JP2016159695A - Automatic steering gear for vessel - Google Patents

Abstract

PROBLEM TO BE SOLVED: To reduce a route error caused by an external disturbance component without executing any integral control for the route error.SOLUTION: A feedback control unit for outputting a command steering angle is configured as an estimator 131 for estimating an azimuth error, a route error and a tidal current, an azimuth control feedback gain device 132A constituting an azimuth control loop, and a route control feedback gain device 132B constituting a route control loop including the azimuth control loop, and the estimator 131 inputs a correction amount based on the tidal current estimation error of a sway direction to the azimuth control feedback gain device 132.SELECTED DRAWING: Figure 3

Description

  The present invention relates to a marine vessel automatic steering apparatus for a route control system.

  An automatic marine vessel steering system includes an azimuth control system (HCS) that controls a rudder angle and follows a heading azimuth to a reference azimuth, and a route control system that controls a rudder angle and follows a hull position on a planned route. (TCS: Track Control System). When the rudder angle is changed, the hull generates a yaw angular velocity and a sway velocity, so the route control system is designed in consideration of these movements. Here, the yaw angular velocity is detected by the sensor as a heading obtained by integrating the yaw angular velocity. However, the sway velocity is an unknown amount that is not normally detected, and is an under-actuated amount that cannot use an actuator that directly controls the sway velocity.

Also, the target ship of the ship automatic steering system has offset components in the yaw angular velocity and the sway velocity due to the wind received by the hull and the left / right unbalance of the hull, and these offset components are converted into the steering angle component in the hull model. In conversion, the yaw steering angle offset δ _or and the sway steering angle offset δ _ov are given to the yaw motion model and the sway motion model, respectively, to generate a channel error.

  In addition, as a technology related to route control, the state estimator is divided into a direction control system estimator based on a direction error system and a route control system estimator based on a route error system. In a system estimator, a marine vessel automatic steering device that estimates a state quantity and an estimated tidal current vector of a channel error system is known (see Patent Document 1).

  The estimator includes an azimuth control system estimator, a route control system estimator, and a ground coordinate system tidal current estimator that estimates a tidal current component in the ground coordinate system. In addition, a marine vessel automatic steering device that estimates an estimated tidal current vector from a hull position in a ground coordinate system without using a reference direction is known (see Patent Document 2).

JP 2009-248896 A JP 2013-86745 A

  In route control, it is necessary to correct the route error due to disturbance components including steering angle offset, wave component, and tidal component. Error countermeasures caused by disturbance components need to be integrated and controlled for type 1 servo systems, that is, route errors, but the integrator affects the closed-loop characteristics, so the review of the control system causes a complicated design. There is a problem.

  An embodiment of the present invention is made to solve the above-described problem, and an automatic steering apparatus for a ship that can reduce a channel error caused by a disturbance component without performing integral control on the channel error. The purpose is to provide.

In order to solve the above-described problem, the marine vessel automatic steering apparatus according to the present embodiment includes a trajectory planning unit that outputs a reference orientation and a reference position of a hull, and the orientation and position of the hull from the orientation and position of the hull detected by a sensor. And a feedback control unit that outputs a command steering angle so as to follow the reference azimuth and reference position, the feedback control unit is configured to estimate azimuth error, channel error, and tidal current , An azimuth control system feedback gain unit that constitutes an azimuth control loop, and a route control system feedback gain unit that constitutes a route control loop including the azimuth control loop, and the estimator includes the azimuth control system feedback gain enter the correction amount to the vessel, the amount of correction, gamma correction amount of the steering angle offset [delta] _or the _h, and gamma _t as the correction amount of the route control loop gamma represented by γ _h + γ _t, γ _t is the feedback gain of the orientation control system feedback gain device to _{f 1 F h (1),} v o ^ = - K v δ or ^, where, _{K v} is a lateral flow gain, δ _or ^ is the yaw steering angle offset, _vcR ^ is converted into a reference coordinate component and approximated in the sway direction, and u is the speed of the speed.

で表されることを特徴とする。

It is represented by.

  According to the embodiment of the present invention, it is possible to reduce the channel error due to the disturbance component without performing integral control on the channel error.

It is the block diagram of the ship automatic steering device and the whole control object. It is a block diagram which shows the structure of a feedback control part. It is a block diagram which shows the detailed structure of a feedback control part. It is a figure which shows the coordinate system used with a route control system. It is a figure which shows an azimuth | direction error and a route error. It is a block diagram which shows an azimuth | direction control loop and a route control loop. It is a block diagram which shows an azimuth | direction control loop. It is a figure which shows the specification of an azimuth | direction control loop. It is a diagram showing a root locus of ^GH _h4 Δa (s). It is a figure which shows the asymptotic line of GH _h4 ( ^DELTA ) a (s). It is a diagram showing a root locus of ^GH _h4 Δb (s). It is a block diagram which shows a route control loop. It is a figure which shows the root locus of GH _t (s). It shows a vector representation of r ₁ and _{r 1} '. It is a figure which shows the specification of a route control loop. It is a figure which shows the simulation result in case there is no integration operation | movement and there is no correction amount. It is a figure which shows the simulation result in case of integral operation and no correction amount. It is a figure which shows the simulation result in case there is no integration operation | movement and there exists a correction amount. It is a figure which shows the simulation result in case there is no integration operation | movement, there is a correction amount, and there is a wave component.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. In the following description, each symbol is defined as a variable, modification, and subscript as shown in the following table.

1.1 Configuration of Ship Automatic Steering Device First, a system including the boat automatic steering device of the present invention will be described. FIG. 1 is an overall block diagram of a marine vessel automatic steering apparatus and an object to be controlled. FIG. 2 is a block diagram illustrating a configuration of the feedback control unit. FIG. 3 is a block diagram illustrating a detailed configuration of the feedback control unit. 2 and 3, it means δ _o ^ = δ _or ^.

As shown in FIG. 1, the marine vessel automatic steering device 1 is a device that controls the rudder to track the hull position on the planned route, and includes a trajectory plan unit 11, a trajectory route error calculation unit 12, a feedback control unit 13, The adder 14 and an identifier (not shown) for identifying each parameter are provided. From the guidance system 2, the ship speed U from the planned route and the speed log of the sensors 4 (more precisely, the hull's surge speed u) is input to the trajectory planner 11, and from the trajectory planner 11, the reference orientation ψ _R , the reference orientation A feedforward steering angle δ _FF is output during reference signals such as x _R and y _R and during a course change.

