Muhammad Ahsan,Mostafa M. Salah,Ahmed Saeed
Abstract:This paper presents a novel adaptive fast-terminal neuro-sliding mode control (AFTN-SMC) for a two-link robot manipulator with unknown dynamics and external disturbances. The proposed controller is chattering-free and adaptive to the time-varying system uncertainties. Furthermore, the radial base function neural network (RBFNN) is employed to approximate the unknown state dynamics. The simulations have been completed in MATLAB, which illustrates the successful implementation of the proposed controller. The results showcased the effectiveness of the AFTN-SMC in achieving accurate tracking and stability, even in the presence of uncertainties and parameter variations. The incorporation of the RBFNN in the controller proved to be a valuable tool for approximating the unknown dynamics, enabling accurate estimation and control of the manipulator's behavior. The research presented in this paper contributes to the advancement in control techniques for robot manipulators in diverse industrial and automation applications.
engineering, electrical & electronic,computer science, information systems,physics, applied
What problem does this paper attempt to address?
The problem that this paper attempts to solve is to design a chattering - free adaptive fast - terminal neural sliding - mode controller (AFTN - SMC) for a two - link robot manipulator with unknown dynamic characteristics and external disturbances, which can adapt to the time - varying system uncertainties. Specifically, the main problems and challenges of the research include:
1. **Unknown dynamic characteristics**: The dynamic model of the robot manipulator may contain parts that are not fully known or difficult to model accurately. This makes it difficult to directly apply traditional control methods based on accurate models.
2. **External disturbances**: External disturbances existing in the actual environment (such as wind, vibration, etc.) will affect the movement of the robot, resulting in tracking errors and performance degradation.
3. **Chattering problem**: Although the switching gain signal used in the traditional sliding - mode control method can achieve the desired sliding surface, it will produce chattering, which may damage the actuator and reduce the stability of the system.
4. **Adaptability**: In order to deal with the situation where the system parameters change over time, the controller needs to have adaptive ability to adjust the control strategy in real time to ensure the stability and accuracy of the system.
To solve the above problems, the author proposes the following solutions:
- **Adaptive fast - terminal neural sliding - mode control (AFTN - SMC)**: Combining the advantages of fast - terminal sliding - mode control and adaptive control, it can converge in a finite time, and compensate for unknown dynamic characteristics and external disturbances by introducing an adaptive mechanism.
- **Radial basis function neural network (RBFNN)**: Used to approximate the unknown state dynamics, thereby improving the adaptability and robustness of the controller to complex nonlinear systems.
Through these methods, the paper aims to develop an efficient and stable control algorithm, so that the two - link robot manipulator can still achieve accurate trajectory tracking and good stability in the presence of unknown dynamic characteristics and external disturbances.
### Summary of mathematical formulas
1. **Definition of sliding surface**:
\[
s=\dot{e}+\theta
\]
where \(\theta\) is an integral term, and its dynamics are expressed as:
\[
\dot{\theta}=k_{1}|\dot{e}|^{\alpha_{1}}\text{sign}(\dot{e})+k_{2}|e|^{\alpha_{2}}\text{sign}(e)
\]
Here, \(k_{1}, k_{2}\) are positive sliding - mode gains, and \(\alpha_{1}, \alpha_{2}\) are constants satisfying \(0 <\alpha_{1}<1\) and \(0 <\alpha_{2}<1\).
2. **RBFNN output approximation**:
\[
X(q,\dot{q}) = W^{T}\theta(q,\dot{q})+\mu
\]
where \(\mu\) is the approximation error, \(W^{T}\in\mathbb{R}^{n}\) is the output weight vector of the first layer, and \(\theta(q,\dot{q})\) is a vector composed of activation functions.
3. **Control law**:
\[
u = u_{eq}+u_{adp}
\]
where the equivalent control signal \(u_{eq}\) is:
\[
u_{eq}=(Y(q)^{-1})(W^{T}\theta(q,\dot{q})+\ddot{q}_{d}-\dot{\theta})-k_{3}|s|^{\alpha_{3}}\text{sign}(s)
\]
The adaptive control signal \(u_{adp}\) is:
\[
u_{adp}=\left\{\begin{array}{ll}
-\frac{s}{|s|}(\hat{\beta}_{1}+\hat{\beta}_{2}\|q\|+\hat{\beta}_{3}\|\dot{q}\|^{2}) & \text { if }|s|\geq\delta_{0} \\
-\frac{}{}
\end{array}\right.