Parameter estimation system for ship motion models, method for estimating parameters of motion models, and program

Abstract

This provides a simple method for estimating the parameters of a ship's specific motion model. [Solution] A parameter estimation system for estimating the parameters of a ship's motion model, comprising an estimation unit that uses a navigation log when the ship sails in a predetermined motion pattern and estimates the parameters of a motion model specific to the ship that correspond to the predetermined motion pattern shown in the navigation log.

Description

Technical Field

[0001] The present invention relates to a parameter estimation system for a ship's motion model, a method for estimating parameters of a motion model, and a program.

Background Art

[0002] Conventionally, technologies for controlling ship navigation have been developed based on a ship's motion model.

[0003] For example, in Patent Document 1, regarding automatic steering control, a configuration is disclosed in which control parameters are set by dimensionalizing dimensionless motion parameters when the ship speed changes. Also, in Patent Document 2, regarding a ship's autopilot, a configuration is described in which control parameters of an engine are corrected based on measured values from sensors and the like.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Patent Document 2

Summary of the Invention

Problems to be Solved by the Invention

[0005] Control models used for ship autopiloting require identifying and setting motion model parameters based on the results of various tests and simulations, taking into account various factors such as the ship's shape and specifications. This identification process is costly and time-consuming. In particular, ships are often produced individually to order rather than in mass production. Therefore, it is necessary to identify motion model parameters individually according to the characteristics of each ship, which increases the cost and time required for parameter identification. For example, ship test conditions and environments are diverse, making it difficult to directly use measurement data from one ship to control another. Furthermore, even with the same ship, its shape and characteristics change due to aging and other factors, which can alter the accuracy of the motion model once it has been identified. Therefore, it is necessary to update the motion model for each ship as needed.

[0006] In view of the above issues, the present invention aims to provide a convenient method for estimating the parameters of a ship-specific motion model. [Means for solving the problem]

[0007] To solve the above problems, one embodiment of the present invention has the following configuration. That is, a parameter estimation system for estimating the parameters of a ship motion model is An estimation unit that uses a navigation log of the aforementioned vessel when it navigates in a predetermined motion pattern to estimate the parameters of a motion model specific to the vessel that corresponds to the predetermined motion pattern shown in the navigation log, It has.

[0008] Another embodiment of the present invention has the following configuration: a parameter estimation method for estimating the parameters of a ship motion model, An estimation step of using a navigation log when the vessel navigates in a predetermined motion pattern, and estimating the parameters of a motion model specific to the vessel that corresponds to the predetermined motion pattern shown in the navigation log, It has.

[0009] Another aspect of the present invention has the following configuration. That is, a program that causes a computer to use a navigation log when a ship sails in a predetermined navigation pattern, and estimates parameters of a motion model specific to the ship corresponding to the predetermined motion pattern indicated by the navigation log, an estimation unit, function as.

Advantages of the Invention

[0010] According to the present invention, it becomes possible to easily estimate parameters of a motion model specific to a ship.

Brief Description of the Drawings

[0011] [Figure 1] Block diagram showing a configuration example of a parameter estimation system for a motion model according to a first embodiment of the present invention [Figure 2] Block diagram showing a configuration example of a ship according to a first embodiment of the present invention [Figure 3] Sequence diagram showing a process flow according to a first embodiment of the present invention [Figure 4] Schematic diagram showing an example of a motion pattern of a ship according to a first embodiment of the present invention [Figure 5] Schematic diagram showing input / output related to parameter estimation according to a first embodiment of the present invention [Figure 6] Flowchart of a process according to a first embodiment of the present invention [Figure 7] Sequence diagram showing a process flow according to a modification of a first embodiment of the present invention [Figure 8] Flowchart of a process according to a modification of a first embodiment of the present invention

Modes for Carrying Out the Invention

[0012] Hereinafter, embodiments for implementing the present invention will be described with reference to the drawings and the like. Note that the embodiments described below are one embodiment for explaining the present invention, and are not intended to be construed as limiting the present invention. Also, not all configurations described in each embodiment are essential configurations for solving the problems of the present invention. In each drawing, the same components are denoted by the same reference numerals to indicate the correspondence. Note that, in order to avoid unnecessary redundancy and facilitate the understanding of those skilled in the art, part of the description may be omitted or simplified. For example, detailed descriptions of well-known matters or duplicate descriptions of substantially the same configurations may be omitted.

[0013] <First Embodiment> [System Configuration] FIG. 1 is a schematic diagram showing a configuration example of a parameter estimation system for a motion model (hereinafter simply referred to as "parameter estimation system") according to a first embodiment of the present invention. The parameter estimation system for the motion model is a system for estimating parameters of a motion model related to the automatic navigation of a ship and constructing a control model corresponding to the ship. The parameter estimation system for the motion model may be used in a form mounted on a ship, or may be configured as a separate device and configured to be able to provide a control model based on the motion model for which parameters have been estimated to the ship. The parameter estimation system 100 may be configured by an information processing device such as a PC (Personal Computer), for example, or may be configured by a control device such as an ECU (Electronic Control Unit) mounted on a ship.

[0014] The parameter estimation system 100 includes a control unit 101, a storage unit 102, an input unit 103, an output unit 104, and a communication unit 105. Each part is configured to be able to communicate with each other by an internal bus or the like.

