Navigation monitoring device, navigation monitoring method, and navigation monitoring program

Abstract

To precisely evaluate whether a ship is sailing along a sea route.SOLUTION: A storage section 2 stores zone information 4 indicating zones 11-13 generated by dividing a sea route area 10 having a predetermined width along a sea route into polygonal shapes with respect to a sailing direction. A processing section 3 identifies the zone 12 where a sailing ship 20 is positioned among the zones 11-13 based on a current position of the sailing ship 20. The processing section 3 outputs evaluation information to evaluate whether the sailing ship 20 is sailing along the sea route based on the relation between a ground course direction of the sailing ship 20 and the identified zone 12.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to a navigation monitoring device, a navigation monitoring method, and a navigation monitoring program. [Background technology]

[0002] Generally speaking, larger ships have the tendency to have a harder time suddenly changing course or stopping. For this reason, various technologies have been developed to avoid collisions between ships. For example, it is considered to detect the approach of ships or calculate the risk of collision between ships based on the distance between ships and the course of each ship relative to the ground.

[0003] As a related technique, for example, a route display device has been proposed that determines that a ship has deviated from its route if its own position is not within any of the unit route areas set in the route area.Furthermore, a system has also been proposed that detects the direction of change in the ship's position, and based on information correlating the navigation area with the direction of travel, identifies the direction of travel associated with the navigation area identified from the ship's position, and detects reverse navigation within the navigation area of ​​the ship based on the angle between the direction of change and the direction of travel. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Application Publication No. 60-244882 [Patent Document 2] International Publication No. 2017 / 138126 Summary of the Invention [Problem to be solved by the invention]

[0005] However, in the technology for assessing whether a ship is sailing along a route, there is a problem in that it is difficult to make an accurate assessment, particularly when the route is curved. In one aspect, the present invention aims to provide a navigation monitoring device, a navigation monitoring method, and a navigation monitoring program that are capable of accurately evaluating whether a ship is sailing along a route. [Means for solving the problem]

[0006] In one proposal, a navigation monitoring device is provided that has a storage unit and a processing unit as follows. In this navigation monitoring device, the storage unit stores partition information indicating a plurality of partitions generated by dividing a route area of ​​a predetermined width along the route into polygonal shapes in the navigation direction. The processing unit identifies one partition in which the currently navigating ship is located from among the plurality of partitions based on the current position of the currently navigating ship, and outputs evaluation information that evaluates whether the currently navigating ship is navigating along the route based on the relationship between the direction of the currently navigating ship's course over ground and the one partition.

[0007] Also, one proposal provides a navigation monitoring method in which a computer executes the same processing as that of the above-mentioned navigation monitoring device. Furthermore, in one proposal, a navigation monitoring program is provided that causes a computer to execute the same processing as that of the above-mentioned navigation monitoring device. Effect of the Invention

[0008] On the one hand, it can accurately assess whether a ship is staying on course. [Brief description of the drawings]

[0009] [Figure 1] 1 is a diagram illustrating a configuration example and a processing example of a navigation monitoring device according to a first embodiment. [Diagram 2] FIG. 11 is a diagram illustrating an example of the configuration of a navigation support system according to a second embodiment. [Diagram 3] FIG. 2 is a diagram illustrating an example of a hardware configuration of a navigation monitoring device. [Figure 4] FIG. 2 is a diagram illustrating an example of the configuration of processing functions included in the navigation monitoring device. [Diagram 5] 4 is an example of a flowchart showing an overall procedure of a navigation monitoring process performed by a navigation monitoring device. [Figure 6] 11 is a diagram showing a process of determining a collision risk using an extension vector; FIG. [Figure 7] FIG. 13 is a diagram showing a process of determining a collision risk using OZT. [Figure 8] FIG. 13 is a diagram showing a process of determining a collision risk using DCPA and TCPA. [Figure 9] FIG. 1 is a first diagram showing a method for setting a route segment. [Figure 10] FIG. 2 is a second diagram showing a method for setting a route segment. [Figure 11] FIG. 13 is a diagram illustrating an example of setting a reference angle. [Figure 12] FIG. 11 is a first diagram showing a method for determining a reference angle using a statistical method. [Figure 13] FIG. 11 is a second diagram showing a method for determining a reference angle using a statistical method. [Figure 14] FIG. 11 is a third diagram showing a method for determining a reference angle using a statistical method. [Figure 15] 13 is an example of a flowchart illustrating a calculation process procedure of a navigation state indicator when a reference angle is used. [Figure 16] FIG. 11 is a diagram illustrating a method for calculating a navigation state indicator based on a reference angle. [Figure 17] 13 is an example of a flowchart illustrating a calculation process procedure of a binary decision value when a reference angle is used. [Figure 18] FIG. 13 is a diagram showing an example of setting a boundary side. [Figure 19] 13 is an example of a flowchart illustrating a calculation process procedure of a navigation state indicator when a boundary side is used. [Figure 20] 13A and 13B are diagrams illustrating an example of a calculation process of a navigation state indicator when a boundary side is used. [Figure 21] 13 is an example of a flowchart illustrating a calculation process procedure of a binary decision value when a boundary side is used. [Figure 22] FIG. 13 is a diagram showing an example of a rule for determining whether or not a route is being followed. [Diagram 23]10 is a flowchart illustrating an example of a correction process procedure for a result of a determination of the presence or absence of a collision risk. [Figure 24] 13 is a first example of a flowchart showing a correction process procedure for a collision risk. [Diagram 25] FIG. 13 illustrates an example of a vessel pair pattern. [Figure 26] 13 is a second example of a flowchart showing a correction process procedure for a collision risk. [Figure 27] 11 is a diagram showing an example of a transition of a collision risk output from a determination result correction unit. FIG. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0010] Hereinafter, an embodiment of the present invention will be described with reference to the drawings. First Embodiment FIG. 1 is a diagram showing an example of the configuration and processing of a navigation monitoring device according to a first embodiment. The navigation monitoring device 1 shown in FIG. 1 is a device that evaluates whether a ship in progress is navigating along a route and outputs the evaluation result. By using such an evaluation result, it becomes possible to support the navigation of the ship, for example. This navigation monitoring device 1 has a memory unit 2 and a processing unit 3.

[0011] The memory unit 2 is realized by a storage area of ​​a storage device (not shown) included in the navigation monitoring device 1. The memory unit 2 stores section information 4. The section information 4 is information indicating a plurality of sections generated by dividing a route area of ​​a predetermined width along the route into polygonal shapes in the navigation direction. The upper right corner of FIG. 1 illustrates a state in which sections 11 to 13 have been generated for route area 10. The section information 4 includes, for example, information indicating the range of each of sections 11 to 13.

[0012] The processing unit 3 is realized, for example, as a processor (not shown) included in the navigation monitoring device 1. The processing unit 3 executes the following processing using the current position and ground course of the ship underway. Here, it is assumed that the processing is executed for the ship 20 underway, as shown in FIG. 1.

[0013] The processing unit 3 identifies the section in which the ship 20 is located from among the sections 11 to 13 set in the route area 10 based on the current position of the ship 20 (step S1). Here, it is assumed that the section 12 has been identified as shown in FIG. 1. The processing unit 3 outputs evaluation information for evaluating whether the ship 20 is navigating along the route based on the relationship between the direction of the ship 20's course over the ground and the identified section 12 (step S2). This processing makes it possible to accurately evaluate whether the ship 20 is navigating along the route. Note that this evaluation information may be information that evaluates in binary terms whether the ship 20 is navigating along the route, or may be information that evaluates the degree to which the ship 20 is navigating along the route (or the possibility of navigating along the route).

[0014] In Fig. 1, the direction of the course over ground of the ship 20 is indicated by an arrow 20a. In step S2, it is evaluated whether the ship 20 is sailing along the course based on the relationship (e.g., the angle difference) between a reference angle (the angle of the dashed line 12a shown in Fig. 1) that is defined for the section 12 in which the ship 20 is located and serves as a reference for the sailing direction and the direction of the course over ground of the ship 20.

[0015] Here, when evaluating whether the ship 20 is sailing along the route based on the direction of the course over ground, for example, in an area where the route is curved, the direction of the course over ground faces the outside of the route area 10. Therefore, even if the ship 20 is actually sailing along the route without any problem, it may be erroneously determined that it is not sailing along the route.

[0016] In contrast, in this embodiment, the route area 10 is divided into a plurality of sections, and whether the ship 20 is navigating along the route is evaluated for each section corresponding to the position of the ship 20. By evaluating whether the ship 20 is navigating along the route from the relationship between the sections into which the route area 10 is divided and the direction of the ship 20's course over the ground, it becomes possible to evaluate with almost constant accuracy regardless of whether the route is curved. Therefore, it is possible to accurately evaluate whether the ship 20 is navigating along the route.

[0017] Second Embodiment Next, a system in which the evaluation process by the navigation monitoring device 1 shown in FIG. 1 is applied to determining the risk of collision between ships will be described.

[0018] FIG. 2 is a diagram showing an example of the configuration of a navigation support system according to a second embodiment. The navigation support system shown in FIG. 2 is an information processing system for supporting ship navigation. This navigation support system includes a land facility 100 installed on land. The land facility 100 is, for example, a facility that performs navigation control for ships at sea. In this case, the land facility 100 grasps the position of each ship based on AIS (Automatic Identification System) information and radar detection information received from each ship, and provides each ship with various information related to maritime traffic. The AIS information includes, for example, information such as time, ship position (latitude, longitude), speed, ship name, ship type, ship size, and bow direction.

[0019] In this embodiment, the land facility 100 includes a wireless communication device 101 capable of receiving AIS information from each ship, and a navigation monitoring device 102 that monitors the navigation of the ships using the AIS information. The wireless communication device 101 includes, for example, an antenna for receiving the AIS information. The navigation monitoring device 102 is a device having the processing function of the navigation monitoring device 1 shown in FIG. 1, and is realized, for example, as a server computer. The navigation monitoring device 102 detects pairs of ships at risk of collision based on the AIS information received by the wireless communication device 101, and outputs warning information to notify the ships.

[0020] 2 illustrates ships 201 and 202 sailing on the sea. The ships 201 and 202 are equipped with AIS devices 203 and 204, respectively, that can transmit and receive AIS information. The ships 201 and 202 can transmit AIS information about themselves to other ships and land facilities 100 via the AIS devices 203 and 204, respectively, and can receive AIS information transmitted from other ships via the AIS devices 203 and 204, respectively.

