Research On Speed Sensor-less DTC Control System For Submersible Motor

As blood of modern industry, oil acts as a very important role in the modernization of the country and our normal life. With the expanding exploitation of underground crude oil, the formation energy can not keep the oil well self-spray. Thus, to meet the economic and efficient command required by the exploitation of the underground oil, Submersible pump develops sophisticated and matured as a way of artificial oil exploitation. Apparently it is very important to make submersible motor work efficiently and stably while submersible motor works as a core component of the submersible pump system.The thesis mainly studies direct torque control of submersible motor which has no speed sensor. Direct Torque Control is a variable voltage variable frequency speed control technology with high performance developed after Vector control. The method does not require complex calculations about motor rotor, and it has good dynamic and static performance, good robustness. As submersible motor is mostly a kind of three-phase squirrel-cage asynchronous motor, it is difficult to detect rotor parameters, so it is suitable to apply direct torque control. The thesis mainly studies the double closed loop of direct torque control which are speed closed loop and closed loop of torque and flux.It is difficult to use traditional tachometer generator or to install speed sensor to monitor the speed due to the limitation of the working environment. So the thesis takes a speed identification scheme based on speed sensor-less case for submersible motor. Also, the parameters is easy to change when submersible motor works, so neural network is used to train different models corresponding to different operating phases respectively. Then the models can be used to conduct ’on-line identification’ for different operating phases, alleviating the problem of falling precision in the process of identification caused by changing parameter in different operating phases.Another closed loop is the torque and flux, accurate observation of flux is the prerequisite for calculation of torque, traditional method of flux observation is integral part, However,the existence of DC offset and initial phase result in the inaccuracy of flux. The thesis improves it, weakening the effects of the two above to some extent. In addition, as submersible motor works in high temperature, changes of stator resistance with temperature will also result in inaccuracy of flux observation. The thesis takes temperature into flux observation, eliminating errors caused by stator resistance changes. Finally, the speed identification model and the improved flux observer are applied to the direct torque control simulation model; the simulation results show that the methods above are effective.