Automatic maneuvering device for vessel

Abstract

To provide technology capable of maintaining stability in a closed loop control system even when the parameter of a controlled object is varied from a stable vessel to an unstable vessel.SOLUTION: An automatic maneuvering device for vessel includes: a feedback control unit 13 for carrying out azimuth control to output an azimuth control system feedback steering angle for having bow azimuth of a hull track a reference azimuth; a parameter identification unit 15 for identifying a hull parameter of the hull; and a robust setting unit 16 for setting a proportional gain included in a design parameter on which a feedback gain for calculating an azimuth control system feedback steering angle is based to be greater than a reference value as well as setting an attenuation coefficient included in the design parameter to be greater than a reference value when a nominal value for the hull parameter is not updated by the parameter identification unit 15, and for setting a proportional gain to a reference value as well as setting an attenuation coefficient to a reference value when a nominal value for the hull parameter is renewed by the parameter identification unit 15.SELECTED DRAWING: Figure 9

Description

[Technical Field]

[0001] The present invention relates to a technique for automatically steering a ship. [Background technology]

[0002] Conventionally, there is known an autonomous ship as shown in Non-Patent Document 1. Such an autonomous ship is given a mission to depart port, arrive at a destination in the shortest time possible, perform tasks such as rescue, relief, search and surveillance, and then return to port.

[0003] During the performance of such missions, the autonomous ship is required to be maneuvered by an autopilot instead of the helmsman. In this case, it is desirable to use a self-contained autopilot, as described in Non-Patent Document 2, rather than the conventional adaptive autopilot. The main purpose of an adaptive autopilot is to follow a course or route and to respond to disturbances caused by the oceanographic environment. On the other hand, a self-contained autopilot not only performs the above-mentioned tracking and responds to disturbances, but also has the function of always keeping the control system in an appropriate state based on the operating state. This allows the autonomous ship to focus on its original purpose.

[0004] Robustness is one of the capabilities of a self-contained control system. In tankers and cargo ships, the draft changes during cargo handling operations. As the draft changes, the parameters of the hull model also change. Suppose that the parameters of a course-stabilized ship (simply called a stable ship) change to those of an unstable ship after cargo handling operations. If the autopilot cannot tolerate the parameter changes, it may cause yawing. Considerations regarding the allowable limits for such course-unstable ships are known (see Non-Patent Document 3).

[0005] The present inventors have reported a robust control technique for the case where there is uncertainty in hull parameters (simply called uncertainty) (see Non-Patent Document 4). However, the uncertainty range was in the stable ship region. Therefore, this robust control technique cannot guarantee closed-loop stability in the unstable ship region. [Prior art documents] [Non-patent literature]

[0006] [Non-Patent Document 1] Hideyuki Ando, ​​Capt Satoru Kuwahara, Capt Koji Kutsuna, and Capt Jun Nakamura, Development and demonstration of autonomous ships in japan, In World Maritime Technology Conference (WMTC), 2022. [Non-patent document 2] Fuyuki Hane, Practical Design Methodology for Marine Autopilots, PhD thesis, Osaka University, 2023. [Non-patent document 3] Kazuhiko Hasegawa, Chuji Ishiyama, Hiroki Umeda, "A Study on the Allowable Limits of Course Instability Caused by Autopilot (1st Report)", Journal of the Society of Naval Architects of Japan, Vol. 1980, No. 148, pp. 92-100, 1980. [Non-patent document 4] Fuyuki Hane, Design of Course-Keeping Systems Using Analytical Methods Based on Course-Keeping Control, Journal of the Japan Society of Naval Architects and Ocean Engineers, Vol. 23, pp. 33-44, Jun 2016. [Non-patent document 5] Hayase Minoru, Introduction to System Control Engineering, Ohmsha, 1980. [Non-patent document 6] Francis H. Raven, Automatic Control Engineering, McGraw-Hill, Inc., New York, NY, USA, 3rd edition, 1978. [Non-Patent Document 7] Fuyuki Hane, Comprehensive Identification Method of Ship Motion Parameters, Journal of the Japan Society of Naval Architects and Ocean Engineers, Vol. 20, pp. 27-38, Dec 2014. [Non-patent document 8] Fuyuki Hane, Identification Method of Wave Disturbance Parameters for Autopilots, Proceedings of the Japan Society of Naval Architects and Ocean Engineers Conference, (23): 461-465, 2016. Summary of the Invention [Problem to be solved by the invention]

