In order to solve the problem of lacking velocity feedback information in the velocity closed-loop control system for linear induction motors (LIM) after cancelling the velocity sensor,a velocity observer based on the extended Kalman filtering algorithm has been implemented,considering the end edge effect of LIM. Firstly,based on the mathematical model of three-phase linear induction motor considering edge effect,an extended Kalman filter observer with appropriate gain and covariance update matrix is derived. Based on the vector control system of LIM,the speed parameters identified by the observer are fed back to the speed closed-loop system. Then,the vector control system model of linear induction motor with speed observer is built in Simulink,and the identification speed of observer is compared with the actual speed of motor. Finally,the results show that the closed-loop control using the identification speed can ensure the stable operation of the system. Under three kinds of loads,the error between the predicted speed and the actual speed of the observer is form 0.51% to 2.34%. The various dynamic performance of the system reveals that the LIM vector control system based on the extended Kalman filter observer increases prediction velocity error and decreases thrust error with increasing load. However,the magnetic flux amplitude error slightly increases. Therefore,considering the end-edge effect,the observer based on the extended Kalman filter can replace the velocity sensor to achieve three-phase LIM control under both unloaded and loaded conditions.