JPH07165180A - Depth control system for submersible navigating body - Google Patents

Abstract

PURPOSE:To facilitate the change of the control algorithm, improve maintenance performance and enable the high speed depth conversion by obtaining th depth rudder fine adjustment rudder angle through the fuzzy estimation, and issuing the final depth steering angle instruction by adding the depth rudder fine adjustment rudder angle to the depth rudder main rudder angle. CONSTITUTION:An attitude rudder control part 9 is equipped with a differentiator 15 as main rudder angle determining means for determining the main rudder angle as a rough steering quantity, subtraction calculator 16, main rudder angle determining part 17, fine adjustment steering speed determining part 18 as fine adjustment rudder angle determining means for obtaining the rudder angle instruction through the addition to the main rudder angle, integrator 19 and an adder 20. The attitude rudder controller 9 receives the attitude angle theta obtained from an attitude angle sensor fdr a submersible traveling body, speed V obtained from the speed sensor. and a target attitude angle thetac and outputs the rudder angle instruction Bea which is given to a steering system for controlling the attitutde angle.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an attitude control system for an underwater vehicle, and more particularly to a depth control system for an underwater vehicle which has a built-in pouring / drainage tank and travels underwater.

[0002]

2. Description of the Related Art Conventionally, a depth control system for an underwater vehicle, such as a submersible, which has a rudder for depth control and a rudder for attitude control, has a depth sensor for detecting the depth of the underwater vehicle and an attitude angle. The steering amount of the rudder for depth control and the rudder for attitude control is PID (proportional + integral + derivative) from the information from the attitude angle sensor that detects the speed and the speed sensor that detects speed, and the information on the target depth and the target attitude angle. Most of them are determined by a control law, and as an example of a more advanced control method, a method in which information of depth deviation is taken into consideration when determining a target attitude angle is disclosed in Japanese Patent Laid-Open No. 62-0.
No. 17809.

[0003]

The conventional depth control system for underwater vehicles described above has the following problems.

(1) When the target posture angle is set, it is determined only by the value of the depth deviation, and the target posture angle cannot be adjusted due to the change in speed. For example, if the speed deviation is the same and the speed is high, it is not necessary to increase the target posture angle (on the contrary, there is a risk that the overshoot will increase).
However, if the speed is low, there is a drawback that the time to reach the target depth becomes long unless the target posture angle is increased to some extent.

(2) Since the basic control law is the PID control law, it is not possible to sufficiently deal with the nonlinearity of the dynamic characteristics of the underwater vehicle (particularly with regard to the attitude angle), and the deviation from the target value in the steady state is considered. There is a problem that it cannot be set to zero. This is because the conventional depth control method for an underwater vehicle is basically designed under the assumption that the weight is 0 and the front and rear trim moment is 0 in water.

As described in (3) and (2), since it is not possible to control to match the target posture angle during depth conversion, it is good during depth conversion when the posture angle value does not matter, but during depth conversion. There is a drawback that it cannot be applied when the attitude angle value needs to exactly match the target value.

(4) The conventional depth and attitude control system for an underwater vehicle determines the proportional gain, integral gain and derivative gain when the underwater vehicle has zero underwater weight and the attitude angle is substantially horizontal. However, when controlling a large attitude angle, the optimum gain is not obtained, and in the worst case, control is impossible.

[0008]

The depth control system for an underwater vehicle of the present invention comprises the following elements.

1. Depth rudder control unit (1) Depth change rudder angle inference unit that fuzzy infers the depth rudder rudder angle corresponding to the depth change of the underwater vehicle, and rudder in the opposite direction to prevent overshoot before reaching the target depth Includes a depth rudder rudder inference unit that cuts off, and a depth hold rudder angle inference unit that fuzzy infers the depth rudder rudder angle required to roughly maintain the depth.
Output as one membership function form, and defuzzification is performed to determine the main rudder angle determination part as the steering amount of the depth rudder, and (2) deviation from the target value when the depth is almost at the target value. The depth rudder fine adjustment steering speed for correcting the above to obtain the required rudder angle command is obtained by fuzzy reasoning by inputting the depth deviation and speed, and integrated at time t to obtain the depth rudder fine adjustment rudder angle, Is added to the main rudder angle to give a depth rudder rudder angle command.

2. Target posture angle inference unit Inputs the depth deviation and speed, determines the target value of the posture angle by fuzzy reasoning, and outputs it to the posture angle control unit described later.

