Improved obstacle function agricultural robot safety tracking control method

Abstract

The invention relates to the technical field of tracking control, and discloses an improved obstacle function agricultural robot safety tracking control method, which comprises the following steps: designing a trajectory tracking controller based on an integral obstacle function with adjustment parameters; adjusting the control precision and the convergence speed of the robot system to a set interval by setting values of different adjustment parameters; then the trajectory tracking controller is operated to enable the position of the robot to be always kept within the constraint interval; the invention provides an integral barrier function with adjusting parameters, and in the design process of the controller, the barrier function does not need to derive an expected target value, so that the structure of the controller is simplified. The control precision and the convergence speed of the robot system can be changed by giving different values to the adjustment parameters in the integral obstacle function.

Description

Technical Field

[0001] The present invention relates to the field of tracking control technology, and more specifically, to a safe tracking control method for an agricultural robot with an improved obstacle function. Background Art

[0002] With the continuous innovation of industrial production, robots are playing an increasingly important role in the industrial production process and undertake the main tasks of automated production. The use of robots makes production more intelligent, improves production efficiency and creates higher value. Robots play a variety of roles in automated production, such as transporting goods, assembly and processing, product testing, and in medical services for patient rehabilitation training. In the future, robots will replace manual labor and continue to promote the development of industrial production, and will definitely become an important pillar of industrial production.

[0003] The control system is the brain of the robot, and the control technology used determines the function and performance of the robot. The robot control task is mainly to complete the control of position, posture, motion trajectory, etc. in the workspace. Nowadays, intelligent control technology has been widely used in robot motion control and has achieved many meaningful results. For example, model predictive control can effectively solve the performance requirements of robot systems [1-3]. Sliding mode control has simple design and strong robustness, and is widely used to solve the uncertainty and disturbance problems of robot systems [4-6]. A series of uncertain robot state and output feedback adaptive control are proposed in references [7-9]. Backstepping control technology that can be used for linear and nonlinear systems is a recursive design controller for robots [10-11]. In references [12-14], some excellent robot motion control strategies with performance guarantees are designed using preset performance control methods. Reference

[15] uses H ∞ Control theory has designed a neural network trajectory tracking controller for the robot system. Reference

[16] designed a biquadratic functional optimal control method for the robot system in order to achieve a smaller control error under a smaller control input. Reference

[17] designed a robot self-disturbance rejection control method that does not rely on an accurate model in order to enhance the system's anti-interference ability. However, the above-mentioned references rarely consider the robot system state and error constraints.

[0004] As we all know, in the actual workspace, due to physical conditions, actuator capabilities, and the position of the end of the robot, the state of the robot cannot be arbitrary. Therefore, according to the task requirements in a specific work scenario, the state constraint problem must be considered when designing the controller. Among the existing constraint control technologies, the barrier Lyapunov function (BLF) is the most widely used technology [18-19]. In response to the time-invariant output constraint requirements, references [20-21] combined neural networks with time-invariant logarithmic BLF to design tracking control for the robot system, solving the system output constraint and system uncertainty problems. Reference

[22] used the time-invariant logarithmic BLF to deal with the full state constraint problem of the robot system. However, the time-invariant constraint is only a special case of the time-varying constraint. In actual engineering, the constraint boundary conditions are often time-varying. Therefore, reference

[23] used the time-varying logarithmic BLF to deal with the state constraint problem of the robot system. In addition, in addition to the logarithmic BLF, the tangent function BLF is also used to solve the state constraint requirements of the robot system. Moreover. The tangent BLF has certain advantages over the logarithmic BLF. Its advantages are mainly reflected in that it can be directly used to design controllers for systems with unconstrained requirements. With the deepening of research, scholars have proposed an integral BLF containing a desired target in order to directly constrain the system state. Reference

[24] uses an integral BLF to directly constrain the state of a multi-link robot system. Although the existing integral BLF has the ability to directly constrain the system state, the obstacle function contains three variables: constrained error, constraint boundary and desired target. If the constraint boundary and the desired target value are both functions of time, then when designing the controller, it is necessary to calculate the partial derivatives of these three variables, which is very computationally intensive. In addition, when designing a controller using the existing integral obstacle function, there are too few parameters that can be used to adjust the system performance [25-28]. After the constraint boundary is selected, it is difficult to adjust the system convergence speed and control accuracy only by the gain of the constraint error. Summary of the invention

