A UUV Unsteady Process Control Method Considering Control Input Deadband and Saturation Characteristics

Abstract

The present invention relates to a UUV unstable process control method considering the control input dead zone and saturation characteristics. The variable structure control parameter is obtained after the deep control target value and the output of the observer, and the variable structure control u is output. After the dead zone compensator, the dead zone-free control instruction τ is output. Compared with the prior art, the advantages are as follows: 1. The variable structure control algorithm is used to replace the PID algorithm in the traditional self-disturbance rejection control theory to effectively solve the multi-state control switching problem; 2. The variable structure control parameters are dynamically adjusted by the self-correction algorithm based on the minimum mean square error, so as to ensure the control performance of the UUV at different speeds; 3. A nonlinear adaptive compensator is designed to change the static compensation of the original method, realize the adaptive dynamic compensation controller dead zone, and ensure the attitude stability control under non-ideal steering conditions.

Description

Technical Field

[0001] The present invention belongs to a control method of a UUV, and relates to an unstable process control method of a UUV taking into account a control input dead zone and a saturation characteristic, and more specifically, to an unstable process control method of a UUV that switches from a hovering state to a high-speed motion in a short time during a startup process. Background Art

[0002] As a safe and efficient underwater operation platform, UUV has received extensive attention and application, such as underwater salvage, seabed environmental monitoring, etc. When working, UUV often encounters the situation of switching from hovering state to high-speed motion in a short time. This process involves the rapid conversion of motion state, which is an unstable process and puts high demands on the robustness of the controller.

[0003] The existing methods for the UUV unstable process motion control problem are mainly: 1. PID control; 2. Sliding mode control; 3. Reinforcement learning control. PID control is a linear control method and cannot be applied to nonlinear control problems. Although sliding mode control can switch the sliding surface to achieve dynamic control, jitter will inevitably occur when the sliding surface switches, threatening the stability of the controller. Reinforcement learning control does not require a precise model. It gains learning experience through the interaction between the UUV and the environment, and improves the robustness of the controller through continuous trial and error. However, learning takes a long time and has high requirements on hardware computing power, which is not suitable for practical applications.

[0004] ADRC estimates and compensates for model uncertainty and external disturbances, and has excellent robustness. The control method of the present invention is based on the ADRC theory, and adopts a variable structure control algorithm to replace the PID algorithm in the traditional ADRC theory, effectively solving the multi-state control switching problem, and dynamically adjusts the variable structure control parameters using a self-correction algorithm based on the minimum mean square error, thereby ensuring control performance at different speeds. In addition, considering the dead zone and saturation characteristics of the servo, a nonlinear adaptive compensator is designed to ensure attitude stability control under non-ideal servo conditions. Summary of the invention

[0005] Technical issues to be solved

[0006] In order to avoid the shortcomings of the prior art, the present invention proposes a UUV unstable process control method taking into account the control input dead zone and saturation characteristics, aiming to solve the unstable process control problem of UUV switching from a hovering state to a high-speed motion state in a short time, realize stable control of the unstable process, and improve the robustness of the control system.

[0007] Technical Solution

[0008] A UUV unstable process control method considering control input dead zone and saturation characteristics is characterized by the following steps:

[0009] Step 1: Consider the control input with dead zone nonlinearity and saturation characteristics and establish the state space model of the UUV:

[0010]

[0011] Where x is the UUV system state, x∈R n , A is the system coefficient matrix, A∈R n×n ,b is the control matrix, b∈R n ,u∈R is the control input with dead zone nonlinearity;

[0012] Step 2: Set the control target value Z d Perform nonlinear tracking differential calculation to obtain the tracking signal r1 of the control target and its differential signal r2;

[0013] Step 3: Input r1, r2, z1, z2 and z3 into the variable structure controller based on the minimum mean square error algorithm and calculate the control signal to obtain u:

[0014] Step 4: The dead-zone-free control command τ is the final control input:

[0015]

[0016] in, is the estimated value of the dead zone nonlinear parameter, and its adaptive update law is:

[0017]

[0018] P is a positive definite matrix, which is the solution of the following linear matrix inequality:

[0019] x T PAx+x T A T Px+x T Pbu+(bu) T P<0.