[0021] The processing function of the navigation monitoring device 102 may be installed on a ship. For example, the monitoring device installed on the ship detects pairs of ships that are at risk of collision based on AIS information of ships currently sailing, including the ship itself, and displays information about the detected ships on a display device.

[0022] Fig. 3 is a diagram showing an example of the hardware configuration of a navigation monitoring device. The navigation monitoring device 102 is realized, for example, as a computer as shown in Fig. 3. The navigation monitoring device 102 shown in Fig. 3 has a processor 111, a random access memory (RAM) 112, a hard disk drive (HDD) 113, a graphics processing unit (GPU) 114, an input interface (I / F) 115, a reader 116, a communication interface (I / F) 117, and a network interface (I / F) 118.

[0023] The processor 111 performs overall control of the entire navigation monitoring device 102. The processor 111 is, for example, a central processing unit (CPU), a micro processing unit (MPU), a digital signal processor (DSP), an application specific integrated circuit (ASIC), or a programmable logic device (PLD). The processor 111 may also be a combination of two or more elements of a CPU, an MPU, a DSP, an ASIC, or a PLD.

[0024] The RAM 112 is used as a main storage device of the navigation monitoring device 102. The RAM 112 temporarily stores at least a part of an OS (Operating System) program and application programs to be executed by the processor 111. The RAM 112 also stores various data necessary for processing by the processor 111.

[0025] The HDD 113 is used as an auxiliary storage device of the navigation monitoring device 102. The OS program, application programs, and various data are stored in the HDD 113. Note that other types of non-volatile storage devices such as SSDs (Solid State Drives) can also be used as the auxiliary storage device.

[0026] A display device 114a is connected to the GPU 114. The GPU 114 displays an image on the display device 114a in accordance with an instruction from the processor 111. The display device may be a liquid crystal display or an organic EL (ElectroLuminescence) display.

[0027] An input device 115a is connected to the input interface 115. The input interface 115 transmits a signal output from the input device 115a to the processor 111. The input device 115a may be a keyboard or a pointing device. The pointing device may be a mouse, a touch panel, a tablet, a touch pad, a trackball, or the like.

[0028] Portable recording medium 116a is detachably attached to reader 116. Reader 116 reads data recorded on portable recording medium 116a and transmits it to processor 111. Portable recording medium 116a may be an optical disk, a magneto-optical disk, a semiconductor memory, or the like.

[0029] The communication interface 117 transmits and receives data to and from other devices, such as the wireless communication device 101 . The network interface 118 transmits and receives data to and from other devices via a network 118a.

[0030] The processing functions of the navigation monitoring device 102 can be realized by the above hardware configuration. 4 is a diagram showing an example of the configuration of processing functions of the navigation monitoring device 102. As shown in FIG. 4, the navigation monitoring device 102 includes a storage unit 120, a navigation data acquisition unit 131, a collision risk determination unit 132, a preprocessing unit 133, a navigation state determination unit 134, and a determination result correction unit 135.

[0031] The storage unit 120 is realized as a storage area of ​​a storage device included in the navigation monitoring device 102, such as the RAM 112 or the HDD 113. The storage unit 120 stores a navigation history database (DB) 121, a route segment database (DB) 122, a reference angle database (DB) 123, and a boundary side database (DB) 124.

[0032] The navigation history database 121 stores past navigation data for each of a plurality of ships. The navigation data is mainly data based on AIS information received from the ship. The navigation data includes at least the ship's position (latitude, longitude), course over ground, and speed over ground. The navigation data may also include the ship's destination. The navigation history database 121 stores, for example, route data based on AIS information received from the ship during a specified period in the past.

[0033] Information about route segments set in a route area is stored in the route segment database 122. A route segment is generated by dividing a route area by polygons. A reference angle corresponding to each route segment is stored in the reference angle database 123. The reference angle indicates the direction of the route in the route segment.

[0034] Information indicating a boundary edge corresponding to each route segment is stored in the boundary edge database 124. A boundary edge refers to, among the edges of a route segment, an edge that forms a boundary between an adjacent route segment. At least one of the reference angle and the boundary side is used in the actual processing of the navigation monitoring device 102. Therefore, it is sufficient that the storage unit 120 stores at least one of the reference angle database 123 and the boundary side database 124.

[0035] The processing of the navigation data acquisition unit 131, the collision risk determination unit 132, the pre-processing unit 133, the navigation state determination unit 134, and the determination result correction unit 135 is realized, for example, by the processor 111 executing a predetermined application program.

[0036] The navigation data acquisition unit 131 acquires AIS information received from the ship by the wireless communication device 101, and generates navigation data based on the AIS information. For example, the position and speed of the ship included in the navigation data are extracted from the AIS information. Also, the ground course included in the navigation data is calculated based on, for example, the position and speed information of the AIS information received during a certain period from the present to the past.

[0037] The navigation data acquired by the navigation data acquisition unit 131 is stored in the navigation history database 121. In addition, the navigation data acquired by the navigation data acquisition unit 131 is input to the collision risk determination unit 132 and the navigation state determination unit 134, and is used to evaluate the collision risk between ships currently sailing.

[0038] The collision risk determination unit 132 determines the collision risk between ships sailing on the basis of the navigation data acquired by the navigation data acquisition unit 131. For example, the collision risk determination unit 132 identifies a pair of ships that is at risk of collision on the basis of the extension vector of each ship. Alternatively, the collision risk determination unit 132 may calculate an index indicating the collision risk for the pair of ships on the basis of the Distance to Closest Point of Approach (DCPA) or the Time to Closest Point of Approach (TCPA) between the ships. The collision risk determination unit 132 may detect an Obstacle Zone by Target (OZT) between one ship and another ship.

[0039] The pre-processing unit 133 executes pre-processing for generating information necessary for the determination process of the navigation state determination unit 134. For example, the pre-processing unit 133 sets route segments for the route area, and sets reference angles and boundary sides for each route segment. The pre-processing unit 133 can also determine the reference angle for each route segment by statistically analyzing the navigation record data accumulated in the navigation history database 121.

[0040] The navigation state determination unit 134 determines the navigation state of the ship based on the navigation data of the ship. The navigation state indicates whether the ship is sailing along the route or the degree to which the ship is sailing along the route. Such a navigation state is determined based on at least one of a reference angle and a boundary side corresponding to the route segment including the current position of the ship.

[0041] The determination result correction unit 135 corrects the determination result by the collision risk determination unit 132 using the determination result by the navigation state determination unit 134. This correction reduces the possibility of erroneous determination by the collision risk determination unit 132.

[0042] Figure 5 is an example of a flowchart showing the overall procedure of navigation monitoring processing by a navigation monitoring device. The processing in Figure 5 is to extract ship pairs from multiple ships currently sailing and determine the collision risk for each ship pair. A ship currently sailing refers to a ship from which AIS information is received and navigation data based on the AIS information is acquired.

[0043] First, the collision risk determination unit 132 determines the collision risk for each ship pair based on the navigation data acquired from each ship currently sailing by the navigation data acquisition unit 131 (step S11). In this process, all combinations of ship pairs are extracted from all ships currently sailing, and the collision risk for all extracted ship pairs is determined. As a determination result, binary data indicating the presence or absence of a collision risk is output for each ship pair. In this case, ship pairs determined to have a collision risk may be identified and output from among the ship pairs. Alternatively, as a determination result, a collision risk level, which is a quantitative index indicating the degree of collision risk, is output for each ship pair. These processes by the collision risk determination unit 132 are performed using existing methods.

[0044] In parallel with this collision risk determination process, the navigation state of each ship currently sailing is determined by the navigation state determination unit 134 (step S12). In this process, based on the navigation data acquired from each ship by the navigation data acquisition unit 131, binary data (hereinafter referred to as "binary determination value") indicating whether the ship is following the route is output. Alternatively, based on such navigation data, a quantitative index (hereinafter referred to as "navigation state index") indicating the degree to which the ship is following the route is output. Such processing by the navigation state determination unit 134 is performed using at least one of the reference angle for each route segment registered in the reference angle database 123 and the boundary edge for each route segment registered in the boundary edge database 124.

[0045] Then, as the final processing step, the collision risk determination result determined for each ship pair by the collision risk determination unit 132 is corrected using the navigation state determination result determined for each ship included in the ship pair by the navigation state determination unit 134 (step S13). Particularly in this embodiment, in order to reduce the possibility of erroneously determining that there is a collision risk or that the collision risk is high based on the former determination, a correction is made using the latter determination result.

[0046] <1. Judging the risk of collision> First, a description will be given of the collision risk determination process (step S11 in FIG. 5) performed by the collision risk determination unit 132. The collision risk determination unit 132 determines the collision risk using any one of the following determination methods (1-1) to (1-3).

[0047] (1-1) Judgment method using extension vectors Fig. 6 is a diagram showing a process of determining a collision risk using an extension vector, in which Fig. 6(A) shows a first example of an extension vector of a ship pair, and Fig. 6(B) shows a second example of an extension vector of a ship pair.

[0048] The extension vector is a vector having the same direction (azimuth) as the ship's course (ground course in this embodiment) and a magnitude that is the distance from the ship's current position to a position that the ship will reach after a predetermined time (generally 5 minutes) if it maintains its current speed. Fig. 6(A) illustrates an extension vector V11 of the ship 211 and an extension vector V12 of the ship 212. Fig. 6(B) illustrates an extension vector V13 of the ship 213 and an extension vector V14 of the ship 214.

[0049] In the process of determining the risk of collision using the extension vectors, it is determined that there is a risk of collision when the extension vectors of a pair of ships intersect. In Fig. 6(A), the extension vectors V11 and V12 intersect, and it is determined that there is a risk of collision between the ships 211 and 212. In Fig. 6(B), the extension vectors V13 and V14 intersect, and it is determined that there is a risk of collision between the ships 213 and 214.

[0050] (1-2) Evaluation method using OZT FIG. 7 is a diagram showing the process of determining the risk of collision using OZT. OZT (obstruction zone) refers to an area where a ship may collide with another ship (opponent ship) sailing nearby if it were to change course. OZT is calculated based on the position, course over ground, ship speed, etc. of each ship. For example, the probability that the ship and the opponent ship are simultaneously in the same position is calculated taking into account measurement errors such as ship speed, and an area where this probability is equal to or exceeds a predetermined threshold is determined to be an OZT.

