JCSE Journal of Control Science and Engineering 1687-5257 1687-5249 Hindawi 10.1155/2019/8146901 8146901 Research Article Nonsingular Fast Terminal Sliding Mode Tracking Control for a Class of Uncertain Nonlinear Systems Chen Siyi 1 http://orcid.org/0000-0003-2049-7176 Liu Wei 1 Huang Huixian 1 Matušů Radek College of Information Engineering Xiangtan University Xiangtan 411105 China xtu.edu.cn 2019 3052019 2019 19 01 2019 18 04 2019 30 04 2019 3052019 2019 Copyright © 2019 Siyi Chen et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Aiming at the tracking control problem of a class of uncertain nonlinear systems, a nonsingular fast terminal sliding mode control scheme combining RBF network and disturbance observer is proposed. The sliding mode controller is designed by using nonsingular fast terminal sliding mode and second power reaching law to solve the problem of singularity and slow convergence in traditional terminal sliding mode control. By using the universal approximation of RBF network, the unknown nonlinear function of the system is approximated, and the disturbance observer is designed by using the hyperbolic tangent nonlinear tracking differentiator (TANH-NTD) to estimate the interference of the system and enhance the robustness of the system. The stability of the system is proved by the Lyapunov principle. The numerical simulation results show that the method can shorten the system arrival time, improve the tracking accuracy, and suppress the chattering phenomenon.

1. Introduction

Sliding mode variable structure control is essentially a nonlinear control that the structure changes over time. Its significant advantage is its strong robustness to uncertain parameters and external disturbances. Therefore, it has been widely used in aerospace, robot control, and chemical control . However, the traditional sliding mode control takes the linear sliding mode as the “sliding mode” of the system. The deviation between the system state and the given trajectory converges exponentially but cannot converge to zero in a finite time. Therefore, the nonlinear term is introduced in the design of terminal sliding mode (TSM) control, and the tracking error on the sliding mode surface can converge to zero in a limited time, which makes it widely used in various control systems . Reference  proposed a terminal sliding mode control design scheme for uncertain dynamic systems with pure feedback form. In Reference , a new terminal sliding mode control design is proposed for the n-link rigid manipulator. Reference  proposed a new sliding mode control method for robot terminals. Reference  discussed the design of terminal sliding mode variable structure control for multi-input uncertain linear system. However, TSM has singularity . Nonsingular terminal sliding mode (NTSM) control evolved on the basis of avoiding the TSM singularity problem. It avoids the control singular regions directly in the sliding mode design and preserves the finite time convergence characteristics of TSM . In recent years, NTSM has developed rapidly. In , a continuous nonsingular terminal sliding mode control method was proposed for the suppression of mismatch interference. In , in order to realize the finite time tracking control of the axial position of the nonlinear thrust active magnetic bearing rotor, a robust nonsingular terminal sliding mode control system was proposed. In order to further improve the convergence speed of the sliding mode, a nonsingular fast terminal sliding mode (NFTSM) surface was designed. Reference  studied the fast finite time control of terminal sliding mode with nonlinear dynamics. In , a state-based nonsingular fast-terminal sliding mode controller was designed using direct instantaneous torque control. However, the system was often affected by the uncertainty of the model and the amount of interference, resulting in system chattering and even instability. In order to solve this problem, Reference  proposed a continuous nonsingular fast terminal sliding mode control scheme with extended state observer and tracking differentiator for second-order uncertain SISO nonlinear systems. In , the stability and attitude control of a class of quadrotor systems with uncertainties and unknown disturbances were studied. A nonsingular fast terminal sliding mode attitude control scheme for tracking differentiators and extended state observers was proposed. However, the extended state observer has an initial differential peak problem. Document  proposes an NDO design method for underactuated robot arm control. In , the NDO is combined with the dynamic surface to design the moving wheel inverted pendulum controller. They have a good effect on the estimation of the disturbance, but both of them have the disadvantage of relying on the prior knowledge of the disturbance. However, it is difficult to obtain prior knowledge of disturbance in practice. Literature  proposed a nonlinear disturbance observer based on tracking differentiator, which overcomes the shortcomings of the literature [32, 33] need to know the prior knowledge of disturbance. At the same time, it has the advantages of simple structure, excellent estimation effect on disturbance, and suppression of measurement noise.

In this paper, to realize fast and stable tracking control for a class of second-order uncertain nonlinear systems, a nonsingular fast terminal sliding mode control strategy combining RBF network and disturbance observer is proposed. The contributions of this paper are as follows.

(1) The sliding mode controller is designed by using nonsingular fast terminal sliding mode and second power reaching law, so that the system can converge to zero smoothly in a short time.

(2) RBF neural networks have strong nonlinear fitting ability to map arbitrarily complex nonlinear relationships. At the same time, it has the advantages of simple learning rules and easy computer implementation. Using the universal approximation principle of RBF network, the unknown nonlinear function is approximated to solve the influence of unknown nonlinear function on the robustness of the system.

(3) The nonlinear disturbance observer based on tracking differentiator has the advantages of simple structure, good disturbance estimation effect, and suppressing measurement noise. A hyperbolic tangent nonlinear disturbance observer is designed to estimate the external disturbance and unknown part of the model and compensate the controller. At the same time, an augmented nonlinear tracking differentiator designed in Reference  is used to filter the given signal, eliminating the influence of the given noise on the system.