3. Posture rudder control unit The posture angle deviation calculated from the target posture angle output from the target posture angle inference unit and the posture angle value from the posture angle sensor, and the posture angle rate sensor or the posture angle data are obtained by differentiating processing. A posture rudder command is output by inputting the posture angular velocity and speed.

[0012]

The present invention will be described below with reference to the drawings.

FIG. 1 is an overall block diagram of the present invention. First, a rough operation will be described. The target depth H _c set by the operator subtracts the depth data obtained from the depth sensor by the subtractor 4 and outputs the depth deviation ΔH. The depth speed dH / dt obtained by differentiating the depth data or from the depth rate sensor is input to the depth rudder control unit 1 together with the depth deviation ΔH and the speed data obtained from the speed sensor. The depth rudder control unit 1 uses a main rudder angle determination unit 2 that fuzzy infers the depth rudder angle β _{ef (m)} for performing rough depth control, and a rudder angle for correcting the depth deviation near the target depth. A fine adjustment steering speed inference unit 3 which fuzzy infers an increase / decrease amount (fine adjustment steering speed dβ _{ef (s)} / dt), and dβ _{ef (s)}
/ Dt is integrated at time t to obtain a fine adjustment steering angle β _{ef (s)} , a main steering angle β _{ef (m)} and a fine adjustment steering angle β
It is composed of an adder 5 for adding _{ef (s)} to obtain a depth rudder command, and outputs the depth rudder command to the depth rudder steering mechanism.

The target attitude angle reasoning unit 7 determines the speed V and the depth deviation Δ.
The target posture angle θ _c is inferred by fuzzy inference using H as an input. The target posture angle θ _c determined by the target posture angle inference unit 7 is subtracted from the current posture angle θ obtained from the posture angle sensor by the subtractor 8 to obtain a posture angle deviation Δθ.

The attitude angle deviation Δθ is input to the attitude rudder control unit 9 together with the speed V and the attitude angular velocity dθ / dt obtained by the differential processing of the attitude angle sensor or the attitude angle rate sensor.
The attitude rudder control unit 3 deduces an attitude rudder steering angle command so that the attitude angle of the underwater vehicle matches the target attitude angle based on the input data, and outputs it to the attitude rudder steering mechanism.

Next, each functional block will be described individually.

1. Main rudder angle inference unit 2 A functional block diagram of the main rudder angle inference unit 2 is shown in FIG. The main rudder angle inference unit 2 includes three types of inference units. The depth change rudder angle inference unit 10 fuzzy infers the rough cutting condition of the depth rudder according to the speed when the depth deviation is medium to large. An example of general rules for inference is shown below.

Rule 1: If ΔH is positively (negatively) large,
Fill the bottom rudder (upper rudder) with the depth rudder.

Rule 2: If ΔH is positive (negative) is medium and V is large, the depth rudder should be small and the rudder should be down (upward).

Rule 3: If ΔH is positive (negative) is medium and V is medium, the depth rudder is medium and down rudder (upper rudder).

Rule 4: If ΔH is positive (negative) is medium and V is small, the depth rudder should be largely down (up).

Hereinafter, for simplicity, the rule expression will be as follows.

Rule 1: IFΔH = PL (NL), TH
EN β _{1F1 (m)} = NL (PL) Rule 2: IFΔH = PM (NM) and V = P
L, THEN β _{1F1 (m)} = NS (PS) Rule 3: IFΔH = PM (NM) and V = P
M, THEN β _{1F1 (m)} = NM (PM) Rule 4: IFΔH = PM (NM) and V = P
S, THEN β _{1H (m)} = NL (PL) When the depth deviation is small to medium, the rudder rudder angle inference unit 11 roughly determines the depth rudder for overshoot suppression depending on the depth speed. It infers the degree of cutting and outputs β _{1F2 (m)} in the form of MF.

An example of the rudder rudder angle inference rule is shown below.

Example of Rule for Inferring Rudder Rudder Angle Rule 1: IFΔH = PM (NM) and dH / d
t = PL (NL), THEN β _{1F2 (m)} = PS (NS) Rule 2: IFΔH = PM (NM) and dH / d
t = PM (NM), THEN β _{1F2 (m)} = ZR Rule 3: IFΔH = PS (NS) and dH / d
t = PM (NM), THEN β _{1F2 (m)} = PS (NS) Rule 4: IFΔH = PS (NS) and dH / d
t = PS (NS), THEN β _{1F2 (m)} = ZR When the depth deviation is small, the depth holding rudder inference unit 12 roughly cuts the depth rudder to hold the depth according to the depth speed. The condition is inferred, and β _{1F3 (m)} is output as the MF form. An example of the rules for inferring the depth retention steering angle is shown below.

