| As is well known,induction motors are frequently used in industrial drives nowadays with numerous advantages such as small volume,light weight,low price,slight rotational inertia and simple maintenance requirements.In high-performance AC drivessystem,speed closed-loop is an absolutely indispensable part.In a general sense,speed is examined and measured by speed sensors.Nevertheless,here come a great many inevitable side effects,like higher precision of installation technology,increased hardware cost,decreased reliability,and less applying occasions.Facing to these issues,speed sensorless control technology gradually returns to the stage.In this paper,further researches are mainly focused on two pivotal techniques in sensorless induction motor drives control system,thus,rotor flux and velocity observation.First,the rotor flux observers are studied.In accordance with the rotor flux oriented vector control system,accuracy of observation plays a dominant role to ensure ensure the success of rotor flux orientation of vector control system.Then two different kinds of rotor flux observers,respectively developed from voltage model and current model,are introduced in the following part.In order to analyze the effect of motor parameters on rotor flux observer,two useful methods are derived to compare the ratio of observed rotor flux to the actual one,and the sensitivity of the two models of motor parameters is further discussed.A hybrid rotor flux observer model combined with the advantages of two models above,is deduced to realize full range of accurate rotor flux observation.With a PI controller,which acts as a transition link,this hybrid model can perform well in switching from current model to voltage model.Also,the trapezoidal method is chosen to achieve the discretization of the hybrid model of rotor flux observer,laying the foundation for establishing the digital induction motor sensorless vector control system.Secondly,researches are on speed estimation method.However,the full order observer can not keep steady at low speeds.So a novel way to convert the stability problem of the speed identification system to that of the poles and zeros is brought out in the observed rotor flux oriented coordinate system.The unstable region under the low-speed regenerative state is obtained by applying the Routh-Hurwitz criterion,and an optimized approach to design feedback gain matrix to minimize the instability region is given.In the meanwhile,in order to lower the bad influence of parameters and better the precision of speed identification,this paper presents a method on fuzzy speed adaptive law on the basis of self correction factor by examining and explaining the sensitivity of rotor and stator of speed identification.Further more,a T-S fuzzy state observer is raised,to a greater extent,to realize the stronger robustness of control system.In consideration of the motion equations of induction motor,a five-order T-S fuzzy state observer is produced with the method of sector partition.Because the feedback gain matrix is a key part to guarantee the stability of the T-S fuzzy state observer model,Lyapunov theory and D-stability analysis method are used to transform the stability problem into a series of linear matrix inequalities(LMI),and LMI toolbox in Matlab is used to get the feedback gain matrix of each subsystem.In conclusion,the sensorless vector control systems of induction motor separately based on voltage and current hybrid model flux rotor observer,fuzzy adaptive full-order flux rotor observer and T-S fuzzy state observer are designed in Matlab/Simulink and TMS320F28055 motor controlling platform,and a large number of simulation and experiments are done to analyze the performance of control system.Corresponding researches and results validate the correctness and effectiveness of the theoretical analysis. |
PDF Full Text Request