[0007] The problem to be solved by the present invention is to provide a technique that can maintain the stability of a closed-loop control system even when the parameters of the controlled object change from a stable ship to an unstable ship. [Means for solving the problem]

[0008] One embodiment comprises a feedback control unit that performs azimuth control and outputs a azimuth control system feedback rudder angle that causes the bow heading of the hull to follow a reference heading; a parameter identification unit that identifies hull parameters of the hull; and a robust setting unit that, when the nominal value of the hull parameter is not updated by the parameter identification unit, sets a proportional gain included in design parameters on which a feedback gain for calculating the azimuth control system feedback rudder angle is based to a value greater than the reference value of the proportional gain and sets a damping coefficient included in the design parameters to a value greater than the reference value of the damping coefficient, and, when the nominal value of the hull parameter is updated by the parameter identification unit, sets the proportional gain to the reference value of the proportional gain and sets the damping coefficient to the reference value of the damping coefficient. [Effects of the Invention]

[0009] According to the present invention, it is possible to provide a technique that can maintain the stability of a closed-loop control system even when the parameters of the controlled object change from a stable ship to an unstable ship. [Brief explanation of the drawings]

[0010] [Figure 1] 1 is a table showing the tendency of hull parameters to change with draft changes. [Figure 2] FIG. 1 is a diagram showing Ca characteristics. [Figure 3] FIG. 10 is a diagram showing the relationship between draft and parameter uncertainty. [Figure 4] FIG. 10 is a diagram showing ship motion and heading error. [Figure 5] FIG. 1 is a diagram illustrating a configuration of a closed-loop control system. [Figure 6] FIG. 10 is a diagram showing the root locus of GHh4 Δa(s) by Δa. [Figure 7] FIG. 10 is a diagram showing the root locus of GHh4 Δa(s) by Δb. [Figure 8] 1 is a table illustrating a robust control approach. [Figure 9] 1 is a block diagram showing the configuration of a marine vessel automatic steering device according to an embodiment; [Figure 10] FIG. 4 is a schematic diagram illustrating the operation of a robust control unit according to the embodiment. [Figure 11] FIG. 10 is a diagram showing the minimum values ​​of the damping coefficients ζh4 Δa and ζh7 Δa due to Δa. [Figure 12] 1 is a table showing hull parameters and parameter uncertainties. [Figure 13] 10 is a table summarizing the simulation results. [Figure 14] FIG. 10 is a diagram showing a simulation result when robust control is not performed. [Figure 15] FIG. 10 is a diagram showing the heading, course error, and commanded rudder angle at the start of robust control. [Figure 16] FIG. 10 is a diagram showing the heading, course error, and command rudder angle at the time of updating the nominal value for heading control when robust control is performed. [Figure 17] FIG. 10 is a diagram showing the heading, course error, and command rudder angle at the time of updating the nominal values ​​for course control when robust control is performed. DETAILED DESCRIPTION OF THE INVENTION

[0011] Hereinafter, an embodiment of the present invention will be described with reference to the drawings.

[0012] (1 specification) (1.1 Assumptions) Assumptions regarding the draft change and hull parameters of the hull according to this embodiment are set as follows.

[0013] 1. The hull is not equipped with a draft gauge. 2. Hull parameter (nominal value) K r ,T r ,T r3 ,K v is proportional to the load. Here, K r ,K v are the turning force gain and the cross flow gain, respectively, and T r ,T r3 is the time constant. Also, the absolute value of the hull parameters when fully loaded is larger than when ballast loaded, and T r is K r The trend is more significant. 3. The tendency of hull parameters to change with draft is shown in Fig. 1.

[0014] (1.2 specifications) The specifications will be explained.

[0015] First, the hull parameter uncertainty (hereafter referred to as uncertainty) Δ a ,Δ b From the above assumption 2,

[0016]

number

[0017] (1-C a -1 ) characteristic becomes discontinuous near the origin with respect to Ca, as shown in Figure 2.