[0005] The present invention provides an improved obstacle function agricultural robot safety tracking control method. The present invention designs an improved integral obstacle function. This obstacle function directly constrains the tracking error, eliminates the process of deriving the target value, and simplifies the structure of the controller. However, the present invention adds adjustment parameters to the integral obstacle function. The addition of the adjustment parameters can effectively improve the convergence speed and control accuracy of the robot tracking system when the constraint error gain is fixed. In addition, based on the existing integral obstacle function, we enrich the relevant theories about the integral obstacle function, and theoretically give the proposed obstacle function to design a controller for a system with no constraints, without changing the structure of the designed controller, and solve the technical problems in the related technology.

[0006] The present invention provides a safe tracking control method for an agricultural robot with an improved obstacle function, comprising:

[0007] The following trajectory tracking controller is designed based on the integral barrier function with tuning parameters:

[0008]

[0009] Where K2 = diag(k 21 ,k 22 ,···,k 2n ) is a positive definite parameter matrix;

[0010]

[0011] where p i represents the adjustment parameter, i∈[1,n];

[0012] The controlled object is an n-joint robot system, which is described as:

[0013]

[0014] Among them, M kn =C0(x1,x2)x2+G0(x1);

[0015] By setting different tuning parameters p i The value of adjusts the control accuracy and convergence speed of the robot system to the set range;

[0016] Then run the trajectory tracking controller to ensure that the robot position always remains within the constraint range.

[0017] Furthermore, the robot position always remains within the constraint range means:

[0018] The robot’s position vector x1=q=[q1,q2,···,q n ] TSuccessfully tracked the target trajectory x d =q d =[x d1 ,x d2 ,···,x dn ] T .

[0019] Furthermore, the error signal of the robot tracking system converges exponentially to zero, and the position and its tracking error are always within the constraint boundary, giving the adjustment parameter p i Different values ​​can appropriately adjust the size of the position error.

[0020] Furthermore, adjusting the parameter p i The relationship with position error is as follows:

[0021]

[0022] Furthermore, adjusting the parameter p i Proportional to the convergence speed.

[0023] Furthermore, the integral barrier function with adjustment parameters is as follows:

[0024]

[0025] Where p is the adjustment parameter of the integral barrier function and is greater than zero; e is the constrained system tracking error, and its constraint boundary is represented by k(t); is the integral variable; when the tracking error is in the open interval |e|<k(t), the barrier function V is continuous, differentiable, and positive definite. When e=0, V=0, and when |e|→k(t), V is radially unbounded.

[0026] Furthermore, the trajectory tracking controller can handle systems with constrained requirements and can stabilize systems without constrained requirements.

[0027] Furthermore, the trajectory tracking controller is obtained by combining an integral barrier function with an adjustment parameter with an observer.

[0028] In at least one embodiment of the present invention, an improved obstacle function agricultural robot safety tracking control system is provided, which is used to execute the aforementioned improved obstacle function agricultural robot safety tracking control method.

[0029] At least one embodiment of the present invention provides a robot trajectory tracking controller with an improved obstacle function, which is designed based on an integral obstacle function with adjustment parameters. The trajectory tracking controller is expressed as follows:

[0030]

[0031] Where K2 = diag(k 21 ,k 22 ,···,k 2n ) is a positive definite parameter matrix;

[0032]

[0033] where p i represents the adjustment parameter, i∈[1,n];

[0034] The controlled object is an n-joint robot system, which is described as:

[0035]

[0036] Among them, M kn =C0(x1,x2)x2+G0(x1).

[0037] By setting different tuning parameters p i The value of adjusts the control accuracy and convergence speed of the robot system to the set range.

[0038] The beneficial effects of the present invention are:

[0039] The present invention proposes for the first time an integral barrier function with an adjustment parameter. In the controller design process, this barrier function does not need to be derived from the desired target value, which simplifies the controller structure. By assigning different values ​​of the adjustment parameter in the integral barrier function, the control accuracy and convergence speed of the robot system can be changed.

[0040] 2) In Theorem 1, we give an inequality condition for the proposed integral barrier function, which can be used to easily prove the stability of the system. In addition, by using the squeeze theorem, we show that the proposed integral barrier function can not only handle systems with constrained requirements, but also directly design controllers for systems without constrained requirements.

