression model.

In the orbital state prediction system, the sway prediction unit may exclude outliers from the sway measurement data.

In the track state prediction system, the outliers are the data measured by a railroad vehicle having a traveling speed of a predetermined speed or less among the shaking measurement data, and the railroad vehicle having a traveling speed different from that of normal traveling at a position on the same track. It may be either the measured data or the data when the position on the orbit in the sway measurement data deviates from the actual position.

In the orbital state prediction system, the orbital prediction unit may further use the rate of increase of the sway prediction data to predict the orbital state.

In the orbital state prediction system, the orbital prediction unit may predict the orbital state by further using the error between the sway measurement data and the sway prediction data.

According to the present invention, the state of the orbit can be predicted efficiently and appropriately. As a result, track maintenance can be performed efficiently.

It is a side view in the longitudinal direction of a train which schematically shows the outline of the structure of a train and a track. It is a side view of FIG. 1 in the lateral direction. It is explanatory drawing which shows the outline of the structure of the orbital state prediction system which concerns on this embodiment. It is a flow chart which shows the main process of the method of predicting the state of an orbit in this embodiment. It is explanatory drawing of the sway (vertical sway) of a vehicle. It is explanatory drawing which shows an example which visualized the sway measurement data of a train on a different day for each kilometer. It is explanatory drawing which shows an example which visualized the time-dependent change of a train sway measurement data. It is explanatory drawing of the prediction model. It is explanatory drawing which shows an example of the prediction result of the state of an orbit. An example of the values of the agitation prediction data (prediction result) and the agitation measurement data (actual measurement result), the rate of increase of the agitation prediction data (prediction result), and the average error of the agitation prediction data (prediction result) and the agitation measurement data (actual measurement result). It is explanatory drawing which shows. It is explanatory drawing which shows the case where the sway measurement data (actual measurement result) is higher than the sway prediction data (prediction result). It is explanatory drawing which shows the guideline of correspondence to the said kilometer with respect to the combination of the sway measurement data (predicted value), the increase rate of the sway measurement data, and the average error of the sway measurement data and the sway prediction data.

Hereinafter, embodiments of the present invention will be described with reference to the drawings. In the present specification and the drawings, components having substantially the same functional configuration are designated by the same reference numerals, so that duplicate description will be omitted.

<Train and track composition>
In this embodiment, the state of the track on which the train travels is predicted. First, the basic configuration of trains and tracks will be described. FIG. 1 is a side view in the longitudinal direction of the train 10 which schematically shows the outline of the configuration of the train 10 and the track 20. FIG. 2 is a side view of FIG. 1 in the lateral direction.

The train 10 is formed by connecting one or more railway vehicles 11 (hereinafter, referred to as "vehicles 11") via a connecting portion. In the example of FIG. 1, three cars 11 are connected in the train 10, but the number of cars 11 connected is not limited to this.

The vehicle 11 is provided with a measuring unit 12 for measuring the shaking of the vehicle 11. The measuring unit 12 is provided with an accelerometer, and measures vertical acceleration, horizontal acceleration, front-rear acceleration, and the like as shaking of the vehicle 11. In the following description, the data measured by the measuring unit 12 is referred to as "sway measurement data". The measuring unit 12 is provided on the train 10, for example, the leading vehicle 11. The number and arrangement of the measuring units 12 are not limited to the illustrated example, and can be set arbitrarily.

The track 20 on which the train 10 (vehicle 11) travels means a track structure including a rail 21, sleepers 22, and a trackbed 23.

<Configuration of orbital state prediction system>
Next, the system for predicting the state of the orbit 20 described above will be described. In the present embodiment, the vertical displacement of the track 20 (rail 21) is predicted as the state of the track 20. FIG. 3 is an explanatory diagram showing an outline of the configuration of the orbital state prediction system 30 according to the present embodiment.

The track state prediction system 30 is connected to the measurement unit 12 of the vehicle 11 via the network N. The network N is not particularly limited as long as it can communicate between the orbital state prediction system 30 and the measurement unit 12, but is configured by, for example, the Internet or a wireless LAN. Further, the measurement unit 12 and the orbital state prediction system 30 do not necessarily have to be connected via the network N, and the data acquired from the measurement unit 12 may be input to the orbital state prediction system 30.