[0018] The closed-loop stability is calculated under the above uncertainties using the seventh-order characteristic polynomial D of the closed-loop system. clh Δ (s) Determined from D clh Δ Factorize (s) into a second-order system to find the damping coefficient of the conjugate root. Set the minimum value of the damping coefficient to be equal to or greater than the specification. Therefore, the minimum damping coefficient

[0019]

number

[0020]

number

[0021] (1.3 Parameter Variation) The parameter variations based on draft are explained below.

[0022] The draft changes depending on the load. As the draft gets deeper, |T r | becomes larger, and K r / T r As a result, the uncertainty Δ a is positive (+), Δ b tends to be negative. As the draft gets shallower, the uncertainty Δ a ,Δ b tends to be the opposite.

[0023] As shown in Figure 3, the relationship between draft and parameter uncertainty is characterized by draft changes being limited to the second quadrant (shallower draft) and the fourth quadrant (deeper draft).

[0024] (2 characteristic polynomial) The closed-loop stability is understood from the characteristic root of the characteristic polynomial. Therefore, this section focuses on the Heading Control System (HCS) and explains the following: 1. Determine the control target and the control system, and find the closed-loop control system. 2. The characteristic polynomial is obtained from the closed-loop control system. 3. Find the characteristic roots by factoring the characteristic polynomial into a quadratic expression.

[0025] (2.1 Control Object) The controlled object is assumed to consist of a ship model and a disturbance model.

[0026] As shown in Figure 4, the ship model is modeled using a yaw motion model with a constant surge speed.

[0027]

number

[0028]

number

[0029] The disturbance model is the yaw steering angle offset δ ro and wave component ψ w The tidal current component is assumed to have no effect on the azimuth axis. Therefore, the disturbance model is

[0030]

number

[0031]

number

[0032] (2.2 Control System) The control system is composed of a linear state estimator and state feedback, and forms a closed-loop control system to ensure closed-loop stability and disturbance rejection.

[0033] The control system can be expressed in state space as follows:

[0034]

number

[0035]

number

[0036] The estimator estimates the state quantity of the controlled object using the estimation gain based on the input heading error. The feedback multiplies the estimated value by the feedback gain to output the commanded steering angle. The control gain is made up of the estimation gain and the feedback gain, and is calculated based on given design parameters. Therefore, the design parameters determine the performance of the closed-loop control system.

[0037] (2.3 Closed-loop control system) A closed loop control system is described.

[0038] As shown in Figure 5, the closed-loop control system consists of the controlled object described in Section 2.1 and the control system described in Section 2.2. Therefore, the closed-loop control system is

[0039]

number

[0040]

number

[0041] The ship model has parameter uncertainties, and the effect of these uncertainties can degrade the stability of the closed-loop system.

[0042] (2.4 Characteristic polynomial) Derive the characteristic polynomial of the closed-loop control system.

[0043] The control system is designed using the characteristic polynomial, which is the characteristic matrix A (in bold) of the augmented system. h Δ From equation (15), Α (bold) h Δ of

[0044]

number

[0045]

number

[0046]

number

[0047] The above equation includes a disturbance model and is a seventh-order equation. To qualitatively evaluate the effect of parameter uncertainty on closed-loop stability, a fourth-order characteristic polynomial excluding the disturbance model is used. The seventh-order characteristic polynomial is used to calculate the estimated gain including the effect of the disturbance model.

[0048] (2.4.1 Fourth-order characteristic polynomial) The fourth-order characteristic polynomial is simpler than the seventh-order characteristic polynomial, making it easier to analyze hull parameters. In this fourth-order characteristic polynomial, the hull model and estimator become a second-order system.

[0049] Therefore, the fourth-order characteristic polynomial is, from Section 2.4,

[0050]

number

[0051]

number

[0052] (3) Robust Control In this section, we calculate the parameter uncertainty Δ a ,Δ b The effect of Δ a ,Δ b In this section, we will consider measures to prevent the deterioration of closed-loop stability based on the fourth-order characteristic polynomial. This measure involves reviewing the design parameters of the control gains. The design parameters are the proportional gain K p and the damping coefficient ζ h is used as the feedback gain, and the estimated damping coefficient ζ eh and the estimated coefficient ρ h is used as the estimated gain.

[0053] The closed-loop stability is governed by the stable conjugate roots (dominant roots) closest to the origin in the Laplace plane (see Non-Patent Document 5). Therefore, we focus on the damping coefficient of the dominant roots.

