Research On Direct Torque Control For Wheel Hub Asynchronous Motor

With its high efficiency and large carrying capacity, mine electric wheel plays an important role in the construction of large mines and super large projects. Along with people pursuit of economic efficiency and the development of science & technology, the performance of electric wheel dump truck has also been gradually increase, and the proportion in the total transportation volume is also growing.The core tasks for the electric wheel dump truck control is effective and reliable VVVF for the hub type asynchronous motor. While high performance variable frequency speed regulation need to use a lot of subject knowledge, such as automation, embedded principle, power electronic technology and so on. First this paper introduces the development status of the domestic and foreign of electric wheel, and draw out the research object—three-phase AC asynchronous motor. Then, introduced the concept of space vector and two kinds of coordinate transformation, and established the mathematical model of asynchronous motor in different coordinate system. After comparing the main control scheme, the direct torque control is accepted.A regular hexagon stator flux can be obtained by the traditional direct torque control system with PWM, which causes the torque ripple at low speed and the IGBT switching frequency is not fixed. If this scheme was directly applied to the system control of electric wheel,will cause the electric wheel slippery and noisy when climbing. So, this paper take advantage of the SVPWM and set up a DTC—SVPWM model which speed and flux linkage are both closed loop. Stator flux observer with low speed in direct torque control will be difficult to accurately track the stator flux due to the variation of stator resistance, which will result in poor performance of the system. In order to solve this problem, two different solutions was designed in this paper. Scheme one: design the stator resistance identification method based on DRNN. Method: first deduced a speed observer which only influenced by stator resistance, when the stator resistance changes, the speed of observer measured is not equal to actual speed of the motor. DRNN was used to identify stator resistance from the error, and then synchronously compensate the stator resistance on stator flux observer. Scheme two: using the concept of stator flux inverse model which was proposed by our research group. RBF neural network was used to construct the flux inverse model of stator flux linkage, and a new neural network closed loop flux observer was designed. In this scheme, the model reference adaptive control of stator flux is carried out based on the actual induction motor as reference. Simulation results show the expected effect, and the low speed performance of the system is improved, which shows that the above scheme is effective. Compared with the scheme one can only overcome the adverse effects of stator resistance variation on the stator flux, and the scheme two can overcome the adverse effects of the other parameters on the flux linkage.In the experimental section, an 15 kw asynchronous motor speed control experiment platform,which was built with the group members was introduced, and introduced the design and construction of the experimental system from two aspects of hardware and software respectively. Finally, the correctness of the DTC—SVPWM algorithm is verified on the experimental platform and some experimental results are given at the same time.