| The main target the wind turbine multi-terminal suspension system is to achieve stable axial suspension without pitch.However,the existence of magnetic coupling between the multi-terminal windings,the integrated design of the turbine and suspension windings,has resulted in severe coupling between the input and output.In particular,the structural parameter differences of the multi-terminal suspension converters and the existence of multi degree of freedom(DOF)disturbances caused by time-varying wind speeds and directions make it extremely challenging for the turbine to achieve stable suspension without pitch.Therefore,this paper takes the multi-terminal suspension system as the research object,and takes the multi-terminal decoupling as the breakthrough,and uses the model reference adaptive,intelligent approximation technology,adaptive control and finite-time control technology to comprehensively improve the transient performance and robust performance against strong interference and parameter time-change.Aiming at the problem that the single winding has magnetic field line distortion and the electromagnetic attraction changes with the horizontal position,a two DOF single-winding dynamic model considering magnetic field line distortion is constructed.Considering the single-input control characteristics,the radial part of the model is introduced into the axial suspension dynamic model as disturbance through the decoupling method of model transformation,and the RBF neural network with intelligent approximation ability is used to decouple by online estimation.Since the weak damping,nonlinearity and multi DOF interference characteristics,the MRAC strategy is introduced into the single-terminal suspension control,the convergence time and steady-state performance are comprehensively considered to construct a reference model,and the reference model is constructed by considering the convergence time and steady-state performance.Combining the model reference matching principle and the RBF intelligent approximation principle,an MRAC adaptive control law is designed to drive the system to approach the reference model.Then,a Lyapunov function containing tracking and estimation errors as well as adaptive parameter approximation errors is constructed,verifying the asymptotic convergence of the above errors.The experiment shows that the proposed strategy has obvious advantages in tracking performance and anti-interference,with the maximum performance improved by 55%/60% respectively.Aiming at the strong coupling and nonlinear characteristics of multi-input and multi-output multi-terminal suspension system,comprehensively considering the multi DOF motion mechanism and interference characteristics of the turbine,the suspension winding is reasonably divided into four parts,controlled by four independent suspension controllers,and the four-terminal multi DOF suspension dynamic model of the wind nacelle is constructed.Then,based on the state transition matrix,the multi DOF model is transformed into a multi-terminal suspension control model based on dominant terms,uncertain terms and coupling terms.Therefore,a main tracking controller based on tracking error state feedback is designed.Based on the multi-terminal suspension air gap,a multi-terminal suspension synchronous adaptive controller is constructed for online fast compensation of coupling terms and synchronous interference.The RBF neural network intelligently approximates the uncertain terms,while the transient performance improvement of the multi-parameter adaptive convergence approximation process is realized by the constructed robust control terms,which synergistically solves the influence of strong coupling and nonlinear strong interference on the stable performance of the system.Based on the multi-terminal suspension simulation platform,it is verified that the proposed strategy realizes multi-terminal suspension tracking and synchronous control,and the anti-pitch interference experiment verifies that the strategy returns to steady state within 0.632 s,and the maximum synchronization error is only 0.097 mm.Since the above two strategies can achieve suspension control and achieve stable suspension with disturbance suppression.However,the turbine suspension conditions are harsh,and real-time transient performance urgently needs to be improved.The above two strategies are asymptotic convergence,which seriously restricts the transient convergence speed,so this paper takes the two-terminal suspension system as the research object,and aiming at the problem that the traditional model reference model cannot dynamically adjust the real-time working condition changes,a MRAC strategy based on the synchronization error of closed-loop feedback information is proposed.Based on the finite-time control technology,the MRAC controller is designed,especially the fractional power function with approximation error and synchronization error is introduced,combined with the intelligent approximation of RBF neural network to obtain the unknown model and the online approximation acquisition of external interference,to ensure the independent control and transient performance.Then,a Lyapunov function with model approximation error,synchronization error and parameter approximation error are constructed,and it is verified that the above variables can achieve finite time convergence,and the convergence domain can be optimized and adjusted by controller parameters.The simulation verifies that the proposed strategy has significant advantages in transient convergence speed and anti-interference ability,and the maximum fluctuation of air gap after interference application is only 0.019 mm,the synchronization error is0.046 mm,and the recovery time is 0.028 s.Based on the multi-terminal suspension system experimental platform,the Experimental studies on multi reference air gap tracking,pitch suppression,and axial interference application are carried out,and the experimental performance is compared with PID control and adaptive control strategy.The experiment results show the proposed strategy has obvious advantages in transient performance improvement and interference suppression.In the axial interference experiment,the system recovery time,synchronization error and maximum air gap fluctuation value are increased by 63%/57%/75%,respectively.In the pitch interference experiment,the system recovery time,synchronization error and maximum air gap fluctuation value are increased by 87%/43%/52%,respectively.These results verify that the proposed strategy has excellent tracking performance,powerful robust performance and synchronization performance. |
PDF Full Text Request