[0015] The control unit 101 is responsible for controlling the operation of the parameter estimation system 100. The control unit 101 is composed of, for example, a CPU (Central Processing Unit) and a GPU (Graphics Processing Unit), and provides various functions by reading and executing various programs and data stored in the memory unit 102.

[0016] The memory unit 102 is a storage device for storing programs, data, and other information necessary for executing various control processes and functions performed by the control unit 101. The memory unit 102 is composed of volatile / non-volatile storage devices such as RAM (Random Access Memory), ROM (Read Only Memory), HDD (Hard Disk Drive), and flash memory.

[0017] The input unit 103 consists of a mouse, keyboard, microphone, etc., and accepts input from the user. The output unit 104 consists of a display, speaker, etc., and outputs various types of data. The output from the output unit 104 may be visual, auditory, or tactile, such as images, sounds, or vibrations. Alternatively, the input unit 103 and the output unit 104 may be integrated using a touch panel display or the like.

[0018] The communication unit 106 is a communication interface for communicating with external devices via a network (not shown). The communication unit 106 may be configured to support multiple communication standards depending on the network configuration. The network may consist of, for example, the Internet, an intranet, a wireless LAN (Local Area Network), or a WAN (Wide Area Network). Note that the communication standards and wired / wireless connections related to the network are not particularly limited, and the network may be configured by combining multiple communication standards.

[0019] Figure 2 is a block diagram showing an example of the configuration of a vessel according to this embodiment. The vessel 200 is configured to enable automatic navigation using a control model constructed based on the estimation results of the parameters of the motion model according to this embodiment. The vessel 200 includes a control unit 201, a memory unit 202, an operation unit 203, a drive unit 204, an output unit 205, a sensor unit 206, an antenna unit 207, and a camera 208.

[0020] The control unit 201 is responsible for controlling the ship 200. For example, it consists of a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), dedicated circuits, etc., and provides various functions by reading and executing various programs and data stored in the memory unit 202.

[0021] The memory unit 202 is a storage device for storing programs, data, and other information necessary to execute various control processes and functions provided by the control unit 201. The memory unit 202 is composed of volatile / non-volatile storage devices such as RAM (Random Access Memory), ROM (Read Only Memory), HDD (Hard Disk Drive), and flash memory.

[0022] The control unit 203 receives various operations related to the navigation of the vessel 200. The control unit 203 may include an accelerator lever, steering wheel, joystick lever, shift lever, various switches, etc. Operations here may include navigation-related operations such as steering and acceleration / deceleration, as well as operations related to various parts that make up the vessel 200. The drive unit 204 performs operations related to the navigation of the vessel 200 based on instructions from the control unit 201 and operations from the control unit 203. The drive unit 204 may include, for example, an engine, motor, thruster, propeller, pump, etc. The type of drive unit 204 is not particularly limited, and any configuration that enables automatic navigation using the parameter estimation and control model of the motion model described later is acceptable.

[0023] The output unit 205 consists of a display, speaker, etc., and outputs various types of data. The output from the output unit 205 may be visual, auditory, or tactile, such as images, sounds, or vibrations. For example, the output unit 205 may display various data acquired by the sensor unit 206, antenna unit 207, and camera 208 on a predetermined user interface screen. The output unit 205 may also output information for navigation using map information, etc.

[0024] The sensor unit 206 consists of multiple sensors for acquiring information about the surroundings and interior of the vessel 200. The sensor unit 206 may include multiple types of sensors, such as an anemometer, acceleration sensor, temperature sensor, humidity sensor, LiDAR, radar, gyro sensor, position sensor, rotation sensor, and sonar. Multiple sensors of a single type may also be provided. The various sensors included in the sensor unit 206 may be configured to acquire information about the entire vessel 200, as well as information about specific parts of the vessel 200. The antenna unit 207 is a unit for transmitting and receiving information to and from the outside. The antenna unit 207 may include multiple types of antennas, such as a GNSS (Global Navigation Satellite System) antenna and an AIS (Automatic Identification System) antenna. The camera 208 is an imaging device for acquiring images of the surroundings of the vessel 200. The direction of images that can be acquired by the camera 208 is not particularly limited, and it may be configured to acquire images in all directions.

[0025] [Parameter estimation for motion models] This section describes the parameter estimation of the motion model related to the automatic operation of the vessel 200 in this embodiment. Figure 3 is a sequence diagram showing the flow of constructing a control model including parameter estimation of the motion model according to this embodiment. In the parameter estimation of the motion model according to this embodiment, a navigation log 310 acquired when the vessel 200 is actually sailing is used. In this embodiment, the navigation log 310 is composed of self-position information 311 and drive information 312. The self-position information 311 is information obtained by self-position estimation and includes the current position (latitude, longitude, altitude), azimuth angle, and speed. The method of self-position estimation is not particularly limited, and known methods may be used. For example, the self-position estimation method described in Japanese Patent Application Publication No. 2023-041501 by the present applicant may be used.

[0026] The drive information 312 includes the propeller rotation speed (hereinafter simply referred to as "rotation speed") and the rudder angle. The information included in the navigation log 310 described here is just an example, and other information may be included. For example, the drive information 312 may further include torque, acceleration, and the direction of propeller rotation. Each piece of information included in the navigation log 310 is acquired by the sensor unit 206, antenna unit 207, and camera 208 mounted on the vessel 200, and is recorded in association with the time information at the time of acquisition. The navigation log 310 may be recorded during manual navigation or during automatic navigation. In this example, it will be described as being recorded during automatic navigation.