The numerical simulation results show that the designed control method can effectively shorten the convergence time, eliminate the noise of the given signal and the chattering phenomenon in the controller, and improve the control tracking accuracy.

2. Problem Description

Consider the following second-order uncertain nonlinear system:(1)x˙1=x2x˙2=fx+bxu+dty=x1

where x=[x1,x2]TR2 is the state variable; f(x) and b(x) are the unknown nonlinear functions; uR is the control input; yR is the system output; d(t) is the slow time-varying interference, and d(t)D, constant D>0.

3. Main Results

The control objective of the system is to design a robust controller that enables accurate and fast tracking of the desired input signal even in the presence of model uncertainties and external disturbances. In order to reach the target, an NFTSM control scheme combining RBF network and disturbance observer is designed. The structure diagram of the system controller is shown in Figure 1.

Structure diagram of the system controller.

3.1. Nonsingular Fast Terminal Sliding Mode Control

In order to solve the problem of singularity and slow convergence of traditional terminal sliding mode control, a novel nonsingular fast terminal sliding mode control method is proposed.

3.1.1. Sliding Mode Design

x d is set the given signal and is second derivable. The position and velocity tracking errors of the system are defined as e1=x1-xd and e2=x2-x˙d, respectively. The designed nonsingular fast terminal sliding mode is(2)s=e1+α1e1γsgne1+α2e2q/psgne2

where α1, α2 are positive real numbers; q, p are odd integers, and 1<q/p<2, γ>q/p. e1 and e2 do not have negative exponential terms, ensuring that the sliding mode s based controller does not have singularity. When the system error state reaches the sliding mode (s=0), the following equation can be obtained:(3)e1+α1e1γsgne1+α2e2q/psgne2=0

Substitute e˙1=e2 into (3), the following equation can be obtained:(4)e˙1=-1α2p/qe1+α1e1γsgne1p/q=-1α2p/qe1p/q1+α1e1γ-1p/q

Assume the time taken from e1(tr)0 to e1(tr+ts)=0 is ts, and the following equation can be obtained through the time integrated on both sides of (4):(5)e1tre1tr+ts1e1p/qde1=-trtr+ts1α2p/q1+α1e1γ-1p/qdτ-trtr+ts1α2p/qdτ

Simplification of Inequality (5) (6) t s α 2 p / q q q - p e 1 t r 1 - p / q

Therefore, the system error can converge to zero for a limited time.

Remark 1.

When the system error state is far from the equilibrium point, the dominant role of e1 in the sliding mode s causes the system trajectory to converge quickly. When the system error state approaches the equilibrium point, the dominant role of α1e1γsgne1 in the sliding mode s causes the system trajectory to converge rapidly. Therefore, NFTSM can achieve fast convergence of the whole state trajectory.

3.1.2. Reaching Law Design

According to the sliding mode variable structure principle, the sliding mode reachability condition only ensures that the moving point at any position in the state space can reach the switching surface within a limited time, and there is no restriction on the specific trajectory of the reaching motion. The reaching law can improve the dynamic quality of reaching movement. The following second power reaching law is adopted in this paper.(7)s˙=-k3sc2sgns-k4sc3sgns

where k3>0, k4>0, 0<c2<1, c3>1. When the system state is far from the sliding mode (s>1), -k4sc3sgns plays a leading role; when the system approaches the sliding mode (s<1), -k3sc2sgn(s) dominates. When s=0, s˙=0, the speed of the state reaching the sliding mode is reduced to zero, and the smooth transition of the sliding mode is realized, which greatly weakens the system chattering phenomenon. Compared with the traditional power reaching law, exponential reaching law, and fast power reaching law, the reaching law (7) has faster convergence speed and better motion quality.

3.1.3. Controller Design and Stability Analysis Lemma 2 (see [<xref ref-type="bibr" rid="B36">36</xref>]).

Set xMRn, x˙=f(x), f:RnR as the continuous functions defined in the equilibrium point region M. Assume that a continuous function V:MR satisfies the following conditions:

(1) V is positive definite.

(2) V˙ is negative definite except for the equilibrium point.

(3) Real number k>0, α>0 and region NM make V˙+kVα0, and the function x˙=f(x) converges at balanced zero point within finite time.

Theorem 3.

For the system shown in (1), design the sliding mode according to (2), using the NFTSM controller shown in (9). The system state s converges to the following areas in a limited time:(8)sminDk31/c2,Dk41/c3s˙minD,k3Dk4c1/c2+mink4Dk3c2/c1,D

Sliding mode controller is designed as(9)u=-bx-1fx+1α2pqe22-q/psgne21+α1γe1γ-1+k3sc2sgns+k4sc3sgns-x¨d

where α1, α2, k3, k4 are normal numbers, 0<c2<1, c3>1; q, p are odd integers, and 1<q/p<2, γ>q/p.

Proof.