The orbital state prediction system 30 includes a communication unit 31, a data processing unit 32, a data visualization unit 33, an agitation prediction unit 34, an orbit prediction unit 35, a storage unit 36, and an output unit 37.

The communication unit 31 is a communication interface that mediates communication with the network N, and performs data communication with the measurement unit 12.

The data processing unit 32 processes the sway measurement data measured by the measuring unit 12.

The data visualization unit 33 visualizes the sway measurement data processed by the data processing unit 32.

The sway prediction unit 34 predicts future sway (hereinafter referred to as “sway prediction data”) from the sway measurement data measured by the measurement unit 12 using a prediction model.

The orbit prediction unit 35 predicts the state of the orbit 20 from the sway measurement data measured by the measurement unit 12 based on a predetermined threshold value of the sway.

The storage unit 36 stores various data processed by the orbital state prediction system 30. Specifically, the sway measurement data input from the measurement unit 12 to the orbital state prediction system 30, the sway measurement data processed by the data processing unit 32, the sway measurement data visualized by the data visualization unit 33, and the sway prediction unit 34. The prediction model used in the above, the sway prediction data predicted by the sway prediction unit 34, the prediction result of the state of the orbit 20 predicted by the orbit prediction unit 35, and the like are stored. Further, various programs for controlling the orbital state prediction system 30 are stored in the storage unit 36.

The output unit 37 outputs each data stored in the storage unit 36 to the outside of the orbital state prediction system 30.

The components of the orbital state prediction system 30 can be composed of a circuit (hardware), a central processing unit such as a CPU, and a program (software) for operating these. Then, this program controls each unit 31 to 37 to execute data processing described later. In this case, the program may be stored in, for example, a storage unit 36, or may be a computer-readable hard disk (HD), flexible disk (FD), compact disk (CD), magnetic optical desk (MO), or the like. It may be stored in a storage medium that can be read by a computer such as various types of memory. Further, the program can be stored in the storage medium or the like by downloading it via a communication network such as the Internet.

<Orbital state prediction method>
Next, a method of predicting the state of the orbit 20 performed by using the orbit state prediction system 30 configured as described above will be described with reference to a specific example. In the present embodiment, during the commercial running of the vehicle 11 which is a commercial vehicle, the measuring unit 12 measures the sway of the vehicle 11 to predict the future sway, and further predict the state of the track 20. FIG. 4 is a flow chart showing the main steps of the method of predicting the state of the track 20.

[S1: Shake measurement process]
First, the sway of the vehicle 11 is measured using the measuring unit 12 provided on the vehicle 11. The measured sway measurement data is input to the orbital state prediction system 30 via the network N. The sway measurement data input to the orbital state prediction system 30 is stored in the storage unit 36.

The sway measurement data includes data on vertical sway, horizontal sway, and front-back sway. In the present embodiment, the data on vertical sway is used among these sway. Further, as shown in FIG. 5, as the vertical sway of the vehicle 11, the value of the total amplitude in the sway waveform, that is, the value of the difference (absolute value) between the plus and minus peaks of the vertical sway is used.

[S2: Data processing process]
Next, the data processing unit 32 integrates and processes the sway measurement data. Specifically, (1) distance measurement of the sway measurement data, (2) filtering of the sway measurement data, (3) alignment of the sway measurement data, and the like are performed.
(1) Distance conversion of the sway measurement data is a process of converting the sway measurement data, which is the data on the time axis, into the spatial axis data associated with the mileage of the vehicle 11.
(2) The sway measurement data filtering process is a process of bandpass filtering the sway measurement data, which is the spatial axis data, on the spatial frequency axis to generate filtered data.
(3) The alignment of the sway measurement data is a process of aligning the position on the orbit 20 with the actual position in the sway measurement data.