[0054] (3.1 Δ a (Measures against Δ a If ρ is given, the estimated coefficient is sufficiently large (ρ h ≫1), then from equation (20),

[0055]

number

[0056] Δ a Check the root locus of (see Non-Patent Document 6). Δ a The open loop transfer function with gain is

[0057]

number

[0058] In Figure 6, which shows the root locus of the above equation, the cross marks D h (s) extreme r h1,h2 =-ζ hωh ±ω h (1-ζ h 2 ) 0.5 indicates D eh2 The pole of (s) is far enough to the left. Its representative root is located at a radius ω from the pole around the origin. h As the object moves up, its damping coefficient changes. a If it is >0, the damping coefficient is decreased, and Δ a If <0, increase the damping coefficient.

[0059] Δ a The damping coefficient of is given by (23):

[0060]

number

[0061]

number

[0062] (3.1.1 Design parameter K p ,ζ h ) The design parameter for the feedback gain is the proportional gain K p and the damping coefficient ζ h It is equivalent to ζ h4 Δa To increase the margin of error, the design parameter K p ,ζ h Review.

[0063] 1.K p From the reference value 1 to 1 <K p <2, where K p Use =1.5. 2. Zeta h From the reference value 1 / √2 to 1 / √2<ζ h < 1, where ζ h =0.9 is used.

[0064] The effectiveness of the above measures is verified through a numerical example. a =-2,K r =0.03s ?1 ,T r = 100s, the improvement is as follows:

[0065]

number

[0066] This effect is observed for a second-order estimator, and is expected to be reduced further with a fifth-order estimator.

[0067] (3.2 Δ b (Measures against Δb If is given, then from equation (21) c t3 If is omitted, the following equation is obtained, and the estimated coefficients affect the closed-loop stability.

[0068]

number

[0069] Δ b From the above equation, Δ b The open loop transfer function with gain is

[0070]

number

[0071] In Figure 7, which shows an example of the root locus of the above equation, the cross marks indicate D. h (s) extreme r h1,h2 and D eh (s) extreme r eh1,eh2 =-ζ eh ω eh ±ω eh (1-ζ eh 2 ) 0.5 The circles indicate zeros, the dashed lines indicate asymptote lines, and the dashed lines indicate the damping coefficient specifications. The number of asymptote lines is the number of poles in the denominator minus the number of zeros in the numerator, which is three in Figure 7, and Δ b Here, b1<0, b2<0. From the situation where the representative root moves to the right (unstable direction), ζ h4 Δb C to maximize degradation b >0 is from the uncertainty specification in Section 1.2

[0072]

number

[0073] (3.2.1 Design parameter ζ eh,ρh ) Among the design parameters of the estimator, the damping coefficient is ζ eh = 1 / √2, and the estimated coefficient ρ h is adjusted to meet the specifications. h The value of is not calculated from the fourth-order characteristic polynomial, but from the seventh-order characteristic polynomial. The former does not include the disturbance model, so ρ h becomes smaller.

[0074] The estimated coefficients are calculated numerically. The parameter uncertainty conditions are set using Equation (30), and the calculation procedure is as follows:

[0075] 1.C b The characteristic polynomial containing the formula is factorized into a second-order system to find the minimum value of the damping coefficient. 2. The estimated coefficients are found by matching the damping coefficients to the specifications, which requires convergent calculations.

[0076] In numerical calculations, an appropriate initial value is required to factorize a seventh-order polynomial. In this solution, the initial value is determined by using the quadratic estimated coefficient of the fourth-order characteristic polynomial. Therefore, the above procedure can be performed in two stages as follows:

[0077] Step 1: Quadratic estimated coefficient ρ of the fourth-order characteristic polynomial h2 Solving a quartic equation does not require an initial value and can be done algebraically. Step 2: Estimated coefficient ρ of the seventh-order characteristic polynomial h =ρ h7 The initial value of the seventh degree equation is ρ h2 Set from.

[0078] For specific calculation methods, please refer to Non-Patent Document 4.