[0020] The control target value Z d Perform nonlinear tracking differential calculations:

[0021]

[0022] Where k is the tuning parameter of the nonlinear tracking differentiator.

[0023] The control signal u:

[0024] u=c1[r2-z2-p1(Z-z1)]+c2[k2 (Z-r1)-2kr2-z3-p2(Z-z1)+K as (uu s )]+ρe(Z-z1) / c2b

[0025] Where: z1, z2 and z3 are the outputs of the anti-saturation nonlinear extended state observer, z1 is the estimated value of the UUV output state variable, z2 is the differential estimated value of the output state variable and z3 is the estimated value of the interference term:

[0026]

[0027] Among them, the depth value Z is the input value of the anti-saturation nonlinear extended state observer, that is, the depth value of UUV; p1, p2 and p3 are the parameters of the nonlinear extended state observer, K as >0 is the anti-saturation compensation coefficient.

[0028] The adjustment parameter k of the nonlinear tracking differentiator is 0.1.

[0029] Beneficial Effects

[0030] The present invention proposes a UUV unstable process control method that considers the control input dead zone and saturation characteristics. The variable structure control parameter is obtained after the deep control target value and the output of the observer are interacted, and the variable structure control u is output. After the dead zone compensator, the dead zone-free control instruction τ is output. Compared with the prior art, the above technical solution conceived by the present invention mainly has the following technical advantages:

[0031] 1. Use variable structure control algorithm to replace the PID algorithm in traditional active disturbance rejection control theory to effectively solve the problem of multi-state control switching;

[0032] 2. Dynamically adjust the variable structure control parameters using a self-correction algorithm based on minimum mean square error to ensure the control performance of the UUV at different speeds;

[0033] 3. A nonlinear adaptive compensator is designed, which changes the static compensation of the original method and realizes the adaptive dynamic compensation controller dead zone to ensure attitude stability control under non-ideal servo conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 Block diagram of variable structure control for automatic disturbance rejection;

[0035] Figure 2 Pitch angle response curve;

[0036] Figure 3 Depth response curve;

[0037] Figure 4 Heel steering response;

[0038] Figure 5 Longitudinal velocity profile;

[0039] Figure 6 UUV axial speed. DETAILED DESCRIPTION

[0040] The present invention will now be further described with reference to the embodiments and the accompanying drawings:

[0041] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not limited to the present invention.

[0042] Step 1: Establish a UUV mathematical model, and its state space form is expressed as follows:

[0043]

[0044] Where x is the UUV system state, x∈R n , A is the system coefficient matrix, A∈R n×n ,b is the control matrix, b∈R n ,u∈R is the control input with dead zone nonlinearity, expressed as follows:

[0045]

[0046] Where δ∈R and δ>0, δ is the dead zone nonlinear parameter, and equation (11) estimates the parameter. τ is the control command when there is no dead zone, τ∈R. Considering the control input saturation problem, the control input with saturation characteristics is defined as

[0047]

[0048] Among them, u s is the control input with saturation characteristics, u is the control input with dead zone nonlinearity, u - and u + are the upper and lower bounds of the control input respectively.

[0049] Step 2: Set the depth control target value Z d =-50m is input into the nonlinear tracking differentiator to obtain the tracking signal r1 and its differential signal r2 of the control target. The algorithm of the nonlinear tracking differentiator is:

[0050]

[0051] Where k is the adjustment parameter of the nonlinear tracking differentiator, k = 0.1

[0052] Step 3: Design the variable structure control switching function as:

[0053] s=c1e1+c2e2 (5)

[0054] Among them, c1 and c2 are control coefficients, e1 = r1-z1, e2 = r2-z2, and the differential of s is

[0055]

[0056] The self-correction algorithm based on the minimum mean square error is

[0057]

[0058] Among them, e=r2-z2-p1(Z-z1), ρ is the liquid density.