Define the following Lyapunov function:(10)V=12s2

Find the time derivative for (10):(11)V˙=se˙1+α1γe1γ-1sgne1e˙1+α2qpe2q/p-1sgne2e˙2=se2+α1γe1γ-1sgne1e2+α2qpe2q/p-1sgne2fx+bxu+dt-x¨d

Substitute (9) into the above equation:(12)V˙=se˙1+α1γe1γ-1e˙1+α2qpe2q/p-1e˙2=se2+α1γe1γ-1e2+α2qpe2q/p-1fx+bx-bx-1fx+1α2pqe22-q/psgne21+α1γe1γ-1+k3sc2sgns+k4sc3sgns-x¨d+dt-x¨d=sα2qpe2q/p-1-k3sc2sgns-k4sc3sgns+dt=-α2qpe2q/p-1k3sc2+1+k4sc3+1-sdt-α2qpe2q/p-1k3sc2+1+k4sc3+1-Ds

Deform (12):(13)V˙-α2qpe2q/p-1k3sc2+1-α2qpe2q/p-1k4sc3+1+α2qpe2q/p-1Ds=-α2qpe2q/p-1k3sc2+1-sα2qpe2q/p-1k4sc3-D

Because α2 is a normal number, c3>1; q, p are odd integers, and 1<q/p<2, so α2q/pe2q/p-10. If k4sc3-D0, then (13) can be further changed to(14)V˙-α2qpe2q/p-1k3sc2+1=-α2qpe2q/p-1k3Vc2+1/2

Also, because k3>0, 0<c2<1, that is, α2q/pe2q/p-1k3>0, c2+1/2>0, (14) satisfies Lemma 2, and the system converges on the finite time of the equilibrium zero and can ensure the convergence of the following regions in a finite time.(15)sDk41/c3

Similarly, equation (12) can be transformed into(16)V˙-α2qpe2q/p-1k4sc3+1-sα2qpe2q/p-1k3sc2-D

Because α2q/pe2q/p-10, if k3sc2-D0, then (16) can be further changed to(17)V˙-α2qpe2q/p-1k4sc3+1=-α2qpe2q/p-1k4Vc3+1/2

Also, because α2q/pe2q/p-1k3>0, c3>1, c3+1/2>0, (17) satisfies Lemma 2, and the system converges on the finite time to the equilibrium zero and can guarantee the convergence of the following regions in a finite time.(18)sDk31/c2

Combining equations (15) and (18), the system state s converges to the following areas within a limited time:(19)sminDk31/c2,Dk41/c3Substituting the previous equation into (7), we can get(20)s˙k3sc2sgns+k4sc3sgnsk3sc2+k4sc3k3·minDk31/c2,Dk41/c3c2+k4·minDk31/c2,Dk41/c3c3=minD,k3Dk4c1/c2+mink4Dk3c2/c1,D

Theorem 3 is proved.

3.2. RBF Network Approximation <inline-formula><mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" id="M100"><mml:mi>f</mml:mi><mml:mo stretchy="false">(</mml:mo><mml:mi mathvariant="bold-italic">x</mml:mi><mml:mo stretchy="false">)</mml:mo></mml:math></inline-formula> and <inline-formula><mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" id="M101"><mml:mi>b</mml:mi><mml:mo stretchy="false">(</mml:mo><mml:mi mathvariant="bold-italic">x</mml:mi><mml:mo stretchy="false">)</mml:mo></mml:math></inline-formula>

The nonlinear functions f(x) and b(x) are required according to the control law (9), but the actual control needs to be obtained based on empirical knowledge, and sometimes it is not available. Therefore, it is approximated by the approximation principle. Because the RBF neural network has strong nonlinear fitting ability, it can map arbitrarily complex nonlinear relationships. At the same time, it has the advantages of simple learning rules and easy computer implementation. Therefore, using the universal approximation principle of RBF network, the unknown nonlinear function is approximated to solve the problem of the acquisition of unknown nonlinear function in the control law and the influence of unknown nonlinear function on the robustness of the system.

3.2.1. RBF Neural Network Structure

The RBF network is a 3-layer forward network with a simple structure and is suitable for real-time control. The structure of the RBF network with multiple inputs and single output is shown in Figure 2.

The RBF network with multiple inputs and single output.

In the RBF network, x=x1x2xnT is the network input, and hj is the Gaussian function, which represents the output of the jth neuron in the hidden layer; namely,(21)hj=exp-x-cj22gj2,j=1,2,···,m

where cj=[cj1,cj2,···,cjm] is the center of the Gaussian function of the node j; gj is the width of the Gaussian function of the node j. The weight of the network is taken as θ=[θ1,θ2,···,θm]T.

Output of RBF network is(22)ymt=θ1h1+θ2h2+···+θmhm

3.2.2. Adaptive Approximation of RBF Network

For the unknown nonlinear functions f(x) and b(x) in (9), the RBF network adaptive approximation is adopted. The RBF network input and output algorithm is(23)hj=exp-x-cj22gj2,j=1,2,···,mfx=WThfx+εf,bx=VThbx+εb

where x is the network input; hj is the Gaussian function; W and V are the ideal network weight vectors of f(x) and b(x), respectively; εf and εb are the network approximation errors, εfεMf, εbεMb. hf(x), hb(x) are vector functions, and the elements of them are all Gaussian function. We can set x=x1x2T, and then the output of the RBF network is(24)f^x=W^Thfx,b^x=V^Thbx

Design control law is(25)u=-b^x-1f^x+1α2pqe22-q/psgne21+α1γe1γ-1+k3sc2sgns+k4sc3sgns-x¨d

where α1, α2, k3, k4 are normal numbers, and 0<c2<1, c3>1. q, p are odd integers, and 1<q/p<2, γ>q/p. f^(x) and b^(x) are the RBF network output values.

