JP6632497B2 - Ship automatic steering system - Google Patents

Description

  The present invention relates to an automatic steering device for a ship having an estimator.

  An automatic steering device for a ship is a device that controls the rudder of a ship so that a heading direction detected by a gyro sensor follows a fixed set direction, and has a function of holding a hand and changing a course of a hand. Is controlled by a feedback controller composed of a state feedback controller for performing PD control and an estimator for performing state estimation and disturbance elimination.

  The hull model to be controlled by the automatic marine vessel steering system has a parameter uncertainty of the nominal value due to a change in the load. Occurs. This error may degrade the closed-loop stability due to the state feedback of the estimated value, and may cause yawing to meander the hull.

With respect to such an error of the estimated value, the inventors of the present application disclose in Patent Document 1 the parameter error due to the parameter uncertainty by actively considering the parameter uncertainty of the nominal value of the hull parameter. An automatic steering device for ships to be reduced and a design method thereof are proposed. Here, λ _e2 is a characteristic polynomial for estimating the state quantity of the hull model in which the wave model is omitted, and ζ _e and ω _e are an attenuation coefficient and a natural frequency for estimating the state quantity, respectively. Polynomial

とし、ω_ｅ＝ρω_ｆを操舵系周波数ω_ｆのρ倍（ρ＞１で推定係数と呼ぶ）に設定し、ζ_ｅを１／√２に設定するように推定器を設計している。これによって、パラメータ不確かさによる推定角速度の推定誤差が低減される。

The estimator is designed so that ω _e = ρω _f is set to ρ times the steering system frequency ω _f (referred to as an estimation coefficient when ρ> 1), and _{ｅ e} is set to 1 / √2. Thereby, the estimation error of the estimated angular velocity due to the parameter uncertainty is reduced.

  Further, in this Patent Document 1, for a control object including a hull model, a wave model, and two disturbance models of a steering angle offset model, an estimation error is equivalent to an estimation error when only a hull model is used. It is proposed to correct the eigenfrequency of the estimator from the disturbance model characteristics. According to this, the control target of the estimator is equivalently only the hull model, and the disturbance model can be ignored.

Also, in Patent Document 2, the inventors focus on the estimation coefficient ρ = ω _e / ω _f with _respect to the representative root of the steering loop, and design a control gain from the relationship between the attenuation coefficient of the representative root and the parameter uncertainty. ing. Here, by estimating the effect of parameter uncertainty based on a closed-loop system without an estimator and a closed-loop system with an estimator, an estimation coefficient ρ that relates the eigenfrequency of the estimator and the eigenfrequency of the steering system is set. are doing.

In addition, based on a closed-loop characteristic polynomial including an estimator including two disturbance models in Patent Document 3, the inventors of the present invention have proposed a characteristic polynomial D when a parameter uncertainty Δ of a hull parameter is added to this characteristic polynomial. _{From c} ^Δ , when the parameter uncertainty Δ is the first specification value (Δ _spec ), the smallest attenuation coefficient の う ち among the attenuation coefficients obtained from the characteristic root of D _c ^Δ (s) is _{ｅ e} ^*, and this ζ _e ^* is determined the estimated coefficient ρ satisfying the value of the second specification value (zeta _spec), estimator by setting the natural frequency omega _e hull model, marine autopilot with increased loop stability Has been proposed.

Further, in this Patent Document 3, the parameter uncertainty Δ is defined as the uncertainty (Δ _b ) of the ratio between the turning force gain of the hull parameter and the time constant, so that the parameter uncertainty having a high effect on the closed-loop stability is obtained. It has been proposed to improve the reliability more than before in consideration of the estimated root caused.

  Further, as a water jet propulsion ship that obtains propulsion by a nozzle that injects a water flow, there is a water jet propulsion ship in which nozzles are provided in a pair on the left and right. In changing the course of a water jet propulsion ship, the nozzle has a dead zone within a predetermined angle inclined right and left from the straight traveling direction of the ship. There is known a technique in which a nozzle is inclined outward by a certain nozzle offset so as to have a C-shape, thereby equivalently canceling out the dead zone.