[0079] (3.3 Control Gain Setting) The control gains include feedback gains and estimated gains, which are determined by the design parameters. The design parameters are selected to prevent the deterioration of closed-loop stability due to parameter uncertainty. Therefore, the effect of parameter uncertainty under the worst-case conditions is as follows:

[0080] 1. Design parameter K p ,ζ h corresponds to the case where the hull parameter fluctuation changes from a stable ship to an unstable ship, and is set as explained in Section 3.1, and the feedback gain is calculated as described later in Appendix Section A.1. 2. Design parameter ζ eh ,ρ h corresponds to the case where the fluctuation changes from an unstable ship to a stable ship, and is set as explained in Section 3.2, and the estimated gain is calculated as described later in Appendix Section A.2.

[0081] (3.4 Robust Control) Robust control in response to draft changes is implemented based on the following three approaches.

[0082] The first is to adjust the control gain as described above. However, this measure is temporary rather than permanent because it increases the feedback gain. The second method is to obtain and use updated values ​​for the hull parameters by hull parameter identification (see Non-Patent Document 7). If the updated values ​​are appropriate, the parameter uncertainty will decrease, and the feedback gains can be returned to their standard values. The third limitation is the steering mode. Steering modes are heading control and course control. Compared to heading control, course control controls sway ship motion, so it tends to be more affected by parameter uncertainty. Course control is performed once an updated value for the nominal value is obtained.

[0083] These robust control approaches are summarized in Figure 8.

[0084] (3.5 Automatic Steering Systems for Ships) The configuration of a marine vessel automatic steering device according to this embodiment, which implements the countermeasures for draft changes described above, will be described below. Fig. 9 is a block diagram showing the configuration of the marine vessel automatic steering device according to this embodiment. Fig. 10 is a schematic diagram showing the operation of the robust control unit according to this embodiment.

[0085] As shown in Figure 9, the automatic steering device for a vessel 1 according to this embodiment has a hull 2 ​​as a control object, and includes a reference signal generating unit 11, an error calculating unit 12, a feedback control unit 13, a feedforward control unit 14, a parameter identifying unit 15, and a robust setting unit 16.

[0086] The hull 2 ​​is equipped with a propulsion device, a rudder, and sensors. The sensors include a speed log that detects the surge speed u of the hull 2, a gyrocompass that detects the heading ψ of the hull 2, and a GNSS sensor that detects the hull position (x, y) obtained from a global navigation satellite system (GNSS) such as GPS.

[0087] The reference signal generator 11, error calculator 12, feedback controller 13, feedforward controller 14, and parameter identifier 15 constitute a control system for controlling the rudder angle of the hull 2. As will be described in detail later, the robust setting unit 16 sets the steering mode and design parameters in the control system according to the state of the control system.

[0088] The reference signal generator 11 generates a reference signal based on the planned route included in the setting information input from outside the control system. R and the reference position (x R ,y R The error calculation unit 12 calculates the time series signal of the reference direction ψ R and the heading error ψ e and the reference position (x R ,y R ) and the course error y based on the ship position (x,y) e Calculate the route error y e is the reference position (x R ,y R) and a line passing through the hull position (x, y) and extending in the sway direction from the hull 2.

[0089] The feedback control unit 13 performs heading control and route control. When performing heading control, the feedback control unit 13 uses the heading error ψ calculated by the error calculation unit 12 as e The feedback steering angle of the azimuth control system is set to a feedback steering angle δ FB When performing route control, the feedback control unit 13 outputs the route error y calculated by the error calculation unit 12. e The route control system feedback rudder angle is set to feedback rudder angle δ FB When both heading control and course control are performed, the feedback rudder angle obtained by adding the feedback rudder angle of the course control system to the feedback rudder angle of the heading control system is output as feedback rudder angle δ FB Output as

[0090] The feedforward control unit 14 calculates the reference direction ψ R Feedforward steering angle δ based on FF This feedforward steering angle δ FF and feedback steering angle δ FB The command rudder angle δc obtained by adding the above is output to the hull 2.

[0091] The parameter identification unit 15 identifies the parameters of the hull model and the wave model by a known method based on the time series data of the heading ψ, the hull position (x, y), and the surge speed u detected by the sensors. For the parameter identification of the hull model, please refer to Non-Patent Document 7. For the parameter identification of the wave model, please refer to Non-Patent Document 8. The parameter identification unit 15 also identifies the update flag

[0092]

number

[0093] The update flag is set to 0 when the nominal value has not been updated, and is set to 1 when the nominal value has been updated. The update flag is also set to 0 as its initial value when the ship 2 departs port.