[0027] In step S301, the parameter estimation system 100 acquires a navigation log 310. Here, the navigation log 310 may only acquire navigation logs within a predetermined range for estimating the parameters of the motion model, or the system may acquire the navigation log 310 and then accept a range to be used from the user. The parameter estimation system 100 may also perform predetermined preprocessing on the acquired navigation log 310. This preprocessing may include a process to derive values ​​to be used in each of the subsequent steps from the information contained in the navigation log 310.

[0028] In step S302, the parameter estimation system 100 uses the acquired navigation log 310 to determine the motion pattern of the vessel 200. Figure 4 shows examples of motion patterns. Five examples are shown here, but many more motion patterns may be handled. Figure 4(a) shows the motion pattern of moving straight at a constant speed. Figure 4(b) shows the motion pattern of turning right at a constant speed. Figure 4(c) shows the motion pattern of a U-turn by turning right at a constant speed. Figure 4(d) shows the motion pattern of turning right after moving straight and then fully opening the throttle. Figure 4(e) shows the motion pattern of meandering from a left turn to a right turn at a constant speed. Note that multiple motion patterns may be determined from the acquired navigation log 310. Alternatively, multiple motion patterns may be combined and used as a new motion pattern.

[0029] In step S303, the parameter estimation system 100 sets the gain (weight) for each parameter included in the calculation formula used for parameter estimation, based on the determined motion pattern. In other words, the behavior of the ship 200 changes according to the motion pattern of the ship 200. Therefore, the weight of each parameter included in the motion model is adjusted according to the motion pattern. The gain for each parameter set here may be predefined in correspondence with the motion pattern, or it may be set by the user. Examples of parameters for which weights are set will be described later.

[0030] In step S304, the parameter estimation system 100 uses the navigation log 310 and a calculation formula with gain settings to estimate the parameters of the motion model. In this embodiment, the known finite-time optimal estimation method MHE (Moving Horizon Estimation) is used for parameter estimation. MHE is a known method and is an estimation method for nonlinear models that can estimate the state at a predetermined time using measured values ​​at a finite time. An example of a motion model parameter estimation method using MHE will be described later. Note that the estimation method is not limited to MHE, and estimation methods using other nonlinear models may be applied.

[0031] In step S305, the parameter estimation system 100 constructs a control model for controlling the vessel 200 using the estimated parameters. The control model is constructed by determining control values ​​to guide the vessel 200 along a desired navigation path, based on each parameter of the estimated motion model. The constructed control model is maintained in association with the vessel 200 and its motion pattern. If the parameter estimation system 100 and the vessel 200 are configured as separate devices, the control model is maintained in a manner that allows it to be provided from the parameter estimation system 100 to the vessel 200.

[0032] In step S306, the vessel 200 performs predictive control related to automatic navigation using the control model constructed by the parameter estimation system 100.

[0033] In step S307, the vessel 200 performs automatic navigation by driving the drive unit 204 according to control values ​​based on predictive control. When the vessel 200 performs automatic navigation, it may acquire a navigation log 310 in a timely manner and store it so that it can be provided to the parameter estimation system 100.

[0034] Figure 5 is a conceptual diagram illustrating the parameter estimation of the above motion model. During parameter estimation, gain (weight) settings for each parameter are applied to the MHE calculation formula, corresponding to the motion pattern. Then, input parameters are input to the MHE calculation formula, which has the gain settings applied to each parameter. In this example, the input parameters used are the speed, turning speed, and rudder angular velocity of the ship 200. These input parameters may be the information contained in the navigation log 310, or different values ​​may be derived based on the information in the navigation log 310. The estimated parameters obtained are the mass, added mass, moment of inertia (yaw), added moment of inertia (yaw), resistance / drag proportional to speed and acceleration, rotational speed vs. thrust characteristics, rotational speed, and rudder angular velocity of the ship 200 at a certain time (future). Note that the estimated parameters are just examples and are not limited to these; other parameters may be estimated.

[0035] Furthermore, there are no particular limitations on the time to be estimated or the range (time interval) used as input. For example, the MHE equation of motion may be constructed to estimate the parameters 0.5 seconds later by using the information shown in the past navigation log 310 for 20 seconds from the current time as input parameters.

[0036] Here, we will explain an example of the method used in steps S302 to S304. In this example, we will use the ENU coordinate system as the coordinate system. The ENU coordinate system is a three-dimensional coordinate system consisting of the E axis (East), N axis (North), and U axis (Up), with the current position of the ship 200 set as the origin. The E axis corresponds to the X direction, the N axis to the Y direction, and the U axis to the Z direction, with the North, East, and Up directions being positive directions. Also, the thrust that moves the hull in the positive direction of each axis is considered positive. The negative direction of the X axis (stern direction) is defined as a rudder angle of 0.

[0037] In this embodiment, equations (1) to (9) below are used to estimate the parameters using continuous-time model predictive control (MPC). While a relatively simple model is shown here as an example, the embodiment is not limited to this, and more complex models may be used. Furthermore, this embodiment assumes a small vessel with a relatively small size. Therefore, the following model may be appropriately adjusted according to the characteristics of the vessel.