All of the above treatments (1) to (3) are performed by using a known method. For example, the above-mentioned (1) distance measurement of sway measurement data and (2) processing of filtering of sway measurement data can be performed by, for example, the method described in Japanese Patent Application Laid-Open No. 2012-21790. Further, the alignment of the above-mentioned (3) sway measurement data can be performed by, for example, the method described in Japanese Patent Application Laid-Open No. 2020-3247.

[S3: Data visualization process]
Next, the data visualization unit 33 visualizes the sway measurement data processed in step S2 for each position on the orbit 20 (hereinafter, referred to as “km”). FIG. 6 is an explanatory diagram showing an example of visualizing the sway measurement data of the train 10 on different days for each kilometer. In FIG. 6, the sway measurement data from N days to N + 5 days is shown, but the sway measurement data to be visualized is not limited to this.

By visualizing the sway measurement data of the train 10 for each kilometer in this way, it is possible to determine whether or not the peaks (maximum value) and valleys (minimum values) in the waveform of the sway measurement data match on different days. Can be done. Then, if these peaks and valleys match, it can be determined that the sway measurement data is appropriate, and the process proceeds to the subsequent prediction process.

In addition, the data visualization unit 33 visualizes changes over time in the sway measurement data of the trains 10 of a plurality of trains. FIG. 7 is an explanatory diagram showing an example of visualizing the change with time of the sway measurement data of the train 10. The vertical axis of FIG. 7 shows the sway measurement data, and the horizontal axis shows the time. In the example of FIG. 7, the sway (vertical sway) of the vehicle 11 tends to increase with time.

[S4: Upset prediction process]
Next, the sway prediction unit 34 predicts future sway from the sway measurement data processed in step S2 using the prediction model. As the prediction model, a simple regression model constructed by simple regression analysis was used. As shown in FIG. 7, the sway of the vehicle 11 increases with time. Therefore, the present inventors have assumed that future sway can be predicted by obtaining the inclination when the sway of the vehicle 11 is on the vertical axis and the passage of time is on the horizontal axis as shown in FIG. , A simple regression model was used as the prediction model.

Specifically, the prediction model is represented by the following equation (1). Then, the future sway (sway prediction data) of the vehicle 11 is calculated.
y = ax + b ... (1)
However, y: shaking of the vehicle 11 (sway prediction data), x: elapsed time from the reference date, a, b: coefficient

[S5: Orbit prediction process]
As shown in FIG. 8, a predetermined threshold value for the sway of the vehicle 11 is set in advance. The threshold value indicates whether the state of the orbit 20 is normal or abnormal. Then, when the sway prediction data predicted in step S4 is equal to or less than the threshold value, it is predicted that the state of the future orbit 20 is normal. In such a case, maintenance of the track 20 is not necessary. On the other hand, when the sway prediction data is larger than the threshold value, it is predicted that the state of the orbit 20 in the future is abnormal. In such a case, maintenance of the track 20 is required. As described above, the state of the orbit 20 is predicted.

FIG. 9 is an explanatory diagram showing an example of the prediction result of the state of the orbit 20. At point A, the orbit 20 is in a normal state even 28 days after the reference date (N day). On the other hand, at point B, the orbit 20 becomes abnormal 21 days after the reference date, and at point C, the orbit 20 becomes abnormal 28 days after the reference date.

The above steps S1 to S5 are performed to predict the state of the orbit 20. Here, as described above, in the conventional track evaluation method described in Patent Document 1, it is necessary to newly provide the vehicle 11 with a measuring device for measuring the track displacement. On the other hand, in step S1 of the present embodiment, the sway measurement data is acquired by using the existing measurement unit 12 provided in the normal vehicle 11. Therefore, the orbital state can be predicted more easily and cheaply than before.

Further, here, in the method described in Patent Document 1, the measurement target (data source) is the track displacement, whereas in the present embodiment, the measurement target is the shaking of the vehicle 11. Hereinafter, the difference in the measurement targets will be described. First, the sway of the vehicle 11 is speed-dependent, whereas the track displacement is not speed-dependent. Further, while the sway of the vehicle 11 uses the value of all amplitudes as described above, the track displacement uses the measured value. Therefore, if the measurement target is different in the sway and the orbital displacement, the prediction of the state of the orbital 20 is also different.