[0051] 7, an OZT 221 is detected between the vessel 215 and the other vessel 216. When such an OZT 221 is detected, it may be determined that there is a risk of collision between the pair of vessels 215, 216.

[0052] (1-3) Judgment method using TCPA and DCPA Fig. 8 is a diagram showing a process of judging the risk of collision using DCPA and TCPA. In the methods shown in (1-1) and (1-2) above, binary data indicating the presence or absence of a risk of collision is output. In contrast, there is a method of calculating a quantitative index (collision risk) indicating the degree of the risk of collision by using DCPA or TCPA.

[0053] In this method, the closest point of approach (CPA) where the other ship comes closest to the ship is obtained from the relative motion based on the navigation record data of each ship pair. Then, the closest distance of approach (DCPA) from the ship to the CPA and the closest time of approach (TCPA) indicating the time it takes for the other ship to reach the CPA are calculated. Figure 8 illustrates an example in which a CPA 222 is detected between ship 217 and its other ship 218, and distance D is calculated as the DCPA.

[0054] At least one of DCPA and TCPA can be used to calculate the collision risk. Basically, the shorter the DCPA and the shorter the TCPA, the higher the collision risk is output. In addition, as a technique for calculating the collision risk by combining DCPA and TCPA, for example, "Takashi Kubota & Nobuyoshi Kawaguchi (2009), Study on Navigator's Decision-Making and Action Model II - Overall Subjective Collision Risk -, Journal of the Japan Institute of Navigation, March 2010, pp. 21-26" and "Chin, Hoong Chor & Debnath, Ashim (2009). Modeling perceived collision risk in port water navigation. Safety Science, 47(10), pp. 1410-1416." are available.

[0055] (1-4) Problems with the above judgment method The determination methods shown in (1-1) to (1-3) above have a problem in that the determination accuracy deteriorates, particularly when each of the pair of ships is navigating a winding route.

[0056] For example, Fig. 6(A) illustrates a case where ships 211 and 212 are sailing in the same direction in a curved route region 210. In the state shown on the left side of Fig. 6(A), the extension vector V11 of the ship 211 and the extension vector V12 of the ship 212 intersect, so it is determined that there is a possibility of collision between the ships 211 and 212. However, if the ships 211 and 212 are both sailing along the route, as in the example on the right side of Fig. 6(A), the ships 211 and 212 will not collide, and the state in which the extension vectors V11 and V12 intersect will soon be resolved.

[0057] Moreover, Fig. 6(B) illustrates a case where ships 213, 214 are sailing in opposition to each other in the same route region 210. In the state shown on the left side of Fig. 6(B), the extension vector V13 of the ship 213 and the extension vector V14 of the ship 214 intersect, so it is determined that there is a possibility of collision between the ships 213, 214. However, when the ships 213, 214 are both sailing along the route, as in the example on the right side of Fig. 6(B), the ships 213, 214 pass each other without colliding, and the state in which the extension vectors V13, V14 intersect is soon resolved.

[0058] Since the extension vector is a vector that extends linearly in the course direction, when a pair of ships simultaneously approach a bend in the sea route, their extension vectors may temporarily cross each other even if they are navigating safely without colliding. For this reason, a pair of ships that are navigating safely without colliding may be erroneously detected as having a collision risk. In particular, the determination method using the extension vector has high accuracy in detecting the risk of collision between a ship on a sea route and a ship crossing the sea route. However, at a bend in the sea route, it is difficult to distinguish between a case where a collision risk occurs due to a ship crossing the sea route and a case where both ships are along the sea route.

[0059] Fig. 7 also illustrates a case where ships 215, 216 are sailing in opposite directions in the same route region 210. In the state shown on the left side of Fig. 7, an OZT 221 is detected between the ships 215 and 216. However, when the ships 215, 216 are both sailing along the route, the ships 215, 216 pass each other without colliding, as in the example on the right side of Fig. 7.

[0060] OZT is assumed to be the case when one ship changes course at the current time, and the subsequent movement of that ship is estimated as a straight line movement. As in the example of Figure 7, when OZT is calculated between two ships approaching a bend in the sea route, a dangerous area is detected where the ship changes course from its current location. However, if each ship is following the sea route, it will not enter the detected dangerous area because it will change course after proceeding a little further. Therefore, dangerous areas that do not require caution may be detected, and the detection results may lead to a false judgment that there is a risk of collision.

[0061] Furthermore, Fig. 8 illustrates an example in which ships 217, 218 are sailing in opposite directions in the same route region 210. In the state shown on the left side of Fig. 8, a distance D is calculated as the DCPA between the ships 217 and 218, but this distance D is short and a relatively high value is calculated as the collision risk. However, when the ships 217, 218 are both sailing along the route, the ships 217, 218 pass each other without colliding, as in the example on the right side of Fig. 8.

[0062] DCPA and TCPA are calculated on the assumption that the course and speed will be maintained from the current position. For this reason, as in the example of Figure 8, the values ​​tend to be small in curved sections of the route, and the collision risk is likely to be calculated as high.

[0063] As described above, the determination methods shown in (1-1) to (1-3) have the problem that when a pair of ships are navigating a curved route, it is easy to erroneously determine that there is a collision risk, or to calculate the collision risk to be higher than it actually is. Therefore, in this embodiment, the determination accuracy is improved by correcting the determination result by the collision risk determination unit 132 using the determination result by the navigation state determination unit 134. The navigation state determination unit 134 calculates the determination result of whether or not each ship is following the route, or a navigation state index that is a quantitative index showing the degree to which the ship is following the route.

[0064] For example, even if a collision risk is erroneously determined as being present for a pair of ships as described above, it can be re-determined that there is no collision risk if each ship is along the sea route. Also, even if the collision risk for a pair of ships is calculated to be higher than the actual risk as described above, the collision risk can be corrected to be lower if the navigation condition index for each ship is high (they are highly along the sea route). In this way, the possibility of erroneous determination of collision risk or of a collision risk being calculated to be higher than the actual risk is reduced.

[0065] <2. Determining navigation status> Next, the process of determining the navigation state by the navigation state determination unit 134 (step S12 in FIG. 5) will be described. This determination process can be broadly divided into a method using a reference angle and a method using a boundary side. Either of these methods uses multiple route segments set in the route area. First, a method of setting the route segment will be described. The process of setting the route segment is executed by the pre-processing unit 133.

[0066] FIG. 9 is a first diagram showing a method for setting a route segment. A route segment is a polygonal area generated by dividing a route area in the navigation direction. In this embodiment, the route segment is a rectangle as an example, but the polygonal shape of the route segment is not particularly limited. For example, the route segment may be a rectangle in a straight section of the route, and the route segment may be a pentagon or hexagon in a bent section of the route. In addition, the "route area" here refers to the area of ​​the route in which a ship can navigate. In other words, the route has a width, and the entire area in the width direction becomes the route area.

[0067] Furthermore, a route may be navigable in only one direction or in both directions. In the former case, a route in one direction and a route in the opposite direction are generally set in parallel. In this case, a route segment is set for each route in each direction.

[0068] The route area 220 illustrated in the upper part of Fig. 9 is divided into a route area 221 in the forward direction (from the top to the bottom in Fig. 9) and a route area 222 in the reverse direction (from the bottom to the top in Fig. 9). In this case, route segments are set individually for each of the route areas 221, 222. In the example shown in the lower part of Fig. 9, route segments DV1 to DV5 are set in order in the route direction in the route area 221, and route areas DV11 to DV15 are set in order in the route direction in the route area 222.

[0069] The pre-processing unit 133 registers information indicating the position and range of each set route segment in the route segment database 122. By referring to the route segment database 122, it becomes possible to identify the route segment in which the ship is located based on the ship's position information (latitude, longitude).

[0070] Fig. 10 is a second diagram showing a method for setting route segments. Route segments are set, for example, by dividing a route area at a fixed distance along the route direction. In areas where the route is curved, it is desirable to set route segments more finely than in areas where the route is straight. This is to prevent the angle difference between the reference angle set in the route segment and the course over the ground of the navigating ship from becoming too large, which deteriorates the accuracy of judging the navigation state.

[0071] As an example, when a route method is divided at a certain distance, a route segment and other route segments located on both sides of the route segment are connected in a bent state (not in a straight line state). In the example on the left side of FIG. 10, route segments DV21-DV23 are set at a bent part of the route, and both route segments DV21 and DV23 are connected to the route segment DV22 in a bent state. When such a route segment DV22 exists, the route area is divided so that new route segments are generated on both sides of one end of the route segment DV22 and on both sides of the other end of the route segment DV22. On the right side of FIG. 10, route segments DV22a and 22b are generated on both sides of the boundary between the route segment DV22 and the route segment DV21, and route segments DV22c and 22d are generated on both sides of the boundary between the route segment DV22 and the route segment DV23. Along with this, the length of the route segment DV21 is reduced to become the route segment DV21a, and the length of the route segment DV23 is also reduced to become the route segment DV23a.

[0072] As an example of a rule for setting route segments in detail in a bending region, a route segment is not connected to both other route segments located on both sides in a bending state (non-linear state). In the example on the left side of FIG. 10, route segments DV21 to DV23 are set at a bending portion of the route. Of these, route segments DV21 and DV23 are both connected to route segment DV22 in a bending state. In this case, as shown on the right side of FIG. 10, route segment DV22 is further divided into two route segments DV22a and DV22b that are linearly connected to each other.

[0073] (2-1) Determining navigation status based on reference angle Next, a process for determining the navigation state based on the reference angle will be described. In this process, a reference angle set for each route segment is used. First, a method for setting the reference angle will be described. The process for setting the reference angle is executed by the pre-processing unit 133.

[0074] Fig. 11 is a diagram showing an example of setting a reference angle, Fig. 11(A) shows an example of setting a reference angle for a one-way route, and Fig. 11(B) shows an example of setting a reference angle for a two-way route.

[0075] The reference angle is an azimuth angle that is the reference for the navigation direction in a route segment. The reference angle is the reference for determining whether a ship navigating the route segment is along the route, and is also the reference for calculating the degree to which the ship is along the route (navigational state index). It can also be said that the reference angle indicates the direction of the average course over the ground that a ship navigating along the route takes within the route segment.

[0076] As shown in Fig. 11(A), in a route where a ship can only navigate in one direction, a reference angle is set for each route segment in only one direction. In Fig. 11(A), the solid arrows shown in each of the route segments DV31 to DV34 indicate the direction corresponding to the set reference angle.