[0038]

number

[0039]

number

[0040]

number

[0041]

number

[0042]

number

[0043] Of the parameters shown above, the following values ​​indicate the state obtained by self-position estimation (UKF (Unscented Kalman Filter) / EKF (Extended Kalman Filter)). Self-position estimation uses values ​​obtained by fusing information from sensors such as GNSS (Global Navigation Satellite System) and IMU (Inertial Measurement Unit). This information is obtained and derived from the navigation log 310, which includes self-position information 311 and drive information 312 shown in Figure 3.

[0044]

number

[0045] Furthermore, among the parameters listed above, the following values ​​are eigenvalues ​​for each vessel. These are the parameters that are estimated by MHE in this embodiment.

[0046]

number

[0047] The thrust model in the vessel 200 differs depending on the number and arrangement of actuators (not shown) included in the drive unit 204 provided in the vessel 200. Here, we will explain using an example of a thrust model for a single outboard motor that does not consider fluid forces. The actuator configuration of a vessel can also include inboard and outboard motors, inboard motors, etc., and thrust models corresponding to these may also be used. The outboard motor has a propeller (not shown) that generates thrust in the X direction and a steering mechanism that rotates the thrust direction of the propeller around the Z axis, and these are mounted together on the outside of the hull.

[0048] The thrust generated by the actuator (not shown) installed on the ship 200 and applied to the hull is calculated using the following equation (10).

[0049]

number

[0050] Of the above values, L x , L y This is determined based on values ​​obtained by measuring the installation position of the outboard motor. Combining equation (10) above with equations (1) to (9) which show the motion model of the hull, we derive the following equation (11).

[0051]

number

[0052] Here, the thrust generated by the propeller (not shown) of the vessel 200 is generally assumed to be estimated by accumulating information based on propeller characteristics, engine characteristics, and torque sensors installed on the propeller shaft. However, since this method is difficult to apply to small vessels, in this embodiment, the thrust generated by the propeller is treated as a function of rotational speed and converted from the rotational speed. For parameter estimation, a Generalized Logic Function (GLF) is used and defined as follows. In equations (12) to (14) shown below, A, K, B, v, C, and M are coefficients, and since they depend on the configuration of actuators such as outboard motors, parameter estimation is necessary.

[0053]

number

[0054] In this example, normalization is performed to eliminate differences between vessels, such as bound / scale, when estimating parameters in MHE, thereby simplifying the setup. This conversion process is not performed within MPC, but rather with respect to the outboard motor thrust F obtained in MPC. P The value is converted inversely to rotational speed r and output to the actuator.

[0055] The above motion model is defined as the state equation of a nonlinear system by the following equation (15).

[0056]

number

[0057] Furthermore, in this example, the optimization problem handled by MPC is given by equation (16) below. In this case, by treating x(t)=v(t), MPC is performed to carry out tracking control. Note that the calculation flow of nonlinear MPC is publicly known, so the details are omitted here.

[0058]

number

[0059] Then, MHE is used as a method for estimating the state. In this example, the optimization problem can be defined as shown in equation (17) below.

[0060]

number

[0061] By solving this optimization problem, the parameters of the motion model are estimated from N state data. The state data uses the velocity and angular velocity shown in equations (5) and (6), with rotational speed and rudder angle information input as u. At this time, the accuracy of the state data is important because it is difficult to predict the model if the observation noise v is large. For this reason, it is desirable that the accuracy of the self-position estimation of the ship, i.e., the self-position information 311, be as high as possible.

[0062] Furthermore, when estimating, the parameters to be given more weight are set as weights, as shown in equation (18).

[0063]

number

[0064] Similarly, the observation noise v and process noise w are weighted as shown in equations (19) and (20) below.

[0065]

number

[0066] The weights described above are set according to the characteristics of the state, so they do not change much depending on the ship's motion pattern. On the other hand, the weights P for each parameter shown in equation (21) below PSince the degree of influence changes depending on the motion pattern of the ship 200, it is necessary to determine the motion pattern and adjust accordingly. Therefore, the motion pattern is determined in step S302, and the gain is set in step S303 based on the result.

[0067]

number

[0068]

number

[0069] Next, we will explain examples of parameter estimation corresponding to each motion pattern.

[0070] (In the case of linear motion) For example, in the case of linear motion that accelerates in stages, the ship's motion is considered under the following conditions for equation (11) above.

[0071]

number

[0072] Furthermore, when a ship has only one outboard motor, the actuator is generally located in the center of the hull, so L y Let = 0. The ship's motion at this time is given by equation (22) below, based on equation (11) above.

[0073]

number

[0074] F p As mentioned above, the conversion is performed using GLF, so if we take this into consideration and further simplify equation (22), we get the following equation (23).

[0075]

number

[0076] In this way, the number of parameters that need to be estimated is limited, making the estimation easier. At this point, the following weight setting for the remaining U becomes important.

[0077]

number

[0078] Furthermore, since straight-line motion allows for navigation using the engine's output to its maximum potential, it becomes easier to estimate the following parameters related to the GLF, which are the conversion coefficients between thrust and rotational speed.