Find the time derivative for (2):(26)s˙=e˙1+α1γe1γ-1e2+α2qpe2q/p-1e˙2=e2+α1γe1γ-1e2+α2qpe2q/p-1fx+bxu+dt-x¨d=e2+α1γe1γ-1e2+α2qpe2q/p-1fx+b^xu+bx-b^xu+dt-x¨d

Substitute (25) into (26):(27)s˙=e2+α1γe1γ-1e2+α2qpe2q/p-1fx+b^x-b^x-1f^x+1α2pqe22-q/psgne21+α1γe1γ-1+k3sc2sgns+k4sc3sgns-x¨d+bx-b^xu+dt-x¨d=α2qpe2q/p-1fx-f^x-k3sc2sgns-k4sc3sgns+bx-b^xu+dt=α2qpe2q/p-1f~x-k3sc2sgns-k4sc3sgns+b~xu+dt=α2qpe2q/p-1W~Thfx-εf-k3sc2sgns-k4sc3sgns+V~Thbx-εbu+dt

where W~=W^-W, V~=V^-V, and(28)f~=f-f^=WThfh-W^Thfx-εf=W~Thfx-εfb~=b-b^=VThbx-V^Thbx-εb=V~Thbx-εb

Design the Lyapunov function as(29)L=12s2+12r1W~TW~+12r2V~TV~

where r1>0, r2>0.

Find the time derivative for (29) and substitute it into (27):(30)L˙=sα2qpe2q/p-1W~Thfx-εf-k3sc2sgns-k4sc3sgns+V~Thbx-εbu+dt+1r1W~TW^˙+1r2V~TV^˙=W~Tα2qpe2q/p-1shfx+1r1W^˙+V~Tα2qpe2q/p-1shbxu+1r2V^˙+α2qpe2q/p-1s-εf-k3sc2sgns-k4sc3sgns-εbu+dt

Take the adaptive law as(31)W^˙=-r1α2qpe2q/p-1shfxV^˙=-r2α2qpe2q/p-1shbxu

Substitute (31) into (30):(32)L˙=α2qpe2q/p-1s-εf-k3sc2sgns-k4sc3sgns-εbu+dt=-α2qpe2q/p-1k3sc2+1+k4sc3+1+sεf+εbu-dt-α2qpe2q/p-1k3sc2+1+k4sc3+1+sεf+εbu-D

Since εf and εb, the RBF network approximation errors, are very small real numbers, δ is selected as δεf+εbu-D. According to Theorem 3, the system state s converges to the following region within a finite time:(33)sminδk31/c2,δk41/c3s˙minδ,k3δk4c1/c2+mink4δk3c2/c1,δ

3.3. Hyperbolic Tangent Nonlinear Disturbance Observer

For system (1), the interference of the system is not only the model unknown interference generated by the unknown nonlinear function, but also the external disturbance. However, from (33), an important condition for the system state s to reach the convergence region in a finite time is that the disturbance amount d(t) in (1) is d(t)D and D is affected by the disturbance amount d(t). When d(t) is larger, D becomes larger, which affects the stability of the system. In order to further eliminate the influence of the interference amount on the control system, the nonlinear disturbance observer using the tracking-differentiator has the advantages of simple structure, good interference estimation effect, suppression of measurement noise, etc., and a hyperbolic tangent nonlinear disturbance observer is designed. The interference in (1) is estimated and feedforward compensation is performed to improve the anti-interference ability of the system.

Lemma 4 (see [<xref ref-type="bibr" rid="B37">37</xref>]).

Consider the following system Σ1:(34)z˙1=z2tz˙2=-a1tanhb1z1t-a2tanhb2z2t

If a1, a2, b1, b2 are all positive real numbers, the system Σ1 is progressively stable at the origin (0,0). It satisfies(35)limtz1t=0limtz2t=0

An improved disturbance observer is designed by using the hyperbolic tangent tracking differentiator proposed in .(36)x^˙2=fx+bxu+d^d^˙=-R2a1tanhb1x^2-x2+a2tanhb2d^R

In the formula, R, a1, a2, b1, b2 are all positive real numbers; x^ and d^ are, respectively, estimates of x and d. If T>0, R>0, there are(37)limR0Tx^2-x2dt=0

When R, it can be obtained that |d^˙|=-R2a1tanhb1x^2-x2+a2tanhb2d^/R gets closer to infinity, so d^ changes faster than f(x)+b(x)u; that is, limRdfx+bxu+d^/dt=d^˙. Therefore, (36) satisfy Lemma 4, and the designed disturbance observer is progressively stable.

For (1), the following control laws are used:(38)u=-b^x-1f^x+kd^t+1α2pqe22-q/psgne21+α1γe1γ-1+k3sc2sgns+k4sc3sgns-x¨d

where α1, α2, k3, k4, k are normal numbers, 0<c2<1, c3>1. q and p are odd integers, and 1<q/p<2, γ>q/p. f^(x) and b^(x) are the output values of the RBF network. d^(t) is the output value of the hyperbolic tangent nonlinear disturbance observer. The system state s will converge to a smaller region for a limited time.