Japanese Patent No. 4603832 Japanese Patent No. 4820621 Japanese Patent No. 5682009

  The above-described control system using the automatic steering device for ships targets various hulls from small boats to large ships, but when this control system is applied to the above-described water jet propulsion ship, the turning of the water jet propulsion ship It has been confirmed that a problem occurs in the process of obtaining the estimation coefficient because the force gain is reduced by half. That is, according to the above-mentioned automatic steering device for a boat, there is a problem that it is not possible to cope with a hull having a small turning force gain, not limited to a water jet propulsion boat.

  An embodiment of the present invention has been made in order to solve the above-described problems, and has as its object to provide an automatic steering device for a boat that can cope with a hull having a small turning force gain.

In order to solve the above-described problems, the automatic steering device for a marine vessel of the present embodiment is an automatic steering apparatus for a marine vessel that controls a control target including a hull and a steering device so as to make the heading follow the reference direction, A feedback controller that outputs a command steering angle based on the reference direction and the heading direction to the control target, wherein the feedback controller is based on a deviation between the reference direction and the heading direction and the command steering angle. An estimator that outputs an estimated deviation and an estimated angular velocity as an estimated value; multiplies the estimated deviation by a proportional gain K _P set to a predetermined first specified gain value; and outputs the proportional gain to the estimated angular velocity. multiplied by the derivative gain K _D that depends on the value of K _P and a state feedback controller for outputting the command steering angle, configuration from the state feedback controller and the estimator Closed loop characteristic polynomial is the subscript _h azimuth control, subscript _e is the estimator, with modifications ^- represent respectively the nominal value, Laplace operator and s, D (s) the characteristic polynomial, the ζ damping coefficient, omega the natural angular _{frequency, ζ eh = ζ h, ω} e = ρ 2 ω eh, ρ 2 secondary estimation coefficient, the turning force gain _{K r,} hull time constant _{T r, T r3} steering sensitivity time constant, delta _b, delta _b ', a _{d b} a scalar quantity parameter uncertainty respectively,

で表され、
前記状態フィードバック制御器は、前記微分ゲインＫ_Ｄの値が負値である場合、前記第１仕様ゲイン値より大きく設定された第２仕様ゲイン値を前記比例ゲインＫ_Ｐに設定し、前記推定器は、前記第２仕様ゲイン値に設定された比例ゲインＫ_Ｐと所定値に設定されたパラメータ不確かさΔ_ｂ’とに基づいて、前記閉ループの特性多項式により前記推定係数ρ_２を算出することを特徴とする。

Represented by
The state feedback controller, the case where the value of the derivative gain K _D of a negative value, setting the second specification gain value that is larger than said first specification gain value to the proportional gain K _P, the estimator Calculates the estimation coefficient ρ ₂ by the closed-loop characteristic polynomial based on the proportional gain K _P set to the second specification gain value and the parameter uncertainty Δ _b ′ set to a predetermined value. Features.

  According to the embodiment of the present invention, it is possible to cope with a hull having a small turning force gain.

It is a block diagram showing the system including the automatic steering device for ships concerning an embodiment. FIG. 3 is a block diagram illustrating a configuration of a feedback controller. It is a block diagram which shows a 2nd-order estimator. It is a diagram showing an asymptote of ^GH _h4 Δb (s). It is a diagram showing a root locus of ^GH _h4 Δb (s). It is a flowchart which shows operation | movement of an estimation coefficient calculation process. It is a diagram showing a root locus of the representative root of ρ _{2 =} 4. It is a figure showing a simulation result.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

1 Automatic Steering Device for Ship 1.1 Overall Configuration First, a system including the automatic steering device for ship of the present invention will be described. FIG. 1 is a block diagram illustrating a system including the automatic steering device for a boat according to the embodiment.