[0079]

number

[0080] On the other hand, the influence on parameters necessary for movement in the Y direction (left-right direction of the ship) and turning motion is small, and estimating with respect to these parameters is likely to converge to incorrect parameters. Therefore, the following weights are adjusted to suppress this influence.

[0081]

number

[0082] By taking these factors into account and setting weights, parameter estimation is performed using MHE to estimate parameters related to straight-line movement and conversion coefficients for thrust and rotational speed.

[0083] (For low-speed and high-speed turns) This section explains the motion patterns of turning movements. Here, "high speed" and "low speed" are not particularly limited, but are arbitrarily set based on predetermined criteria for each vessel. For example, a speed faster than a specified speed may be treated as "high speed," and a speed slower than that specified speed may be treated as "low speed."

[0084] To improve accuracy, it is preferable to estimate the turning motion pattern for each parameter at either high or low speeds. In the case of high-speed turns, the Coriolis force and drag shown in equations (24) and (25) below, which are elements included in equation (1) above, have a significant influence.

[0085]

number

[0086]

number

[0087] Therefore, relatively speaking, it becomes more difficult to capture the change in added mass shown in equation (26) below, which is one of the elements included in equation (1) above.

[0088]

number

[0089] Considering the reasons above, by estimating the parameters of the added mass described below using the turning motion pattern at low speeds, it becomes possible to estimate more accurately even at high speeds.

[0090]

number

[0091] Furthermore, the following parameters related to speed and turning in the Y direction are also important.

[0092]

number

[0093] On the other hand, the following parameters related to the square of the velocity are best estimated when considering turning motion patterns at high speeds.

[0094]

number

[0095] In this case, the parameters proportional to the velocity, as described below, also have a significant influence, so it is desirable to give them importance and use them accordingly.

[0096]

number

[0097] (Estimated flow) By performing parameter estimation in the following order, it becomes possible to estimate the model parameters of the hull and thrust in a short amount of time. (1) A pattern of repeatedly accelerating and decelerating in stages while moving in a straight line. (2) A pattern of steady circular turns, figure-eight movements, and meandering at a sufficiently low speed. (3) High-speed steady-state circular turns, figure-eight motion, and meandering patterns

[0098] Furthermore, by combining the following methods, it becomes possible to estimate parameters more accurately. • Within each pattern, P P The estimation is performed iteratively, gradually changing the weights. • After performing the estimation in the order above, P P Change the weights and re-estimate the patterns (1) to (3) above. To cancel out the effects of wind and current, data from the opposite direction is used in estimation for linear motion. In other words, in addition to estimating using a linear motion pattern in a certain direction, estimation is also performed using a linear motion pattern in the opposite direction.

[0099] [Processing flow] Figure 6 is a flowchart of the process including parameter estimation of the motion model according to this embodiment. This process flow is realized, for example, by the control unit 101 of the parameter estimation system 100 reading and executing programs and various data stored in the storage unit 102. Here, the parameter estimation system 100 and the ship 200 are described as separate devices. And, among the steps shown in Figure 5, we will focus on the steps processed by the parameter estimation system 100.

[0100] Before this processing flow begins, it is assumed that the target vessel 200 has completed a voyage and that the voyage log 310 is available. The contents of the voyage log 310 are not particularly limited, but it is assumed that data corresponding to the motion pattern of interest has been acquired, for example, as shown in Figure 4.

[0101] In step S601, the parameter estimation system 100 acquires a navigation log 310. At this time, the parameter estimation system 100 may perform preprocessing such as filtering on the acquired navigation log 310. Preprocessing may include deriving values ​​to be used in each of the subsequent steps from the information contained in the navigation log.

[0102] In step S602, the parameter estimation system 100 determines the motion pattern shown in the navigation log 310 acquired in step S601. The type of motion pattern to be determined may be predetermined.

[0103] In step S603, the parameter estimation system 100 sets gains (weights) for each parameter corresponding to the motion pattern determined in step S602. The gains set here may be predetermined to correspond to the motion pattern. At this time, gains for parameters that are identified independently of the motion pattern may also be set.

[0104] In step S604, the parameter estimation system 100 estimates various parameters of the motion model at a desired time by inputting each input parameter shown in the navigation log 310 to the MHE calculation formula, for which the gain was set in step S603. The parameter estimation is based on the calculation formulas described above.

[0105] In step S605, the parameter estimation system 100 constructs a control model using the motion model parameters estimated in step S604. At this time, a Low Pass Filter (LPF) or the like may be applied to suppress abrupt changes in the estimated parameter values. This makes it possible to improve the stability of the behavior in navigation control using the control model.

[0106] In step S606, the parameter estimation system 100 records the control model constructed in step S605 in association with the motion pattern. Then, this processing flow is terminated.

[0107] In the example above, the motion pattern is determined from the navigation log 310, and the gain setting corresponding to that motion pattern is performed (steps S602 to S603). However, if the motion pattern shown in the navigation log 310 is predetermined, these steps may be omitted, and the process in step S604 may be performed using a calculation formula that has been predetermined to set the gain corresponding to that motion pattern.

[0108] [Differentiation] The above embodiment shows an example using a navigation log obtained during the automatic navigation of the vessel 200. As a variation, a configuration in which a navigation log is recorded during manual navigation will be described.

