Strict feedback system neural network control method based on lumped composite estimation
A neural network control and feedback system technology, which is applied in the field of strict feedback system neural network control based on lumped composite estimation, and can solve the problems of unknown dynamics and time-varying disturbances of uncertain feedback systems.
Active Publication Date: 2018-12-21
NORTHWESTERN POLYTECHNICAL UNIV
View PDF12 Cites 6 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0005] In order to solve the problem of unknown dynamics and time-varying disturbances in uncertain feedback systems with external disturbances, the present invention proposes a neural network control method for strict feedback systems based on lumped composite estimation. This method is based on the framework of backstepping method and adopts The neural network is used to estimate the nonlinear function of the system, and the disturbance observer is used to process the compound disturbance formed by the neural network approximation error and the time-varying disturbance, and then the aggregate prediction error is formed based on the two estimated information, and applied to the neural network and the disturbance In the update law of the observer, the control input is finally fed forward into the dynamic model of the system
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment
[0113] refer to figure 1 , the neural network control of strict feedback system based on lumped composite estimation in the present invention is applied to a third-order strict feedback system, and is realized by the following steps:
[0114] (a) Consider a third-order strict feedback system dynamics model:
[0115]
[0116] (b) According to formula (1), define the tracking error as e 1 =x 1 -y r , where y r =sin(t) represents a reference signal.
[0117] Step 1: Design virtual control volume for
[0118]
[0119] in Represents the estimated value of the optimal weight of the neural network, Represents the neural network basis function vector, Denotes the derivative of the reference signal, Denotes the estimated value of the compound disturbance, k 1 = 3, L f1 =1.
[0120] Design a first-order filter as
[0121]
[0122] where τ 2 =0.05 is the filter parameter.
[0123] Design compensation signal z 1 for
[0124]
[0125] where z 2 given in ...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The invention relates to a strict feedback system neural network control method based on lumped composite estimation, belongs to the field of intelligent control methods, and is used for solving the problem of unknown dynamic states and time-varying disturbance in an uncertain feedback system with external interference. The method is based on a frame of an inverse step method, and a neural networkis adopted to estimate a nonlinear function of the system; a disturbance observer is used for processing the composite disturbance formed by the neural network approximation error and the time-varying disturbance; by utilizing the neural network estimation and disturbance observation estimation of the online data, the lumped composite estimation error is constructed, and an updating law of the neural network and the disturbance observer is designed; and finally input is designed and controlled based on neural network estimation and disturbance observation estimation. According to the method,neural network learning and disturbance observation are organically combined, so that an effective way for processing the strict feedback system control with time-varying interference is provided.
Description
technical field [0001] The invention relates to a neural network control method, in particular to a neural network control method for a strict feedback system based on lumped composite estimation, and belongs to the field of intelligent control methods. Background technique [0002] Backstepping control is widely used in strict feedback systems, but there is a "complexity explosion" problem in traditional design, so dynamic surface design and command filtering design are applied to reduce design complexity. For the design of nonlinear strict feedback systems, due to the existence of nonlinearity, intelligent control technology has been widely concerned. At present, most existing intelligent control researches update weights based on tracking error, which only guarantees the stability of the closed-loop system, and it is difficult to achieve the expected nonlinear estimation effect. [0003] "Composite Learning Control of MIMO Systems With Applications" (B Xu, YShou, "IEEE T...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More Patent Type & Authority Applications(China)
IPC IPC(8): G05B13/04
Inventor 许斌寿莹鑫
Owner NORTHWESTERN POLYTECHNICAL UNIV