[0041] 3) The tracking controller designed by combining the observer with the proposed obstacle function can make the system signal exponentially tend to the equilibrium point while ensuring that the robot position constraints are always met. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 is a schematic diagram of the tracking effect of joint 1 in the simulation result of the present invention;

[0043] Figure 2 is a schematic diagram of the tracking error of joint 1 in the simulation result of the present invention;

[0044] Figure 3 is a schematic diagram of the tracking effect of joint 2 in the simulation result of the present invention;

[0045] Figure 4 is a schematic diagram of the tracking error of joint 2 in the simulation result of the present invention;

[0046] Figure 5 is a schematic diagram of speed tracking error in the simulation results of the present invention;

[0047] Figure 6 is a schematic diagram of robot system control input in the simulation result of the present invention;

[0048] Figure 7 is a schematic diagram of system uncertainty in the simulation results of the present invention;

[0049] Figure 8 It is the basic flow chart of the improved obstacle function agricultural robot safety tracking control method of the present invention. DETAILED DESCRIPTION

[0050] The subject matter described in the present invention will now be discussed with reference to example embodiments. It should be understood that the discussion of these embodiments is only to enable those skilled in the art to better understand and implement the subject matter described in the present invention, and the functions and arrangements of the elements discussed may be changed without departing from the scope of protection of the present specification. Various examples may omit, replace or add various processes or components as needed. In addition, the features described in some examples may also be combined in other examples.

[0051] In order to better explain the problem and the controller proposed by the present invention, the improved obstacle function, robot system, controller design, and controller stability analysis are described below:

[0052] The description of the improved barrier function is as follows:

[0053] In order to improve the control accuracy of the robot tracking system and deal with position constraints, this section first proposes the following integral barrier function with adjustment parameters:

[0054]

[0055] Where p is the adjustment parameter of the integral barrier function and is greater than zero. e is the constrained system tracking error, and its constraint boundary is represented by k(t). is the integral variable. From the definition of the integral function (1), it can be seen that when the tracking error is in the open interval |e|<k(t), the barrier function V is continuous, differentiable, and positive definite. When e=0, V=0, and when |e|→k(t), V is radially unbounded.

[0056] Next, we will give the theorem related to the barrier function (1) mentioned above.

[0057] Theorem 1: In the open interval |e|<k(t), the barrier function V defined in (1) satisfies the following inequality conditions:

[0058]

[0059] Step 1: In step 1, we prove the inequality on the left side of inequality (2) First, define a helper function:

[0060]

[0061] Taking the derivative of the auxiliary function L(e) with respect to the tracking error e, we can obtain:

[0062]

[0063] According to the partial derivative of the auxiliary function L(e), when e<0 When e>0 And L(0)=0 always holds. Therefore, we can infer that the inequality is established.

[0064] Step 2: In step 2, we prove the inequality on the right side of inequality (2) It is true. Similar to step 1, define an auxiliary function:

[0065]

[0066] Taking the derivative of the auxiliary function N(e) with respect to the tracking error e, we can obtain:

[0067]

[0068] According to the partial derivative of the auxiliary function N(e), when e<0 When e>0 And N(0)=0 is always true. Therefore, we can infer that the inequality is established.

[0069] Therefore, the inequality This is always true in the open interval |e|<k(t).

[0070] Note 1: According to the proof of Theorem 1, we can get the inequality It is always satisfied in the open interval |e|<k(t). Therefore, when the constraint boundary k(t) tends to infinity, that is, the system is equivalent to being unconstrained, the limit of the right side of inequality (2) is obtained:

[0071]

[0072] According to the squeeze theorem:

[0073] Therefore, the barrier function with adjustment parameters proposed for the first time in the present invention can be transformed into a standard quadratic control Lyapunov function. Therefore, this barrier function can not only handle systems with constrained requirements, but also directly use the controller designed by it to stabilize systems without constrained requirements.