  The sensors 4 of the hull 3 include a speed log that detects the surge speed u of the hull 3, a gyrocompass that detects the heading ψ of the hull 3, and a hull position (x, y) from a satellite positioning system (GNSS) such as GPS. GNSS sensor for detecting

Detection signals from the sensors 4 such as the heading azimuth ψ and the hull position (x, y) are input to the orbital route error calculation unit 12, and the orbital channel error calculation unit 12 receives the reference azimuth ψ _R and the reference azimuth x _R. , Y _R and the detection signal are compared, and an azimuth error ψ _e , a channel error x _e , y _{e and the} like are output.

The closed loop system of the marine vessel automatic steering apparatus 1 includes a control target including a hull model and a disturbance model, and a feedback control unit 13. As shown in FIG. 2, the feedback control unit 13 includes an estimator 131 and a control system feedback gain unit 132. The estimator 131 receives the heading error ψ _e and the route errors x _e and y _e from the orbital route error calculation unit 12.

As shown in FIG. 3, the estimator 131 includes an azimuth control system estimator 131A, a route control system estimator 131B, and a ground coordinate system power flow estimator 131C. The azimuth control system estimator 131A estimates the azimuth error and outputs an estimated value of the state quantity related to the azimuth such as the estimated azimuth error ψ _e ^ from which the disturbance is removed and the estimated turning angle error r ^. The route control system estimator 131B outputs an estimated value of the state quantity related to the route, such as the estimated route error y _e ^ in the lateral direction. The ground coordinate tidal current estimator 131C performs calculations in the ground coordinate system and outputs estimated values related to tidal currents such as estimated tidal current vectors u _c ^ and v _c ^.

The control system feedback gain unit 132 includes a direction control system feedback gain unit 132A and a route control system feedback gain unit 132B. The azimuth control system feedback gain unit 132A multiplies the estimated azimuth error ψ _e ^ and the estimated turning angle error r ^ by a feedback gain. The route control system feedback gain unit 132B multiplies the estimated route error y _e ^ by a feedback gain. Azimuth control system feedback gain units 132A and route control system feedback gain device 132B according to the result summed with a feedback steering angle [delta] _FB is output.

Steering engine of the hull 3, to move the steering angle that is proportional to the command steering angle [delta] _C by marine autopilot 1 having the configuration described above, the hull 3, results in turning angular velocity by the steering angle, orientation, position changes . Along with the generation of the turning angular velocity, a skew angle (lateral velocity) is generated.

2. Formulation 2.1 Coordinate System Here, the route coordinate system will be described. FIG. 4 is a diagram showing a coordinate system used in the route control system. FIG. 5 is a diagram showing a heading error and a navigation error.

As shown in FIG. 4, the coordinate system used in route control system, earth fixed coordinate system O-XY, it consists of a reference coordinate system _{O R -X} R _Y R hull fixed coordinate system _{G-X B Y} B, both This is a right-handed three-axis orthogonal coordinate system. In these coordinate systems, the Z-axis is positive in the direction of gravity, and the rotation polarity is positive in the right-handed screw direction. Since the coordinate system uses two dimensions, the X axis and the Y axis, the Z axis is omitted in FIG. The earth fixed coordinate system has the X axis facing north, the hull fixed coordinate system has the X axis in the bow direction, and the reference coordinate system has the X axis in the forward direction.

Planning route in route control system is set as a path connecting the waypoints shown as W _P in FIG. 5, consisting of straight route and curved route. The reference route is placed on the planned route, the origin is fixed at an appropriate position on the straight route, and the reference route is placed on the intersection of the line connecting the turning center and the hull position on the curved route. Determined. 4B, u is the surge speed, v is the sway speed, r is the turning angular speed, ψ is the heading, ψ _R is the reference direction, δ is the rudder angle, x and y are the hull positions, x _R and y _R is the reference channel, R is the turning radius, x _C and y _C are the turning center, U is the water velocity, U = √ (u ² + v ² ), β is the skew angle, sin β = v / U, and the suffix _c is The tidal component is shown. From these, the azimuth error ψ _e and the channel errors x _e and y _e are as follows.

ここで、ｄｉｓｔ_ｘ＝ｘ’−ｘ_ｃ、ｄｉｓｔ_ｙ＝ｙ’−ｙ_ｃ、Ｍは座標変換行列、

Where dist _x = x′−x _c , dist _y = y′−y _c , M is a coordinate transformation matrix,

2.2 Control Object Here, the control object will be described. The controlled object consists of a hull model and a disturbance model. The hull model uses a response model by the present inventor who improved Nomoto's response model with a constant surge speed.

に定める。ここでｓはラプラス演算子、Δ_ｃ（ｓ）、δ_ｃは命令舵角、Ｋ_ｒは旋回力ゲイン、Ｋ_ｖは横流れゲイン、Ｔ_ｒ，Ｔ_ｒ３，Ｔ_ｖは時定数を示す。本実施の形態においては、安定船を対象とし、δ_ｃは操舵機応答が方位制御応答に比べて十分に速いものとしてδ_ｃ≒δとおく。この船体モデルは野本の船体モデルと比較して、Ｔ_ｒ３を有し、Ｔ_ｖ＝Ｔ_ｒという特徴をもつ。（５）式を次式のように変更する。

Stipulated in Here, s is a Laplace operator, Δ _c (s), δ _c is a command steering angle, K _r is a turning force gain, K _v is a lateral flow gain, and T _r , T _r3 , and T _v are time constants. In the present embodiment, a stable ship is targeted, and δ _c is set as δ _c ≈δ assuming that the steering response is sufficiently faster than the direction control response. This hull model has T _r3 and T _v = T _{r as} compared with Nomoto's hull model. (5) The equation is changed to the following equation.

ここでＣ_Ｔ３＝１−Ｔ_ｒ３／Ｔ_ｒである。上式の状態空間表現を次式に示す。

Here, C _T3 = 1−T _r3 / T _r . The state space expression of the above equation is shown in the following equation.