[0077] On the other hand, as shown in Fig. 11(B), in a route where a ship can navigate in both directions, a reference angle is set for each route segment for each direction. In Fig. 11(B), solid arrows and dashed arrows are shown for each of the route segments DV41 to DV44. The solid arrows indicate the direction corresponding to the reference angle for the forward direction, and the dashed arrows indicate the direction corresponding to the reference angle for the reverse direction.

[0078] The pre-processing unit 133 determines the reference angle of the route segment based on, for example, the shape information of the route segment. For example, from among the sides of the route segment, a side that contacts an adjacent route segment at the entrance side of the route and a side that contacts an adjacent route segment at the exit side of the route are identified. Then, the azimuth angle of a line connecting the midpoint of the former side to the midpoint of the latter side is determined as the reference angle. As another method, a straight line passing through the center of the route in the width direction may be determined by an approximation calculation in the route segment, and the azimuth angle of the line may be determined as the reference angle.

[0079] The preprocessing unit 133 registers the reference angle thus set for each route segment in the reference angle database 123 in association with the route segment. By referring to the reference angle database 123, it becomes possible to identify the reference angle set for the route segment from the identification information of the route segment.

[0080] Furthermore, the pre-processing unit 133 may determine the reference angle of a route segment by a statistical method based on navigation record data of ships that have passed through the route segment in the past, as illustrated in the following Figures 12 to 14. Note that, as an example, in Figures 12 to 14, a "reference angle when along the route" and a "reference angle when not along the route" are set for each route segment, but only the former reference angle may be set.

[0081] Fig. 12 is a first diagram showing a method for determining a reference angle using a statistical method. Fig. 12(A) shows an example of a histogram in the case of a one-way route, and Fig. 12(B) shows an example of a histogram in the case of a two-way route.

[0082] In the method shown in FIG. 12, ships that have passed through a sea route segment in the past are classified as along-route ships and off-route ships. An along-route ship refers to a ship that has passed through a sea route segment and then navigated without deviating from the sea route domain. On the other hand, an off-route ship refers to a ship that has passed through a sea route segment and then deviated from the sea route domain. The latter ship also includes a ship that has been located in the sea route segment and then deviated from the same sea route segment to outside the sea route domain.

[0083] The pre-processing unit 133 acquires the azimuth of the ground course for each ship along the route when it was in the route segment. If a ship is detected multiple times in the route segment as it moves, the azimuth of the ground course at each detection time is acquired. The pre-processing unit 133 calculates a histogram of the azimuth angles of the ground course acquired in this way (histogram along the route).

[0084] Similarly, for each ship that is not along the sea route, the pre-processing unit 133 acquires the azimuth angle of the course over the ground when the ship was in the sea route segment. The pre-processing unit 133 calculates a histogram of the acquired azimuth angle of the course over the ground (a histogram when not along the sea route).

[0085] 12(A), for a route segment in one direction, one histogram is calculated for the case along the route and one histogram for the case not along the route. The pre-processing unit 133 determines the azimuth angle with the highest frequency in the former histogram as the reference angle θ1 for the case along the route, and determines the azimuth angle with the highest frequency in the latter histogram as the reference angle θ2 for the case not along the route.

[0086] As shown in Fig. 12(B), for route segments in both directions, histograms for along-route and off-route are calculated for the forward and reverse directions, respectively. That is, the pre-processing unit 133 calculates a histogram for along-route and a histogram for off-route based on the route record data of a ship proceeding in the forward direction. The pre-processing unit 133 calculates the azimuth angle with the highest frequency in the former histogram as the reference angle θ for along-route. 1a The azimuth angle with the highest frequency in the latter histogram is determined as the reference angle θ 2a In addition, the pre-processing unit 133 calculates a histogram for the case along the route and a histogram for the case not along the route based on the route record data of the ship proceeding in the opposite direction. The pre-processing unit 133 determines the azimuth angle with the highest frequency in the former histogram as the reference angle θ 1b The azimuth angle with the highest frequency in the latter histogram is determined as the reference angle θ 2b is decided.

[0087] In such a case, the pre-processing unit 133 registers two reference angles for the case along the route and two reference angles for the case not along the route in the reference angle database 123 in association with the route segments.

[0088] Fig. 13 is a second diagram showing a method for determining a reference angle using a statistical method. Even within the same route segment, the optimal reference angle may differ depending on the location. Therefore, each route segment may be further divided into smaller sub-segments, and a histogram may be calculated for each sub-segment using the method shown in Fig. 12(A) or 12(B), and the reference angle may be determined based on each histogram.

[0089] FIG. 13 shows an example in which the route segment DV51 is divided into smaller grids. For each grid, the pre-processing unit 133 calculates a histogram for when the route is along the route and a histogram for when the route is not along the route based on the navigation record data of ships that have passed through the grid in the past. The determination of whether the route is along the route or not is made based on the manner in which the ship passes through the route area after the route segment DV51, as in the case of FIG. 12(A) or FIG. 12(B). For each grid, the pre-processing unit 133 determines a reference angle for when the route is along the route and a reference angle for when the route is not along the route based on the calculated histogram.

[0090] Fig. 14 is a third diagram showing a method for determining a reference angle using a statistical method. The reference angle may be determined as a continuous quantity according to position information (latitude, longitude) within a route segment. Here, as an example, a case will be described in which the reference angle is determined by calculation using the position within the route segment as an input.

[0091] The pre-processing unit 133 classifies ships that have passed through the route segment in the past into ships that are along the route and ships that are not along the route, using a method similar to that of FIG. 12(A) or FIG. 12(B). The pre-processing unit 133 extracts, from the navigation history database 121, position information when each ship along the route was detected within the route segment and the azimuth of the course over the ground at that time. As shown in FIG. 14, the pre-processing unit 133 plots each ship along the route on a coordinate space having the azimuth of the course over the ground and the position information of the ship as coordinate axes. Based on these plot positions in the coordinate space, the pre-processing unit 133 generates a calculation formula that takes the position information of the ship as input and outputs the azimuth. A curve 231 shown in FIG. 14 indicates the relationship between input and output in such a calculation formula. Such a calculation formula is generated, for example, by kernel density estimation.

[0092] This formula is used to find the reference angle when following a route, and the preprocessing unit 133 registers information (such as coefficients) indicating this formula in the reference angle database 123 in association with the route segment. This makes it possible to find a reference angle according to the ship's position in the route segment in which the ship is located, based on the position information of the ship currently sailing.

[0093] A calculation formula for calculating the reference angle when not along the sea route is also generated by the same method as above. That is, the pre-processing unit 133 extracts the position information when each ship that did not follow the sea route was detected in the sea route segment and the azimuth angle of the course over the ground at that time from the navigation history database 121. As shown in FIG. 14, the pre-processing unit 133 plots each ship that did not follow the sea route on a coordinate space having the azimuth angle of the course over the ground and the ship's position information as coordinate axes. Based on these plot positions in the coordinate space, the pre-processing unit 133 generates a calculation formula that inputs the ship's position information and outputs the azimuth angle. A curve 232 shown in FIG. 14 indicates the input / output relationship in such a calculation formula. As a result, a calculation formula for calculating the reference angle when not along the sea route is generated, and information indicating the calculation formula is registered in the reference angle database 123 in association with the sea route segment.

[0094] In the coordinate space shown in FIG. 14, for ease of explanation, the location information is shown as one axis, but in reality, the location information has two axes, the latitude and the route. In the above description, the reference angle is set for each route segment, but it may also be set for each of various attributes related to the ship. For example, a reference angle may be set for each ship type for each route segment. Also, a reference angle may be set for each route segment for each destination of the ship that passed through the route segment. Furthermore, a reference angle may be set for each route segment for each time period when the ship passed through the route segment.

[0095] Next, a procedure for the process of determining the navigation state using the reference angle will be described. As described above, in this determination process, either a navigation state indicator indicating the degree to which the ship is along the route or a binary determination value indicating whether the ship is along the route is output as a determination result.

[0096] (2-1-1) Calculation of navigation status index First, the calculation process of the navigation state index when the reference angle is used will be described. As described above, the navigation state index indicates the degree to which the ship is along the route. In other words, the navigation state index can be the degree to which the ship is estimated to be sailing along the route, or the possibility that the ship will continue to sail along the route for a certain period of time from now.

[0097] Fig. 15 is an example of a flowchart showing a calculation process procedure of a navigation state index when a reference angle is used. The process of Fig. 15 is executed for each ship that is sailing. When the navigation data acquisition unit 131 acquires navigation data for the ship, the process of Fig. 15 is executed.

[0098] [Step S21] The navigation data acquisition unit 131 calculates the ship's course over the ground based on the most recently acquired navigation data and navigation data acquired about the same ship during a period from the present time up to a predetermined time ago.

[0099] [Step S22] The navigation state determination unit 134 refers to the route segment database 122 and identifies the route segment to which the ship belongs (including the ship's current position) from the ship's position information (latitude, longitude) based on the latest navigation data.

[0100] [Step S23] The navigation state determination unit 134 determines whether the identified route segment is a segment of a route that is navigable in both directions. If it is a segment of a route that is navigable in only one direction, the process proceeds to step S24. If it is a segment of a route that is navigable in both directions, the process proceeds to step S25.

[0101] [Step S24] The navigation state determination unit 134 obtains, from the reference angle database 123, a reference angle corresponding to the identified route segment. [Step S25] The navigation state determination unit 134 acquires reference angles for two navigation directions (forward or reverse) corresponding to the identified route segment from the reference angle database 123. The navigation state determination unit 134 calculates the angle difference between the two acquired reference angles and the azimuth angle of the ground course calculated in step S21, and acquires the reference angle with the smaller calculated angle difference as the reference angle corresponding to the navigation direction of the ship.

[0102] When a reference angle is registered for each subdivision in a route segment as shown in Fig. 13, the reference angle corresponding to the subdivision identified based on the ship's position information is acquired in steps S24 and S25. When the reference angle is defined as a continuous quantity as shown in Fig. 14, the reference angle is calculated based on the ship's position information in steps S24 and S25.

[0103] [Step S26] The navigation state determination unit 134 calculates the angle difference between the acquired reference angle and the azimuth angle of the ground course calculated in step S21. If the angle difference has already been calculated in step S25, the already calculated angle difference may be acquired.