[0109] Figure 7 is a sequence diagram showing the flow of constructing a control model, including parameter estimation of the motion model, in this modified example. The configuration of the navigation log 720 and the steps S704 to S709 are the same as those of the navigation log 310 and steps S301 to S307 shown in Figure 3. Here, the method of generating the navigation log is different.

[0110] In step S701, the vessel 200 instructs its operator to perform manual operations via the output unit 205, etc. The instructions here are presented based on a predetermined motion pattern.

[0111] In step S702, the vessel 200 receives control commands for manual operation via the control unit 203.

[0112] In step S703, the vessel 200 operates the drive unit 204 based on the control command received in step S702 to perform navigation. At this time, the vessel 200 sequentially records the information acquired by the sensor unit 206, etc., as a navigation log 720. It may also be determined whether the predetermined motion pattern corresponding to the instruction presented in step S701 matches the motion pattern shown in the navigation log 720 obtained as a result of the drive control in step S703. If the motion patterns do not match, the operator may be instructed again to perform manual steering.

[0113] Then, using the navigation logs obtained through manual operation in this manner, the parameters of the motion model are estimated in the same manner as in Figure 3, and a control model is constructed.

[0114] Figure 8 is a flowchart of the process, including parameter estimation of the motion model, based on the processing sequence shown in Figure 7. Here, the parameter estimation system 100 and the ship 200 are described as being configured as a single unit. Therefore, this processing flow is realized, for example, by the control unit 201 of the ship 200 reading and executing programs and various data stored in the memory unit 202.

[0115] Before this processing flow is initiated, it is assumed that information regarding the motion pattern has been pre-registered and that the system is configured to allow manual steering instructions to be presented to the user.

[0116] In step S801, the vessel 200 determines a maneuvering pattern for manual operation. This determination may prioritize maneuvering patterns corresponding to motion patterns with less accumulated navigation logs 720, or it may be determined by focusing on specific motion patterns.

[0117] In step S802, the vessel 200 instructs the operator to operate the vessel based on the maneuvering pattern determined in step S801. This instruction may include rudder angle, acceleration / deceleration, steering timing, etc. The method of instruction is not particularly limited; it may be displayed on a screen (not shown) or instructed by voice.

[0118] In step S803, the vessel 200 receives the operator's maneuvering instructions. The vessel 200 then operates the drive unit 204 based on the received maneuvering instructions and proceeds to navigation. After manual navigation based on the maneuvering pattern is completed, the process proceeds to step S805.

[0119] In step S804, the vessel 200 acquires a navigation log 720 based on the navigation in step S803. At this time, the parameter estimation system 100 may perform preprocessing such as filtering on the acquired navigation log 720. Preprocessing may include deriving values ​​to be used in each of the subsequent steps from the information contained in the navigation log 720.

[0120] In step S805, the vessel 200 determines the motion pattern based on the navigation log 720 acquired in step S804. The result of this determination may be recorded in association with the navigation log 720.

[0121] In step S806, the vessel 200 determines whether the maneuvering pattern determined in step S801 matches the motion pattern determined in step S805. This determination may be based on, for example, the similarity of the navigation routes. If the patterns are determined to match (step S806: YES), the vessel 200 proceeds to step S807. On the other hand, if the patterns do not match, that is, if the navigation is not being carried out according to the maneuvering instructions (step S806: NO), the vessel 200 returns to step S802 and repeats the process. At this point, it may also return to step S801 and execute a different maneuvering pattern.

[0122] In step S807, the ship 200 sets the gain for each parameter corresponding to the motion pattern determined in step S805. The gains set here may be predetermined to correspond to the motion pattern.

[0123] In step S808, the ship 200 estimates the parameters of the motion model at a desired time by inputting each input parameter shown in the navigation log 720 to the MHE calculation model, whose gain was set in step S807. The parameter estimation is based on the calculation formulas described above.

[0124] In step S809, the ship 200 constructs a control model using the motion model parameters estimated in step S808. At this time, a Low Pass Filter (LPF) or the like may be applied to suppress abrupt changes in the estimated parameter values. This makes it possible to improve the stability of the behavior in navigation control using the constructed control model.

[0125] In step S810, the ship 200 records the control model constructed in step S809 in association with the motion pattern. Then, this processing flow is terminated.

[0126] Furthermore, the vessel 200 can automatically navigate using the control model constructed in step S810, based on the operator's instructions. The system may also be configured to repeatedly feed back the above-described processing flow, thereby improving the accuracy of the control model.

[0127] Furthermore, when manual navigation is being performed, the control value obtained by the manual control may be compared with the control value obtained by the control model based on the motion model whose parameters have been estimated, and the control value obtained by the manual control may be adjusted based on the difference. Alternatively, the difference may be notified to the operator.

[0128] In summary, this embodiment makes it possible to easily estimate the parameters of a ship-specific motion model. Furthermore, by using MHE, complex processing involving many parameters becomes possible. In addition, it is possible to improve the robustness of parameter estimation.

[0129] <Other Embodiments> The specifications, characteristics, size, and shape of the vessel to which the above-described motion model parameter estimation process can be applied are not particularly limited. For example, the method according to the present invention can be applied even to vessels with a small amount of accumulated navigation logs.

[0130] Furthermore, in the present invention, the functions of one or more embodiments described above can also be realized by supplying a program or application to a system or device using a network or storage medium, and one or more processors in the computer of that system or device reading and executing the program.