Take a derivative of function (29) according to the Lyapunov, combining with (26), (31), and (38):(39)L˙=se2+α1γe1γ-1e2+α2qpe2q/p-1fx+b^x-b^x-1f^x+kd^t+1α2pqe22-q/psgne21+α1γe1γ-1+k3sc2sgns+k4sc3sgns-x¨d+bx-b^xu+dt-x¨d+1r1W~TW^˙+1r2V~TV^˙=sα2qpe2q/p-1W~Thfx-εf-kd^t-k3sc2sgns-k4sc3sgns+V~Thbx-εbu+dt+1r1W~TW^˙+1r2V~TV^˙=W~Tα2qpe2q/p-1shfx+1r1W^˙+V~Tα2qpe2q/p-1shbxu+1r2V^˙+α2qpe2q/p-1s-εf-k3sc2sgns-k4sc3sgns-εbu-kd^t+dt=α2qpe2q/p-1s-εf-k3sc2sgns-k4sc3sgns-εbu-kd^t+dt=-α2qpe2q/p-1k3sc2+1+k4sc3+1+sεf+εbu+kd^t-dt-α2qpe2q/p-1k3sc2+1+k4sc3+1+sεf+εbu+kd^t-D

Setting δεf+εbu+kd^(t)-D, according to Theorem 3, the system state s converges to the following region within a finite time:(40)sminδk31/c2,δk41/c3s˙minδ,k3δk4c1/c2+mink4δk3c2/c1,δ

Theorem 5.

For the system shown in (1), design the sliding mode according to (2), using the NFTSM controller shown in (38). The system state converges to the following areas in a limited time:(41)e1minδk3α1c21/γc2,δk4α1c31/γc3e2minδk3α2c2p/qc2,δk4α2c3p/qc3

Proof.

According to Theorem 3 and (40), system state enters sminδ/k31/c2,δ/k41/c3 region, and we can get(42)e1+α1e1γsgne1+α2e2q/psgne2=σwhere σΩs=ssminδ/k31/c2,δ/k41/c3.

Equation (42) can be changed to(43)e1+α1e1γsgne1+α2-σe2q/psgne2e2q/psgne2=0

When α2-σ/e2q/psgn(e2)>0, (43) satisfies the NFTSM form described in (2), and the system trajectory will converge to(44)α2σe2q/psgne2

Therefore, state e2 can converge rapidly to the region within a finite time ts:(45)e2σα2p/qminδ/k31/c2,δ/k41/c3α2p/qminδk3α2c2p/qc2,δk4α2c3p/qc3(46)tsα2-σ/e2q/psgne2p/qqq-pe1tr1-p/q

Using the same approach, (42) can be changed to(47)e1+α1-σe1γsgne1e1γsgne1+α2e2q/psgne2=0

When α1-σ/e1γsgne1>0, (47) satisfies the NFTSM form described in (2), and the system trajectory will converge to(48)α1σe1γsgne1

Therefore, state e1 can converge rapidly to the region within a finite time ts:(49)e1σα1γminδ/k31/c2,δ/k41/c3α1γminδk3α1c21/γc2,δk4α1c31/γc3

Theorem 5 is proved.

Remark 6.

As the amount of disturbance d(t) in (1) becomes larger, D and d^(t) become larger. It can be ensured that the system can still converge with a smaller region. Owing to the fact that the control rate is nonsingular and continuous and there is no switching term in the general control rate, the chattering caused by frequent switching is avoided, so that the proposed control rate u can get lower chattering.

4. Simulations

In order to verify the feasibility and effectiveness of the control method in this paper, the controlled object is taken as a single-stage inverted pendulum system, and its dynamic equation is(50)x˙1=x2x˙2=fx+bxuwhere f(x)=gsinx1-mlx22cosx1sinx1/mc+mp/l4/3-mpcos2x1/mc+mp, b(x)=cosx1/(mc+mp)/l4/3-mpcos2x1/mc+mp. x1 and x2 are, respectively, the swing angle and the swing speed, u is the control input, mc is the weight of the trolley, mp is the weight of the pendulum, l is the length of the pendulum, and g is the acceleration of gravity.

The performance of system (50) contorted by the proposed method is compared with the NTSM control method  and the NFTSM control method  which are based on the following exponential reaching law.

(1) The sliding mode, the reaching law, and the controller of the NTSM control method based on the exponential reaching law are as follows:(51)s=e1+1βe2q/p,s˙=-ws-ρsgns,u=-bx-1fx+βpqe22-q/p+ws+ρsgns-x¨d

(2) The sliding mode, the reaching law, and the controller of the NFTSM control method based on the exponential reaching law are as follows:(52)s=e1+α1e1γsgne1+α2e2q/psgne2,s˙=-ws-ρsgns,u=-bx-1fx+1α2pqe22-q/psgne21+α1γe1γ-1+ws+ρsgns-x¨d

(3) The sliding mode, the reaching law, and the controller of the NFTSM control method based on the second power reaching law (SPNFTSM) are as follows:(53)s=e1+α1e1γsgne1+α2e2q/psgne2,s˙=-k3sc2sgns-k4sc3sgns,u=-bx-1fx+1α2pqe22-q/psgne21+α1γe1γ-1+k3sc2sgns+k4sc3sgns-x¨d

(4) The sliding mode, the reaching law, and the controller of the NFTSM control method based on the second power of RBF network (RBFSPNFTSM) are as follows:(54)s=e1+α1e1γsgne1+α2e2q/psgne2,s˙=-k3sc2sgns-k4sc3sgns,u=-b^x-1f^x+1α2pqe22-q/psgne21+α1γe1γ-1+k3sc2sgns+k4sc3sgns-x¨d