[0104] [Step S27] The navigation state determination unit 134 calculates a navigation state index, which indicates the degree to which the ship is along the route, based on the angle difference calculated in step S26. The calculation method will be described with reference to the following FIG.

[0105] Fig. 16 is a diagram showing a method of calculating a navigation state index based on a reference angle, Fig. 16(A) to (C) respectively showing first to third methods of calculating the index. In the first index calculation method shown in FIG. 16(A), the angle difference (θ 11 A navigation state index is calculated based on the maximum value of the angle difference (step S27a). For example, a method of calculating the navigation state index according to the following formula (1) is considered. Navigation condition index=(90-|θ 11 |) / 90 ···(1) The "90" in equation (1) stands for |θ 11 |The normal maximum value (maximum angle) of |θ is 90 degrees. 11 If |>90, the navigation status index should be set to 0.

[0106] Here, in formula (1), the maximum value of the angle difference is 90 degrees, but the maximum value of the angle difference may differ depending on the route segment and may differ depending on the left and right directions relative to the route. For example, at a bend that turns left, if the ship is heading left of the reference angle, it is highly likely that the ship is changing course to turn along the route, so the navigation state index will be high. On the other hand, if the ship is heading right of the reference angle, it is highly likely that the ship is about to deviate from the route, so the navigation state index will be low. Therefore, for example, the navigation state index can be calculated in the following way.

[0107] For example, the parameter Q1 ([%]) is set. In addition, from the histograms of the enveloping angles of the course over the ground when following the route shown in Figures 12 and 13, the maximum angle difference L1 in the counterclockwise direction and the maximum angle difference R1 in the clockwise direction are calculated by the following formulas (2a) and (2b), respectively, assuming that the angle is clockwise from the reference angle to be positive (+) and counterclockwise to be negative (-). L1={(100-Q1) / 2}% quantile ···(2a) L2={(100-Q1) / (2+Q)}% quantile...(2b) According to equations (2a) and (2b), for example, if Q1=95%, L1 is the 2.5% quantile and L2 is the 97.5% quantile.

[0108] L1, R1, θ for each route section 11 Then, the navigation status index is calculated by the following equations (3a) and (3b). θ 11 If ≦0: Navigation status index = (L1-θ 11 ) / L1 (3a) θ 11 If >0: Navigation status index = (R1-θ 11) / R2 ···(3b) Note that the navigation state indicator may be output as a signed indicator. For example, when the azimuth of the ship's course relative to the ground is inclined to the right side of the navigation direction with respect to the reference angle, a plus sign is added to the navigation state indicator, and when it is inclined to the left side, a minus sign is added to the navigation state indicator.

[0109] In the second index calculation method shown in FIG. 16(B), the navigation state indicator is calculated using the angle difference θ calculated in step S26 of FIG. 15 (step S27b). In this step S27b, for each ship that has passed through the route section in the past and then sailed along the route, the angle difference between the azimuth of the course relative to the ground in the route section and the reference angle is the above angle difference θ 11 and the past navigation data stored in the navigation history database 121. In this step S27b, for each ship that has passed through the route section in the past and then sailed along the route, the probability that the angle difference between the azimuth of the course relative to the ground in the route section and the reference angle becomes the above angle difference θ 11 is obtained, and the navigation state indicator is calculated based on this probability. As an example, the following calculation method can be used.

[0110] Based on the navigation history database 121, ships that have passed through the corresponding route section in the past and then sailed along the route are identified. Then, a histogram of the angle difference (when the course relative to the ground is clockwise from the reference angle, it is +, and when it is counterclockwise, it is -) between the course relative to the ground recorded by the identified ship in the route section and the reference angle corresponding to this route section is created. Assume that this histogram follows a specific probability distribution. For example, assume that it is a normal distribution represented by the mean μ 11 and the standard deviation σ 11 of the angle difference calculated from this histogram.

[0111] On this normal distribution, the probability P 11 that the angle difference is within |θ 11 -μ 11 | is obtained. This probability P 11 takes a smaller value as θ 11 is closer to the mean. Also, parameters Pa and Pb are set. However, 0 ≦ Pa < Pb ≦ 1. At this time, the navigation state indicator is calculated by the following equations (4a) to (4c). P 11 When P ≤ Pa: Navigation status indicator = 1 ···(4a) When Pa < P 11 and P < Pb: Navigation status indicator = 1 - (P 11 - Pa) / (Pb - Pa) ···(4b) When Pb ≤ P 11 : Navigation status indicator = 0 (4c) Also in the third indicator calculation method shown in FIG. 16(C), similar to the second indicator calculation method, the angle difference θ calculated in step S26 of FIG. 15 11 and the past navigation data stored in the navigation history database 121 are used to calculate the navigation status indicator. However, in the third indicator calculation method, the respective reference angles for the case of following the route and the case of not following the route, calculated by the methods shown in FIGS. 12 to 14, are used. And among the ships that have passed through the route section in the past, the ships that subsequently navigated along the route and the ships that deviated from the route are considered.

[0112] For example, in step S26 of FIG. 15, the reference angle for the case of following the route and the reference angle for the case of not following the route are acquired from the reference angle database 123. And the angle difference θ 11 between the azimuth of the course over ground and the reference angle for the case of following the route is calculated (step S26a), and at the same time, the angle difference θ 12 between the azimuth of the course over ground and the reference angle for the case of not following the route is calculated (step S26b). Then, by inputting the calculated angle differences into a learning model as follows, the navigation status indicator is calculated (step S27c).

[0113] The learning model is created, for example, as follows. For each of the ships that have passed through the route section in the past, the navigation result flag F1 and the angle differences θ 11 , θ 12 are calculated. The navigation result flag F1 is a flag that is TRUE when the corresponding ship subsequently navigated along the route and FALSE when it subsequently deviated from the route. The angle differences θ 11 , θ 12are calculated in the same manner as in steps S26a and S26b, respectively. Then, a learning model (such as a support vector machine) is created in which the navigation result flag F1 is an objective variable and the angle difference θ is an objective variable.

[0114] In step S27c, the angle difference calculated in step S26a and the angle difference calculated in step S26b are input into the above-mentioned learning model, and among the probabilities output from the learning model, the navigation result flag F1 is TRUE, and is output as a route condition index.

[0115] The above first to third index calculation methods are merely examples, and other calculation methods can also be used. The navigation state index calculated by the above method may be corrected in the following manner, taking into account the connection between route segments. For example, the angle difference between the reference angle of a route segment and the reference angle of the adjacent route segment (the previous route segment) ahead of the route direction is calculated, and an angle range is set centered on the reference angle of the previous route segment according to the angle difference. The larger the calculated angle difference, the larger the angle range is set. Then, when the navigation state index is provisionally calculated by the above method for a ship located in the former route segment, it is determined whether the azimuth angle of the ship's course over the ground is included in the angle range, and if it is included in the angle range, the provisionally calculated navigation state index is corrected to a larger value. This correction is performed by multiplying a coefficient exceeding 1, for example, 1.2.

[0116] When the angular difference between the reference angle of a route segment and the reference angle of the immediately preceding route segment is large, the ship's course is likely to change significantly in the former route segment. Therefore, in the former route segment, even if the angular difference between the azimuth angle of the ship's ground heading and the reference angle is large, there is a high possibility that the ship will continue to navigate along the route. By the above correction, the navigation state indicator in such a case is corrected to a large value. Conversely, when the angular difference between the reference angle of a route segment and the reference angle of the immediately preceding route segment is small, the ship's course does not change very much in the former route segment. By the above correction, the navigation state indicator in such a case is corrected to a large value only when the angular difference between the azimuth angle of the ground heading and the reference angle is small. By such correction, the calculation accuracy of the navigation state indicator can be improved.

[0117] (2-1-2) Calculation of binary decision value Fig. 17 is an example of a flowchart showing a process for calculating a binary judgment value when a reference angle is used. Similar to the process of Fig. 15, the process of Fig. 17 is executed for each ship currently sailing.

[0118] When the navigation data regarding the ship is acquired by the navigation data acquisition unit 131, the processing of steps S21 to S27 in Fig. 15 is executed. As a result, the navigation state index regarding the ship is calculated, and the following processing is executed.

[0119] [Step S31] The navigation state determination unit 134 determines whether the calculated navigation state index is equal to or greater than a predetermined threshold. If the navigation state index is equal to or greater than the threshold, the process proceeds to step S32. If the navigation state index is less than the threshold, the process proceeds to step S33.

[0120] [Step S32] The navigation state determination unit 134 determines that the ship is along the route, and outputs a binary determination value of “1”, which indicates that the ship is along the route. [Step S33] The navigation state determination unit 134 determines that the ship is not along the route, and outputs “0” as a binary determination value indicating that the ship is not along the route.

[0121] According to the process of determining the navigation state using the reference angle described in (2-1) above, the route area is divided into route segments, and the navigation state is determined based on the relationship between the reference angle set for the route segment in which the ship is located and the direction of the ship's course over the ground. This makes it possible to determine with almost constant accuracy whether the ship is following the route, or the degree to which it is following the route, regardless of whether the route is curved. Therefore, the navigation state can be determined with high accuracy.

[0122] (2-2) Determining navigation status based on boundary edges Next, a process for determining a navigation state based on a boundary edge will be described. In this process, a boundary edge set for each route segment is used. First, a method for setting the boundary edge will be described. The process for setting the boundary edge is executed by the pre-processing unit 133.

[0123] Fig. 18 is a diagram showing examples of boundary side setting, in which Fig. 18(A) shows an example of setting in the case of a one-way route, and Fig. 18(B) shows an example of setting in the case of a two-way route. The boundary edge refers to an edge of a route segment that is a boundary between the route segment and an adjacent route segment. The boundary edges include an entrance boundary edge, which is an edge where a ship enters from an adjacent route segment, and an exit boundary edge, which is an edge where a ship enters an adjacent route segment.

[0124] Fig. 18(A) shows an example of a route segment in a unidirectional route. In Fig. 18(A), the route direction is from top to bottom in the figure (forward direction), and route segments DV61, DV62, and DV63 are set along the route direction. In this way, a pair of an entrance side boundary edge and an exit side boundary edge is set for each route segment in a unidirectional route. Information indicating these entrance side boundary edges and exit side boundary edges is registered in the boundary edge database 124 in association with the route segments.