As shown in FIG. 1, the system according to the present embodiment includes an automatic steering device 1 for a boat and a boat 2 to be controlled by the system. Here the ship 2, steering gear 21, heading ψ of the hull 22 and the hull 22 ^- is a combination of the gyrocompass 23 for detecting a. Automatic marine steering system 1, heading [psi to set the azimuth [psi _R ^- comprises a feedback controller 10 for outputting a command steering angle [delta] _C to follow the to the steering device 21.

1.2 Feedback Controller The configuration of the feedback controller will be described. FIG. 2 is a block diagram illustrating a configuration of the feedback controller.

As shown in FIG. 2, the feedback controller 10 includes a state feedback controller 11 and an estimator 12. The estimator 12 outputs the estimated deviation _{ｅ e} ＾ and the estimated angular velocity r 基 づ い based on the deviation ψ _e between the set azimuth ψ _R and the estimated azimuth ψ ＾ and the command steering angle δ _C output from the feedback controller 10. I do. State feedback controller 11 multiplies a proportional gain K _p the estimated deviation [psi _{e ^,} multiplied by the derivative gain K _d to the estimated angular velocity r ^, adds them, and outputs a command steering angle [delta] _C.

2. Control target Here, the control target will be described. The control target is composed of a hull model and a disturbance model, but the disturbance model is omitted in the present embodiment. The hull model uses the response model of the present inventor who improved Nomoto's response model,

に定める。ここでｓはラプラス演算子、Δ_ｃ（ｓ）、δ_ｃは命令舵角を示し、Ｋ_ｒは旋回力ゲイン、Ｔ_ｒは船体時定数、Ｔ_ｒ３は舵感度時定数を示す。この船体モデルは野本の船体モデルと比較して、Ｔ_ｒ３を有するという特徴をもつ。

Set forth in Where s is the Laplace _{operator, Δ c (s), δ} c represents a command steering angle, _{K r} is the turning force gain, _{T r} hull time _{constant, T r3} denotes a steering sensitivity time constant. This hull model has a feature that it has _Tr3 as compared with the hull model of Nomoto.

3 Design Hull motion characteristics are based on maneuverability index or hull motion parameters (hull parameters). Hull _parameter, and _{K r, T r, T r3} , these values obtained from the known identifier which is not shown. Here, for _simplicity, and _{T r »T r3, T r3 ≒} 0. In the case of a small boat, especially a water jet propulsion boat whose dead zone has been corrected as described later, both _Kr and _Tr are small, and the differential gain of the azimuth control loop is small, and the value may be negative. Although the negative differential gain itself does not matter, the negative differential gain causes a problem when calculating the estimation coefficient.

The estimated gain k (k ₁ , k ₂ ) is used in the estimator 12 as shown in FIG. 3 and is a value for adjusting the effect of parameter uncertainty, as disclosed in Patent Document 3, and is a characteristic polynomial described later. Is calculated from When the differential gain is positive, all the coefficients of the characteristic polynomial are positive. On the other hand, when the differential gain is negative, a part of the coefficient becomes negative, and the property of the characteristic polynomial changes. Therefore, it is necessary to cope with the change in this property.

3.1 Differential gain characteristics If the differential gain is _Tr3 = 0,

になる。ここでＫ_Ｐは比例ゲイン、Ｋ_Ｄは微分ゲイン、ζ_ｈは減衰係数、ω_ｈは固有角周波数、添字_ｈは方位制御を示し、設計パラメータは、Ｋ_Ｐ≧１，ζ_ｈ＝１／√２になる。

become. Here _{K P} is a proportional gain, _{K D} is the differential gain, zeta _h is an attenuation coefficient, omega _h the natural angular frequency, subscript _h indicates the orientation control, the design _{parameters, K P ≧ 1, ζ h} = 1 / √ It becomes 2.

From the above equation, the differential gain has a tendency to be proportional to _{ｈ h} , √K _P for the design parameters, and to √ (K _r / T _r ), 1 / K _r for the hull parameters. Therefore, the differential gain increases as the proportional gain increases. Next, the effect of the sign of the differential gain on the characteristic polynomial will be described.

