Automatic ship handling system

Abstract

This system provides an automated ship navigation system that can ensure route tracking while keeping computational complexity low. [Solution] An automatic ship handling system that generates a target path (R) based on at least obstacle information and constraints on the ship's behavior, evaluates the degree of agreement between the predicted trajectory (r) of the ship (20) after a predetermined time and the target path (R) when a predetermined amount of control is given to the ship (20) for a predetermined time longer than the control cycle, and determines the amount of control, wherein the predetermined time is divided into a plurality of periods, each including at least a first period before the change time and a second period after the change time, the amount of control includes a first amount of control which is the amount of control in the first period and a second amount of control which is the amount of control in the second period, the ratio of the first period to the second period is changed by changing the change time to predict the predicted trajectory (r), the degree of agreement between the predicted trajectory (r) and the target path (R) is evaluated, and the amount of control is determined.

Description

Technical Field

[0001] The present invention relates to an automatic ship maneuvering system.

Background Art

[0002] Conventionally, in a ship maneuvering apparatus used for automatic berthing of a ship, a controller sets the nearest position in the current bow direction among positions along the berthing site as a first target position, determines a position separated by an offset amount from the first target position in a direction perpendicular to the edge of the berthing site as a second target position, and after controlling a propulsion device to make the ship reach the second target position, it is known to control the propulsion device to make the ship reach the first target position.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of Invention

Problems to be Solved by the Invention

[0004] In the control for causing a ship to follow a target route in a conventional ship maneuvering apparatus, the target route is generated based on obstacle information and the behavior constraints of the hull, and the operation amount is determined for each unit time longer than the control cycle which is the minimum controllable unit of the ship. Therefore, it is required to improve the following accuracy by suppressing the calculation amount, and there is room for improvement in this respect.

[0005] Accordingly, the present invention provides an automatic ship maneuvering system capable of ensuring route followability while suppressing the calculation amount.

Means for Solving the Problems

[0006] To solve the above problems, the present invention adopts the following configurations. (1) Embodiment 1 of the automatic ship handling system according to the present invention generates a target path (R) based on obstacle information and ship behavior constraints, evaluates the degree of agreement between the predicted trajectory (r) of the ship (20) after a predetermined time and the target path (R) when a predetermined amount of operation is given to the ship (20) for a predetermined time longer than the control cycle, and determines the amount of operation (Q), comprising at least a first period (T1) before the change time (ta) and a second period (T2) after the change time (ta) The method is characterized by dividing the period into multiple periods (T1 and T2), the manipulated variable (Q) including a first manipulated variable (Q1) which is the manipulated variable in the first period (T1), and a second manipulated variable (Q2) which is the manipulated variable in the second period (T2), the ratio of the first period (T1) to the second period (T2) is changed by changing the change time (ta) to predict the predicted trajectory (r), the degree of agreement between the predicted trajectory (r) and the target path (R) is evaluated, and the manipulated variable (Q) is determined.

[0007] According to the automatic ship handling system of the present invention, when a predetermined control amount is given to the ship for a predetermined time longer than the control cycle, the predetermined time is divided into multiple parts based on the change time, and the path taken when the predetermined control amount is given to the first and second halves of each period can be evaluated. In other words, the control amount is determined by evaluating the degree of agreement between the predicted trajectory and the target path, so the control amount can be optimized. Therefore, it is possible to increase the number of predicted trajectory options and improve tracking performance while reducing the computational load due to control.

[0008] (2) Aspect 2 of the present invention may be characterized in that the automatic ship handling system of Aspect 1 optimizes and calculates the change time (ta).

[0009] This configuration allows for the optimization of the predicted trajectory, resulting in a control quantity that closely matches the target path.

[0010] (3) A third aspect of the present invention may be characterized in that, in the automatic ship handling system of aspect 2, the degree of agreement between the end (ra) of the predicted trajectory (r) and a point on the target path (R) that is closest to the end (ra) is evaluated.

[0011] This configuration allows for the optimization of the manipulated variable, as it evaluates the degree of agreement between the predicted trajectory's endpoint and the nearest point on the target path.

[0012] (4) A fourth aspect of the present invention is an automatic ship handling system of the third aspect, wherein the target path (R) includes a target speed and a target heading, the predicted trajectory (r) includes at least the predicted speed and predicted heading at the end (ra), and the degree of agreement between the X coordinate, Y coordinate, speed and heading of the end (ra) of the predicted trajectory (r) and a point on the target path that is closest to the end (ra) is evaluated.