(5) The sliding mode, the reaching law, and the controller of the NFTSM control method based on the second power of RBF network and disturbance observer (RBFDOSPNFTSM) are as follows:(55)s=e1+α1e1γsgne1+α2e2q/psgne2,s˙=-k3sc2sgns-k4sc3sgns,u=-b^x-1f^x+kd^t+1α2pqe22-q/psgne21+α1γe1γ-1+k3sc2sgns+k4sc3sgns-x¨d

4.1. Dynamic Performance Analysis

The initial state of the inverted pendulum system is set as [π/60,0], and the system model parameters and controller parameters are shown in Tables 1 and 2, respectively. Given signal xd=0, interference d(t)=sinx1+5sin(πt).

System model parameters.

the weight of the trolley (mc) the weight of the pendulum (mp) the length of the pendulum (l) the acceleration of gravity (g)
1kg 0.1kg 0.5m 9.8 m/s2

Controller parameters.

parameter NTSM NFTSM SPNFTSM RBFSPNFTSM RBFDOSPNFTSM
α 1 - 0.1 0.1 0.1 0.1
α 2 - 0.02 0.02 0.02 0.02
γ - 27/19 27/19 27/19 27/19
q p 21/19 21/19 21/19 21/19 21/19
β 10 - - - -
w 1000 1000 - - -
ρ 1 1 - - -
k 3 - - 100 100 100
k 4 - - 1 1 0.1
c 2 - - 0.5 0.5 0.5
c 3 - - 1.5 1.5 1.5
k - - - - 0.291

In order to better analyse the stability of the system, the following two performance evaluation indicators are adopted:

(1) Integral of Error Squared Value (ISE) (56) I S E = 0 t x i 2 τ d τ

(2) Integral of Absolute Value of Error (IAE) (57) Ι Α Ε = 0 t x i τ d τ

The performance index values of the inverted pendulum system for the above two evaluation methods are shown in Table 3. It can be seen from Table 3 that the performance evaluation index value of the state error e1 of the control method used in this paper is smaller than other control methods. Therefore, the system stability of RBFDOSPNFTSM is better than that of other control methods.

Performance index values of the inverted pendulum system.

ISE IAE
e 1 e 1
NTSM 2.0317e-04 0.0229
NFTSM 1.6484e-04 0.0211
SPNFTSM 2.0334e-04 0.0154
RBFSPNFTSM 1.5577e-04 0.0062
RBFDOSPNFTSM 1.5375e-04 0.0050

Figures 36 show the response curves of the pendulum, swing speed, sliding mode, and system phase trajectory, respectively. It can be seen from them that the system state x1, x2 can converge to the equilibrium point at a faster speed under the control of the proposed scheme and can maintain a good stability state accuracy even if a large amount of interference is added. And the three-dimensional curve of the phase trajectory can intuitively show that the control process of the proposed method is smooth and jitter-free, and the system is more robust. Figure 7 shows the control input response curve. It can be seen from Figure 7 that the control method of the present invention controls the input signal to be smooth, which can reduce the wear of the motor. Figure 8 shows that the disturbance observer designed in this paper can estimate the interference value very well.

Swing angle response curve.

Swing speed response curve.

Sliding mode s response curve.

System phase trajectory three-dimensional curve.

Control input response curve.

Disturbance observation result.

4.2. Signal Tracking Analysis after Adding Noise

If the initial states of the inverted pendulum system, system model parameters, controller parameters, and interference remain unchanged, consider the given signal xd=xd1+ηn(t) with the noise signal to analyze the tracking control performance of the system, where η=0.0001, n(t) is a random number between [-1, 1], and xd1 is shown as below:(58)xd1=0.5t-14πsin2πt,0t<10.5,1t<2-0.5t+1.5+14πsin2πt,2t<30,3t<4-0.5t+2+14πsin2πt,4t<5-0.5,5t<60.5t-3.5-14πsin2πt,6t<70,7t<8

Figures 9 and 10 show the swing angle tracking curve and the swing speed response curve of the inverted pendulum system with noise interference. It can be seen from Figures 9 and 10 that the control method of this paper has a poor suppression effect on noise and obvious chattering. Through the analysis of the control method of this paper, it is found that when the first and second derivatives are obtained for the given signal, a differential explosion phenomenon will be generated, and the noise signal in the given signal will be amplified. Therefore, the tracking differentiator needs to be designed to filter the given signal.

Swing angle tracking curve (adding noise).

Swing speed response curve (adding noise).

The augmented nonlinear tracking differentiator (ATD) proposed in  is used to filter the given signal and suppress the influence of the noise signal on the control system. Its expression is as follows:(59)z˙0t=z1tz˙1t=z2tz˙2t=z3tz˙3t=-R4sigz0t-vτdτ;a0,b+sigz1tR;a1,b+sigz2tR2;a2,b+sigz3tR3;a3,bsigz;ai,b=ai1+e-bz-1-0.5

where a0>0, a1>0, a2>0, a3>0, b>0, R>0, and v(t) is an arbitrary bounded input signal. z1, z2, and z3 are estimated values, estimated values of first-order derivative, and estimated values of second-order derivative of v(t), respectively. Proof of stability has been given in . The parameters of the augmented nonlinear tracking differentiator are shown in Table 4.