[0131] Alternatively, it may be implemented by a circuit that performs one or more functions (for example, an ASIC (Application Specific Integrated Circuit) or an FPGA (Field Programmable Gate Array)).

[0132] Although various embodiments have been described above with reference to the drawings, it goes without saying that this disclosure is not limited to these examples. It will be clear to those skilled in the art that various modifications, alterations, substitutions, additions, deletions, and equivalents can occur within the scope of the claims, and these will naturally fall within the technical scope of this disclosure. Furthermore, the components of the various embodiments described above can be combined arbitrarily without departing from the spirit of the invention.

[0133] Thus, the present invention is not limited to the embodiments described above. It is also intended and within the scope of protection to be provided for the combination of each configuration of the embodiments, as well as for modifications and applications by those skilled in the art based on the description in the specification and well-known technology.

[0134] As described above, the following matters are disclosed in this specification:

[0135] (Technology 1) A parameter estimation system (e.g., 100) for estimating the parameters of a motion model of a ship (e.g., 200), An estimation unit (for example, 101) uses a navigation log of the vessel when it navigates in a predetermined motion pattern to estimate the parameters of a motion model specific to the vessel that corresponds to the predetermined motion pattern shown in the navigation log. A parameter estimation system having the following characteristics. This configuration makes it possible, for example, to easily estimate the parameters of a ship-specific motion model.

[0136] (Technology 2) The parameter estimation system according to Technical Reference 1, wherein the estimation unit determines the predetermined motion pattern indicated by the navigation log and sets a gain for the parameters included in the motion model in accordance with the determined predetermined motion pattern. This configuration allows, for example, to set gains in accordance with motion patterns, enabling more accurate parameter estimation of motion models corresponding to the motion patterns of each ship.

[0137] (Technology 3) The parameter estimation system according to Technology 1 or Technology 2, wherein the predetermined motion pattern includes a pattern of gradual acceleration and deceleration in a straight line, steady-state turning, figure-eight motion, and meandering pattern. This configuration makes it possible to estimate the parameters of a ship-specific motion model by considering various motion patterns in a ship.

[0138] (Technology 4) The estimation unit, (1) A pattern of repeatedly accelerating and decelerating in stages while moving in a straight line. (2) A pattern of steady-state turning, figure-eight motion, and meandering at a first speed faster than a predetermined speed. (3) A pattern of steady circular turning, figure-eight motion, and meandering at a second speed slower than the predetermined speed. The parameter estimation system described in Technique 3 estimates the parameters of the motion model in the following order. This configuration allows for more accurate parameter estimation of motion models by defining the estimation order according to the motion pattern.

[0139] (Technology 5) The parameter estimation system according to any one of Techniques 2 to 4, wherein the estimation unit estimates the parameters of a motion model by repeating the process while changing the gain in the same motion pattern. This configuration makes it possible to further improve the accuracy of parameter estimation.

[0140] (Technology 6) The parameter estimation system according to any one of Techniques 2 to 5, wherein the estimation unit estimates the parameters of a motion model using a navigation log moving in a predetermined direction and a navigation log moving in the opposite direction to the predetermined direction in a straight-line motion pattern. This configuration suppresses the influence of external factors such as waves and wind, and makes it possible to further improve the accuracy of parameter estimation for motion models in straight-line motion patterns.

[0141] (Technology 7) The parameter estimation system according to any one of the technologies 1 to 6, wherein the estimation unit estimates a predetermined parameter from among a plurality of parameters included in the motion model of the vessel using a navigation log corresponding to a predetermined motion pattern. This configuration makes it possible to switch the target of parameter estimation depending on the motion pattern, such as during high-speed turns and low-speed turns.

[0142] (Technology 8) The aforementioned navigation log includes self-position information and driving information. The self-position information includes at least one of the current position, azimuth angle, and speed of the vessel. The parameter estimation system according to any one of the following technologies, wherein the drive information includes at least one of the propeller rotation speed and the rudder angle. This configuration makes it possible, for example, to estimate the parameters of a ship's unique motion model using the ship's own position information and propulsion information during navigation.

[0143] (Technology 9) A parameter estimation system described in any of Techniques 1 to 8, which uses MHE (Moving Horizon Estimation) to estimate the parameters of a nonlinear motion model. This configuration allows for parameter estimation of more complex motion models compared to linear models, for example. It also improves the robustness of parameter estimation.

[0144] (Technology 10) A recording unit (e.g., 202) for recording the navigation log of the aforementioned vessel, An instruction unit (e.g., 201, 205) that instructs the manual operation of the vessel based on the predetermined motion pattern, A determination unit (for example, 201) determines whether the navigation log of the manually operated vessel recorded in the recording unit, which is performed in response to the instructions from the instruction unit, matches the predetermined motion pattern, It has, The parameter estimation system according to any one of the technologies 1 to 9, wherein if the determination unit determines that there is no match, the instruction unit again instructs the manual operation of the vessel based on the predetermined motion pattern. This configuration makes it possible to obtain the desired navigation log by, for example, performing manual steering. It also improves the operability for the operator performing manual steering.