[0013] This configuration allows for the optimization of control variables by evaluating the degree of agreement between the X, Y coordinates, ship speed, and heading of the nearest point on the target path, taking into account information on the predicted ship speed and predicted heading at the end of the predicted trajectory.

[0014] (5) Aspect 5 of the present invention may be characterized in that, in the automatic navigation system of aspect 4, the target speed is allowed to be exceeded at points other than the end (ra) of the predicted trajectory (r).

[0015] This configuration allows for more flexible control. [Effects of the Invention]

[0016] The automated ship handling system according to the present invention can ensure route following while reducing the amount of computation required. [Brief explanation of the drawing]

[0017] [Figure 1] This is a schematic diagram showing the movement of a ship when it is automatically navigating using the automatic ship handling system according to an embodiment of the present invention. [Figure 2] This is a block diagram showing the configuration of the docking route generation system. [Figure 3] This is a plan view showing the arrangement of sensor systems on a ship. [Figure 4] This is a flowchart showing the process of searching for a berthing facility. [Figure 5] This is a schematic diagram for explaining the control processing operation of the predicted trajectory of a vessel. [Figure 6] This is a diagram showing the relationship between a predetermined time and an operation amount in the automatic vessel maneuvering system. [Figure 7] This is a diagram showing the relationship between a predetermined time and an operation amount in the automatic vessel maneuvering system, and is a diagram in which behavior prediction is performed at time intervals. [Figure 8] This shows an example of a predicted trajectory calculated at the time intervals shown in Fig. 7. [Figure 9] This is a block diagram showing a route control unit provided in the automatic vessel maneuvering system. [Mode for Carrying Out the Invention]

[0018] As shown in Fig. 1, the automatic vessel maneuvering system 1 is mounted on a vessel 20 for use. The automatic vessel maneuvering system 1 is a system that generates a target route R based at least on obstacle information D1 and the behavior constraints of the hull, and performs automatic navigation. In this control, when an operation amount is given with a predetermined time longer than a later-described control cycle as a parameter, the degree of matching between the predicted trajectory r of the vessel 20 after the predetermined time and the target route R is evaluated, and control for determining the operation amount is performed by the vessel maneuvering control unit.

[0019] The vessel 20 on which the automatic steering system 1 is installed is not particularly limited and can include, for example, pleasure boats (including pontoons, V-hull boats, and fishing boats), fishing vessels, water jet boats, electric propulsion vessels, and hybrid vessels. The vessel 20 may also be a runabout or a personal watercraft. The vessel 20 is equipped with a propulsion device, such as a screw, which rotates when driven by a power source such as an engine or electric motor. The propulsion device is located at the stern of the vessel 20. The propulsion device can include an outboard motor, an inboard / outboard motor, an inboard motor, a rim drive, a POD, etc. In this embodiment, multiple outboard motors are provided as the propulsion device. The target value of the maneuver calculated by the steering control unit is sent via the integrated control unit to the known FI-ECU of the outboard motor and the controller of the steering device that steers the outboard motor relative to the hull. Based on the received target value of the maneuver, the FI-ECU and controller control the actuator for driving the throttle valve on the outboard motor and the actuator of the steering device.

[0020] The automated ship handling system 1 comprises a sensor system, an HMI (Human-Machine Interface) system, and a ship handling control unit. The ship handling control unit controls the sensor system, HMI system, input devices, and propulsion devices to generate a docking route for automatic navigation of the ship when docking. The HMI system includes an MFD (Multifunction Display).

[0021] The ship handling control unit comprises a first control unit and a second control unit. The first control unit includes a space recognition unit for recognizing available space, a self-position recognition unit for recognizing its own position, an image processing unit for processing images acquired by cameras, etc., an obstacle recognition unit for recognizing obstacles, a hull behavior model generation unit, a state management unit, a route generation unit for generating a route, and a follow control unit for following the generated route. The second control unit includes a berth recognition unit, a trailer recognition unit, a short-range lateral distance unit, an obstacle detection unit, a server, a UI display unit, and an IVI function unit.