Parameters of the augmented nonlinear tracking differentiator.

a 0 a 1 a 2 a 3 b R
10 10 10 1 10 25

Figures 1114 are the response curves of the swing angle, swing speed, sliding mode, and system phase trajectory of the inverted pendulum system with noise interference and filtered by the tracking differentiator. It can be seen from Figures 1114 that the RBFDOSPNFTSM control method for filtering of the given signal by the augmented nonlinear tracking differentiator (ATDRBFDOSPNFTSM) has advantages in tracking accuracy, suppression of chattering, and the dynamic quality. Figure 15 shows the control input response curve. It can be seen from Figure 15 that the control method adding the ATDRBFDOSPNFTSM has basically no chattering phenomenon and has advantages in actual energy consumption and mechanical wear of operation.

Swing angle tracking curve (adding noise).

Swing speed response curve (adding noise).

Sliding mode s response curve (adding noise).

System phase trajectory curve (adding noise).

Control input response curve (adding noise).

5. Conclusions

This paper proposes a nonsingular fast terminal sliding mode control scheme combining RBF network and disturbance observer to solve the tracking control problem of uncertain nonlinear systems. Firstly, the sliding mode controller is designed by using nonsingular fast terminal sliding mode and second power reaching law to solve the problem of singularity and slow convergence of traditional terminal sliding mode control. Then, the unknown nonlinear function of the system is approximated by the RBF network, and the hyperbolic tangent disturbance observer is used to estimate the interference of the system and enhance the robustness of the system. Finally, the given signal is filtered by the augmented nonlinear tracking differentiator, eliminating the effects of noise in a given signal on the system. Through the numerical simulation results, the dynamic performance of the inverted pendulum system and the signal tracking of adding noise are analyzed. The control method can eliminate the noise pollution of a given signal, shorten the system arrival time, improve the tracking accuracy, and suppress the chattering phenomenon.

Data Availability

The data used to support the findings of this study are available from the corresponding author upon request.

Conflicts of Interest

The authors declare that there are no conflicts of interest regarding the publication of this paper.