[0074] The description of the robot system is as follows:

[0075] The control object of the present invention is an n-joint robot system, which is described as follows [22,29]:

[0076]

[0077] in, represents the total uncertainty of the system, J T (q) is the Jacobian matrix, f(t) represents the unknown force on the end of the robot arm, is the inertia matrix, G0(q)∈R n and denote the gravity matrix and the Coriolis centripetal force matrix respectively and It is established, q, and The subtables represent the robot joint positions, velocities and accelerations, and the control input of the robot system is represented by τ(t).

[0078] Next, to facilitate the design of the controller, we perform the following coordinate transformation:

[0079]

[0080] Considering equation (4), system (3) can be rewritten as follows:

[0081]

[0082] Among them, M kn =C0(x1,x2)x2+G0(x1).

[0083] The control objective of the present invention is to design a trajectory tracking controller for a robot system (5) using the proposed obstacle function (1) with adjustment parameters, and to ensure that the robot's position vector x1 = q = [q1, q2, ···, q n ] T Successfully tracked the target trajectory x d =q d =[x d1 ,x d2 ,···,xdn ] T , so that the robot position always remains within the constraint interval, that is, |x1|<k c (t) = [k c1 (t),k c2 (t),···,k cn (t)] T At the same time, different adjustment parameters p can be set i The value of is used to adjust the control performance of the robot tracking system.

[0084] Assumption 1: System uncertainty f un is a bounded, differentiable, slow-varying signal, that is, there exists a positive constant F M Make |f un |≤F M and

[0085] Assumption 2: There exists a satisfying The vector X1 of the inequality condition makes the target trajectory satisfy |x d |≤X1. In order to ensure that the position constraint is always satisfied, we set the position error boundary condition as: k q (t) = k c (t)-X1.

[0086] The controller design and stability analysis are described as follows:

[0087] In order to achieve the control goal of the present invention, the integral barrier function with adjustment parameters first proposed in the present invention is used to design the robot tracking controller. First, we define the robot position error and velocity error as: e1 = [e 11 ,e 12 ,···,e 1n ] T =x1-x d and e2=[e 21 ,e 22 ,···,e 2n ] T =x2-α, the desired speed α will then be designed. The integral barrier function with adjustment parameters about the position error is designed as follows:

[0088]

[0089] Among them, p i is the adjustment parameter of the barrier function, and the error constraint boundary is: k q =k c -X1=[k q1 ,k q2 ,···,k qn ]T ,|x d |≤X1.

[0090] Taking the derivative of formula (6), we get:

[0091]

[0092] Next, we design the virtual position control law. First, we take the derivative of the position error and get:

[0093]

[0094] With the help of backstepping technology, the virtual position control law is designed as follows:

[0095]

[0096] in,

[0097]

[0098] I2=[I 21 ,I 22 ,···,I 2n ] T =I diag I dot I′ (11)

[0099] in,

[0100]

[0101] Note 2: From (11), we can see that the denominator of the virtual position control law contains the tracking error e 1i , which is likely to cause singularity in the control law. However, according to L’Hôpital’s rule, when e 1i When it approaches 0, we find its limit and get:

[0102]

[0103] Therefore, the control law we designed will not have singularity.

[0104] Substituting (8) into (7), we obtain:

[0105]

[0106] Considering (9), (10), (11), (12), (13) and (14), equation (15) can be rewritten as:

[0107]

[0108] In order to design the control law τ, we first differentiate the velocity tracking error:

[0109]

[0110] Since the velocity tracking error derivative contains system uncertainties, we introduce a disturbance observer to observe it

[30] :

[0111]

[0112] Among them, k f =diag(k f11 ,k f22 ,···,k fnn ) is a positive parameter of the observer, δ f ∈R n is the state variable of the observer, f un The estimated value of is defined as

[0113] Next, we define the second Lyapunov function:

[0114]

[0115] Combining (16) and (17), we can derive (19) to obtain:

[0116]

[0117] Among them, M′ kn =C0(x1,x2)α+G0(x1).

[0118] According to Lyapunov stability theory, we design the following controller:

[0119]

[0120] Where K2 = diag(k 21 ,k 22 ,···,k 2n ) is a positive definite parameter matrix;

[0121]

[0122] Next, we analyze the stability of the system through the following Theorem 2.

[0123] Theorem 2: Considering the robot system (3) with position constraints and system uncertainty, assuming that the initial position error lies in the open interval |e 1i (0)|<k qi(0), assumptions 1-2 are satisfied. Then, under the designed controller (9), (21) and observer (18), the error signal of the robot tracking system can converge to zero asymptotically and the position constraint boundary will never be violated under different adjustment parameters.