ここでｘ_ｒ＝［ψ_ｒｘ］^Ｔ、添字^Ｔは転置行列、Δ_ａ，Δ_ｂはパラメータ不確かさを示し、Δ_ａ，Δ_ｂは想定値（添字^〜）とノミナル値の差として、

Here, x _r = [ψ _rx ] ^T , subscript ^T is a transposed matrix, Δ _a and Δ _b are parameter uncertainties, and Δ _a and Δ _b are differences between an assumed value (subscript ^to ) and a nominal value,

になる。Ｔ_ｒ３のパラメータ不確かさはＴ_ｒ３／Ｔ_ｒ＜≒０．１として影響を与え難いため省く。またノミナル値を初期値の速度変化によって定めると

become. Parameter uncertainty of T _r3 is omitted since it is difficult to affect the _{T r3 / T r <≒ 0.1} . Also, if the nominal value is determined by the speed change of the initial value,

になる。ここで添字_０は初期値、Ｋはゲイン、Ｔは時定数を示す。

become. Here, the subscript ₀ indicates an initial value, K indicates a gain, and T indicates a time constant.

Disturbance model, the steering angle offset [delta] _or, [delta] _ov, wave component [psi _w and tidal component _u c shown in FIG. 4, and _{v c.} Here, δ _or , δ _ov are obtained by converting the angular velocity acting around the azimuth axis (yaw direction) and the velocity acting in the sway direction induced by wind, hull characteristics, etc., into steering angles, respectively. Generate a navigation error. ψ _w is a directional conversion of the output of the narrow band filter to which white noise ν is input, and generates an invalid rudder in δ _c . u _c and v _c generate a route error. Therefore, the disturbance model

とする。ここでｘ_ｗ＝［ξ ψ_ｗ］^Ｔ、ξは変数、Ｋ_ｗ，ζ_ｗ，ω_ｗはそれぞれ波浪モデルのゲイン、減衰係数、固有周波数として、

And Where x _w = [ξ ψ _w ] ^T , ξ is a variable, K _w , ζ _w , and ω _w are the wave model gain, attenuation coefficient, and natural frequency, respectively.

2.3 Route error model A route error model is derived. FIG. 6 is a block diagram showing a direction control loop and a route control loop. The ground speed of the hull is

になる。ここで添字_ｇは対地速度成分を示す。これにより参照座標系における航路誤差の速度成分は次式になる。

become. Here, the subscript _g indicates a ground speed component. As a result, the speed component of the channel error in the reference coordinate system is as follows.

ここでｕ_Ｒ，ｖ_Ｒは参照速度を示し、航路保持の横方向成分はゼロになり、ψ_ｅを微小として近似ｃｏｓψ_ｅ≒１，ｓｉｎψ_ｅ≒ψ_ｅを用い、

Wherein _u R, _{v R} represents a reference speed, lateral component of the route holding becomes zero, the approximation cos _e ≒ 1 the [psi _e as minute, using sinψ _e ≒ _{ψ e,}

In the present embodiment, the marine vessel automatic steering apparatus 1 does not control the surge direction, and constructs a navigation control system from the azimuth control loop and the navigation control loop as shown in FIG. Azimuth control loop controls the [psi _e, route control loop controls the v _e, u _e is not controlled. 6 uses the equations (1), (2), and (15), and uses ψ _R = 0 for the sake of simplification. Moreover, in this channel maintenance control system, the steering angle offset and the tidal component are handled as a disturbance model, and the wave component is omitted because the average value is zero.

2.4 Steady error Steady error of heading and route in this embodiment is derived from the offset component of the disturbance model. Figure 6 shows the transfer characteristics of the control system.

になる。ここで、Γ（ｓ），γは修正項を示す。Ｆ_ｈ（ｓ）は比例ゲインｆ_ｐと微分ゲインｆ_ｄからなる方位フィードバック、Ｆ_ｔ（ｓ）は航路ゲインｆ_ｙ、積分ゲインｆ_ｉからなる航路フィードバックであり、それぞれ、

become. Here, Γ (s) and γ indicate correction terms. F _h (s) is azimuth feedback comprising a proportional gain _{f p} derivative gain _{f d, F} t (s) is a route feedback consisting route gain _{f y,} the integral gain _{f i,} respectively,

また、Ｖ（ｓ）は（１７）、（１８）式より

Further, V (s) is obtained from the equations (17) and (18).

になる。ここで

become. here

よって、方位偏差と航路偏差は（１７）〜（２０）式、（２３）式より

Therefore, the heading deviation and the route deviation are obtained from the equations (17) to (20) and (23).

になる。定式の分母である閉ループ特性は安定として、Ｆ_ｈ（ｓ）は固定して、Ｆ_ｔ（ｓ）は積分制御の有無による影響を調べる。方位偏差、航路誤差のそれぞれの定常値を求めると

become. The closed loop characteristic, which is the denominator of the formula, is stable, F _h (s) is fixed, and F _t (s) is examined for the influence of the presence or absence of integral control. Finding the steady values of heading deviation and route error

になる。方位偏差は横流れ速度成分による斜航角を持ち、航路誤差は積分制御の有無により異なる。積分制御は外乱オフセットに対して１型サーボ特性をもつ。よって、修正量が航路誤差をゼロにすることができるかが重要となる。

become. The heading deviation has a skew angle due to the transverse flow velocity component, and the channel error varies depending on the presence or absence of integral control. The integral control has type 1 servo characteristics against disturbance offset. Therefore, it is important whether the correction amount can make the route error zero.

3. Direction Control Loop Here, the direction control loop will be described. FIG. 7 is a block diagram showing an azimuth control loop.

As shown in FIG. 7, the azimuth control loop has zero azimuth error ψ _e = ψ−ψ _R in a control target composed of a hull model having parameter uncertainty and a disturbance model including a steering angle offset and a wave component. It is configured as a control system that converges. For simplicity, we put a zero constant reference azimuth [psi _R. In addition, since it is handled around the azimuth axis, it is not affected by tidal currents. This control system includes an azimuth control system estimator 131A (deterministic observer) and an azimuth control system feedback gain unit 132A (state feedback), and ensures closed-loop stability and disturbance elimination. At this time, the state space expression of the azimuth control loop is as follows.

ここでｘ_ｈ＝［ｘ_ｒ ｘ_ｗ δ_ｏｒ］^Ｔ、添字＾は推定値、添字^￣は検出値、ψ^￣＝−ψ−ψ_ｗは検出方位、Ｋ_ｈは推定ゲイン、Ｆ_ｈはフィードバックゲイン、Ｏ_ｉ×ｊはｉ行ｊ列のゼロ行列を用いて

Where _{x h = [x r x w} δ or] T, the subscript ^ is the estimated value, the subscript ^¯ detection value, ψ ^¯ = -ψ-ψ _w are detected azimuth, _{K h} is the estimated gain, _{F h} feedback gain , O _{i × j} using a zero matrix of i rows and j columns

3.1 Feedback Gain The feedback gain of the azimuth control loop is obtained by giving the design parameter of the specification to the characteristic polynomial of the nominal value using the azimuth control system estimator 131A. The characteristic polynomial is obtained by substituting equation (21) into the denominator of equation (25).