Depth-keeping steering angle inference rule example Rule 1: IFΔH = PS (NS) and dH / d
t = ZR, THENβ _{1F3 (m)} = NS (PS) Rule 2: IFΔH = ZR and dH / dt = Z
R, THEN β _{1F3 (m)} = ZR Conclusion The integration unit 13 outputs β of the three inference units 10 to 12.
_{1F1 (m)} , β _{1F2 (m)} , β _{1F 3 (m)} into one MF form β _{1F (m)}
The processing is summarized as. Conclusion Output β of the integration unit 13
_{Since 1F (m)} is in the form of MF and cannot be used for control as it is, after performing the center of gravity calculation processing by the defuzzification unit 14, the rough rudder angle (main rudder angle) β of the depth rudder β
Output _{ef (m)} .

The fine adjustment steering speed inference unit 3 is provided for zeroing the steady depth deviation near the target depth, which cannot be controlled by the main steering angle. The depth deviation ΔH and the speed V are input to the target depth. The amount of increase or decrease of the rudder angle of the depth rudder for correcting the deviation from, that is, the steering speed is inferred. An example of inference rules is shown below.

Example of Rule for Fine Tuning Steering Speed Inference Rule 1: IFΔH = PL (NL), THEN dβ
_{ef (s)} / dt = ZR Rule 2: IFΔH = PM (NM), THEN dβ
_{ef (s)} / dt = ZR Rule 3: IFΔH = PS (NS) and V = P
L, THEN dβ _{ef (s )} / dt = NS (PS) Rule 4: IFΔH = PS (NS) and V = P
M, THEN dβ _{ef (s )} / dt = NM (PM) Rule 5: IFΔH = PS (NS) and V = P
S, THEN dβ _{ef (s )} / dt = NL (PL) In the fine adjustment steering speed inference unit, after the defuzzification processing is performed as in the main steering angle inference unit, the fine adjustment steering speed dβ
Output _{ef (s)} / dt. This dβ _{ef (s)} / dt becomes the fine adjustment steering angle β _{ef (s)} after time integration by the integrator 4.
To obtain the depth rudder command, the adder 5 adds the main rudder angle β _{ef (m)} .

The target posture angle inference unit 7 determines the target value of the posture angle by fuzzy inference using the depth deviation ΔH and the speed V as inputs. An example of the target posture angle inference rule is shown below.

Example of Target Posture Angle Inference Rule Rule 1: IFΔH = PL (NL), THEN θ _c =
NL (PL) Rule 2: IFΔH = PM (NM) and V = P
L, THEN θ _c = NS (PS) Rule 3: IFΔH = PM (NM) and V = P
M, THEN θ _c = NM (PM) Rule 4: IFΔH = PM (NM) and V = P
S, THEN θ _c = NL (PL) Rule 5: IFΔH = PS (NS), THEN θ _c =
ZR Rule 6: IFΔH = ZR, θ _c which is inferred THEN theta _c = ZR target posture angle estimating unit 7 is input to the subtracter 8 to obtain a Δθ required in the attitude steering control unit.

Next, the attitude rudder control section 9 of FIG. 1 will be described.

FIG. 3 is a block diagram of the attitude rudder control section 9, and generally, a differentiator 15, a subtractor 16 and a main rudder angle determination as main rudder angle determining means for determining the main rudder angle as the steering amount. And a fine adjustment steering speed determining unit 1 as a fine adjustment steering angle determining unit that obtains a fine adjustment steering angle for finely adjusting the steering amount and adds it to the main steering angle to obtain a steering angle command.
8, an integrator 19 and an adder 20.

The steering angle given to the steering system for controlling the attitude angle is input with the attitude angle θ acquired from the attitude angle sensor of the underwater vehicle, the speed V acquired from the speed sensor, and the target attitude angle θ _c as inputs. Outputs the command β _ea .

The posture angle θ is input to the differentiator 15 to obtain the posture angular velocity dθ / dt, and the target posture angle θ
in combination with _c used to calculate the attitude angle deviation Δθ (= θ _c -θ).

Posture angular velocity dθ obtained by the differentiator 15
/ Dt is input to the main steering angle determination unit 17 in order to determine the main steering angle β _{ea (m)} as the steering angle together with the attitude angle deviation Δθ and the speed V. Also, the attitude angle deviation Δ
θ and the speed V are the fine adjustment steering speed dβ _{ea (s )} required to obtain the fine adjustment angle for performing the fine adjustment near the target posture angle.
It is input to the fine adjustment steering speed determination unit 18 to determine / dt.