[0022] The ship's control unit is one or more electronic control units installed in a ship. The ship's control unit is configured as a computer, including circuits, a CPU, ROM, and RAM. The ROM stores the program for operating the ship's control unit. Through the cooperation of the above hardware and software, the ship's control unit can function as a component of the ship's control unit.

[0023] The sensor system includes a LiDAR (ambient field sensor), IMU, camera, trailer recognition camera, GNSS antenna, and anemometer.

[0024] LiDAR is a sensor that measures the distance to surrounding objects and recognizes the external environment. Specifically, LiDAR detects the presence or absence of surrounding objects by emitting pulsed light and reflecting the reflected light. If an object is present, LiDAR detects the orientation and distance of the object based on the direction of the pulsed light when the reflected light is received and the time until the light is received. Based on this detection result, LiDAR acquires point cloud data representing the objects present in the surroundings. LiDAR sensors are evenly distributed around the entire 360°. For example, they are placed at four locations (one each in the front, back, left, and right directions) at 90° intervals in the circumferential direction.

[0025] The IMU is a sensor that detects the movement of the ship's hull. The IMU is positioned near four LiDARs. For example, four LiDARs are arranged around the IMU.

[0026] The cameras are designed to capture images of the area around the ship's hull. The cameras also have the function of displaying the captured images and identifying the images to measure the distance to nearby objects. For example, cameras are positioned on the front, rear, left, and right sides of the ship 20.

[0027] The trailer recognition camera detects trailer tabs provided on the trailer. The trailer recognition camera is positioned, for example, in front of the vessel 20. Known April tags may be used for the trailer tabs, and the relative positional relationship with the trailer can be recognized by recognizing the trailer tabs with the trailer recognition camera. In this embodiment, the trailer recognition camera and the camera are provided separately, but the camera may be used to recognize the trailer tabs.

[0028] The GNSS antenna detects the position and orientation of the vessel 20. The GNSS antenna receives GNSS radio waves from satellites and performs known positioning calculations to obtain the current absolute position (latitude, longitude) of the vessel 20. While GNSS positioning can be done using a single antenna, using known DGNSS positioning or RTK (Real-time Kinematic) positioning is preferable because it allows for highly accurate acquisition of the vessel 20's position. The GNSS antenna is mounted on the vessel 20 in a manner that avoids interference from objects above it. In addition to the GNSS antenna, the sensor system also includes a compass sensor. Furthermore, the absolute position of the vessel 20 is measured using the GNSS antenna, and the ship speed (ground speed) V is calculated based on the time-series positioning results. The compass sensor obtains the orientation of the vessel's bow. The compass sensor can be, for example, a magnetic compass sensor or a satellite compass.

[0029] An anemometer is a sensor that detects the direction and strength of the wind. It is mounted on the ship 20 so that it receives the wind from the front.

[0030] The ship handling control unit receives surrounding environment recognition data acquired by LiDAR, position data of the ship 20 acquired by the GNSS antenna, and direction data of the ship 20 acquired by the direction sensor.

[0031] The berth recognition unit creates a map of obstacles around the vessel 20 based on point cloud information acquired by LiDAR and attitude information of the vessel 20 acquired by IMU. In other words, obstacle map information is created. This allows the system to search for spaces larger than the vessel 20 at the berthing point. Furthermore, berth recognition can be achieved by recognizing the berth facility using a camera and then extracting available space near the berth facility.

[0032] The movement of the vessel 20 based on control by the automatic steering system 1 may be fully automatic, or it may be assisted by teaching the system to operate the throttle lever or steering wheel to follow the target path R.

[0033] Here, in the vessel 20, the direction from the bow 20a toward the stern is defined as the longitudinal direction X1, and the direction perpendicular to the longitudinal direction X1 in a plan view is defined as the left-right direction X2, and the explanation follows.

[0034] As shown in Figure 2, the docking route generation system 1 includes a sensor system 40A, an HMI system 40B (Human-Machine Interface system), and a ship handling control unit 50 (input device). The ship handling control unit 50 controls the sensor system 40A, the HMI system 40B, the input device 50, and the propulsion system to generate a docking route for the ship 20 to automatically navigate when docking. The HMI system 40B includes an MFD (Multifunction Display) 47.

