Online tuning gain scheduling MIMO neural PID control of the 2-axes pneumatic artificial muscle (PAM) robot arm

Ánh, Hồ Phạm Huy

doi:10.1016/j.eswa.2010.02.131

Cited by 67 publications

(35 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To compare different structures shown in Figure 5, some related analysis functions are deduced as following functions (12) to (15) …”

Section: Feedback Gains Of Imposed Axesmentioning

confidence: 99%

“…There are many new methods and attempts made in the field of MIMO PID controller parameter tuning, such as closed-pole assignment method, 5 dynamics relative iterative method, 13 IMC method, 14 direct synthesis method, 15 loop transfer recovery method, 16 intelligent methods, etc. Gu¨ndesa nd Ozguler 17 gave sufficient conditions for PID stabilizability of MIMO plant.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A new method using pole placement technique to tune multi-axis PID parameter for matched servo dynamics

Gao

Tao

Mei

et al. 2012

Proceedings of the Institution of Mechanical Engineers, Part C:

View full text Add to dashboard Cite

A new method which employs pole placement technique of modern control theory to tune proportional-integralderivative parameter for simultaneously achieving multi-axis matched dynamics and single axis stability is addressed in this article. This method has the advantage of simplicity in tuning, intuition in principle, and easy in program. Four different cascade proportional-integral-derivative control structures are investigated for applying the matching method appropriately and better understanding of its limitation. The deduction that the feedback gain ratios of position to speed loop of the same type interpolation axes imposed with the same desired poles are approximately equal is obtained. Experimental results show that the servo feed system designed or tuned by the proposed method can easily achieve matched dynamics. Furthermore, this method can be also employed to multi-axis motion system which is composed with both linear and rotating axes.

show abstract

“…To compare different structures shown in Figure 5, some related analysis functions are deduced as following functions (12) to (15) …”

Section: Feedback Gains Of Imposed Axesmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

A new method using pole placement technique to tune multi-axis PID parameter for matched servo dynamics

Gao

Tao

Mei

et al. 2012

Proceedings of the Institution of Mechanical Engineers, Part C:

View full text Add to dashboard Cite

show abstract

“…Most of the works in this category are combining linear control methods with higher-level computation, which addresses the nonlinearities through modulation of control parameters, in [12] Ho Pham Huy Anh has proposed an online tuning gain scheduling MIMO neural dynamic DNN PID control architecture to a nonlinear 2-axes pneumatic artificial muscle robot arm, and later in [13] Ho Pham Huy Anh and Kyoung Kwan Ahn investigated the possibility of applying a hybrid feed-forward inverse nonlinear autoregressive with exogenous input (NARX) fuzzy model PID controller to a nonlinear pneumatic artificial muscle robot arm, the inverse NARX fuzzy model was identified by a modified genetic algorithm (MGA) based on input/output training data gathered experimentally from the PAM system. Despite the slow updating of control parameters the above strategies seem giving good performances compared to traditional PID, moreover it is still possible to achieve more performances and robustness by combining non linear control strategies with artificial intelligence (neural network, fuzzy logic, genetic algorithm).…”

Section: Introductionmentioning

confidence: 99%

Intelligent twisting sliding mode controller using neural network for pneumatic artificial muscles robot arm

Boudoua

Hamerlain

2015

2015 International Workshop on Recent Advances in Sliding Modes (RASM)

View full text Add to dashboard Cite

In this note we present a novel intelligent twisting sliding mode controller using neural network, achieving chatter reduction for the control of pneumatic artificial muscles robot arm. The system is highly non-linear and somehow difficult to model therefore resorting to robust control is required. Thanks to their property as universal approximators, in this work a two layer NN with on line adaptive learning law is used to reconstruct unknown and unmodeled robot dynamics, and the realisation of a two sliding mode is achieved through the design of a nonlinear sliding surface. The stability of the overall system is guaranteed by Iyapunov method. Experimental results are presented and discussed.978-1-4799-8947-8/15/$31.00 ©2015 IEEE

show abstract

“…al., 1992, Nauck andKruse 1993]. On the other hand, according to the fuzzy model adopted, there are two types of fuzzy models that can be integrated with a neural network to form a FNN [Anh, 2010]. These two models are the TS-model [Takagi and Sugeno, 1985] and the Mamdani-model [Lee, 1990a and 1990b].…”

mentioning

confidence: 99%

Adaptive Functional-Based Neuro-Fuzzy-PID Incremental Controller Structure

Fahmy

Ghany

2014

IJEEI

View full text Add to dashboard Cite

This paper presents an adaptive functional-based Neuro-fuzzy-PID incremental (NFPID) controller structure that can be tuned either offline or online according to required controller performance. First, differential membership functions are used to represent the fuzzy membership functions of the inputoutput space of the three term controller. Second, controller rules are generated based on the discrete proportional, derivative, and integral function for the fuzzy space. Finally, a fully differentiable fuzzy neural network is constructed to represent the developed controller for either offline or online controller parameter adaptation. Two different adaptation methods are used for controller tuning, offline method based on controller transient performance cost function optimization using Bees Algorithm, and online method based on tracking error minimization using back-propagation with momentum algorithm. The proposed control system was tested to show the validity of the controller structure over a fixed PID controller gains to control SCARA type robot arm. 25Function-based FPID controllers (F-FPID) [Tang et. al., 2001]. In this paper, the controller functionally performs fuzzy derivative and fuzzy integral functions, so that no calculations are required outside the FLC. The proposed controller employs only two inputs (present and previous errors), so that the design procedure is simpler. Additionally, most fuzzy logic based PID controllers in literature adapt the triangular membership function shape for simplicity of implementation, despite it's linear nature. In this paper, other types of membership functions, such as Gaussian, and sigmoid membership functions are utilized in the design of the controller to allow tuning for the controller membership functions as well as online adaptation of the controller based on the Back-propagation with momentum (BP) learning algorithm.Since the early 1990s, FNN have attracted a great deal of interest because such systems are more flexible and transparent than either NN or FLS alone. Different types of FNN have been presented in the literature [Shing and Jang, 1993]. For tracking control where the reference and dynamics is always changing, the inverse dynamic control is the best to be utilized, despite very difficult to be implemented mathematically. Consequently, researcher tends to use neural network and neuro-fuzzy systems to avoid complex mathematical formulation [Sinthipsomboon et. al., 2011]. In [Anh and Pham, 2010] a gain-scheduling neural PID controller is utilized with 2-axis robotic structure for varying the parameters of the neural PID controller to include information from the robot dynamics. The FNN types can be identified based on the structure of the FNN, the fuzzy model employed and the learning algorithm adopted [Ahn and Anh, 2009]. On the other hand, the most commonly used and successful approach is the feed-forward and recurrent structure model, while using the BP as the learning algorithm [Yuan et. al., 1992, Nauck andKruse 1993]. On the other hand, accord...

show abstract

Online tuning gain scheduling MIMO neural PID control of the 2-axes pneumatic artificial muscle (PAM) robot arm

Cited by 67 publications

References 7 publications

A new method using pole placement technique to tune multi-axis PID parameter for matched servo dynamics

A new method using pole placement technique to tune multi-axis PID parameter for matched servo dynamics

Intelligent twisting sliding mode controller using neural network for pneumatic artificial muscles robot arm

Adaptive Functional-Based Neuro-Fuzzy-PID Incremental Controller Structure

Contact Info

Product

Resources

About