ここでＤ_ｈ（ｓ）は方位制御ループの特性多項式、

Where D _h (s) is the characteristic polynomial of the orientation control loop,

ここでζ_ｈは減衰係数、ω_ｈは固有周波数、ｆ_ｐとζ_ｈを図８に示す設計パラメータとし、上式に与えると

Here zeta _h is an attenuation coefficient, omega _h is the design parameters shown in Figure 8 the natural frequency, f _p and zeta _h, when applied to the above equation

が求まる。本実施の形態において、制御システムは方位制御ループのω_ｈを基準に設定される。フィードバックＦ_ｈ（ｓ）はｒの利用を想定しているため、（７）式のｒ_ｘの利用に変更する。命令舵角δ_ｃ＝−ｆ_ｄｒ−ｆ_ｐψ_ｅに（７）式のｒを代入すると次式を得る。

Is obtained. In the present embodiment, the control system is set based on ω _{h of the} azimuth control loop. Since the feedback F _h (s) assumes the use of r, the feedback F _h (s) is changed to use of r _x in equation (7). Substituting r in the equation (7) into the command steering angle δ _c = −f _d r−f _p ψ _e , the following equation is obtained.

これより、フィードバックゲインは舵角オフセットδ_ｏｒの修正量γ_ｈ＝−δ_ＯＲ＾を加えると次式になる。

Accordingly, the feedback gain is obtained by adding the correction amount γ _h = −δ _OR ^ of the steering angle offset δ _or .

3.2 Estimated Gain The estimated gain of the azimuth control loop is set by matching the characteristic polynomial of the azimuth control system estimator 131A with the specified characteristic polynomial. Ie

と定める。ここでｄｅｔは行列式、Ｉは適当な単位行列、Ｄ_ｅｈ（ｓ），Ｄ_ｅｗ（ｓ），Ｄ_ｅｏ（ｓ）はそれぞれ方位推定、波浪推定、舵角オフセット推定に対応する特性多項式であり、

It is determined. Here, det is a determinant, I is an appropriate unit matrix, D _eh (s), D _ew (s), and D _eo (s) are characteristic polynomials corresponding to bearing estimation, wave estimation, and steering angle offset estimation, respectively. ,

推定ゲインの設計パラメータは図８に示される。推定ゲインＫ_ｈは推定係数ρ_ｈを与えると、補足として後述する推定ゲインの計算方法から求まる。

The design parameters for the estimated gain are shown in FIG. Assuming that the estimated gain K _h is provided with an estimation coefficient ρ _h , the estimated gain K _h is obtained from an estimated gain calculation method described later as a supplement.

3.3 Estimated coefficient ρ _h
A method for setting the estimation coefficient ρ _h will be described. As described above, when ρ _h is obtained, the estimated gain K _h is determined. The hull model has modeling errors due to approximation. In addition, the hull parameters have fluctuation components due to changes in speed, draft and trim conditions. When such parameter fluctuation components and modeling errors are aggregated as parameter uncertainties and incorporated in the controlled object, the closed loop stability changes. In this case [rho _h is set to satisfy the specifications of the closed loop stability by parameter uncertainty.

  As described above, the feedback gain in the bearing control loop is obtained based on the nominal value of the hull parameter, and the estimated gain is obtained based on the closed loop stability considering the parameter uncertainty. The robustness of the closed-loop stability against parameter uncertainty is high in the control system excluding the estimator, but is reduced in the control system including the estimator. Since an estimator that ensures disturbance rejection is essential, it is necessary to use a closed-loop control system that includes parameter uncertainty for the estimation coefficient.

Parameter uncertainty is given to the controlled object to derive the characteristic polynomial of the closed-loop control system. Based on the characteristic polynomial, the parameter uncertainty that has a large effect on the closed loop stability (damping coefficient) is selected. An estimation coefficient ρ _h that satisfies the design parameter of the attenuation coefficient due to the parameter uncertainty is obtained. In the azimuth control loop of equation (28), if the characteristic matrix Α _h ^Δ of the expansion system is assumed to have an estimation error of η = x _h −x _h ^,

になる。ここで、

become. here,

よって、Ａ_ｈ ^Δの特性多項式をノミナル項とパラメータ不確かさ項とに分離し、Δ_ａ，Δ_ｂを同時に加えないとすれば、

Therefore, if the characteristic polynomial of A _h ^Δ is separated into a nominal term and a parameter uncertainty term, and Δ _a and Δ _b are not added simultaneously,

になる。ここでｆ_１＝Ｆ_ｈ（１），ｆ_２＝Ｆ_ｈ（２）を用いて、

become. Here, using f ₁ = F _h (1) and f ₂ = F _h (2),

上式は項数が多く取扱いが難しいため、制御対象を船体モデルのみとする。このとき、推定器は２次系に、閉ループは４次系になり、特定多項式は次式になる。

Since the above equation has many terms and is difficult to handle, the control target is only the hull model. At this time, the estimator becomes a quadratic system, the closed loop becomes a quartic system, and the specific polynomial becomes the following equation.

なお、２次系の推定ゲインは後述する（８６）式を用いる。

The estimated gain of the secondary system uses equation (86) described later.

The scale of the closed loop stability is defined as a minimum attenuation coefficient using a characteristic polynomial, and the influence of Δ _a and Δ _{b on} the attenuation coefficient will be described below. First, in the case of delta _a, if the estimated coefficient and ρ _h »1, (37) formula

と近似できる。パラメータ不確かさの範囲を次式と定める。

Can be approximated. The range of parameter uncertainty is defined as

これより、減衰係数は（３９）式より

From this, the attenuation coefficient is calculated from the equation (39).