3.2 Characteristics of Characteristic Polynomial The azimuth control loop becomes a seventh-order system when a disturbance component is included, but becomes a fourth-order system when the disturbance component is removed from the control target for simplification. As disclosed in Patent Document 3, the fourth-order azimuth control loop can grasp the basic characteristics of the closed-loop control system, and includes a secondary-system hull motion model and its estimator. The characteristic polynomial of this closed loop is

で与えられる。ここでｓはラプラス演算子、Ｄ（ｓ）は特性多項式、添字_ｅは推定器を意味し、

Given by Where s is a Laplace operator, D (s) is a characteristic polynomial, subscript _e is an estimator,

ここで修飾^〜は船体２２の旋回の度に同定器により同定され更新されるノミナル値を示し、ζ_ｅｈ＝ζ_ｈ，ω_ｅ＝ρ_２ω_ｅｈであり、ρ_２は２次推定係数を示す。Δ_ｂ，Δ_ｂ’，ｄ_ｂはそれぞれパラメータ不確かさのスカラー量を示し、Ｔ_ｒ３＝０の場合はｆ_１＝Ｋ_Ｐ，ｆ_２＝Ｋ_Ｄ，Ｃ_Ｔ３＝１。なお、不確かさパラメータを船体パラメータの旋回力ゲインと時定数との比のパラメータ不確かさ（Δ_ｂ）とする点については、特許文献３を参照されたい。

Here, the modification ^〜 indicates a nominal value that is identified and updated by the identifier every time the hull 22 turns, _eh = ζ _h , ω _e = ρ ₂ ω _eh , and ρ ₂ indicates a secondary estimation coefficient. . _{Δ b, Δ b ', d} b each represent a scalar parameter _{uncertainty, f} 1 ₌ K _P For _{T r3 = 0, f 2 =} K D, C T3 = 1. Note that the point of the uncertainty parameters Parameter uncertainty of the ratio of the turning force gain and the time constant of the hull parameters (delta _b), refer to Patent Document 3.

Check the coefficients of b ₁ and b _{0 in} the equation (2). Assuming that the estimated coefficients satisfy the following conditions, the estimated gains k ₁ and k ₂ are as follows:

になる。これよりｂ_０＜０になり、ｂ_１の符号はｆ_２に依存する。すなわち

become. From this, b ₀ <0, and the sign of b ₁ depends on f ₂ . Ie

になる。上式よりｆ_２≦‐ｆ_１（ｋ_１／ｋ_２）＜０になるから、ｂ_１の符号変化をｆ_２＝０の基準によって判定すればよい。さらにその基準は推定係数を含んだｋ_１，ｋ_２を用いないので、簡単に実現できる。よって、特性多項式は

become. Since f ₂ ≦ −f ₁ (k ₁ / k ₂ ) <0 from the above equation, the sign change of b ₁ may be determined on the basis of f ₂ = 0. Further, since the criterion does not use k ₁ and k _{2 including the} estimation coefficient, it can be easily realized. Therefore, the characteristic polynomial is

と置き換えることができる。上式でΔ_ｂが一定ならば、ｆ_２≦０の特性多項式はｆ_２＞０のそれに比べて係数が負になっているため、不安定根を生じやすい。

Can be replaced by If the above equation with delta _b is constant, the characteristic polynomial of f ₂ ≦ 0 Since coefficients than that of f _2> 0 is in the negative, prone to instability roots.

  The asymptote of the root locus of the above equation changes as shown in FIG. 4, and the sign of the zero point also changes. Ie

になる。ここでｚはゼロ点を示す。よって、（２）式の根軌跡の一例は図５に示すように、不安定根の特性が異なることを確認できる。

become. Here, z indicates a zero point. Therefore, it can be confirmed that the example of the root locus of the equation (2) has different characteristics of the unstable root as shown in FIG.