[0094] As shown in FIG. 10, the robust setting unit 16 includes a switch SW, a setting changeover unit RF, a parameter setting unit DP, and a steering mode setting unit PM.

[0095] The switch SW selects whether or not to set the control system by the robust setting unit 16, and outputs On or Off as a switch output according to the ON / OFF information included in the setting information. Here, On indicates that the control system is set by the robust setting unit 16, and Off indicates that the control system is not set by the robust setting unit 16.

[0096] The setting switching unit RF sets the robust flag uf based on the switch output and the update flag uf.

[0097]

number

[0098] The parameter setting unit DP sets the feedback gain F for calculating the azimuth control system feedback steering angle according to the value of the robust flag rf output from the setting switching unit RF. h (See Section A.1 below for details) is the design parameter based on which the proportional gain K p and the damping coefficient ζ h is set as follows:

[0099]

number

[0100] In this way, when the control system is set by the robust setting unit 16 and the nominal value is not updated, the parameter setting unit DP sets the proportional gain K p and the damping coefficient ζ h and are set to values ​​greater than the reference values ​​and within the ranges shown in Section 3.1.1. In addition, when the control system is not set by the robust setting unit 16 or when the nominal value is updated, the parameter setting unit DP sets the proportional gain K p and the damping coefficient ζ h are set to the reference value. h is a proportional gain K set by the parameter setting unit DP by the feedback control unit 13. p and the damping coefficient ζ h The calculation shall be based on the following.

[0101] The steering mode setting unit PM sets the control system, specifically the steering mode of the feedback control unit 13, according to the value of the robust flag rf output from the setting switching unit RF, as shown in the following equation.

[0102]

number

[0103] In this way, when the robust setting unit 16 sets the control system and the nominal value is not updated, the steering mode setting unit PM limits the steering modes that can be implemented by the feedback control unit 13 to only azimuth control. In addition, when the robust setting unit 16 does not set the control system or the nominal value is updated, the steering mode setting unit PM releases the restriction on the steering modes that can be implemented by the feedback control unit 13.

[0104] (4) Numerical Verification The effectiveness of the marine vessel automatic steering device according to this embodiment will be verified by simulation.

[0105] (4.1 Δ a Damping coefficient characteristics due to Δ a The damping coefficient ζ h4 Δa ,ζ h7 Δa In Figure 11, which shows the minimum value of C a = -2, and the subscript h4,h7 are the fourth and seventh degree characteristic polynomials, respectively. a The damping coefficient by has the following characteristics:

[0106] ζ h4 Δa is ζ h7 Δa Compared to , it is about 0.1 to 0.15 larger due to the reduction in order. The influence of hull parameters is r ×T r tend to depend on Specifications ζ spec Δ ≦ζ h7 Δa To satisfy r ×T r ≧3}∩ζ h4 Δa ≥ 0.55 is required.

[0107] (4.2 Effect of robust control) (4.2.1 Effects of Δa measures) The following Δ a The results of the measures show that the measures are effective.

[0108] ·Hull parameters and parameter uncertainty Δ a ,Δ bIn Figure 12, the target ship is an unstable ship, and the nominal values ​​of the control system are set at the start and at the time of update. The values ​​at the time of update are parameters identified from a single course change response (for details of parameter identification, see Non-Patent Document 7). Δ a ,Δ b In this case, the updated value is significantly reduced compared to the initial value due to the effect of parameter identification. In Figure 13, which summarizes the simulation results, h is the estimated coefficient, and the angular frequency ω h [rad·s -1 ] is the value when min{ζh}. Figures 14 to 17, which correspond to Figure 13, show the heading ψ and course error y e and commanded rudder angle δ c 13, the prior art shows a case where robust control is not performed, and the present invention shows a case where robust control is performed. - At the start of the conventional technology, min{ζh}=0.14, and it can be seen from FIG. 14 that closed-loop stability cannot be ensured. - At the start of the present invention, min{ζh}=0.41, and it can be seen from FIG. 15 that closed-loop stability can be ensured.

[0109] (4.2.2 Nominal Value Update) The results of updating the nominal value can be summarized as follows: The effect of updating the nominal value is effective.

[0110] In Figure 12, the effect of nominal value update is to reduce parameter uncertainty, and the closed-loop stability is significantly improved compared to the initial state. In the update of the present invention in FIG. a ,Δ b 16 and 17, it can be seen that the influence of the two is reduced, and min{ζh} = 0.71 for both.