[0035] The ship handling control unit 50 comprises a first control unit 50A and a second control unit 60A. The first control unit 50A includes a space recognition unit 51 for recognizing empty space, a self-position recognition unit 52 for recognizing its own position, an image processing unit 53 for processing images acquired by cameras 43, etc., an obstacle recognition unit 54 for recognizing obstacles, a hull behavior model generation unit 55, a state management unit 56, a route generation unit 57 for generating a route, and a follow control unit 58 for following the generated route. The second control unit 60A includes a berth recognition unit 61, a trailer recognition unit 62, a short-range distance unit 63, an obstacle detection unit 64, a server 65, a UI display unit 66, and an IVI function unit 67.

[0036] The ship handling control unit 50 is one or more electronic control units provided by the ship 20. The ship handling control unit 50 is configured as a computer including a circuit, CPU, ROM, and RAM. The ROM stores a program for operating the ship handling control unit 50. Through the cooperation of the above hardware and software, the ship handling control unit 50 (50A, 60A) can be made to function as parts of the ship handling control unit 50.

[0037] As shown in Figures 2 and 3, the sensor system 40A includes a LiDAR 41 (external environment recognition sensor), an IMU 42, a camera 43, a trailer recognition camera 44, a GNSS antenna 45, and an anemometer 46.

[0038] LiDAR41 is a sensor that measures the distance to surrounding objects and recognizes the external environment. Specifically, LiDAR41 detects the presence or absence of surrounding objects by emitting pulsed light and using reflected light. If an object is present, LiDAR41 detects the orientation and distance of the object based on the direction of the pulsed light when the reflected light is received and the time until the light is received. Based on this detection result, LiDAR41 acquires point cloud data representing the objects present in the surroundings. LiDAR41 is evenly distributed around the entire 360°. In Figure 2, it is placed at four locations (one each in the front, back, left, and right directions) at 90° intervals in the circumferential direction.

[0039] The IMU42 is a sensor that detects the movement of the ship's hull. The IMU42 is positioned near the four LiDAR41s. In Figure 2, the four LiDAR41s are arranged around the IMU42.

[0040] Camera 43 is a camera that captures images of the area around the hull. Camera 43 also has the function of displaying the captured images or identifying images to measure the distance to nearby objects. In Figure 2, cameras 43 are positioned on the front, rear, left, and right sides of the ship 20.

[0041] The trailer recognition camera 44 detects trailer tabs provided on the trailer. As shown in Figure 2, the trailer recognition camera 44 is positioned in front of the vessel 20. Known April tags may be used as trailer tabs, and the relative positional relationship with the trailer can be recognized by recognizing the trailer tabs with the trailer recognition camera 44. In this embodiment, the trailer recognition camera 44 and camera 43 are provided separately, but the trailer tabs may also be recognized using camera 43.

[0042] The GNSS antenna 45 detects the position and orientation of the vessel 20. The GNSS antenna 45 receives GNSS radio waves from satellites and performs known positioning calculations to obtain the current absolute position (latitude, longitude) of the vessel 20. GNSS positioning can be done using a single antenna, but using known DGNSS positioning or RTK (Real-time Kinematic) positioning is preferable because it allows for highly accurate acquisition of the vessel 20's position. The GNSS antenna 45 is mounted on the vessel 20 in a manner that avoids interference from objects above it. In addition to the GNSS antenna 45, the sensor system 40A is also equipped with an orientation sensor (not shown). The GNSS antenna 45 is used to measure the absolute position of the vessel 20, and the ship speed (ground speed) V is calculated based on the time-series positioning results. The orientation sensor obtains the orientation of the vessel 20's bow. The orientation sensor can be, for example, a magnetic orientation sensor or a satellite compass.

[0043] The anemometer 46 is a sensor that detects the direction and strength of the wind. It is mounted on the vessel 20 so that it receives wind from the front.

[0044] The ship handling control unit 50 receives ambient external recognition data acquired by the LiDAR 41, position data of the ship 20 acquired by the GNSS antenna 45, and direction data of the ship 20 acquired by the direction sensor.

[0045] The berth recognition unit 61 creates a map of obstacles around the vessel 20 based on point cloud information acquired by the LiDAR 41 and attitude information of the vessel 20 acquired by the IMU 42. In other words, obstacle map information is created. This makes it possible to search for a space larger than the vessel 20 at the berth location. In addition, berth recognition can be achieved by recognizing the berth facility 30 with the camera 43 and then extracting empty space near the berth facility 30.