[0124] Proof: Substituting (20) into (20) we obtain:

[0125]

[0126] Substituting (21) into (22), we obtain:

[0127]

[0128] Among them, I n×n is an n-order identity matrix, and the control parameters must satisfy and

[0129] According to Theorem 1 proposed by the present invention, we can change the above formula again:

[0130]

[0131] in,

[0132]

[0133] Next, we prove that Theorem 2 is true mathematically. Solving the differential equation (24) yields 0≤V2≤V2(0)e -ρt . Combined with the definition of V1, we can deduce:

[0134]

[0135] Solving the above inequality, we get:

[0136]

[0137] Therefore, we can obtain |e 1i |<k qi (t). According to the relationship between the error constraint boundary and the position constraint boundary in Assumption 2, we know that |x1|<k c (t) is always satisfied. According to Theorem 1, we can further obtain:

[0138]

[0139] Considering (19) and (25), the following inequality holds:

[0140] In summary, we can conclude that the error signal of the robot tracking system can be exponentially asymptotically

[0141] Converges to zero, the position and its tracking error are always within the constraint boundary. And according to the above formula, given the adjustment parameter p i Different values ​​can be used to appropriately adjust the size of the position error.

[0142] Note 3: Given the proof of Theorem 2, the position error always satisfies the inequality condition Therefore, the barrier function method with adjustment parameters we proposed can not only constrain the error or position, but also make the constrained error asymptotically approach zero. In addition, by adjusting the parameter p i , the convergence speed and convergence accuracy of the constrained error can be appropriately adjusted.

[0143] The present invention provides the following simulation experiment to illustrate the effectiveness of the aforementioned obstacle function tracking control method with adjustable parameters, and the simulation experiment is conducted based on a dual-joint robot [22,29]. The position vector of the selected robot and the corresponding system matrix are shown as follows:

[0144]

[0145] G0(q)=[G 11 G 12 ] T

[0146]

[0147] in,

[0148] For the physical meaning of the main parameters and variables of the robot system, readers are referred to references [22,29].

[0149] The initial value of the robot system is set as:

[0150]

[0151] The desired circular trajectory is set as:

[0152] x d =[0.15sin(t),0.15cos(t)] T

[0153] The system uncertainty is set as:

[0154] f un =M0[0.4sin(0.5t),0.6cos(0.5t)] T

[0155] +C0[0.3cos(0.5t),0.3sin(0.5t)] T

[0156] In order to ensure the position constraint |q|<k c (t)=[0.24+0.05cos(t),0.24+0.05sin(t)] T It is never violated. According to Assumption 2 and the selected expected trajectory, we select the expected trajectory boundary as X1 = [0.15, 0.15] T and k q (t) = k c (t)-X1=[0.09+0.05cos(t),0.09+0.05sin(t)] T is the error constraint boundary. The control parameters and observer parameters are selected as k 11 =k 12 =1, K2 = diag(2,2), and k f =diag(20,20). In order to verify the effectiveness of the adjustment parameters introduced in the present invention, we simulated the adjustment parameters p. i To set different values:

[0157] p1=p2=0.1,2,5,10.

[0158] Therefore, the barrier function with adjustment parameters proposed for the first time in the present invention can be transformed into a standard quadratic control Lyapunov function. Therefore, this barrier function can not only handle systems with constrained requirements, but also directly use the controller designed by it to stabilize systems without constrained requirements.

[0159] The simulation results of robot tracking control based on improved integral BLF with adjustment parameters are shown in Figure 1-7 shown. Figure 1-4 The tracking effect of robot joints 1 and 2 and the curves of their tracking errors changing with time are given. Figure 5 The velocity tracking error curves of robot joints 1 and 2 are given. Figure 6 The control input curves under different tuning parameters are depicted. Figure 7 A diagram showing the effect of uncertainty on the observer's observation system.