3.3 Parameter uncertainty specifications Parameter uncertainty specifications

とする。なお、ｆ_１＝μ^−１Ｃ_Ｔ３Ｋ_Ｄ，μ＞０，Ｃ_Ｔ３＞０の関係より、ｆ_２をＫ_Ｄに置き換えても良い。上式からパラメータ不確かさのマージンが半分となるが、船舶用自動操舵装置１は制御対象とする対象船が限定されるので、パラメータ変動は小さいとするためである。

And ^{_{Incidentally, f 1 = μ -1 C T3}} K D, μ> 0, the relationship of _{C T3>} 0, may be replaced by _{f 2} to _{K D.} Although the parameter uncertainty margin is halved from the above formula, the parameter variation is small because the target ship to be controlled is limited in the automatic marine vessel steering device 1.

4. Estimation coefficient calculation processing Next, the estimation coefficient calculation processing will be described. FIG. 6 is a flowchart illustrating the operation of the estimation coefficient calculation process. The estimation coefficient calculation process, during navigation of the ship, in order to calculate the appropriate secondary estimation coefficient [rho ₂ in the estimator, a process of the hull parameter in the feedback controller is executed each time it is updated.

6, first, the state feedback controller 11 sets the first specification gain value set in advance in the proportional gain K _P (S101), calculates the differential gain K _D by (1) ( S102), determines whether the value of the derivative gain _{K D} of a positive (S103).

If the value of the derivative gain _{K D} of a positive (S103, YES), the estimator 12 sets the first specification value which is previously set in the parameter uncertainty _{Δ b '(S104), (} 2) expression of the characteristic calculating a second estimation coefficient [rho ₂ by polynomial (S105).

On the other hand, when the value of the derivative gain _{K D} of a negative (S103, NO), the second specification gain value set to a value greater than the first specification gain value set in the proportional gain _{K P} (S106), the proportional gain _K based on the _P (1) equation by calculating the derivative gain _{K D} (S107), determines whether the value of the derivative gain _{K D} of a positive (S108).

If the value of the derivative gain _{K D} of a positive (S108, YES), the estimator 12 sets the first specification value to the parameter uncertainty delta _b '(S109), 2-order by the characteristic polynomial of the expression (2) It calculates an estimated coefficient ρ ₂ (S105).

On the other hand, when the value of the derivative gain _{K D} of a negative (S103, NO), sets the second design value set to a value smaller than the first design value to the parameter uncertainty _{Δ b '(S110), (} 2 ) the characteristic polynomial of equation to calculate the second order estimation coefficient [rho ₂ (S105).

The calculation of the secondary estimation coefficient ρ ₂ is performed so that the minimum attenuation coefficient _{ｅ e} among the attenuation coefficients obtained from the characteristic roots of the characteristic polynomial of the equation (2) satisfies the specification ζ _Δ. The natural frequency ω _e of the hull model of the estimator 12 is set by ρ ₂ . For details of the method of calculating the estimation coefficient ρ, refer to Patent Document 3.

とする。ここでＴ_ｒ３は初期値。 5 Verification Numerical verification will be performed for a water jet propulsion ship with dead zone correction applied to the control target. A water-jet propulsion ship has a water-jet propulsion device including a pair of nozzles, and these nozzles have a dead zone within a predetermined angle inclined right and left from a straight traveling direction in which the ship goes straight. To correct this dead zone, the operating positions of the pair of nozzles are tilted outward by a nozzle offset, which is a correction angle corresponding to the dead zone, so as to form a C-shape, thereby equivalently canceling out the dead zone. Is to be done. Hull parameters of the water jet propulsion ship

And Here, _Tr3 is an initial value.

Table 1 shows the differential gain of this ship. According to the table, the differential gain becomes negative when the dead zone is corrected. However, _{Tr and the} like remain unchanged before and after the dead zone correction.