[0046] Next, we will describe in detail an example of the processing flow by the docking route generation system 1. Figure 4 is a flowchart of the process for searching for docking facilities 30 (berths).

[0047] As shown in Figure 4, when processing by the docking route generation system 1 begins, the berth search mode processing flow starts in step S1, and the flow is completed when the docking mode is activated in step S2.

[0048] In step S1, the berth search mode is implemented by receiving the destination coordinates in step S11. This is done, for example, by using a nautical chart or autopilot in the ship handling control unit 50, which includes an input device. Next, the route (path T0) is set in step S12, and automatic navigation is started in step S13. Then, the process proceeds to step S20, where candidate berths (docking facilities 30) are determined.

[0049] In step S20, the berth candidate determination begins with step S21, when the LiDAR 41 acquires a point cloud of information around the ship 20. This point cloud is then converted into a 2D map through two-dimensional processing, and the height of the target position is processed (step S22), generating an obstacle map (grid map) that shows occupied areas occupied by objects and unoccupied areas not occupied by objects. Next, in step S23, the availability of empty space is determined. In step S23, the size of the unoccupied area is determined from the grid map generated in step 22. Unoccupied areas with a width of at least a first threshold and a length of at least a second threshold, i.e., empty space, are extracted along the area occupied by objects. Then, in step S24, the berth determination is performed using the camera 43 in conjunction with image recognition from the camera 43. If an object near the empty space extracted in step S23 is a docking facility, the empty space is determined to be a berth. Next, the coordinates and orientation of the berth center (docking point P0) are set (step S25), and in step S26, the intermediate point P2 is set.

[0050] Next, in step S27, it is determined whether the intermediate point P2 and the docking point P0 can be connected. If the intermediate point P2 and the docking point P0 can be connected in step S27 (step S27: YES), the process proceeds to step S28. On the other hand, if the intermediate point P2 and the docking point P0 cannot be connected in step S27 (step S27: NO), the process proceeds to step S50, which is the resetting process for the intermediate point P2. Step S50 will be described later.

[0051] In step S28, it is determined whether the intermediate point P2 and the current location can be connected. If the intermediate point P2 and the current location can be connected in step S28 (step S28: YES), the process proceeds to step S29 and the location is recognized as a berth candidate. On the other hand, if the intermediate point P2 and the current location cannot be connected in step S28 (step S28: NO), the process proceeds to step S50, which is the resetting process for the intermediate point P2.

[0052] In the operation flow for determining a berth candidate in step S20, a berth candidate is recognized in step S29. The operation flow for determining a berth candidate in step S20 is continuously executed in berth search mode S1. In other words, the berth candidate determination in step S20 is actually continued until the user specifies a berth in step S30. Furthermore, there is not limited to just one berth candidate; two or more berth candidates may be displayed on the MFD47 if they satisfy the berth candidate determination criteria from S20 to S29.

[0053] The operation flow for berth specification in step S30 begins with receiving the user's berth specification input in step S31, followed by the start of fixed-point holding in step S32. After that, the user's docking direction specification is received (step S33), and the operation flow for berth specification in step S30 is completed.

[0054] In the automatic ship handling system 1, the degree of agreement between the predicted trajectory r of the ship 20 and the target path R after a predetermined time when a predetermined control amount is given to the ship 20 for a predetermined time longer than the control cycle is evaluated, and the control amount Q is determined. Here, the control amount Q corresponds to the number of times the operation is switched, i.e., the amount of computation by the control process. Here, the control period is defined as a control every Δk (in microseconds). If we define the control period as K0 for the first control, and K1, K2, K3, ... for subsequent controls, the time step of Δk will be shorter than Δt. For example, when the control period is 50 milliseconds, the manipulated variable will be controlled 20 times per second.

[0055] Figure 5 is a schematic diagram illustrating the control process operation for predicting the ship's trajectory. Figure 5 shows multiple (in this case, four) predicted trajectories r1, r2, r3, and r4 calculated by the control process. As shown in Figure 1, the automatic navigation system 1 includes a point that evaluates the degree of agreement between the endpoint ra of multiple predicted trajectories r and the nearest point on the target path R to that endpoint ra. In this control process, as shown in Figure 5, the target path R is made to be a straight line, and perpendiculars are found from each endpoint ra (ra1, ra2, ra3, ra4) of the four predicted trajectories r1, r2, r3, and r4 to the straight target path R. These perpendiculars represent the shortest distance L from each endpoint ra to the target path R, and the control variable Q is set to select the predicted trajectory r with the smallest shortest distance L. For example, in Figure 5, the shortest distance L from the endpoint ra2 of the predicted trajectory with sign r2 to the target path R is small, so this predicted trajectory r2 is selected.