[0160] from Figure 1 and 3 It can be seen that under different adjustment parameters, joint 1 and joint 2 do not exceed the time-varying constraint boundary. However, as the adjustment parameter value increases, the corresponding speed of reaching the desired target trajectory is faster. Figure 2 and 4It can be seen from the tracking error curve that with the increase of the adjustment parameter value, the overshoot of the dynamic control performance index increases. Therefore, the adjustment parameter of the improved integral barrier function proposed in the present invention cannot be increased blindly, otherwise it will cause a large overshoot. The convergence speed and control accuracy of the system can only be appropriately adjusted within a certain range. Figure 5 It can also be seen from the speed tracking error that when the adjustment parameter is large, the corresponding overshoot is also large. However, from the local enlarged diagram of the speed tracking error, it can be seen that the steady-state error under the maximum adjustment parameter is the smallest, which corresponds to our conclusion in Theorem 2. Figure 6 The system control input curves under different adjustment parameters are shown. It can be seen that in the initial stage, in order to quickly track the desired trajectory, when the adjustment parameter is the largest, the corresponding control input is also the largest. The curve of the designed disturbance observer observing the system uncertainty is depicted in Figure 7 It can be seen that the observation effect is satisfactory.

[0161] This study first proposed an integral barrier function with adjustment parameters to deal with the state constraint problem of nonlinear systems. And the inequality conditions related to this barrier function are given to prove the stability of the system. According to the given Theorem 1, we prove from a mathematical point of view that the proposed barrier function can not only handle systems with constraint requirements but also systems without constraint requirements. At the same time, the adjustment parameters contained in the barrier function can appropriately adjust the control accuracy and convergence speed of the system. The derivation process of Theorem 2 also proves this conclusion. Then, we simulated and verified the proposed improved obstacle function robot tracking control method. The simulation results show that under different adjustment parameters, the position and tracking error constraint boundaries are never broken, and the adjustment parameters can achieve the effect of adjusting the control accuracy. In future work, we will change the barrier function proposed in this invention into an asymmetric form to make it more general.

[0162] References of the present invention:

[0163] [1]LIH,FREIR,WENSINGP.ModelHierarchyPredictiveControlofRoboticSystems[J].IEEERoboticsandAutomation Letters,2021,6(2):3373-3380.

[0164] [2]WangY.Robotalgorithmbasedonneuralnetworkandintelligentpredictivecontrol[J].JournalofAmbientIntelligenceand HumanizedComputing,2020,11:6155-6166.

[0165] [3]Elsisi M,Mahmoud K,Lehtonen M,Darwish M.Effective Nonlinear ModelPredictive Control Scheme Tuned by ImprovedNNforRoboticManipulators[J].IEEEAccess,2021,9:64278-64290.

[0166] [4]Xian J,Shen L,Chen J,Feng W.Continuous Sliding Mode Control ofRobotic Manipulators Based on Time-Varying Disturbance Estimation andCompensation[J].IEEEAccess,2022,10:43473-43480.

[0167] [5]Zhang X,Quan Y.Adaptive Fractional-Order Non-Singular FastTerminal Sliding Mode Control Based on Fixed Time Disturbance Observer forManipulators[J].IEEEAccess,2022,10:76504-76511.

[0168] [6]Zhang L,WangY,Hou Y,Li H.Fixed-Time Sliding Mode Control forUncertain Robot Manipulators[J].IEEEAccess,2019,7:149750-149763

[0170] [7]Dong Y,He W,Kong L,Hua X.Impedance Control for Coordinated Robotsby State and Output Feedback[J].IEEE Transactions on Systems,Man,andCybernetics:Systems,2021,51(8):5056-5066.

[0171] [8]Yin W,Sun L,Wang M,Liu J.Nonlinear state feedback position controlfor flexible joint robot with energy shaping[J].Robotics andAutonomousSystems,2018,99:121-134.

[0173] [9]Kong L,He W,Dong Y,Cheng L,Yang C,Li Z.Asymmetric Bounded NeuralControl for an Uncertain Robot by State Feedback and Output Feedback[J].IEEETransactions on Systems,Man,and Cybernetics:Systems,2021,51(3):1735-1746.

[0175]

[10] Pan Y,Wang H,Li X,Yu H.Adaptive Command-Filtered BacksteppingControl of Robot Arms With Compliant Actuators[J].IEEE Transactions onControl Systems Technology,2018,26(3):1149-1156.