のとき最小値となる。後述する検証の条件より−１／３ω_ｈＴ_ｒ≒−０．２９，ζ_ｈ４ ^Δａ≒０．４２になる。よって、Δ_ａの影響は、推定係数をρ_ｈ≫１とすればρ_ｈに無関係になる。また、（３７）式においてΔ_ａをゲインに、閉ループ伝達関数を

At the minimum. From the verification conditions described later, −1 / 3ω _h T _r ≈−0.29 and ζ _h4 ^Δa ≈0.42. Thus, the influence of the delta _a becomes independent of the [rho _h if the estimated coefficients and [rho _h >> 1. Also, the gain delta _a in equation (37), a closed loop transfer function

におくと、特性多項式はＤ_ｈ４ ^Δａ（ｓ）＝１＋Δ_ａＧＨ_ｈ４ ^Δａ（ｓ）になる。上式の根軌跡の一例を図９に示す。図９において、×は極、○はゼロ点を示し、漸近線の本数（上式で分母の極数から分子のゼロ点数を差し引いた数）は推定器の次数に関係なく実軸上の１本になる。

In other _words , the characteristic polynomial is D _h4 ^Δa (s) = 1 + Δ _a GH _h4 ^Δa (s). An example of the root locus of the above equation is shown in FIG. In FIG. 9, x indicates a pole, ○ indicates a zero point, and the number of asymptotes (the number obtained by subtracting the number of zeros of the numerator from the number of poles of the denominator in the above formula) is 1 Become a book.

Next, the case of delta _b, the estimated coefficient [rho _h is the results of the delta _a, determined from a specific polynomial in delta _b. (38) the gain of the delta _b in formula, the closed-loop transfer function

におく。図１０に示すように、上式の漸近線は３本になり、Δ_ｂの正負で閉ループは不安定になる。減衰係数はΔ_ｂ＜０の場合には１以外に指定することができないが、Δ_ｂ＞０の場合には０から１まで指定することができる。パラメータ不確かさの範囲を次式と定める。

Put it in. As shown in FIG. 10, asymptote in the above equation becomes three, closed loop positive and negative delta _b becomes unstable. The attenuation coefficient cannot be specified other than 1 when Δ _b <0, but can be specified from 0 to 1 when Δ _b > 0. The range of parameter uncertainty is defined as

（４２）式の根軌跡を図１１に示す。

FIG. 11 shows the root locus of the equation (42).

Here, the specifications regarding the setting of the estimation coefficient ρ _h are summarized.

<1> Parameter uncertainty and design parameters are as follows.

<2> Factor the seventh-order characteristic polynomial of equation (36) to obtain an estimated coefficient ρ _h that matches the minimum attenuation coefficient with the design parameter.

ここで添字^＊はΔ_ｂを含んだ場合を示す。なお、上式の算法については後述する。

Here subscript ^* indicates a case that contains the Δ _b. The calculation method of the above formula will be described later.

4). Next, the framework of the route control loop and the setting of its gain will be described. FIG. 12 is a block diagram showing a route control loop.

  As shown in FIG. 12, the framework of the route control loop is the same as the direction control loop. The state space representation of the route control loop is

ここで、ｘ_ｔ＾＝［ｙ_ｅ＾ ｖ_ｅｏ＾］^Ｔ、ｖ_ｅｏ＾はモデル化、検出信号と積分計算などの誤差を吸収するオフセット成分、ｘ_ｔ∫はｙ_ｅ＾の積分量、Ｋ_ｔは推定ゲイン、Ｆ_ｔはフィードバックゲイン、γ_ｔは修正量を示し、

Where x _t ^ = [y _e v _eo ^] ^T , v _eo ^ is an offset component that absorbs errors such as modeling, detection signal and integration calculation, x _{t ∫} is the integral of y _e ^, K _t is an estimated gain, F _t is a feedback gain, γ _t is a correction amount,

システム行列Ａ_ｔと入力行列Ｂ_ｔに制御対象のパラメータを含まない構成を採用し、推定器がパラメータフリーになる。また、図１２においては省略されているが、航路制御ループには潮流成分を推定する対地座標系潮流推定器１３１Ｃが含まれる。

Employs a configuration that does not include the parameters of the control object in the system matrix A _t as input matrix B _t, estimator is a parameter-free. Although omitted in FIG. 12, the navigation control loop includes a ground coordinate system tidal current estimator 131C that estimates tidal current components.

4.1 Feedback gain The feedback gain in the route control loop is obtained in the same way as in the direction control loop.

とする。ここで、ｆ_ｙは航路ゲイン、ｆ_ｉは積分ゲインを示す。ｆ_ｙは航路制御ループを安定化し、ｆ_ｉを除いて求める。ｆ_ｉは航路制御ループを外乱のオフセット成分に対して１型サーボ系に構成し、ｆ_ｙを用いて求める。

And Here, f _y represents the channel gain, and f _i represents the integral gain. f _y stabilizes the route control loop is obtained with the exception of f _i. f _i is obtained by configuring the route control loop as a type 1 servo system with respect to the offset component of the disturbance and using f _y .

4.1.1 Channel Gain The channel gain is determined by giving the design parameter attenuation coefficient ζ _t to the characteristic polynomial D _t (s) of the channel control loop. D _t (s) is obtained by rearranging the denominator of equation (25).

ここでｚはゼロ点、ａ_ｔは係数、ζ_ｔは減衰係数、ω_ｔは固有周波数、

Where z is zero, _{a t} is a coefficient, zeta _t attenuation coefficient, omega _t is the natural frequency,

（４８）式において、ｓの係数を比較すると次式になる。

In the equation (48), when the coefficient of s is compared, the following equation is obtained.

（４８）式からω_ｔに関して解くと、３次方程式は次式になる。

Solving for ω _t from equation (48), the cubic equation becomes:

上式で係数の符号は正負あり、解の符号についても同様である。よって、上式にζ_ｔを与えるとω_ｔの解は正の最小値として求まり、その最大値はω_ｈになる。これよりａ_ｔ，ｆ_ｙは次式から求まる。

In the above equation, the sign of the coefficient is positive and negative, and the sign of the solution is the same. Therefore, when ζ _t is given to the above equation, the solution of ω _t is obtained as a positive minimum value, and the maximum value is ω _h . This from _a t, _{f y} is obtained from the following equation.

4.1.2 Design parameter ζ _t
The design parameter ζ _t is set based on the characteristics of the route control loop. The characteristic polynomial is characterized by C _t <0 and z <0 resulting from sway hull motion. Therefore, when the channel gain is increased, the conjugate root moves to the unstable side and makes the closed loop unstable. When the case of no zero point is added to the equation (48), the following equation is obtained.

上式と等価な閉ループ伝達関数はｆ_ｙをゲインとして

The closed-loop transfer function equivalent to the above equation has _fy as the gain

になる。ここでＤ_ｔ（ｓ）＝１＋ｆ_ｙＧＨ_ｔ（ｓ）である。

become. Here, D _t (s) = 1 + f _y GH _t (s).