[0056] Figure 6 shows the relationship between a predetermined time T and the manipulated variable Q in the automatic ship handling system 1. In Figure 6, the horizontal axis represents time t, and the vertical axis represents the manipulated variable Q. As shown in Figure 6, the automatic ship handling system 1 performs model predictive control at finite predetermined time intervals T. In this embodiment, the predetermined time T is divided into predetermined periods T1 and T2, which include at least a first period T1 before the change time ta and a second period T2 after the change time ta. When divided into two periods, the predetermined time is the sum of the first period T1 and the second period T2. The predicted trajectory r is calculated assuming that an operation is performed using manipulated variable Q1 in the first period T1 and an operation is performed using manipulated variable Q2 in the second period T2. Then, by changing the change time ta, the ratio of the first period T1 to the second period T2 is changed to predict the predicted trajectory r as shown in Figure 1, the degree of agreement between the predicted trajectory r and the target path R is evaluated, and the manipulated variable is determined. Here, the change time ta is calculated by optimization.

[0057] Furthermore, when evaluating the degree of agreement between the predicted trajectory r and the target path R, it is preferable to evaluate the degree of agreement between the endpoint ra of the predicted trajectory r and the point on the target path R that is closest to the endpoint ra.

[0058] Furthermore, the target path R preferably includes coordinate information, target speed, and target heading.

[0059] Furthermore, it is preferable that the predicted trajectory r includes at least the predicted ship speed, predicted heading, and predicted coordinates at the terminal ra. In this case, the degree of agreement between the X coordinate, Y coordinate, ship speed, and heading of the terminal ra of the predicted trajectory r and the nearest point on the target path R to the terminal ra is evaluated. Then, it is permissible to exceed the target speed at points other than the terminal ra of the predicted trajectory r. For the speed, a value slightly lower than the maximum speed that should be achieved along the path is set as the target speed. For example, the cruising speed stipulated within the marina is set as the target speed. In this case, if the ship speed at the terminal ra of the predicted trajectory r is below the target speed, the highest evaluation is uniformly given, and only if the speed exceeds the target speed is the lowest evaluation given. In this case, as shown in Figure 5, by comparing at least the terminal ra with the nearest point, all predicted paths along the path except the final docking point will have no difference from the target in terms of X coordinate. In this way, the automatic navigation system 1, by handling the target speed and X coordinate as described above, is an optimal system for docking operations where "accuracy" is required rather than "speed".

[0060] Figure 7 shows the relationship between a predetermined time T and the control input Q in the automatic ship handling system 1, illustrating an example of behavior prediction using a time step. Figure 8 shows an example of a predicted trajectory r calculated with the time step Δt shown in Figure 7.

[0061] The predicted trajectory r is generated by setting an upper speed limit and varying the speed within that limit. For example, if the predetermined time (n seconds) shown in Figure 6 is 18 seconds, then restrictions on acceleration and deceleration are set for this predetermined time of 18 seconds. Specifically, the restrictions on acceleration and deceleration can be such that deceleration occurs only in the first half, acceleration occurs only in the second half, or the timing of the acceleration / deceleration switch can be set to 1-second increments. For example, as shown in Figures 7 and 8, the predetermined time n for calculating the predicted trajectory r can be set to 18 seconds, and the time step Δt can be set in advance to 3 seconds. In this case, the control is performed by accumulating calculations of the destination point (the end point ra of the predicted trajectory r) every 3 seconds. Shortening the time step Δt improves the prediction accuracy, but increases the amount of computation.