[0125] On the other hand, FIG. 18(B) shows an example of a route segment in a two-way route. In this case, a combination of an entrance side boundary edge and an exit side boundary edge is set for each route direction for each route segment. In the example on the left side of FIG. 18(B), the route direction is from top to bottom in the figure (forward direction), and route segments DV61, DV62, and DV63 are set along the route direction. Then, for each of the route segments DV61 to DV63, an entrance side boundary edge and an exit side boundary edge for the forward direction are determined. Also, in the example on the right side of FIG. 18(B), the route direction is from bottom to top in the figure (reverse direction), and route segments DV63, DV62, and DV61 are set along the route direction. Then, for each of the route segments DV61 to DV63, an entrance side boundary edge and an exit side boundary edge for the reverse direction are determined. Information indicating the combination of the entrance-side boundary edge and the exit-side boundary edge for each direction determined in this manner is registered in the boundary edge database 124 in association with the route segment.

[0126] In the above description, boundary edges are set for each route segment, but they may also be set for various attributes related to the ship. For example, a boundary edge may be set for each ship type for each route segment. Also, a boundary edge may be set for each route segment for each destination of a ship that has passed through the route segment. Furthermore, a boundary edge may be set for each route segment for each time period in which a ship has passed through the route segment.

[0127] Next, the procedure for the process of determining the navigation state using the boundary side will be described. In this determination process, as in the case of using the reference angle, either a navigation state indicator indicating the degree to which the ship is along the route or a binary determination value indicating whether the ship is along the route is output as a determination result.

[0128] (2-2-1) Calculation of navigation status index Fig. 19 is an example of a flowchart showing a calculation process procedure of a navigation state index when a boundary side is used. The process of Fig. 19 is executed for each ship that is sailing. When the navigation data acquisition unit 131 acquires navigation data for the ship, the process of Fig. 19 is executed.

[0129] [Step S41] The navigation data acquisition unit 131 calculates the ship's course over the ground based on the most recently acquired navigation data and navigation data acquired about the same ship during a period from the present time up to a predetermined time ago.

[0130] [Step S42] The navigation state determination unit 134 refers to the route segment database 122 and identifies the route segment to which the ship belongs (including the ship's current position) from the ship's position information (latitude, longitude) based on the latest navigation data.

[0131] [Step S43] The navigation state determination unit 134 determines whether the identified route segment is a segment of a route that is navigable in both directions. If it is a segment of a route that is navigable in only one direction, the process proceeds to step S44. If it is a segment of a route that is navigable in both directions, the process proceeds to step S45.

[0132] [Step S44] The navigation state determination unit 134 acquires, from the boundary edge database 124, a boundary edge corresponding to the identified route segment. [Step S45] The navigation state determination unit 134 determines the navigation direction (forward or reverse) of the ship in the identified route segment. For example, the navigation state determination unit 134 acquires the boundary edge set in the identified route segment from the boundary edge database 124. Based on the ship's most recent navigation trajectory, the navigation state determination unit 134 determines from which side of the boundary edge the ship entered the navigation segment, and determines the ship's navigation direction from the determination result. Of the boundary edges acquired from the boundary edge database 124, the navigation state determination unit 134 selects the boundary edge corresponding to the determined navigation direction as the boundary edge to be used.

[0133] [Step S46] The navigation state determination unit 134 calculates a navigation state index, which indicates the degree to which the ship is following the route, from the relationship between the course over the ground calculated in step S41 and the boundary side corresponding to the route segment. Here, an example of a method for calculating the navigation state index will be described with reference to the following Figure 20.

[0134] Fig. 20 is a diagram showing an example of a calculation process of a navigation state index when a boundary edge is used. In the process shown in Fig. 20, the angle between the azimuth angle of the ground course and the exit-side boundary edge (approach angle θout) and the position where the direction of the ground course intersects with the exit-side boundary edge (passing position Dout) are calculated based on the current position information and ground course of the ship and the exit-side boundary edge corresponding to the route section. Then, the approach angle θout and the passing position Dout are input to a learning model generated in advance to calculate a navigation state index.

[0135] The learning model is generated by a learning process by the pre-processing unit 133, for example, as follows. In this learning process, navigation data on a ship that has passed through a corresponding route segment and crossed the exit-side boundary edge to enter the adjacent next route segment, among the navigation data stored in the navigation history database 121, is referenced. Based on this navigation data, the pre-processing unit 133 executes machine learning using the approach angle θout and the passing position Dout as explanatory variables and information on whether the ship subsequently continued to navigate along the route or subsequently deviated from the route as a target variable. This generates a learning model that outputs the probability that the ship subsequently continues to navigate along the route.

[0136] In step S46, the approach angle θout and the passing position Dout are input to the learning model, and the output value of the probability of continuing navigation along the route is output as the navigation state index. In the calculation process of this navigation state index, if the direction of the course over the ground does not intersect with the exit side boundary edge, "0" is calculated as the navigation state index.

[0137] (2-2-2) Calculation of binary decision value Fig. 21 is an example of a flowchart showing a process for calculating a binary judgment value when a boundary side is used. The process in Fig. 21 is executed for each ship that is sailing. When the sailing data acquisition unit 131 acquires sailing data for the ship, the process in Fig. 21 is executed.

[0138] In the process shown in Fig. 21, step S46a is executed instead of step S46 among the process steps shown in Fig. 19. That is, when the navigation data acquisition unit 131 acquires navigation data about the ship, steps S41 to S45 in Fig. 19 are executed. Then, the following process of step S46a is executed.

[0139] [Step S46a] The navigation state determination unit 134 determines whether the ship is along the route based on at least the ground course calculated in step S41 and the information on the boundary edges corresponding to the route section, and outputs a binary determination value indicating the determination result.

[0140] In this step S46a, for example, the navigation state index calculated in step S46 of Fig. 19 may be compared with a predetermined threshold value, and a binary judgment value may be output based on the comparison result. In this case, if the navigation state index is equal to or greater than the threshold value, it is determined that the ship is along the route, and a binary judgment value of "1" indicating that the ship is along the route is output. On the other hand, if the navigation state index is less than the threshold value, it is determined that the ship is not along the route, and a binary judgment value of "0" indicating that the ship is not along the route is output.

[0141] Also, in step S46a, it may be determined whether the ship is along the route according to a determination rule such as that shown in the following Fig. 22, and a binary determination value indicating the determination result may be output. Note that in steps S44 and S45, only the exit side boundary edge or both the entrance side and exit side boundary edges are acquired as the boundary edge corresponding to the route segment, depending on the determination rule used in step S46a.

[0142] Fig. 22 is a diagram showing examples of rules for determining whether a ship is along a route, and shows two types of rules R1 and R2. When judgment rule R1 is used, the navigation state judgment unit 134 judges that the ship is along the route if the direction of the ship's course over the ground intersects with the exit boundary edge. On the other hand, judgment rule R2 uses the ship's navigation trajectory up to the last moment in addition to the course over the ground. When judgment rule R2 is used, the navigation state judgment unit 134 judges that the ship is along the route if the conditions of judgment rule R1 are met and the ship's navigation trajectory passes through the entrance boundary edge.

[0143] The lower part of FIG. 22 shows examples of route patterns when a ship passes through a route segment, and the determination results when the determination rules R1 and R2 are applied to each route pattern. In route pattern PT1, the ship enters the route segment from the entrance boundary edge, and the current course direction crosses the exit boundary edge. In this case, both judgment rules R1 and R2 judge that the ship is along the route, and a binary judgment value of "1" is output.

[0144] In route pattern PT2, the ship enters the route segment from the entrance boundary edge, but the current course direction does not intersect with the exit boundary edge. In this case, both judgment rules R1 and R2 judge that the ship is not along the route, and a binary judgment value of "0" is output.

[0145] In route pattern PT3, the ship enters the route segment from an edge other than the entrance boundary edge, but the current direction of the course over ground intersects with the exit boundary edge. In this case, judgment rule R1 judges that the ship is along the route, and outputs a binary judgment value of "1". On the other hand, judgment rule R2 judges that the ship is not along the route, and outputs a binary judgment value of "0". As in this example, judgment rule R2 makes it possible to accurately judge that the ship is not along the route in cases where the ship appears to be along the route from the heading direction, but has entered the route segment by crossing the route.

[0146] According to the process of determining the navigation state using the boundary edge described in (2-2) above, the route area is divided into route segments, and the navigation state is determined based on the relationship between the boundary edge set in the route segment in which the ship is located and the ship's course over the ground. This makes it possible to determine with almost constant accuracy whether the ship is following the route, or the degree to which it is following the route, regardless of whether the route is curved. Therefore, the navigation state can be determined with high accuracy. <3. Correction of collision risk assessment results> Next, the correction process of the judgment result by the judgment result correction unit 135 (step S13 in FIG. 5) will be described. In this correction process, the collision risk judgment result by the collision risk judgment unit 132 is corrected by the navigation state judgment result by the navigation state judgment unit 134. In this correction process, for example, any one of the following processes (3-1) to (3-3) is executed.

[0147] (3-1) Correction of collision risk determination results Here, as shown in (1-1) and (1-2) above, a case will be described in which the collision risk determination unit 132 outputs binary data indicating the presence or absence of a collision risk for a ship pair. In this case, the collision risk determination unit 132 determines the presence or absence of a collision risk for each ship pair extracted from the ships currently sailing, and identifies ship pairs determined to have a collision risk from among these ship pairs. In addition, as shown in (2-1-2) and (2-2-2) above, the navigation state determination unit 134 outputs a binary determination value indicating whether or not the ship is along the route for each ship currently sailing. Then, the determination result correction unit 135 finally determines the presence or absence of a collision risk between the ship pairs based on the binary determination value for each ship included in the ship pairs determined to have a collision risk.

[0148] 23 is an example of a flowchart showing a correction process procedure for the determination result of the presence or absence of a collision risk. At the start of the process shown in FIG. 23, it is assumed that information indicating each ship pair determined to have a collision risk by the collision risk determination unit 132 is registered in the ship pair information stored in the storage unit 120.

[0149] [Step S51] The determination result correction unit 135 selects one pair from among the ship pairs determined by the collision risk determination unit 132 to have a collision risk. [Step S52] The judgment result correction unit 135 acquires a binary judgment value indicating whether or not each ship included in the selected ship pair is along the route from the navigation state judgment unit 134. The judgment result correction unit 135 judges whether both ships included in the ship pair are along the route based on the acquired binary judgment value. If both ships are along the route, processing proceeds to step S53, and if at least one ship is not along the route, processing proceeds to step S54.