FIG. 7 shows the results of calculating b ₁ , b ₀ , z and ω, の of the representative root with respect to the estimation coefficient ρ ₂ using the equation (7), K _P = 1.5, K _D = −4.9. Show. 7 that becomes _b 1> 0, z> 0 in [rho ₂ ≧ 4, zeta, against ω is uncertainty parameter _{d b,} inverse proportion, respectively, in proportion, approximately constant value [rho ₂ ≧ 4 . At this time, we can not meet the specifications _d b = 2. FIG. 8 shows the root locus of the representative root of ρ ₂ = 4. FIG. 8 (b), at _{0 ≦ Δ b ≦ 2K r /} T r, becomes Δ _b = _{2K r} / _T r or _d b = 2 when zeta ≒ 0.2, results and 7 of (d) Almost match. Therefore, the nominal value of the ship

に見積もると、本船の２次推定係数をρ_２≒４にすれば、ζ≒０．４になり仕様を満足することができる。

If the secondary estimation coefficient of the ship is ρ ₂ ≒ 4, it will be ζ ≒ 0.4, which satisfies the specifications.

  The embodiments of the present invention have been presented by way of example and are not intended to limit the scope of the invention. These new embodiments can be implemented in other various forms, and various omissions, replacements, and changes can be made without departing from the spirit of the invention. These embodiments and their modifications are included in the scope and gist of the invention, and are also included in the invention described in the claims and their equivalents.

DESCRIPTION OF SYMBOLS 1 Automatic steering device for ships 10 Feedback controller 11 State feedback controller 12 Estimator

Claims (3)

An automatic steering device for a ship that controls a controlled object including a hull and a steering gear so as to make the heading follow the reference heading,
A feedback controller that outputs a command steering angle based on the reference direction and the heading direction to the control target,
An estimator configured to output an estimated deviation and an estimated angular velocity as an estimated value based on the deviation between the reference azimuth and the heading and the command steering angle; with multiplying the set proportional gain K _P to specifications gain value, and a state feedback controller for outputting the command steering angle is multiplied by a derivative gain K _D that depends on the value of the proportional gain K _P to the estimated angular velocity With
In the closed-loop characteristic polynomial composed of the estimator and the state feedback controller, the suffix _h represents the azimuth control, the suffix _e represents the estimator, the modification ^〜 represents the nominal value, s represents the Laplace operator, and D (s ) Is a characteristic polynomial, ζ is a damping coefficient, ω is a natural angular frequency, _{ｅ eh} = ζ _h , ω _e = ρ ₂ ω _eh , ρ ₂ is a second-order estimation coefficient, K _r is a turning force gain, and _Tr is a hull. _{constant, T r3} steering sensitivity time constant, delta _b, delta _b ', a _{d b} a scalar quantity parameter uncertainty respectively,

Represented by
The state feedback controller, the case where the value of the derivative gain K _D of a negative value, setting the second specification gain value that is larger than said first specification gain value to the proportional gain K _P,
The estimator, when the value of the derivative gain K _D of a positive value, based on the parameters uncertainty which is set to a predetermined first specified value delta _{b ',} the secondary estimation coefficient by the characteristic polynomial of the closed loop calculating a [rho _2, wherein when the value of the derivative gain K _D of a negative value, based on said second specification is set to the gain value proportional gain K _P and the parameter uncertainty which is set to a predetermined value delta _{b '} And calculating the _second estimation coefficient ρ2 using the closed-loop characteristic polynomial. The estimator, when the value of the derivative gain K _D of a negative value, setting the second specification value set smaller than the first specified value to the parameter uncertainty delta _{b ',} the parameter uncertainty delta _{2. The} automatic steering apparatus for a boat according to claim 1, wherein the second estimation coefficient ρ ₂ is calculated by _b ′ based on the characteristic polynomial of the closed loop. 3. The estimator, if the value of the second specification gain value to set the proportional gain K derivative gain calculated based on _P K _D is a negative value, the parameter uncertainty the second specification value delta _b 3. The automatic steering apparatus for a boat according to claim 2, wherein the second estimation coefficient ρ ₂ is calculated by the closed loop characteristic polynomial based on the parameter uncertainty Δ _b ′. 4.

Images