The fine adjustment steering speed dβ _{ea (s)} / dt output from the fine adjustment steering speed determination unit 18 is input to the integrator 19,
The fine adjustment steering angle β _{ea (s)} is output by the integration processing.

The fine rudder angle β _{ea (s)} is determined by the main rudder angle determination unit 1
7 is added to the main steering angle β _{ea (m)} output to obtain the final steering angle command β _ea (= β _{ea (m)} + β _{ea (s)} ).

Next, the operations of the main rudder angle determination unit 17 and the fine adjustment steering speed determination unit 18 will be described. The main rudder angle determination unit 17 is a part that determines a main rudder angle for roughly approaching the target posture angle when the amount of change in the posture angle is large, and corresponds to, for example, an operation in which a person roughly performs steering. The main steering angle determination unit 17 determines the main steering angle β _{ea (m)} by fuzzy inference using the attitude angle deviation Δθ, the attitude angular velocity dθ / dt and the speed data V as input values.

As shown in FIG. 4, the main rudder angle determining unit 17 approaches the target attitude angle to some extent, and the attitude angle changing rudder angle inferring unit 21 which fuzzy infers the rudder angle corresponding to the attitude angle change. A steering rudder angle inference unit 2 for fuzzy inference of a steering angle corresponding to a situation where the steering is turned in the opposite direction to prevent overshooting against the target posture angle.
2 and three types of rudder angle inference units for inferring the steering amount corresponding to the scene in which the target attitude angle is held, and the inference result of the steering amount corresponding to each scene is obtained. It is output in the form of a membership function (hereinafter referred to as MF). These outputted conclusions MF are the conclusion integration unit 24.
Are output to one conclusion MF. The conclusion MF put together into one is the MF that represents the final steering angle inference result in the main steering angle determination unit 17, and cannot be used for control as it is. Therefore, it is input to the defuzzification unit 25 and the center of gravity is calculated. After that, it is converted to the main steering angle β _{ea (m)} and output.

The fine adjustment steering speed determining unit 18 determines the fine adjustment steering speed for obtaining the fine adjustment angle for correcting the deviation from the target value when the attitude angle is almost the target value. As shown in FIG. 5, using the attitude angle deviation Δθ and the speed V as input, the fine adjustment steering speed dβ _{ea (s)} / dt is fuzzy inferred by the fine adjustment steering speed inference unit 26 as in the main steering angle determination unit 17. Then, after the conclusions are integrated by the conclusion integrating unit 27, the center of gravity is calculated by the defuzzifying unit 28 to obtain the result. However, the rules for reasoning are different.

Next, the rules of each inference unit will be described. An example of rules of the main rudder angle determination unit 17 is shown in FIG.
It shows in (D). Further, FIG. 7 shows a rule example of the fine adjustment steering speed inference unit 18.

FIG. 6A shows an example of a posture angle changing rudder angle inference rule, where the number of rules N1 is 11. See also FIG.
(B) shows an example of a rudder rudder angle inference rule, and the rule number N
2 is 4. Further, FIG. 6C shows an example of the attitude angle holding steering angle inference rule, and the rule number N3 is 5.
Further, (D) of FIG. 6 shows the meanings of the labels used in (A) to (C) of FIG.

Now, the inference rules will be described by taking, for example, FIG. 1-No. 4 indicates the following contents, respectively.

No. 1: IFΔθ = PM and dθ
/ Dt = PL, THEN β _{ea (m)} = NS (If the attitude angle deviation is positively medium and the attitude angular velocity is positively large, make the main steering angle small and negative.) Or No. 2: IFΔθ = NM and dθ / dt = N
L, THEN β _{ea (m)} = PS (If the attitude angle deviation is negative and medium and the attitude angular velocity is negative, set the main steering angle small and positive.) Or No. 3: IFΔθ = PM and dθ / dt = P
M, THEN β _{ea (m)} = ZR (If the attitude angle deviation is positive and medium and the attitude angular velocity is positive and medium, set the main steering angle to almost 0.) or No. 4: IFΔθ = NM and dθ / dt = N
M, THEN β _{ea (m)} = ZR (If the attitude angle deviation is negative and medium and the attitude angular velocity is negative and medium, set the main steering angle to almost 0.) Regarding the rules of other inference units including FIG. Since it is the same, detailed description thereof will be omitted.

Regarding the attitude rudder control unit, Japanese Patent Application No. 4-
Filed under No. 234436.