[0176]

[11] Chang W, Li Y, Tong S. Adaptive Fuzzy Backstepping Tracking Control for Flexible Robotic Manipulator[J]. IEEE / CAA Journal ofAutomatica Sinica, 2021,8(12):1923-1930.

[0177]

[12] Song Z, Sun K. Prescribed performance adaptive control for anuncertain robotic manipulator with input compensation updating law [J]. Journal of the Franklin Institute, 2021, 358(16): 8396-8418.

[0178]

[13] Yang P,Su Y.Proximate Fixed-Time Prescribed Performance TrackingControl of Uncertain Robot Manipulators[J].IEEE / ASME Transactions onMechatronics,2022,27(5):3275-2022.

[0180]

[14] Wang M, Yang A.Dynamic Learning From Adaptive Neural Control ofRobot Manipulators With Prescribed Performance[J].IEEE Transactions onSystems,Man,and Cybernetics:Systems,2017,47(8):2244-2255.

[0181]

[15] Li Xiaohua, Yang Ruifang, Liu Hui, He Shuai. A class of adaptive finite-time bounded H-type manipulator systems. ∞ Tracking control. Control Theory and Applications, 2021, 38(1): 147-156. Li Xiaohua, Yang Ruifang, Liu Hui, He Shuai. Adaptive finite-time bounded-H ∞tracking control for a class of manipulator system[J]. Control Theory&Applications, 2021,38(1):147-156.

[0184]

[16] Liao Liefa, Yang Yiguo. Adaptive radial basis function neural network bi-quadratic functional optimal control for manipulators[J]. Control Theory & Applications, 2020, 37(1): 47-58.

[0186]

[17] Mou Fangli, Wu Dan, Dong Yunfei. Active disturbance rejection control with multilayer perceptron compensating network for robot systems[J]. Control Theory & Applications, 2020, 37(6): 1397-1405.

[0188]

[18] Tee KP, Ge SS, Tay E H. Barrier Lyapunov Functions for the control of output-constrained nonlinear systems[J]. Automatica, 2009, 45(4):918-927.

[0190]

[19] Ren B,Ge S S,Tee K P,Lee T H.Adaptive Neural Control for OutputFeedback Nonlinear Systems Using a Barrier Lyapunov Function[J].IEEETransactions onNeural Networks,2010,21(8):1339-1345.

[0191]

[20] Zhang S,DongY,OuYangY,YinZ,PengK.AdaptiveNeural ControlforRoboticManipulatorsWithOutputConstraints andUncertainties[J].IEEETransactionsonNeuralNetworksandLearningSystems,2018,29(11):5554-5564.

[0192]

[21] He W,DavidA O,Yin Z,Sun C.Neural Network Control ofa RoboticManipulator With Input Deadzone and Output Constraint[J].IEEETransactionsonSystems,Man,andCybernetics:Systems,2016,46(6):759-770.

[0193]

[22] He W,Chen Y,Yin Z.Adaptive Neural Network Control ofan UncertainRobot With Full-State Constraints[J].IEEE TransactionsonCybernetics,2016,46(3):620-629.

[0194]

[23] Lu S M,Li D P,Liu Y J.Adaptive Neural Network Control forUncertain Time-Varying State Constrained Robotics Systems[J].IEEETransactionsonSystems,Man,andCybernetics:Systems,2019,49(12):2511-2518.

[0195]

[24] XiaJ,ZhangY,YangC,WangM,AnnamalaiA.Animprovedadaptiveonlineneuralcontrolforrobotmanipulatorsystems usingintegralBarrierLyapunovfunctions[J].InternationalJournalofSystems Science,2019,50(3):638-651.

[0196]

[25] Zhang J.Integral Barrier Lyapunov Functions-based Neural Controlfor Strict-feedback Nonlinear Systems with Multi-constraint[J].InternationalJournalofControl,AutomationandSystems,2018,16:2002-2010.

[0197]

[26] Li D J,Li J,Li S.Adaptive control ofnonlinear systems with fullstate constraints using Integral Barrier Lyapunov Functionals[J].Neurocomputing,2016,186:90-96.

[0198]

[27] Zhao B, FengZ, Guo J. IntegralbarrierLyapunovfunctions-basedintegratedguidance andcontrol design for strap-down missilewithfield-of-viewconstraint[J].TransactionsoftheInstituteofMeasurementandControl,2021,43(6):1464-1477.