The root locus of the above equation is shown in FIG. In FIG. 13, when there is a zero point, it is indicated by a solid line, when there is no zero point, it is indicated by a broken line, (a) shows a wide area, and (b) shows an enlargement near the origin. The conjugate root moves from the pole x. From (a), when there is no zero point, it converges to an asymptotic line ± 60 degrees, and when there is a zero point, it goes to the zero point ○ and a positive value. On the other hand, from (b), the conjugate root with zero point moves to the real axis side from the conjugate root without zero point. In the case where there is a zero point in the equation (52), to obtain the characteristic root sensitivity with respect to _fy ,

になる。ここで

become. here

これより、感度はｆ_ｙに関するｓの微係数に相当し

From this, the sensitivity corresponds to the derivative of s with respect to _fy.

になる。ここで

become. here

The sensitivity at the pole of the conjugate root (f _y = 0) is obtained. The pole corresponds to the root of the azimuth control loop, and its conjugate root is selected by the following equation.

ここで減衰係数ζ_ｒ１＝ζ_ｈと偏角θ_ｒ１との関係はζ_ｒ１＝−ｃｏｓθ_ｒ１になり、ａ_ｒ１＝ω_ｈ ^２，θ_ｒ１＝（３／４）π，ｒ_１を（５５）式に代入すると

The relationship between the deflection angle theta _r1 and where the damping coefficient ζ _r1 = ζ _h becomes _{ζ r1 = -cosθ r1, a r1} = ω h 2, θ r1 = (3/4) π, the _{r 1} (55) When assigned to an expression

になる。ここで

become. here

になる。ｚ＝−∞は（５２）式でゼロ点なしの場合に相当する。よって、極での感度は（５４）式より次式になる。

become. z = −∞ corresponds to the case where there is no zero point in the equation (52). Therefore, the sensitivity at the pole is expressed by the following equation from the equation (54).

ここでθ＝θ_ｎ−θ_ｄであり、このとき、感度の偏角は次式になる。

Here, θ = θ _n −θ _d . At this time, the declination of sensitivity is expressed by the following equation.

よって偏角はゼロ点の存在により−４５度より下向きになる。したがって、極ｒ_１の位置変化は航路ゲインを微小として

Therefore, the declination becomes lower than -45 degrees due to the presence of the zero point. Therefore, the change in the position of the pole r ₁

に近似できる。上式の関係を図１４に示す。ｒ_１’の減衰係数はθ＜−π／４からζ_ｔ’＞ζ_ｈになる。また、共役根はｆ_ｙが大きくなると不安定になる傾向をもつ。この結果、ζ_ｔはζ_ｈから増加し最大値に達してその後減少しζ_ｈを通過する。図１３（ｂ）中、ζ_ｔはζ_ｈとの比較において、小の黒丸が最大、中の黒丸が等価、大の黒丸が減少を示す。設計パラメータである減衰係数は、共役根と実数根のバランスを考慮して次式の条件を選ぶ。

Can be approximated. The relationship of the above equation is shown in FIG. The attenuation coefficient of r ₁ ′ changes from θ <−π / 4 to ζ _t ′> ζ _h . In addition, the conjugate root tends to become unstable as _fy increases. As a result, ζ _t increases from ζ _h , reaches a maximum value, and then decreases and passes through ζ _h . In FIG. 13B, in comparison with ζ _h , ζ _t indicates that the small black circle is maximum, the middle black circle is equivalent, and the large black circle is decreased. For the damping coefficient, which is a design parameter, the following equation is selected in consideration of the balance between the conjugate root and the real root.

4.1.3 Integral gain Integral gain is obtained using the already obtained channel gain, taking into account closed-loop stability mainly with disturbance rejection. Here, disturbance elimination means a response time until disturbance is suppressed. In addition, due to the fact that the closed loop stability decreases in proportion to the integral gain, the transient phenomenon tends to increase as the response time is advanced. The characteristic polynomial is changed from a cubic expression to a quartic expression by introducing an integral element, and a conjugate root is newly derived near the origin. The integral gain is calculated by designating a damping coefficient ζ _i as a design parameter.

f _i is obtained by giving the design parameter ζ _i to the characteristic polynomial D _ti (s) of the route control loop. D _ti (s) is derived from the denominator of equation (25).

になる。ここで添字_ｔｉ，_ｉは、それぞれ、Ｄ_ｔ（ｓ）の共役根、実根から派生したものを示す。上式において、ｓの係数を比較すると

become. Here, the subscripts _ti and _i indicate those derived from the conjugate root and real root of D _t (s), respectively. In the above equation, when comparing the coefficients of s

になる。上式をω_ｉに関して解くと、４次方程式は次式になる。

become. When the above equation is solved with respect to ω _i , the quartic equation becomes the following equation.

上式にζ_ｉを与えれば、ω_ｔの解は正の最小値として求まる。その最大値はＤ_ｔ（ｓ）の共役根付近になる。よってω_ｔｉ，ζ_ｔｉとｆ_ｉは次式から求まる。

If ζ _i is given to the above equation, the solution of ω _t is obtained as a positive minimum value. The maximum value is near the conjugate root of D _t (s). Therefore, ω _ti , ζ _ti and f _i are obtained from the following equations.

The closed loop root arrangement by feedback is first determined by f _p and ζ _{h for} the two roots of the azimuth control loop, then determined by ζ _t = ζ _{h for} the second root of the route control loop, and finally integrated characteristics in the route control loop. The four roots with are determined by ζ _i .

4.1.4 Closed-loop stability Closed-loop stability (robustness) against parameter uncertainty in the route control loop will be described. The parameter uncertainty uses the speed change of equation (10). The route control loop omits the parameter-free estimator and uses only feedback. In the case of an azimuth control loop, if a speed change is applied to equation (31)

になり、減衰係数は速度を修正しなければ，ζ_ｈ≠ζ_ｈ０になる。また、航路制御ループの場合、（４９）式に速度変化を適用すると

If the speed is not corrected, the damping coefficient becomes ζ _h ≠ ζ _h0 . In the case of a route control loop, if a speed change is applied to equation (49)

になる。上式を（５０）、（５１）式に代入すると

become. Substituting the above equation into equations (50) and (51)

になる。よってｆ_ｙは速度修正に無関係で一定値になり、減衰係数はζ_ｔ＝ζ_ｔ０を保つ。ただしζ_ｈ＝ζ_ｈ０の場合。

become. Therefore f _y becomes a constant value independent of the speed correction, the attenuation factor is keeping ζ _{t =} ζ _t0. However, when ζ _h = ζ _h0 .

  On the other hand, when the integral gain is used, the speed change is applied to the equations (65) and (66).