[0111] (Appendix A Control Gain) The control gain consists of a feedback gain and an estimated gain.

[0112] (A.1 Feedback Gain) The feedback gain is calculated by comparing the characteristic polynomial of the closed loop with its specifications and giving the design parameters. The characteristic polynomials of both are

[0113]

number

[0114]

number

[0115] The feedback gain is the design parameter K p ,ζ h Therefore, from equations (35) and (36) and the above equation, the following equation can be obtained:

[0116]

number

[0117] K p ,K d The feedback control is the turning angular velocity r=ψ · Since it is assumed that r x Change to using r and r x The relationship between

[0118]

number

[0119]

number

[0120] Therefore, the feedback gain F his expressed as follows:

[0121]

number

[0122] (A.2 Estimated Gain) (A.2.1 Second-order estimation gain) The second-order estimation gain corresponds to an estimator that uses only the hull model. The characteristic polynomial of the second-order estimation gain is

[0123]

number

[0124]

number

[0125] Comparing the coefficients of s from equations (42) and (43), the second-order estimated gain is given by

[0126]

number

[0127] (A.2.2 5th order estimation gain) The fifth-order estimation gain corresponds to the case where the estimator includes a hull model, a wave model, and a rudder angle offset model. The characteristic polynomial of the fifth-order estimation gain corresponds to its determinant, and when it corresponds to the specifications, it is

[0128]

number

[0129] The characteristic polynomials of the specifications are as follows:

[0130]

number

[0131] D eh In (s), ω eh ,ζ eh is the design parameter ζ eh ,ρ h where ρ h is a design parameter called an estimation coefficient. D ew In (s), the wave components are removed by a notch filter, so the notch filter condition is

[0132]

number

[0133]

number

[0134]

number

[0135] D eo In (s), ω eo is the estimated coefficient ρ ho Using this, the following equation is obtained:

[0136]

number

[0137] ρ ho ≪ρ h Therefore, δ^ ro ψ^ e ,r^ x It estimates slowly compared to

[0138] Estimated gain K (bold) h The characteristic polynomial of the estimator is derived by expanding the right-hand side of equation (46).

[0139]

number

[0140]

number

[0141]

number

[0142] On the other hand, the characteristic polynomial of the specification is obtained by expanding the right-hand side of equation (47)

[0143]

number

[0144] Therefore, the estimated gain can be calculated by matching the above coefficients H (bold) and E (bold).

[0145]

number

[0146] The embodiments of the present invention are presented as examples and are not intended to limit the scope of the invention. This novel embodiment can be embodied in various other forms, and various omissions, substitutions, and modifications can be made without departing from the spirit of the invention. This embodiment and its modifications are included within the scope and spirit of the invention, and are also included in the inventions described in the claims and their equivalents. [Explanation of symbols]

[0147] 1 Automatic steering system for ships 2. Hull 11 Reference signal generation section 12 Error calculation section 13 Feedback control section 14 Feedforward control section 15 Parameter Identification Unit 16 Robust setting section

Claims

1. a feedback control unit that performs azimuth control by outputting a azimuth control system feedback rudder angle to make the hull's heading follow the reference heading; a parameter identification unit that identifies hull parameters of the hull; a robust setting unit that, when the nominal value of the hull parameter is not updated by the parameter identification unit, sets a proportional gain included in design parameters on which a feedback gain for calculating the azimuth control system feedback rudder angle is based to a value greater than the reference value of the proportional gain and sets a damping coefficient included in the design parameters to a value greater than the reference value of the damping coefficient, and, when the nominal value of the hull parameter is updated by the parameter identification unit, sets the proportional gain to the reference value of the proportional gain and sets the damping coefficient to the reference value of the damping coefficient; An automatic steering device for a vessel.

2. The feedback control unit further performs route control to output a route control system feedback rudder angle that causes the hull position of the hull to follow the route, 2. The marine vessel automatic steering device according to claim 1, wherein the robust setting unit further restricts the feedback control unit to only perform the heading control when the nominal value of the hull parameter is not updated by the parameter identification unit, and releases the restriction on the feedback control unit when the nominal value of the hull parameter is updated by the parameter identification unit.

Images