[28] ZhangD, MaP, DuY, Chao T. Integralbarrier Lyapunov function-based three-dimensional low-order integrated guidance and control design with seeker's field-of-viewconstraint [J]. Aerospace Science and Technology, 2021, 116: 106886.

[0199]

[29] LiuYJ,Lu S,TongS.NeuralNetworkControllerDesignforanUncertainRobotWithTime-VaryingOutputConstraint[J].IEEETransactionsonSystems,Man,andCybernetics:Systems,2017,47(8):2060-2068.ZhengZ,Fer oskhanM.PathFollowingofaSurfaceVesselWithPrescribedPerformanceinthePresenceofInputSaturationandExternalDisturbances[J].IEEE / ASMETransactionsonMechatronics,2017,22(6):2564-2575.

[0202] The above describes an embodiment of the present invention, but this embodiment is not limited to the above-mentioned specific implementation mode. The above-mentioned specific implementation mode is merely illustrative and not restrictive. Under the guidance of this embodiment, ordinary technicians in this field can also make more forms of equivalent embodiments, all of which are within the protection of this embodiment.

Claims

1. An improved obstacle function agricultural robot safety tracking control method, characterized in that: include: The following trajectory tracking controller is designed based on the integral barrier function with tuning parameters: Where K2 = diag(k 21 ,k 22 ,···,k 2n ) is a positive definite parameter matrix; where p i represents the adjustment parameter, i∈[1,n]; The controlled object is an n-joint robot system, which is described as: Among them, M kn =C0(x1,x2)x2+G0(x1); By setting different tuning parameters p i The value of adjusts the control accuracy and convergence speed of the robot system to the set range; Then run the trajectory tracking controller to ensure that the robot position always remains within the constraint range.

2. The improved obstacle function agricultural robot safety tracking control method according to claim 1 is characterized in that: The robot position always remains within the constraint range means: The robot’s position vector x1=q=[q1,q2,···,q n ] T Successfully tracked the target trajectory x d =q d =[x d1 ,x d2 ,···,x dn ] T .

3. The improved obstacle function agricultural robot safety tracking control method according to claim 1 is characterized in that: The error signal of the robot tracking system converges exponentially to zero, and the position and its tracking error are always within the constraint boundary. The adjustment parameter p is given i Different values ​​can appropriately adjust the size of the position error.

4. The improved obstacle function agricultural robot safety tracking control method according to claim 3 is characterized in that: Adjust the parameter p i The relationship with position error is as follows:

5. The improved obstacle function agricultural robot safety tracking control method according to claim 1 is characterized in that: Adjust the parameter p i Proportional to the convergence speed.

6. The improved obstacle function agricultural robot safety tracking control method according to claim 1 is characterized in that: The integrated barrier function with tuning parameters is as follows: Where p is the adjustment parameter of the integral barrier function and is greater than zero; e is the constrained system tracking error, and its constraint boundary is represented by k(t); is the integral variable; when the tracking error is in the open interval |e|<k(t), the barrier function V is continuous, differentiable, and positive definite. When e=0, V=0, and when |e|→k(t), V is radially unbounded.

7. The improved obstacle function agricultural robot safety tracking control method according to claim 1 is characterized in that: Trajectory tracking controllers can handle systems with constrained demands and can stabilize systems with unconstrained demands.

8. The improved obstacle function agricultural robot safety tracking control method according to claim 1 or 7, characterized in that: The trajectory tracking controller is obtained by combining an integral barrier function with a tuning parameter with an observer.

9. An improved obstacle function agricultural robot safety tracking control system, characterized in that: It is used to execute the improved obstacle function agricultural robot safety tracking control method as described in any one of claims 1-8.

10. A robot trajectory tracking controller with improved obstacle function, characterized in that: Based on the design of an integral barrier function with tuning parameters, the trajectory tracking controller is expressed as follows: Where K2 = diag(k 21 ,k 22 ,···,k 2n ) is a positive definite parameter matrix; where p i represents the adjustment parameter, i∈[1,n]; The controlled object is an n-joint robot system, which is described as: Among them, M kn =C0(x1,x2)x2+G0(x1); By setting different tuning parameters p i The value of adjusts the control accuracy and convergence speed of the robot system to the set range.

Images