になる。よってｆ_ｉはζ_ｉを一定に保つため、速度修正が必要になる。

become. Therefore, f _i needs to be corrected to keep ζ _i constant.

  In the control gain excluding the proportional gain, both the differential gain and the integral gain are easily affected by the speed change because of the time dimension, and the route gain is not easily affected because it is not the time dimension. Therefore, the control gain of the route control loop should be handled by the route gain without using the integral gain in order to ensure the closed loop stability. However, for that purpose, it is necessary that the navigation error shown in the equation (27) is corrected to almost zero by the correction amount.

4.2 State Estimator The route control loop includes a route control system estimator 131B and a ground coordinate system power flow estimator 131C. The route control system estimator 131B removes disturbance components (position jump, wave component, etc.) included in the detection signal and extracts an appropriate signal. The ground coordinate system tidal current estimator 131C includes a factor for moving the hull in addition to the tidal current component, and operates not only while maintaining the route but also during curved navigation. These estimators have the same configuration, and the estimated gain is obtained by giving a design parameter of the natural angular frequency.

4.2.1 Route control system estimator The route error estimation formula is given by equation (46).

になる。ここでＫ_ｔ＝［ｋ_ｔ１ ｋ_ｔ２］^Ｔは航路推定ゲイン、ｙ_ｅ ⁻は航路誤差検出量。上式の特性多項式を次式と定める。

become. Here, K _t = [k _t1 k _t2 ] ^T is the channel estimation gain, and y _e ⁻ is the channel error detection amount. The above characteristic polynomial is defined as:

ここでω_ｅｔ，ω_ｅｔｏは仕様の角周波数を示す（詳細は図１５を参照）。これより航路推定ゲインは次式になる。

Here, ω _et and ω _eto indicate the angular frequency of the specification (see FIG. 15 for details). From this, the route estimation gain is as follows.

4.2.2 Ground coordinate system tidal current estimator Since tidal current components change during navigation, it is required to always be able to estimate them in navigation control. Since the tidal current component is a function of the reference direction as shown in the equation (16), it becomes non-linear as it is and is difficult to handle. Therefore, if tidal current component estimation (tidal current estimation) is composed of ground coordinates, the tidal current component can be described in a line format that does not include a direction. Since the tidal current estimation is obtained from the moving speed of the hull position, the moving speed due to wind and hull imbalance cannot be separated.

  The estimated value of ground speed is given by the following equation from equation (14).

  When the tidal current estimation formula is constructed from the above formula, the x system and the y system have the same shape without mutual interference, and are separated as shown in the following formula.

ここでｘ⁻，ｙ⁻は船体位置検出量、ｋ_ｃ１，ｋ_ｃ２は潮流推定ゲインを示す。よって潮流推定式は（７０）式と同形になるので、潮流推定ゲインも同様に定まる。上式の特性多項式を次式と定める。

Here, x ⁻ and y ⁻ indicate hull position detection amounts, and k _c1 and k _c2 indicate tidal current estimation gains. Therefore, since the tidal current estimation formula has the same form as the equation (70), the tidal current estimation gain is similarly determined. The above characteristic polynomial is defined as:

ここでω_ｅｃ，ω_ｅｃｏは仕様の角周波数を示す（詳細は図１５を参照）。これより潮流推定ゲインは次式になる。

Here, ω _ec and ω _eco indicate the angular frequency of the specification (refer to FIG. 15 for details). From this, the tidal current estimation gain is as follows.

  Next, the tidal current estimation error resulting from the hull velocity component is shown. The steady value of tidal current estimation is the following equation from equation (73).

  Estimated values of ground hull velocity and tidal velocity are defined as

ここでΔは誤差項を示す。上式を（７８）式に代入すると、潮流推定誤差は

Here, Δ indicates an error term. Substituting the above equation into equation (78), the tidal current estimation error is

として求まる。ここでΔ_ｕ，Δ_ｖはそれぞれｓｕｒｇｅ速度の誤差、ｓｗａｙ速度の誤差を示す。

It is obtained as Here, Δ _u and Δ _v indicate the error of the surge speed and the error of the sway speed, respectively.

4.2.3 Correction amount The correction amount of the route control loop will be described. This correction amount is input from the estimator 131 to the direction control system feedback gain unit 132A. When the equation (79) is converted to a reference coordinate component and approximated, the tidal current estimation error is

になる。このとき修正量にｖ_ｃＲ＾を用いると（２７）式から

become. At this time, using v _cR ^ as the correction amount,

になる。ここで

become. here

よって、航路誤差をゼロにする修正量は次式から得られる。

Therefore, the correction amount that makes the route error zero is obtained from the following equation.

As mentioned above, trend estimation cause errors due to the offset component delta _v of sway velocity. On the other hand, the route error is not caused by unknown v _o but caused by v _o ^, but can be canceled by adding v _o to the correction amount. Therefore, in this embodiment, a method using the correction amount is employed without using the integral control. The control gain of the route control loop is shown in FIG.

5. Verification This embodiment is verified by simulation. Conditions for this simulation are shown below.

<Planned route>
The planned route is initial heading: 40 deg, amount of change of course: 10 deg, ship speed: U = 10 kn. Ignore the needle change response.

<Control target>
The hull model as a control object uses the following secondary system.

The hull parameters are marner class L = 160.93 m, T ₁ = 177.0 s, T ₂ = 11.6 s, T _r3 ^* = 27.8 s, T _v3 ^* = 5.9 s, K _r ^* = 0.123 1 / s, K _v ^* = − 9.8 m / s, and when identified by the first order model, T _r ^˜ = 85.6 s, T _r3 ^˜ = 5.99 s, K _r ^˜ = 0.0745 1 / s , K _v ^˜ = 4.85 m / s.

The disturbance parameters are offset component δ _or = δ _ov = 3 deg, wave component ψ _w = a _w sin 2π / cycle × t, a _w = 1 deg, cycle 13 seconds, tidal component U _c = 5 kn, ψ _c = 0 deg.

<Control system>
The hull parameters are set as u ⁻ = U, K _r = 0.0267 1 / s, K _v = −2.67 m / s, T _r = 36.1 s, T _r3 = 3.6 s.

The feedback gains (see FIGS. 8 and 15) are: ω _h = 0.0323 rad / s (period 194 seconds), f ₁ = 1.41, f ₂ = 19.3 s, ω _t = 0.0217 rad / s, _{f t = 0.00197, a t =} 0.0150rad / s, ω i = 0.00588rad / s, ω ti = 0.0229rad / s, f i = 0.00257 1 / s, ζ ti = 0.766 .