Further, according to the present embodiment, the prediction model is used in step S4 to predict future sway (sway prediction data) from the sway measurement data, and further, in step S5, from the sway prediction data based on a predetermined threshold of sway. Predict the state of orbit 20. Therefore, the state of the track 20 can be appropriately predicted, and as a result, the maintenance of the track 20 can be efficiently performed.

Further, since the simple regression model is used as the prediction model of the present embodiment, the future sway prediction in step S4 is simple and easy for the user to understand. In addition, by comparing the slopes of the prediction models, it is possible to easily overview how much the sway is increasing.

The prediction model is not limited to the simple regression model. For example, as the prediction model, a multiple regression model constructed by multiple regression analysis may be used, or an autoregressive model constructed by time series analysis may be used.

<Other Embodiments>
Next, other embodiments will be described.

In step S4 of the above embodiment, it is preferable to exclude outliers from the sway measurement data. The outliers can be obtained by referring to the sway measurement data visualized in step S3, for example.

Specifically, when the inventors diligently examined this book, it was found that the outliers were one of the following three data. That is, the outliers are among the sway measurement data.
(1) Data measured by a vehicle 11 having a running speed (knitting speed) equal to or lower than a predetermined speed.
(2) Data measured by a vehicle 11 having a traveling speed different from that of normal traveling in the same kilometer.
(3) Data when the kilometer in the sway measurement data deviates from the actual kilometer,
Is one of.

By excluding these outliers, the accuracy of predicting future turmoil can be further improved.

As described above, in the present embodiment, since the measurement target is the shaking of the vehicle 11 and there is a speed dependence, the same value may be large even if the measurement is performed on the same day. Further, it is the sway of the vehicle 11 measured in the commercial train, and the data allows the position deviation larger than the track displacement. Therefore, excluding outliers as described above is useful for achieving highly accurate predictions. On the other hand, when the measurement target is the orbital displacement as in the conventional case, the variation of the measurement data is small, and it is not necessary to exclude the outliers.

In step S5 of the above embodiment, the state of the orbit 20 may be predicted by further using the rate of increase of the sway prediction data and the average error of the sway prediction data and the sway measurement data in addition to the sway prediction data. .. FIG. 10 shows the values of the agitation prediction data (prediction result) and the agitation measurement data (actual measurement result), the rate of increase of the agitation prediction data (prediction result), and the average of the agitation prediction data (prediction result) and the agitation measurement data (actual measurement result). It is explanatory drawing which shows an example of an error.

The rate of increase in the sway measurement data is the slope of the sway of the vehicle 11 in the prediction model. The sway prediction data is a value of future sway assuming that the sway increases linearly, and if this value is larger than a predetermined threshold value, it is predicted that the state of the future orbit 20 is abnormal. However, even if the sway prediction data is a high value larger than the predetermined threshold value, if the rate of increase is small, that is, if the sway remains high, the response is different, such as paying close attention to the situation rather than immediately preparing the data. For example, it is assumed that a point including a tunnel or a turnout corresponds to this. On the other hand, even if the sway prediction data is low, if the rate of increase is high, it can be a target to watch the situation.

The average error between the sway measurement data and the sway prediction data is the average value of the error between the measured value and the predicted value. When the average error between the sway measurement data and the sway prediction data is large, that is, when the sway measurement data (measured result) is higher than the sway prediction data (prediction result) as shown in FIG. 11, the vehicle It is determined that the point where the rate of increase in the agitation of 11 is rapidly increasing (rapid extension point). Then, a point where the sway measurement data (actual measurement value) is large and the average error is positive (the sway measurement data is larger than the sway prediction data) can be a point of caution.

FIG. 12 is an explanatory diagram showing a guideline for responding to the kilometer with respect to the combination of the sway measurement data (predicted value), the rate of increase of the sway measurement data, and the average error of the sway measurement data and the sway prediction data. Specifically, the measures for whether or not maintenance is necessary are as described above. Then, by using the guideline shown in FIG. 12, the state of the orbit 20 can be predicted and an appropriate response can be taken.

