Research On Adaptive Tracking Control Of A Climbing Robot For Wooden Columns

There are many timber architectures of great antiquity in the world,such that the interior of wood components is inevitably damaged by worms,decay,and so on.Therefore,it is necessary to detect the health condition for these buildings.In order to solve the defects of the existing methods,such as low detection accuracy,complicated operation,damage to wood components and difficulty in three-dimensional imaging,the paper introduces a climbing robot for wooden columns by combining CT imaging technique,which has the advantages of high climbing accuracy,large load capacity and high reliability.The robot with the structure of the ambient main,ancillary and rotating bodies,can transport the CT scanner to the desired position accurately for CT scanning,and then reconstruct the three-dimensional image of the internal structure for wooden columns.Consequently,a new technique is provided for the health detection of timber architectures.Furthermore,in view of the nonlinearity of the climbing robot,the research of nonlinear adaptive control is achieved for the control objectives of accurate climbing and high reliability.Firstly,aiming at the control input with deadzone and saturation of the motors on the climbing robot,the adaptive fuzzy funnel control is studied for nonlinear systems with input deadzone and saturation via fuzzy logic systems and funnel control,which can guarantee the precise climbing of the robot.Then,based on prescribed performance control,an improved infinite-time and finite-time adaptive adjustable prescribed performance tracking control schemes are presented to ensure the safety of the robot during the climbing process.Moreover,due to the wireless network transmission of the signals for the climbing robot,a novel event-triggered adaptive tracking controller is developed for nonlinear systems with unknown virtual control coefficients,which can reduce the data exchange of communication without degenerating the control performance.Finally,taking the unknown virtual control coefficient caused by the change of robot load as the research object,an adaptive tracking control strategy is proposed for nonlinear systems with unknown input nonlinearities and virtual control coefficients,which can eliminate the ill-defined problem of controllers caused by too large amplitudes of virtual control coefficients in applications.The above control methods are not only proved theoretically for their stability,but also applied to a general nonlinear system and the rotator of the climbing robot,respectively.This paper consists of seven chapters and the outline is stated as:In Chapter 1:For the health detection of timber architectures,the necessity of developing the climbing robot with CT scanner for wooden columns is discussed,which is followed by the summary of climbing robots at home and abroad.Besides,according to the problems for control of the climbing robot,the research of related control theories is reviewed.In Chapter 2:The structure of the climbing robot with CT scanner is described,the control system and problems to be solved are presented,which not only provide the research object for the following nonlinear control methods,but also construct the platform for practical applications.In Chapter 3:The problem of adaptive fuzzy funnel control is addressed for nonlinear systems with unknown input deadzone and saturation.For the first time,a smooth function is introduced to approximate the unknown input nonlinearity with deadzone and saturation,simultaneously.By utilizing fuzzy logic systems and funnel control,an adaptive fuzzy funnel controller is designed,which handles the control inputs with deadzone and saturation of the driving motors and guarantees the climbing precision of the robot.In Chapter 4:The adaptive prescribed performance tracking control is researched for uncertain nonlinear systems.An infinite-time and a finite-time adjustable performance functions are given,whose parameters can be adjusted in real time according to the change of tracking errors.For uncertain strict-feedback nonlinear systems with zero dynamics,two adaptive adjustable prescribed performance tracking controllers are defined,which ensures the climbing reliability of the robot.In Chapter 5:An event-triggered adaptive tracking control approach is proposed for nonlinear systems with uncertain visual control coefficients.Instead of Nussbaum gain technology,fuzzy logic or neural network,a new method is introduced to deal with unknown visual control coefficients.Furthermore,a novel event-triggered control strategy is developed and an event-triggered adaptive tracking controller is constructed,which can mitigate the waste of wireless network resources effectively.In Chapter 6:The problem of adaptive tracking control is concerned for nonlinear systems with unknown input nonlinearity and virtual control coefficients,which are transformed into known nonzero smooth functions and unknown bounded functions.By virtue of the smooth function introduced to approximate input nonlinearity in Chapter3 and combining Nussbaum gain technology and an auxiliary variable,a new adaptive tracking controller is developed,which can be applied to robot systems with large ranges of virtual control coefficients.In Chapter 7:The conclusion and further remarks are presented,respectively.