The estimated gains are ρ _h2 = 3.52 (secondary system), ρ _h = 4.43, ω _w = 0.524 rad / s (period 12 seconds), and ζ _w = 0.1.

<Parameter uncertainty>
Parameter uncertainty is

になる。

become.

The result of the simulation under the above conditions will be described. Note that the correction amount γ _h of the azimuth control loop is valid. FIG. 16 is a diagram illustrating a simulation result when there is no integration operation and no correction amount. FIG. 17 is a diagram illustrating a simulation result when the integration operation is performed and the correction amount is not present. FIG. 18 is a diagram illustrating a simulation result when there is no integration operation and there is a correction amount. FIG. 19 is a diagram illustrating a simulation result when there is no integration operation, there is a correction amount, and there is a wave component.

  When there is no integration operation and no correction amount, as shown in FIG. 16, although it is a stable transient response, a channel error occurs due to a tidal component. When the direction is 40deg

  When the integration operation is performed and there is no correction amount, as shown in FIG. 17, the channel error converges to zero by the integration control of the channel control loop.

In the case where there is no integration operation and there is a correction amount, as shown in FIG. 18, the channel error becomes substantially zero due to the correction amount γ _t of the route control loop, and the transient response is faster than in the case where there is an integration operation and there is no correction amount. Even if a wave component is further added to this condition, the channel error becomes almost zero as shown in FIG. Therefore, the power flow correction using the correction amount γ _t is almost the same as in the case of the integral control. Therefore, according to the present invention, it is possible to reduce the channel error due to the disturbance offset component without using integral control with high sensitivity of parameter uncertainty and large influence on the closed loop stability.

6). Calculation of estimated gain Here, the calculation of estimated gain will be supplemented. The characteristic polynomial of the estimator is obtained by expanding the left side of equation (33).

になる。ここで、Ｈ＝［ｈ_４ ｈ_３ ｈ_２ ｈ_１ ｈ_０］^Ｔ，推定ゲインＫ_ｈ＝［ｋ_１ ｋ_２ ｋ_３ ｋ_４ ｋ_５］^Ｔにおくと

become. Here, when H = [h ₄ h ₃ h ₂ h ₁ h ₀ ] ^T and the estimated gain K _h = [k ₁ k ₂ k ₃ k ₄ k ₅ ] ^T

  On the other hand, the characteristic polynomial of the specification expands its right side

になる。ここで

become. here

  Therefore, the estimated gain is expressed by the following equation when the coefficients of s are compared.

An estimated gain K _h2 = [k ₁₂ k ₂₂ ] T of the secondary system is obtained.

上式からｓの係数を比較すると、２次系の推定ゲインは次式になる。

Comparing the coefficient of s from the above equation, the estimated gain of the secondary system becomes the following equation.

7). Calculation Method of Estimated Coefficient ρ _h Here, a supplementary explanation will be given regarding the calculation method of the estimated coefficient ρ _h . When calculating the estimation coefficient ρ _h from the seventh order equation, an appropriate initial value is required. In the calculation of the estimation coefficient ρ _h , first, a solution of a quartic equation satisfying the specification is obtained, and then an initial value of a seventh equation is set from the solution.

Condition: near ζ _h ^Δ , subscript 2 indicates the case of a secondary estimator,
<1> in the vicinity of the minimum of ζ ^* is ζ _h ^Δ, is approximately proportional with respect to ρ _h.
<2> The initial value ρ _h20 = 3, 7th order expression of the quaternary expression ρ _h0 , ω ₀ ^*

に定める。なお、刻みは１とおく。

Stipulated in The increment is set to 1.

Ρ _h is obtained through two steps: a solution method using convergence calculation is used.
_Find ρ _h2 , ω ₂ ^* where quaternary expression ζ ₂ ^* = ζ _h ^Δ .
Find ρ _h and ω ^{* such} that the seventh order equation ζ ^* = ζ _h ^Δ .

[Rho] _h solution by convergence calculation: A solution method of algebraic equations is used.
<1> A range of ρ _h sandwiching ζ _h ^Δ is determined by an initial value and a step.
<2> ρ _h that satisfies ζ ^* = ζ _h ^Δ is determined by convergence calculation within that range.

Solving algebraic equations: Factor the given equation and find the minimum attenuation coefficient and its natural frequency from the conjugate root.
Quaternary Equations Quaternary equations are factored into two quadratic equations by an algebraic solution.
Seventh-order equation Initial values ω ₀ ^* , ζ ₀ ^* and the Bairstow method are converted to the second-order and fifth-order equations, and the initial value ρ ₀ ω ^* and the Newton-Raphson method are converted to the first-order and fourth-order equations, respectively. Factor. The above equation is used as the quartic equation.

  The embodiments of the present invention are presented as examples and are not intended to limit the scope of the invention. These novel embodiments can be implemented in various other forms, and various omissions, replacements, and changes can be made without departing from the scope of the invention. These embodiments and modifications thereof are included in the scope and gist of the invention, and are included in the invention described in the claims and the equivalents thereof.

DESCRIPTION OF SYMBOLS 1 Ship automatic steering apparatus 11 Trajectory plan part 13 Feedback control part 131 Estimator 132 Control system feedback gain unit 132A Direction control system feedback gain unit 132B Route control system feedback gain unit

Claims (1)

A trajectory planning unit that outputs the reference direction and reference position of the hull, and feedback control that outputs a command rudder angle so that the direction and position of the hull follow the reference direction and reference position from the direction and position of the hull detected by the sensor. An automatic steering device for a ship provided with a section,
The feedback control unit includes an estimator for estimating an azimuth error, a lane error, and a tidal current, an azimuth control system feedback gain unit that constitutes an azimuth control loop, and a lane control system that constitutes a lane control loop including the azimuth control loop A feedback gain unit,
The estimator inputs a correction amount to the direction control system feedback gain unit,
The correction amount is expressed as γ = γ _h + γ _t, where γ _h is the correction amount of the steering angle offset δ _or , and γ _t is the correction amount of the route control loop.
γ _t is the feedback gain F _h (1) of the azimuth control system feedback gain unit, f ₁
v _o ^ = − K _v δ _or ^, where K _v is the lateral flow gain, δ _or ^ is the yaw steering angle offset,
sway direction estimation error approximated by converting v _cR ^ to a reference coordinate component,
Let u be the surge speed,

A marine vessel automatic steering device characterized by the following:

Images