[0150] [Step S53] The judgment result correction unit 135 excludes the ship pair selected in step S51 from the above-mentioned ship pair information. [Step S54] The judgment result correction unit 135 judges whether all ship pairs judged to have a collision risk by the collision risk judgment unit 132 have been selected in step S51. If there are unselected ship pairs, the process proceeds to step S51, and one of the unselected ship pairs is selected. On the other hand, if there are no unselected ship pairs, the process proceeds to step S55.

[0151] [Step S55] The judgment result correction unit 135 finally outputs the ship pairs remaining in the ship pair information as ship pairs that pose a collision risk. For example, the judgment result correction unit 135 generates display information in which information indicating these ship pairs is drawn, and transmits the display information to the navigation monitoring device 102 or another device (for example, a device mounted on a ship that is sailing), and causes the device to display an image based on the display information.

[0152] According to the above process, even if the collision risk determination unit 132 determines that there is a collision risk for a certain ship pair, if the navigation state determination unit 134 determines that both ships in the ship pair are along the route, it is determined that there is no collision risk. This reduces the possibility of erroneously determining that there is a collision risk for a ship pair navigating a bend in the route, and improves the accuracy of determining the collision risk.

[0153] In the process of Fig. 23, the ship pairs determined to have a collision risk by the collision risk determination unit 132 were re-determined using a binary determination value indicating whether or not they are along the route. However, other processes may be performed as correction using the binary determination value. For example, when binary data indicating whether or not there is a collision risk for the ship pair is output from the collision risk determination unit 132, if a Yes is determined in step S52 for the ship pair determined to have a collision risk, the binary data may be determined.

[0154] In addition, the correction of the OZT determination result shown in (1-2) above can be performed, for example, as follows: For example, when OZT is detected between a ship and another ship, a binary determination value for the two ships is obtained from the navigation state determination unit 134. When it is determined from the binary determination value that both ships are along the route, it is determined that the OZT was erroneously detected, and the determination result is changed to "no OZT."

[0155] (3-2) Correction for collision risk (part 1) Here, we will explain the case where the collision risk determination unit 132 calculates the collision risk for a ship pair, as shown in (1-3) above. In this case, the collision risk determination unit 132 calculates the collision risk indicating the degree of collision risk for all ship pairs extracted from the ships currently sailing. In addition, the navigation state determination unit 134 calculates a navigation state index indicating the degree to which the ship is along the route for each ship currently sailing, as shown in (2-1-1) and (2-2-1) above. Then, the determination result correction unit 135 corrects the collision risk calculated for each ship pair using the navigation state index calculated for each ship included in the ship pair.

[0156] FIG. 24 is a first example of a flowchart showing a correction process procedure for the collision risk. [Step S61] The determination result correction unit 135 selects one pair from among the ship pairs for which the collision risk determination unit 132 has calculated the collision risk.

[0157] [Step S62] The judgment result correction unit 135 acquires the navigation state index calculated for each ship included in the selected ship pair from the navigation state judgment unit 134. Based on the acquired pair of navigation state indexes, the judgment result correction unit 135 calculates a route deviation degree indicating the degree to which the selected ship pair is not along the route. The route deviation degree takes a value between 0 and 1, and a larger value indicates a higher degree of deviation from the route.

[0158] [Step S63] The judgment result correction unit 135 corrects the collision risk calculated for the selected ship pair by the collision risk judgment unit 132 with the course deviation degree calculated in step S62, and outputs the corrected value as the final collision risk for this ship pair. In this correction, for example, the collision risk is multiplied by the course deviation degree.

[0159] [Step S64] The judgment result correction unit 135 judges whether all ship pairs for which the collision risk has been calculated by the collision risk judgment unit 132 have been selected in step S61. If there are unselected ship pairs, the process proceeds to step S61, and one of the unselected ship pairs is selected. On the other hand, if there are no unselected ship pairs, the process of FIG. 24 ends.

[0160] Here, the route deviation degree calculated in step S62 will be described. (3-2-1) First method for calculating route deviation A first calculation method is a method of performing calculation based on the comparison result of the navigation condition indexes acquired from the navigation condition determination unit 134 in step S62. For example, when both acquired navigation condition indexes of each ship are high, a low value is output as the route deviation degree, and the collision risk is corrected to a low value. As an example, information that maps the combination of navigation condition indexes of each ship and the route deviation degree is prepared in advance, and based on this information, the higher both navigation condition indexes of each ship are, the lower the value of the route deviation degree is output.

[0161] Additionally, the following method can also be considered as a method based on the comparison results. In this method, the lower of the acquired navigation condition indexes for each ship is selected. If the selected navigation condition index is C0, then (1-C0) is calculated as the route deviation degree. In step S63, the collision risk is corrected by multiplying the collision risk by the route deviation degree. In this method, when the navigation condition index of at least one of the ships in the ship pair is low, the collision risk is corrected to maintain a high value.

[0162] (3-2-2) Second method of calculating route deviation A second calculation method is a calculation method that takes into account the ship's course direction. In this second calculation method, in step S62, the signed navigation state indicator described in Fig. 16(A) is acquired from the navigation state determination unit 134. If right-hand traffic is practiced on the route, the signed navigation state indicator takes a positive value when the azimuth angle of the ship's ground course is inclined to the right of the navigation direction with respect to the reference angle, and takes a negative value when it is inclined to the left.

[0163] Here, the signed navigation state indicator for one ship in the ship pair is denoted as C1, and the signed navigation state indicator for the other ship is denoted as C2. The determination result correction unit 135 calculates each of the signed navigation state indicators C1 and C2 by the difference value between the reference angle and the azimuth angle of the course over the ground (θ 11 This conversion is performed by subtracting the absolute value of the signed navigation state indicator from 1 and adding the sign of the signed navigation state indicator to the subtraction result. If the converted values ​​of the signed navigation state indicators C1 and C2 are C1a and C2a, respectively, the determination result correction unit 135 calculates the degree of route deviation based on these values.

[0164] There are several possible methods for this calculation, and the following method can be applied, for example. In one method, the greater of the absolute value of C1a |C1a| and the absolute value of C2a |C2a| is calculated as the route deviation degree. In another method, the absolute value of the difference |C1a-C2a| is calculated as the route deviation degree. In step S63, the collision risk is corrected by multiplying the collision risk by the above route deviation degree.

[0165] Fig. 25 is a diagram showing examples of ship pair patterns. A ship pair pattern is a pattern of the course-over-ground direction of each ship included in a ship pair, and five ship pair patterns PT11 to PT15 are shown as examples in Fig. 25. Using Fig. 25, we will explain how the collision risk is corrected when the course deviation degree calculated by the second calculation method is used.

[0166] In Figure 25, in ship pair pattern PT11, the navigational state indexes (absolute values ​​of signed navigational state indexes) of each ship are all relatively high. On the other hand, in ship pair pattern PT12, the navigational state index of one ship is relatively high, but the navigational state index of the other ship is relatively low. In this case, the ship pair pattern PT12 has a higher degree of course deviation compared to ship pair pattern PT11, and the collision risk is corrected to maintain a high value.

[0167] In addition, in all of the vessel pair patterns PT13 to PT15, the navigational condition indexes of both vessels are relatively low. However, in the vessel pair pattern PT13, the ground course directions of both vessels are tilted to the right with respect to the reference angle. In this case, the course deviation degree is a relatively low value, and the collision risk is corrected to be low.

[0168] On the other hand, in vessel pair pattern PT14, the course over ground directions for both vessels are tilted to the left with respect to the reference angle. Furthermore, in vessel pair pattern PT15, the direction of inclination of the course over ground directions with respect to the reference angle differs for each vessel. In vessel pair patterns PT14 and PT15, the degree of course deviation is higher than in vessel pair pattern PT13, and the collision risk is corrected to maintain a high value.

[0169] As described above, even if the navigation condition index of each ship is low, if both ships are facing to the right of the reference angle, as in ship pair pattern PT13, the possibility of the ships colliding with each other is low. In such cases, the collision risk is corrected to a low value. However, if at least one ship is facing to the left of the reference angle, as in ship pair patterns PT14 and PT15, there is a possibility that that ship will cross the waterway, increasing the possibility of colliding with the other ship. Therefore, in such cases, the collision risk is maintained at a high value.

[0170] In this way, when the second calculation method for the degree of route deviation is used, it becomes possible to output a collision risk that takes into account the ship navigation rule of keeping to the right side of the road, thereby improving the accuracy of the calculation of the collision risk.

[0171] Note that the correction method shown in Fig. 25 is merely an example. In particular, when the navigation state indexes of both ships are relatively low, such as in ship pair patterns PT13 to PT15, it may be preferable to calculate the degree of route deviation using other detection indexes in addition to the navigation state index.

[0172] (3-3) Correction for collision risk (part 2) In the above (3-2), the navigation state determination unit 134 calculates a navigation state index for each ship, and the determination result correction unit 135 corrects the collision risk calculated for the ship pair by the collision risk determination unit 132 using the navigation state index calculated for each ship in the ship pair. In contrast, in (3-3), the navigation state determination unit 134 calculates a binary determination value indicating whether or not each ship is along the route, and the determination result correction unit 135 corrects the collision risk calculated for the ship pair using the binary determination value calculated for each ship in the ship pair.

[0173] FIG. 26 is a second example of a flowchart showing a correction process procedure for the collision risk. [Step S71] The determination result correction unit 135 selects one pair from among the ship pairs for which the collision risk determination unit 132 has calculated the collision risk.

[0174] [Step S72] The judgment result correction unit 135 acquires a binary judgment value indicating whether or not each ship included in the selected ship pair is along the route from the navigation state judgment unit 134. The judgment result correction unit 135 judges whether both ships included in the ship pair are along the route based on the acquired binary judgment value. If both ships are along the route, processing proceeds to step S73, and if at least one ship is not along the route, processing proceeds to step S74.

[0175] [Step S73] The determination result correction unit 135 corrects the collision risk for the selected ship pair calculated by the collision risk determination unit 132 to a constant value. This constant value is set to 0 or a predetermined value close to 0 (a value significantly lower than 1).

[0176] [Step S74] The determination result correcting unit 135 maintains the collision risk calculated by the collision risk determining unit 132 for the selected vessel pair as is. [Step S75] In steps S73 and S74, the collision risk value may suddenly decrease or increase depending on the value of the binary judgment value for each ship included in the ship pair. The judgment result correction unit 135 corrects the collision risk value after correction in step S73, or the collision risk value maintained as is in step S74, so that it becomes continuous over time. Here, the correction in step S75 is referred to as "re-correction."

