An inverse Prandtl–Ishlinskii model based decoupling control methodology for a 3-DOF flexure-based mechanism

Guo, Zhiyong; Tian, Yanling; Liu, Xianping; Shirinzadeh, Bijan; Wang, Fujun; Zhang, Dawei

doi:10.1016/j.sna.2015.04.018

Cited by 50 publications

(14 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…[11][12][13] Among these hysteresis models, the P-I model has been widely utilized due to the simple structure, easy modeling identification, and analytical inverse. 14 Ang et al observed that the hysteresis slope of the PEA was affected linearly by the actuation rate, and they accordingly designed a rate-dependent P-I (RDPI) hysteresis slope model. 15 Qin et al established a novel direct inverse modeling approach to directly obtain the inverse P-I model based on experimental data.…”

Section: Article Scitationorg/journal/rsimentioning

confidence: 99%

Modeling and control methodology for an XYZ micro manipulator

Tian

et al. 2019

Review of Scientific Instruments

View full text Add to dashboard Cite

This paper presents the modeling and control methodology for a piezoactuated compliant XYZ manipulator toward precision positioning. The manipulator was fabricated using a wire electrical discharge machining technique, and the system identification was conducted to obtain the dynamic model based on the frequency response. To reduce the effects of hysteresis, creep, and external disturbances, a feedforward/feedback hybrid controller is proposed, which contains a dynamic dependent Prandtl-Ishlinskii (DDPI) hysteresis model and a novel sliding mode controller. The DDPI hysteresis model has excellent modeling accuracy at high operating frequencies with consideration of the dynamic characteristics of the micromanipulator. The novel sliding mode controller integrated with uncertainty and disturbance estimation (SMCUDE) technique is developed, which has the advantages of fast response, strong robustness, and resistance to chattering. The performance of the DDPI hysteresis model and the novel sliding mode controller is validated and compared using experimental tests. The experimental results indicate that the DDPI model provides better positioning accuracy than the traditional Prandtl-Ishlinskii (P-I) model and the rate-dependent P-I model, and furthermore, the SMCUDE controller can improve the response speed without loss of stability, which demonstrates that precision positioning operations can be implemented by the developed manipulator using the proposed control strategy.

show abstract

Section: Article Scitationorg/journal/rsimentioning

confidence: 99%

Modeling and control methodology for an XYZ micro manipulator

Tian

et al. 2019

Review of Scientific Instruments

View full text Add to dashboard Cite

show abstract

“…This method does not require complicated modeling of thermo-mechanical parameters, and the method only considers the output of the actuator corresponding to the applied input. 7,8 Different models have been proposed to characterize the hysteresis and nonlinearities of smart materials at a single-frequency such as Krasnoselskii-Pokrovskii (KP), Preisach, and Prandtl-Ishlinskii (PI) models. 9 Among these models, PI is more popular in model-based control systems because of its simplicity and analytical inverse.…”

Section: Introductionmentioning

confidence: 99%

Modeling and characterization of the shape memory alloy–based morphing wing behavior using proposed rate-dependent Prandtl-Ishlinskii models

Shakiba

Yousefi‐Koma

Jokar

et al. 2019

Proceedings of the Institution of Mechanical Engineers, Part I:

View full text Add to dashboard Cite

Unique features of shape memory alloys make them a proper actuation choice in various control systems. However, their nonlinear hysteresis behavior negatively affects wide utilization of such materials in structure actuation. In this study, the frequency effect on the hysteresis behavior of a shape memory alloy–actuated structure is experimentally investigated, and also two proposed versions of rate-dependent Prandtl-Ishlinskii (modified rate-dependent Prandtl-Ishlinskii and revised modified rate-dependent Prandtl-Ishlinskii) are presented, which are capable of characterizing this phenomenon. Experimental results show that increasing excitation frequency leads to bigger hysteresis loops. It is also proven that rate-dependency cannot be predicted by generalized Prandtl-Ishlinskii model. In addition, a comparison between the dead zone function-based rate-dependent Prandtl-Ishlinskii model as an only benchmark model and the proposed models have been done that proves the proposed models’ superiority. In addition, genetic algorithm is exploited to identify unknown parameters of all models. Trained models performance is also experimentally evaluated at different input frequencies. Comparison between simulation and experimental results indicates that the proposed models can reliably predict saturated, asymmetric, rate-dependent hysteresis behavior, and minor loops in shape memory alloy–embedded actuators.

show abstract

“…Many studies have been conducted on how to improve the accuracy of the nanopositioning stage by improving the traditional PI model. There are ways to change the operator by modifying the classical PI model into a generalized PI model, resetting the initial value that can eliminate the influence of the polarization by changing the hysteresis characteristics of the symmetric bias modeling and piecewise fitting method [20][21][22][23]. The above researches greatly improve the hysteresis description accuracy of the nanopositioning stage for single-ring linear voltage.…”

Section: Introductionmentioning

confidence: 99%

Compensation of Hysteresis in the Piezoelectric Nanopositioning Stage under Reciprocating Linear Voltage Based on a Mark-Segmented PI Model

Yang

et al. 2019

Micromachines

View full text Add to dashboard Cite

The nanopositioning stage with a piezoelectric driver usually compensates for the nonlinear outer-loop hysteresis characteristic of the piezoelectric effect using the Prandtl–Ishlinskii (PI) model under a single-ring linear voltage, but cannot accurately describe the characteristics of the inner-loop hysteresis under the reciprocating linear voltage. In order to improve the accuracy of the nanopositioning, this study designs a nanopositioning stage with a double-parallel guiding mechanism. On the basis of the classical PI model, the study firstly identifies the hysteresis rate tangent slope mark points, then segments and finally proposes a phenomenological model—the mark-segmented Prandtl–Ishlinskii (MSPI) model. The MSPI model, which is fitted together by each segment, can further improve the fitting accuracy of the outer-loop hysteresis nonlinearity, while describing the inner-loop hysteresis nonlinearity perfectly. The experimental results of the inverse model compensation control show that the MSPI model can achieve 99.6% reciprocating linear voltage inner-loop characteristic accuracy. Compared with the classical PI model, the 81.6% accuracy of the hysteresis loop outer loop is improved.

show abstract

An inverse Prandtl–Ishlinskii model based decoupling control methodology for a 3-DOF flexure-based mechanism

Cited by 50 publications

References 30 publications

Modeling and control methodology for an XYZ micro manipulator

Modeling and control methodology for an XYZ micro manipulator

Modeling and characterization of the shape memory alloy–based morphing wing behavior using proposed rate-dependent Prandtl-Ishlinskii models

Compensation of Hysteresis in the Piezoelectric Nanopositioning Stage under Reciprocating Linear Voltage Based on a Mark-Segmented PI Model

Contact Info

Product

Resources

About