[0046]

As described above, the present invention has the following effects. (1) Since fuzzy inference is used for the control algorithm, it is easy to change the control algorithm and maintainability is improved. (2) Since the posture angle can be specified during depth conversion, high-speed depth conversion is possible. (3) Depth conversion can be performed by setting the steady-state deviation at a large attitude angle to 0, which was not possible in the past. (4) It is possible to set the target posture angle according to the depth deviation and the speed in a form that incorporates the experience and know-how of the operator.

[Brief description of drawings]

FIG. 1 is an overall block diagram of an embodiment of the present invention.

FIG. 2 is a block diagram of a main steering angle inference unit in the embodiment.

FIG. 3 is a block diagram of a posture rudder control unit in the embodiment.

FIG. 4 is a block diagram of a main steering angle determination unit in FIG.

5 is a block diagram of a fine adjustment steering speed determination unit in FIG.

FIG. 6 is a diagram showing an inference rule example of an attitude angle change rudder angle inference unit, a contact rudder rudder angle inference unit and an attitude angle holding rudder angle inference unit in FIG. 4;

7 is a diagram showing an inference rule example of a fine adjustment steering speed determination unit in FIG.

[Explanation of symbols]

1 ... Depth rudder control unit, 2 ... Main rudder angle inference unit, 3 ... Fine adjustment steering speed inference unit, 4 ... Integrator, 5 ... Adder, 6 ... Subtractor, 7 ... Target attitude angle inference unit, 8 ... Subtractor Container, 9 ... Posture rudder control unit, 10 ... Depth change rudder angle inference unit, 11 ... Depth rudder rudder angle inference unit, 12 ... Depth retention rudder angle inference unit, 13 ... Conclusion integration unit, 14 ... Defuzzying unit, 15 ... Differentiator, 16 ... Subtractor, 17 ... Main rudder angle determination unit, 18 ... Fine adjustment steering speed determination unit, 19 ... Integrator, 20 ... Adder, 21 ... Attitude angle change rudder angle inference unit, 22 ... Rudder rudder angle inference unit, 23 ... Posture angle holding rudder angle inference unit, 24 ... Conclusion integration unit, 25 ... Non-fuzzy unit, 26 ... Fine adjustment steering speed inference unit, 27 ... Conclusion integration unit, 28 ... Non-fuzzy Akabe.

Claims (5)

[Claims] 1. A depth change rudder angle inference for fuzzy inference of a rudder angle corresponding to a depth change of an underwater vehicle, and a depth rudder for turning in the opposite direction to prevent overshoot before reaching a target depth. The three types of steering amount inference results, depth rudder rudder angle inference for fuzzy inference of rudder and depth hold rudder angle inference for fuzzy inference of depth rudder angle to perform approximate depth hold Output in the format
Depth rudder main rudder angle determining means for integrating these inference results into one membership function and then defuzzifying to determine the main rudder angle as a steering amount of the rudder for depth control When the depth is almost at the target value, the depth rudder fine adjustment rudder angle for correcting the deviation from the target value to obtain the required rudder angle command is obtained by the fuzzy inference using the depth deviation and the speed as input. Depth rudder control means including fine adjustment rudder angle determination means for adding to the depth rudder main rudder angle to obtain a final depth rudder rudder angle command. Depth control method. 2. The depth change rudder angle inference is inferred by inputting a depth deviation of a difference between a target depth as a target of an underwater vehicle and a depth acquired by a depth sensor, and a speed acquired by a speed sensor, The depth rudder rudder angle inference and the depth hold rudder angle inference are obtained from the depth deviation of the underwater vehicle and the data from the depth sensor as a differential process, or from a depth rate sensor that outputs a velocity in the depth direction. The depth control system for an underwater vehicle according to claim 1, wherein fuzzy inference is performed using the obtained depth velocity as an input. 3. A target posture angle determining means for determining a posture angle target value by fuzzy inference using the depth deviation and the speed as inputs and outputting it to a posture angle control means described later. Depth control method for underwater vehicles. 4. The attitude angle deviation, which is the difference between the attitude angle target value and the attitude angle obtained from the attitude angle sensor, and data obtained from the attitude angle sensor are obtained by differentiating, or the rate of change of the attitude angle with time. The attitude rudder control means for fuzzy inferring a rudder angle command to the attitude control rudder by inputting the attitude angular velocity obtained from the attitude angle rate sensor for detecting the speed and the speed acquired from the speed sensor. Depth control method for mid-water vehicles. 5. The depth rudder control means controls the depth rudder,
The depth control of the underwater vehicle according to claim 4, wherein the posture angle control means for controlling the posture angle of the target value output by the target posture angle determination means controls the optimum posture angle when the depth is changed. method.