[0145] (Technology 11) The parameter estimation system according to any one of the Techniques 1 to 10, wherein the navigation log of the vessel is either a navigation log recorded during manual operation based on the predetermined motion pattern, or a navigation log recorded during automatic operation based on the predetermined motion pattern. This configuration makes it possible to estimate the parameters of a ship's unique motion model using navigation logs that include a predetermined motion pattern, specifically those obtained when the ship is operating under either automatic or fully automatic conditions.

[0146] (Technology 12) The motion model estimation system is a parameter estimation system according to any one of the technologies 1 to 11, which is installed on the ship. This configuration makes it possible, for example, to perform real-time parameter estimation and autonomous navigation.

[0147] (Technology 13) A parameter estimation method for estimating the parameters of a ship motion model, An estimation step of using a navigation log of the vessel when it navigates in a predetermined motion pattern, and estimating the parameters of a motion model specific to the vessel that corresponds to the predetermined motion pattern shown in the navigation log, A parameter estimation method having the following characteristics. This configuration makes it possible, for example, to easily estimate the parameters of a ship-specific motion model.

[0148] (Technology 14) A computer (for example, 100), An estimation unit (e.g., 101) uses a navigation log of a vessel navigating in a predetermined motion pattern to estimate the parameters of a motion model specific to the vessel that corresponds to the predetermined motion pattern shown in the navigation log. A program designed to function as such. This configuration makes it possible, for example, to easily estimate the parameters of a ship-specific motion model. [Industrial applicability]

[0149] The present invention is useful, for example, as an apparatus, system, method, or program for estimating the parameters of a motion model used in the automatic operation of a ship. [Explanation of symbols]

[0150] 100... Parameter estimation system. 101... Control Unit 102...Storage section 103...Input section 104…Output section 105... Communications Department 200…ship 201... Control Unit 202...Storage section 203...Operation unit 204…Drive unit 205...Output section 206...Sensor section 207... Antenna section 208... Camera

Claims

1. A parameter estimation system for estimating the parameters of a ship motion model, An estimation unit that uses a navigation log of the aforementioned vessel when it navigates in a predetermined motion pattern to estimate the parameters of a motion model specific to the vessel that corresponds to the predetermined motion pattern shown in the navigation log. A parameter estimation system having the following characteristics.

2. The parameter estimation system according to claim 1, wherein the estimation unit determines the predetermined motion pattern indicated by the navigation log and sets a gain for the parameters included in the motion model in accordance with the determined predetermined motion pattern.

3. The parameter estimation system according to claim 1, wherein the predetermined motion pattern includes a pattern of repeatedly accelerating and decelerating in a stepwise manner while moving in a straight line, steady-state turning, figure-eight motion, and meandering pattern.

4. The estimation unit, (1) A pattern of repeatedly accelerating and decelerating in stages while moving in a straight line. (2) A pattern of steady-state turning, figure-eight motion, and meandering at a first speed faster than a predetermined speed. (3) A pattern of steady circular turns, figure-eight motion, and meandering at a second speed slower than the predetermined speed. The parameter estimation system according to claim 3, which estimates the parameters of a motion model in the following order.

5. The parameter estimation system according to claim 2, wherein the estimation unit estimates the parameters of a motion model by repeating the process while changing the gain in the same motion pattern.

6. The parameter estimation system according to claim 2, wherein the estimation unit estimates the parameters of a motion model using a navigation log moving in a predetermined direction and a navigation log moving in the opposite direction to the predetermined direction in a straight-line motion pattern.

7. The parameter estimation system according to claim 1, wherein the estimation unit estimates a predetermined parameter among a plurality of parameters included in the motion model of the vessel using a navigation log corresponding to a predetermined motion pattern.

8. The aforementioned navigation log includes self-position information and driving information. The self-position information includes at least one of the current position, azimuth angle, and speed of the vessel. The parameter estimation system according to claim 1, wherein the drive information includes at least one of the propeller rotation speed and the rudder angle.

9. The parameter estimation system according to claim 1, which estimates the parameters of a nonlinear motion model using MHE (Moving Horizon Estimation).

10. A recording unit for recording the navigation log of the aforementioned vessel, An instruction unit that instructs the manual operation of the vessel based on the predetermined motion pattern, A determination unit determines whether the navigation log of the manually operated vessel recorded in the recording unit, which is performed in response to the instructions from the instruction unit, matches the predetermined motion pattern. It has, The parameter estimation system according to claim 1, wherein if the determination unit determines that there is no match, the instruction unit again instructs the manual operation of the vessel based on the predetermined motion pattern.

11. The parameter estimation system according to claim 1, wherein the navigation log of the vessel is either a navigation log recorded during manual operation based on the predetermined motion pattern, or a navigation log recorded during automatic operation based on the predetermined motion pattern.

12. The parameter estimation system described in claim 1 is installed on the ship.

13. A parameter estimation method for estimating the parameters of a ship motion model, An estimation step of using a navigation log of the aforementioned vessel when it navigates in a predetermined motion pattern, and estimating the parameters of a motion model specific to the vessel that corresponds to the predetermined motion pattern shown in the navigation log, A parameter estimation method having the following characteristics.

14. Computers, An estimation unit that uses a navigation log of a vessel navigating in a predetermined motion pattern to estimate the parameters of a motion model specific to the vessel that corresponds to the predetermined motion pattern shown in the navigation log. A program designed to function as such.

Images