Choi H. H. Sliding-mode output feedback control design IEEE Transactions on Industrial Electronics 2008 55 11 4047 4054 10.1109/TIE.2008.2004386 2-s2.0-56349129367 Liu J. Wu Y. Guo J. Chen Q. High-order sliding mode-based synchronous control of a novel stair-climbing wheelchair robot Journal of Control Science and Engineering 2015 2015 16 680809 10.1155/2015/680809 MR3417560 Yuan L. Wu H.-S. Multiple-sliding-mode robust adaptive design for ship course tracking control Control Theory and Applications 2010 27 12 1618 1622 2-s2.0-79951974797 Corradini M. L. Cristofaro A. Nonsingular terminal sliding-mode control of nonlinear planar systems with global fixed-time stability guarantees Automatica 2018 95 561 565 10.1016/j.automatica.2018.06.032 MR3851499 Lu B. Fang Y. Sun N. Continuous sliding mode control strategy for a class of nonlinear underactuated systems IEEE Transactions on Automatic Control 2018 63 10 3471 3478 10.1109/TAC.2018.2794885 2-s2.0-85041668142 Fei J. Feng Z. Adaptive fuzzy super-twisting sliding mode control for microgyroscope Complexity 2019 2019 13 6942642 10.1155/2019/6942642 Qi D. Feng J. Yang J. Longitudinal motion control of auv based on fuzzy sliding mode method Journal of Control Science and Engineering 2016 2016 7 7428361 10.1155/2016/7428361 MR3465046 Baklouti F. Aloui S. Chaari A. Adaptive fuzzy sliding mode tracking control of uncertain underactuated nonlinear systems: a comparative study Journal of Control Science and Engineering 2016 2016 12 9283103 10.1155/2016/9283103 MR3531322 Liu J. K. Sun F. C. Research and development on theory and algorithms of sliding mode control Control Theory & Applications 2007 24 3 407 418 Zbl1150.93330 Zhihong M. Yu X. H. Terminal sliding mode control of MIMO linear systems IEEE Transactions on Circuits and Systems I: Fundamental Theory and Applications 1997 44 11 1065 1070 10.1109/81.641769 MR1488806 2-s2.0-0031268937 Gao W. Chen X. Du H. Bai S. Position tracking control for permanent magnet linear motor via continuous-time fast terminal sliding mode control Journal of Control Science and Engineering 2018 3 1 6 10.1155/2018/3813624 MR3879351 Xu S. S. D. Super-twisting-algorithm-based terminal sliding mode control for a bioreactor system Abstract and Applied Analysis 2014 2014 9 495680 10.1155/2014/495680 MR3216054 Wu Y. Yu X. Man Z. Terminal sliding mode control design for uncertain dynamic systems Systems & Control Letters 1998 34 5 281 287 2-s2.0-0002394836 10.1016/S0167-6911(98)00036-X Zbl0909.93005 Tang Y. Terminal sliding mode control for rigid robots Automatica 1998 34 1 51 56 10.1016/s0005-1098(97)00174-x MR1614164 Zbl0908.93042 Zhao D. Y. Li S. Y. Gao F. A new terminal sliding mode control for robotic manipulators International Journal of Control 2009 82 10 1804 1813 10.1080/00207170902769928 MR2567229 2-s2.0-68949117683 Yu X. Zhihong M. Multi-input uncertain linear systems with terminal sliding-mode control Automatica 1998 34 3 389 392 10.1016/S0005-1098(97)00205-7 MR1623085 Zbl0915.93012 Zhang W. Wang J. Nonsingular terminal sliding mode control based on exponential reaching law Journal of Control and Decision 2012 27 6 909 913 MR2987459 2-s2.0-84865147815 Feng Y. Yu X. Man Z. Non-singular terminal sliding mode control of rigid manipulators Automatica 2002 38 12 2159 2167 10.1016/S0005-1098(02)00147-4 MR2134882 Feng Y. Bao S. Yu X. H. Design method of non-singular terminal sliding mode control systems Journal of Control and Decision 2002 17 2 194 198 Li S. B. Li K. Q. Wang J. Q. Nonsingular and fast terminal sliding mode control method Information and Control 2009 38 1 1 8 Song J. H. Song M. S. Zhou H. B. Adaptive nonsingular fast terminal sliding mode guidance law with impact angle constraints International Journal of Control, Automation, and Systems 2016 14 1 99 114 10.1007/s12555-014-0155-8 2-s2.0-84957943032 Jeong J. W. Yeon H. S. Park K.-B. Extended nonsingular terminal sliding surface for second-order nonlinear systems Abstract and Applied Analysis 2014 2014 6 267510 10.1155/2014/267510 MR3212406 Tran M.-D. Kang H.-J. Nonsingular terminal sliding mode control of uncertain second-order nonlinear systems Mathematical Problems in Engineering 2015 2015 8 181737 10.1155/2015/181737 MR3415821 Wang W. Zhao Q. Zhao Y. Du D. A nonsingular terminal sliding mode approach using adaptive disturbance observer for finite-time trajectory tracking of mems triaxial vibratory gyroscope Mathematical Problems in Engineering 2015 2015 8 205652 10.1155/2015/205652 MR3319219 Wang H. Shi L. Man Z. Continuous fast nonsingular terminal sliding mode control of automotive electronic throttle systems using finite-time exact observer IEEE Transactions on Industrial Electronics 2018 65 9 7160 7172 10.1109/TIE.2018.2795591 2-s2.0-85040926104 Yang J. Li S. Su J. Yu X. Continuous nonsingular terminal sliding mode control for systems with mismatched disturbances Automatica 2013 49 7 2287 2291 10.1016/j.automatica.2013.03.026 MR3063089 2-s2.0-84878874964 Chen S.-Y. Lin F.-J. Robust nonsingular terminal sliding-mode control for nonlinear magnetic bearing system IEEE Transactions on Control Systems Technology 2011 19 3 636 643 10.1109/TCST.2010.2050484 2-s2.0-79955471917 Yang L. Yang J. Nonsingular fast terminal sliding-mode control for nonlinear dynamical systems International Journal of Robust and Nonlinear Control 2011 21 16 1865 1879 10.1002/rnc.1666 MR2845640 Zbl1237.93039 2-s2.0-80053075150 Cheng Y. Lin H. Nonsingular fast terminal sliding mode positioning control in switched reluctance motor Electric Machines and Control 2012 16 9 78 82 2-s2.0-84868650210 He Z. Liu C. Zhan Y. Li H. Huang X. Zhang Z. Nonsingular fast terminal sliding mode control with extended state observer and tracking differentiator for uncertain nonlinear systems Mathematical Problems in Engineering 2014 2014 16 639707 10.1155/2014/639707 MR3275779 Hua C.-C. Wang K. Chen J.-N. You X. Tracking differentiator and extended state observer-based nonsingular fast terminal sliding mode attitude control for a quadrotor Nonlinear Dynamics 2018 94 1 343 354 2-s2.0-85048525757 10.1007/s11071-018-4362-3 Mohammadi A. Tavakoli M. Marquez H. J. Hashemzadeh F. Nonlinear disturbance observer design for robotic manipulators Control Engineering Practice 2013 21 3 253 267 10.1016/j.conengprac.2012.10.008 2-s2.0-84872391084 Huang J. Ri S. Liu L. Wang Y. Kim J. Pak G. Nonlinear disturbance observer-based dynamic surface control of mobile wheeled inverted pendulum IEEE Transactions on Control Systems Technology 2015 23 6 2400 2407 10.1109/tcst.2015.2404897 2-s2.0-84924185183 Bu X. W. Wu X. Y. Chen Y. X. Design of a class of new nonlinear disturbance observers based on tracking differentiators for uncertain dynamic systems International Journal of Control Automation & Systems 2015 13 3 595 602 Shao X. Liu J. Li J. Cao H. Shen C. Zhang X. Augmented nonlinear differentiator design and application to nonlinear uncertain systems ISA Transactions 2017 67 30 46 10.1016/j.isatra.2016.11.011 2-s2.0-85007552795 Marks G. Shtessel Y. Gratt H. Shkolnikov I. Effects of high order sliding mode guidance and observers on hit-to-kill interceptions Proceedings of the AIAA Guidance, Navigation, and Control Conference and Exhibit 2005 San Francisco, Calif, USA American Institute of Aeronautics and Astronautics 10.2514/6.2005-5967 Haijie M. Wei L. Xiaolin F. Nonlinear tracking differentiator design based on hyperbolic tangent Computer Applications 2016 36 s1 305 309