[0177] [Step S76] The judgment result correction unit 135 judges whether all ship pairs for which the collision risk has been calculated by the collision risk judgment unit 132 have been selected in step S71. If there are unselected ship pairs, the process proceeds to step S71, and one of the unselected ship pairs is selected. On the other hand, if there are no unselected ship pairs, the process of FIG. 26 ends.

[0178] Fig. 27 is a diagram showing an example of the transition of the collision risk output from the determination result correction unit. As described above, in steps S73 and S74 in Fig. 26, the collision risk value may suddenly decrease or increase depending on the value of the binary determination value for each ship included in the ship pair.

[0179] As shown in (2-1-2), since the binary judgment value is calculated by comparing the navigation state index with the threshold value, if the navigation state index has a value close to the threshold value for a certain period of time, the binary judgment value may fluctuate even if the navigation state index does not change significantly. Therefore, even if the navigation state index or the collision risk calculated by the collision risk determination unit 132 changes gradually, the final output value of the collision risk may change suddenly.

[0180] Graph 241 in Fig. 27 shows an example of such a case. For example, if the navigation state index of one of the vessels in the vessel pair falls below the threshold at time T1 and the binary judgment value of that vessel changes from "1" to "0," the collision risk that was suppressed to a constant value in step S73 suddenly rises in step S74.

[0181] Furthermore, when the navigation state index has a value close to the threshold value for a certain period of time, the binary judgment value may change frequently. In such a case, the collision risk value may frequently repeat rapid rises and falls in steps S73 and S74. Graph 242 shown in Fig. 27 shows an example of such a case. In this way, if the collision risk value output to the user frequently rises and falls, it is not convenient for the user who wants to be aware of the risk of collision between ships.

[0182] Therefore, in step S75, the collision risk levels output in steps S73 and S74 are re-corrected so that the collision risk levels to be output change continuously. As a method of recorrection, for example, a method of calculating a moving average using the most recent n (n is an integer of 2 or more) values ​​of the collision risk outputted in steps S73 and S74 can be used. Alternatively, a method of performing recorrection using a navigation state index and a threshold value for determining a binary judgment value and a function that determines the amount of change in the collision risk outputted in steps S73 and S74 according to the difference between the navigation state index and the threshold value can be used.

[0183] In addition, the above method basically makes it possible to output the recorrected collision risk without delay. In contrast, if a certain degree of output delay is allowed, a method can be used in which, for example, a predetermined number of collision risks output in steps S73 and S74 are obtained before and after the current time, and the obtained collision risks are smoothed using a low-pass filter or the like.

[0184] The re-correction method may be selected according to the use case. Use cases are broadly divided into online processing (analyzing data that arrives sequentially and outputting the results in real time) and offline processing (analyzing past data). With online processing, if the output of the collision risk is delayed, the time available for the user to respond to the risk is reduced. For this reason, the candidates are methods that can output the collision risk without delay or that only incur a small delay. With offline processing, output delays are not an issue, so the method is selected according to the degree of smoothing required.

[0185] The dotted curves in graphs 241a and 242a shown in FIG. 27 show an example of recorrection using a method that allows output delay. The dotted curve in graph 241a shows the transition of the collision risk after recorrection when the collision risk output in steps S73 and S74 transitions as shown in graph 241. The dotted curve in graph 242a shows the transition of the collision risk after recorrection when the collision risk output in steps S73 and S74 transitions as shown in graph 242. In this way, according to graphs 241a and 242a, it can be seen that the collision risk is recorrected so that the output changes continuously around the time when the collision risk output in steps S73 and S74 changes sharply. This prevents the output of the collision risk from frequently fluctuating greatly, making it easier for the user to recognize the collision risk.

[0186] The processing functions of the devices (e.g., navigation monitoring devices 1, 102) shown in each of the above embodiments can be realized by a computer. In this case, a program describing the processing contents of the functions that each device should have is provided, and the above processing functions are realized on the computer by executing the program on a computer. The program describing the processing contents can be recorded on a computer-readable recording medium. Examples of computer-readable recording media include magnetic storage devices, optical disks, and semiconductor memories. Examples of magnetic storage devices include hard disk drives (HDDs) and magnetic tapes. Examples of optical disks include CDs (Compact Discs), DVDs (Digital Versatile Discs), and Blu-ray Discs (BD, registered trademark).

[0187] When distributing a program, for example, the program is recorded on a portable recording medium such as a DVD or CD and then sold. The program can also be stored in a storage device of a server computer and transferred from the server computer to other computers via a network.

[0188] A computer that executes a program stores, for example, a program recorded on a portable recording medium or a program transferred from a server computer in its own storage device. The computer then reads the program from its own storage device and executes processing according to the program. The computer can also read the program directly from a portable recording medium and execute processing according to the program. The computer can also execute processing according to the received program each time a program is transferred from a server computer connected via a network. [Explanation of symbols]

[0189] 1 Navigation monitoring equipment 2 Storage section 3 Processing section 4. Plot Information 10 Route area Sections 11 to 13 12a Dashed line 20 ships 20a Arrow S1, S2 steps

Claims

1. A storage unit that stores partition information indicating a plurality of partitions generated by dividing a route area of ​​a predetermined width along the route into polygonal shapes in the navigation direction; a processing unit that identifies one of the plurality of sections in which the ship is located based on a current position of the ship, and outputs evaluation information that evaluates whether the ship is navigating along a route based on a relationship between a course-over-ground direction of the ship and the one section; and having A reference angle is defined for each of the plurality of sections as a reference for the navigation direction within the corresponding section, and the reference angle corresponding to each of the plurality of sections is determined based on the direction of the course over ground in the corresponding section of ships that have passed through the corresponding section in the past and have navigated without deviating from the route area after passing through the corresponding section; and outputting the evaluation information based on a relationship between a ground course direction of the ship underway and the reference angle defined in the one section. Navigation monitoring equipment.

2. A storage unit that stores partition information indicating a plurality of partitions generated by dividing a route area of ​​a predetermined width along the route into polygonal shapes in the navigation direction; a processing unit that identifies one of the plurality of sections in which the ship is located based on a current position of the ship, and outputs evaluation information that evaluates whether the ship is navigating along a route based on a relationship between a course-over-ground direction of the ship and the one section; and having A reference angle serving as a reference for a navigation direction within each of the plurality of sections is defined for each of the plurality of sections; In the output of the evaluation information, an angle difference between the direction of the ship's course over the ground and the reference angle defined in the one section is normalized by a predetermined maximum value that the angle difference can take, and the normalized value is calculated as an index showing the degree to which the ship is sailing along the route. Navigation monitoring equipment.

3. In the output of the evaluation information, based on a comparison result between the index and a predetermined threshold, outputting, as the evaluation information, binary data indicating whether the ship is navigating along a route; 3. The navigation monitoring device according to claim 2.

4. The processing unit further comprises: Identifying a first ship and a second ship as the ships currently sailing, calculating vectors from a current position of each of the first ship and the second ship in a direction of a course over ground to a position to be reached after a predetermined time, and determining that there is a high risk of collision between the first ship and the second ship when the vectors intersect; when it is determined that the collision risk between the first vessel and the second vessel is high based on each of the vectors, and if it is determined from the binary data for each of the first vessel and the second vessel that both the first vessel and the second vessel are navigating along the route area, it is finally determined that the collision risk between the first vessel and the second vessel is low; 4. The navigation monitoring device according to claim 3.

5. The processing unit further comprises: identifying a first ship and a second ship as the ships currently sailing, and calculating a collision risk between the first ship and the second ship based on at least one of a closest approach distance and a closest approach time between the first ship and the second ship; correcting the collision risk based on the indicators for each of the first vessel and the second vessel; 3. The navigation monitoring device according to claim 2.

6. A storage unit that stores partition information indicating a plurality of partitions generated by dividing a route area of ​​a predetermined width along the route into polygonal shapes in the navigation direction; a processing unit that identifies one of the plurality of sections in which the ship is located based on a current position of the ship, and outputs evaluation information that evaluates whether the ship is navigating along a route based on a relationship between a course-over-ground direction of the ship and the one section; and having A reference angle serving as a reference for a navigation direction within each of the plurality of sections is defined for each of the plurality of sections; In the output of the evaluation information, Calculating a first angle difference between a direction of a course over ground of the navigating vessel and the reference angle defined in the one section; calculating a probability that an angular difference between a direction of a course over ground and the reference angle defined in the one section will be the first angular difference based on a navigation history of a ship that has passed through the one section in the past and subsequently navigated within the route area without deviating; calculating an index indicating the degree to which the ship is sailing along the route based on the calculated probability; Navigation monitoring equipment.

7. The computer referring to partition information indicating a plurality of partitions generated by dividing a route area of ​​a predetermined width along the route into polygonal shapes in the navigation direction, and identifying one of the plurality of partitions in which the currently navigating vessel is located based on the current position of the currently navigating vessel; outputting evaluation information for evaluating whether the ship underway is sailing along a route based on a relationship between a course direction of the ship underway and the one section; Processing A reference angle is defined for each of the plurality of sections as a reference for the navigation direction within the corresponding section, and the reference angle corresponding to each of the plurality of sections is determined based on the direction of the course over ground in the corresponding section of ships that have passed through the corresponding section in the past and have navigated without deviating from the route area after passing through the corresponding section; and outputting the evaluation information based on a relationship between a ground course direction of the ship underway and the reference angle defined in the one section. Navigation monitoring method.

8. On the computer, referring to partition information indicating a plurality of partitions generated by dividing a route area of ​​a predetermined width along the route into polygonal shapes in the navigation direction, and identifying one of the plurality of partitions in which the currently navigating vessel is located based on the current position of the currently navigating vessel; outputting evaluation information for evaluating whether the ship underway is sailing along a route based on a relationship between a course direction of the ship underway and the one section; Execute the process, A reference angle is defined for each of the plurality of sections as a reference for the navigation direction within the corresponding section, and the reference angle corresponding to each of the plurality of sections is determined based on the direction of the course over ground in the corresponding section of ships that have passed through the corresponding section in the past and have navigated without deviating from the route area after passing through the corresponding section; and outputting the evaluation information based on a relationship between a ground course direction of the ship underway and the reference angle defined in the one section. Navigation monitoring program.

Images