[0059] Input r1, r2, z1, z2 and z3 into the variable structure controller based on the minimum mean square error algorithm and calculate the control signal u as

[0060] u=c1[r2-z2-p1(Z-z1)]+c2[k 2 (Z-r1)-2kr2-z3-p2(Z-z1)+K as (uu s )]+ρe(Z-z1) / c2b(

[0061] 8) Where z1, z2 and z3 are the outputs of the anti-saturation nonlinear extended state observer, which will be introduced in step 5.

[0062] Step 4: Design the dead-zone-free control command τ as follows, which is the final effective control law

[0063]

[0064] in, is the estimated value of the dead zone nonlinear parameter, and its adaptive update law is

[0065]

[0066] Where P is a positive definite matrix and is the solution to the following linear matrix inequality.

[0067] x T PAx+x T A T Px+x T Pbu+(bu) T P<0 (11)

[0068] Step 5: Input the depth value Z into the anti-saturation nonlinear extended state observer to obtain the estimated value z1 of the UUV depth, the estimated value z2 of the vertical velocity, and the estimated value z3 of the interference term. The algorithm of the anti-saturation nonlinear extended state observer is:

[0069]

[0070] Among them, p1, p2 and p3 are the parameters of the nonlinear extended state observer, K as >0 is the anti-saturation compensation coefficient.

[0071] Simulation Analysis

[0072] like Figure 2-Figure 6 As shown in the figure, the UUV is set to start the depth control for the second time, that is, in the vertical direction, it will enter the depth control after 80 seconds of buoyancy. It can be seen from the simulation results that the UUV sails at the specified depth of 50 meters, and the speed of the UUV reaches about 40 knots.

Claims

1. A UUV unstable process control method considering control input dead zone and saturation characteristics, characterized in that Here are the steps: Step 1: Consider the control input with dead zone nonlinearity and saturation characteristics and establish the state space model of the UUV: Where x is the UUV system state, x∈R n , A is the system coefficient matrix, A∈R n×n ,b is the control matrix, b∈R n ,u∈R is the control input with dead zone nonlinearity; Step 2: Control target value Z d Perform nonlinear tracking differential calculation to obtain the tracking signal r1 of the control target and its differential signal r2; Step 3: Input r1, r2, z1, z2 and z3 into the variable structure controller based on the minimum mean square error algorithm and calculate the control signal to obtain u: Step 4: The dead-zone-free control command τ is the final control input: in, is the estimated value of the dead zone nonlinear parameter, and its adaptive update law is: P is a positive definite matrix, which is the solution of the following linear matrix inequality: x T PAx+x T A T Px+x T Pbu+(bu) T P<0。 2. The UUV unstable process control method considering control input dead zone and saturation characteristics according to claim 1 is characterized by: The control target value Z d Perform nonlinear tracking differential calculations: Where k is the tuning parameter of the nonlinear tracking differentiator.

3. The UUV unstable process control method considering control input dead zone and saturation characteristics according to claim 1, characterized in that: The control signal u: u=c1[r2-z2-p1(Z-z1)]+c2[k 2 (Z-r1)-2kr2-z3-p2(Z-z1)+K as (u-u s )]+ρe(Z-z1) / c2b Where: z1, z2 and z3 are the outputs of the anti-saturation nonlinear extended state observer, z1 is the estimated value of the UUV output state variable, z2 is the differential estimated value of the output state variable and z3 is the estimated value of the interference term: Among them, the depth value Z is the input value of the anti-saturation nonlinear extended state observer, that is, the depth value of UUV; p1, p2 and p3 are the parameters of the nonlinear extended state observer, K as >0 is the anti-saturation compensation coefficient.

4. The UUV unstable process control method considering control input dead zone and saturation characteristics according to claim 2 is characterized by: The adjustment parameter k of the nonlinear tracking differentiator is 0.1.

Images