[0062] In the example shown in Figure 6, two manipulated variables Q (Q1, Q2) are set for the first period T1 and the second period T2, based on the change time ta, within a predetermined time T. That is, there are two manipulated variables Q within the predetermined time T. The time of change (change time ta) is a variable. For example, the manipulated variables Q for an outboard motor are engine speed and steering angle. In that case, five parameters are used: change time, engine speed (first half), engine speed (second half), steering angle (first half), and steering angle (second half). For example, the engine speed can be set as a parameter in increments of 500 rpm or 100 rpm, and by setting a lower limit of engine speed (e.g., 700 rpm, not limited to the above) and an upper limit of engine speed (e.g., 2000 rpm, not limited to the above), the number of candidate manipulated variables can be reduced, thereby reducing the amount of calculation required when calculating the predicted trajectory r. For example, the steering angle can be set in increments of 5 degrees or 10 degrees, with a positive number representing left steering and an upper limit of 30 degrees, and a negative number representing right steering and a lower limit of -30 degrees. In this control system, the above combinations of engine speed and steering angle are treated as a single manipulated variable Q, and multiple predicted trajectories r are calculated for each combination of change time ta, engine speed, and steering angle.

[0063] As described above, the method for selecting (evaluating) the predicted trajectory r is to evaluate four points: the position coordinates (X coordinate in the direction of travel and Y coordinate in the width direction), the bearing, and the ship's speed to select the optimal predicted trajectory r. For the endpoint ra of the predicted trajectory r and the nearest point of the target path R to the endpoint ra, the X error (in the direction of travel), Y error (in the width direction), bow bearing error, and ship speed error are evaluated.

[0064] Figure 9 is a block diagram showing the route control unit 30 provided in the automatic ship handling system 1. As shown in Figure 9, the automatic ship handling system 1 is equipped with a route control unit 30 on the ship 20. The route control unit 30 includes a filter 31, an integral compensator 32, a manipulated variable generator 33, a ship behavior prediction model 34, and an optimizer 35. The route control unit 30 calculates target values ​​for the manipulated variables through control using the ship behavior prediction model 34.

[0065] The operation flow for calculating the target value of the manipulated variable Q by the route control unit 30 is as follows: First, in step S1, the actual values ​​of the hull coordinates and the manipulated variable Q are input to the filter 31. The filter 31 removes measurement noise from the input values. Next, in step S2, the input values ​​that have passed through the filter 31 are processed by the integral compensator 32 to feed back into steady-state deviations such as tidal currents. Then, the manipulated variable generator 33 generates a group of manipulated variable candidates based on the input values ​​and target route information (step S3).

[0066] Subsequently, in step S4, the hull behavior prediction model 34 is used to calculate multiple predicted trajectories r(r1, r2, ...) that take into account the effect of wind for each of the candidate control variables Q generated in step S3. At this time, the hull behavior prediction model 34 is input with information on disturbances such as wind speed, wind direction, and currents.

[0067] Next, in step S5, the predictive trajectory r is optimized by the optimizer 35, and the manipulated variable Q with a high degree of agreement with the target path R is selected as the final target value output. Of the multiple predictive trajectories r1, r2, ... shown in Figure 1, the predictive trajectory with sign r1 is the predictive trajectory with a high degree of agreement with the target path R. In this way, in steps S3 to S5, predictive control using the hull behavior prediction model 34 is implemented to handle the nonlinear multi-input multi-output system. At this stage, direct error feedback and wind feedforward are performed. By calculating the predictive trajectory (local path) that takes the effect of wind into account, the system can follow the path with high responsiveness due to feedforward.

[0068] In addition, the path control unit 30 shown in Figure 9 performs control to predict the ship's behavior every 3 seconds by assigning an input variable Q based on information about external disturbances such as wind speed, wind direction, and current. However, it is preferable to perform control to calculate predicted trajectories r1, r2, ... based on the ship's current speed, heading, steering angle, and engine speed as information.

[0069] As described above, the automatic ship handling system 1 in the above embodiment generates a target path R based on at least obstacle information and ship behavior constraints, evaluates the degree of agreement between the predicted trajectory r of the ship 20 after a predetermined time and the target path R when a predetermined amount of control is given to the ship 20 for a predetermined time longer than the control cycle, and determines the amount of control Q. The automatic ship handling system divides the predetermined time into a plurality of periods T, which include at least a first period T1 before the change time ta and a second period T2 after the change time ta, predicts the predicted trajectory r by changing the ratio of the first period T1 to the second period T2 by changing the change time ta, evaluates the degree of agreement between the predicted trajectory r and the target path R, and determines the amount of control Q.