In the above embodiment, the case where the vertical sway is used as the sway of the vehicle 11 has been described, but the horizontal sway and the front-rear sway may be further used.

Although preferred embodiments of the present invention have been described above with reference to the accompanying drawings, the present invention is not limited to such examples. It is clear that a person skilled in the art can come up with various modifications or modifications within the scope of the ideas described in the claims, which naturally belong to the technical scope of the present invention. It is understood as a thing.

The present invention is useful in predicting the state of a track on which a railroad vehicle travels.

10 Train 11 Railroad vehicle 12 Measuring unit 20 Track 21 Rail 22 Sleepers 23 Roadbed 30 Track condition prediction system 31 Communication unit 32 Data processing unit 33 Data visualization unit 34 Sway prediction unit 35 Track prediction unit 36 Storage unit 37 Output unit N network

Claims (16)

It is a method of predicting the state of the track on which a railroad vehicle travels.
A sway measurement process that measures the sway of the railway vehicle using a measuring unit provided on the railway vehicle, and
Using a prediction model, a sway prediction process that predicts future sway from the sway measurement data measured in the sway measurement process,
A method for predicting an orbital state, which comprises an orbital prediction step for predicting an orbital state from the sway prediction data predicted in the sway prediction step based on a predetermined threshold value of the sway. The data processing process for processing the sway measurement data and
The orbital state prediction method according to claim 1, further comprising a data visualization step for visualizing the processed sway measurement data for each position on the orbit. The orbital state prediction method according to claim 1 or 2, wherein the prediction model is constructed by a regression model. The orbital state prediction method according to any one of claims 1 to 3, wherein in the sway prediction step, outliers are excluded from the sway measurement data. The outlier is the data measured by a railroad vehicle having a traveling speed of a predetermined speed or less, the data measured by a railroad vehicle having a traveling speed different from that of normal traveling at a position on the same track, or the above-mentioned sway measurement data. The orbital state prediction method according to claim 4, wherein the position on the orbit in the sway measurement data is one of the data when the position on the orbit deviates from the actual position. The orbital state prediction method according to any one of claims 1 to 5, wherein in the orbital prediction step, the orbital state is predicted by further using the rate of increase of the sway prediction data. The orbital state prediction method according to claim 6, wherein in the orbit prediction step, an error between the sway measurement data and the sway prediction data is further used to predict the orbital state. A program that operates on a computer that controls the orbital state prediction system so that the orbital state prediction method according to any one of claims 1 to 7 is executed by the orbital state prediction system. A readable computer storage medium containing the program according to claim 8. A system that predicts the state of the track on which a railroad vehicle travels.
A storage unit that stores the sway measurement data of the railway vehicle measured using the measurement unit provided on the railway vehicle, and a storage unit.
A sway prediction unit that predicts future sway from the sway measurement data using a prediction model,
An orbital state prediction system comprising: an orbit prediction unit that predicts an orbital state from sway prediction data predicted by the sway prediction unit based on a predetermined threshold value of sway. A data processing unit that processes the sway measurement data,
The orbital state prediction system according to claim 10, further comprising a data visualization unit that visualizes the processed sway measurement data for each position on the orbit. The orbital state prediction system according to claim 10 or 11, wherein the prediction model is constructed by a regression model. The orbital state prediction system according to any one of claims 10 to 12, wherein the sway prediction unit excludes outliers from the sway measurement data. The outlier is the data measured by a railroad vehicle having a traveling speed of a predetermined speed or less, the data measured by a railroad vehicle having a traveling speed different from that of normal traveling at a position on the same track, or the above-mentioned sway measurement data. The orbital state prediction system according to claim 13, wherein the position on the orbit in the sway measurement data is one of the data when the position on the orbit deviates from the actual position. The orbital state prediction system according to any one of claims 10 to 14, wherein the orbital prediction unit predicts the orbital state by further using the rate of increase of the sway prediction data. The orbital state prediction system according to claim 15, wherein the orbital prediction unit predicts the orbital state by further using an error between the sway measurement data and the sway prediction data.

Images