[0070] According to the automatic ship handling system 1 of this embodiment, when a predetermined control amount is given to the ship 20 for a predetermined time longer than the control cycle, the predetermined time is divided into multiple parts by the change time ta, and the path can be evaluated when a predetermined control amount Q is given to the first and second halves of each period. In other words, the control amount Q is determined by evaluating the degree of agreement between the predicted trajectory r and the target path R, so the control amount Q can be optimized. Therefore, it is possible to increase the number of choices for the predicted trajectory r and improve tracking performance while reducing the amount of computation.

[0071] The above-mentioned automatic ship handling system 1 optimizes and calculates the change time t. This configuration allows for the optimization of the predicted trajectory r, resulting in a control variable that closely matches the target path R.

[0072] In the above-described automated navigation system 1, the degree of agreement between the endpoint ra of the predicted trajectory r and the point on the target path R that is closest to the endpoint ra is evaluated. With this configuration, an appropriate control variable Q can be obtained by evaluating the degree of agreement between the predicted trajectory r and the nearest point on the target path R that is the end ra of the predicted trajectory r.

[0073] In the above-described automated ship handling system 1, the target path R includes the target speed and target heading, the predicted trajectory r includes the predicted ship speed and predicted heading at least at the terminal ra, and the degree of agreement between the X coordinate, Y coordinate, ship speed, and heading of the terminal ra of the predicted trajectory r and the point on the target path that is closest to the terminal ra is evaluated. With this configuration, an appropriate control variable Q can be obtained by evaluating the degree of agreement between the X, Y coordinates, ship speed, and heading of the nearest point on the target path and the predicted ship speed and heading at the end ra of the predicted trajectory r.

[0074] In the above-described automated navigation system 1, the system allows exceeding the target speed at points other than the terminal ra of the predicted trajectory r. This configuration allows for more flexible control.

[0075] The configuration in the above embodiment is just one example of the present invention, and various modifications are possible without departing from the spirit of the invention, such as replacing the components of the embodiment with well-known components.

[0076] In the embodiment described above, control is performed to optimize and calculate the change time t, but this optimization control may be omitted.

[0077] It is also possible to omit the evaluation of the degree of agreement between the nearest point on the target path R to the terminal ra of the predicted trajectory r.

[0078] Furthermore, the automatic navigation system 1 is not limited to a configuration in which the target path R includes the target speed and target heading, the predicted trajectory r includes the predicted speed and predicted heading at least at the terminal ra, and the system evaluates the degree of agreement between the X coordinate, Y coordinate, speed, and heading of the predicted trajectory r and a point on the target path that is closest to the terminal ra. [Explanation of Symbols]

[0079] 1. Automatic ship handling system 20 ships 30 Route control unit R Target path r Predicted trajectory ra termination T1 Period 1 T2 Second Period t predetermined time ta change time Q Manipulated amount

Claims

1. An automatic ship handling system that generates a target path (R) based on at least obstacle information and ship behavior constraints, evaluates the degree of agreement between the predicted trajectory (r) of the ship (20) after a predetermined time when a predetermined maneuver is given to the ship (20) for a predetermined time longer than the control cycle, and the target path (R), and determines the maneuver (Q), The predetermined time is divided into a plurality of periods (T1 and T2) which include at least a first period (T1) before the change time (ta) and a second period (T2) after the change time (ta). The manipulated variable (Q) includes a first manipulated variable (Q1) which is the manipulated variable in the first period (T1) and a second manipulated variable (Q2) which is the manipulated variable in the second period (T2). By changing the change time (ta), the ratio of the first period (T1) to the second period (T2) is changed to predict the predicted trajectory (r). An automatic ship handling system that evaluates the degree of agreement between the predicted trajectory (r) and the target path (R), and determines the control variable (Q).

2. The automatic ship handling system according to claim 1, which optimizes and calculates the aforementioned change time (ta).

3. The automatic ship handling system according to claim 2, which evaluates the degree of agreement between the end (ra) of the predicted trajectory (r) and a point on the target path (R) that is closest to the end (ra).

4. The aforementioned target path (R) includes the target speed and the target heading. The predicted trajectory (r) includes at least the predicted ship speed and predicted heading at the end (ra), The automatic ship handling system according to claim 3, which evaluates the degree of agreement between the X coordinate, Y coordinate, ship speed, and bearing of the end (ra) of the predicted trajectory (r) and a point on the target path that is closest to the end (ra).

5. The automatic ship handling system according to claim 4, which allows the target speed to be exceeded at a point other than the end (ra) of the predicted trajectory (r).

Images