Research On Sensorless Control Strategy Of Load Commutated Inverter-fed Synchronous Motor

Due to the advantages of large capacity,low cost,simple control and high reliability,load commutated inverter(LCI)-fed synchronous motor control system is widely used in pumped storage unit,synchronous compensator,gas turbine,blast furnace blower,finishing mill main driving system and other important high power drive application.Among them,the finishing mill main driving system has the highest control requirements for LCI-fed synchronous motor system because of its step load characteristic and large load requirement.However,domestic related industries mostly adopt the products from foreign companies such as Siemens,ABB and General Electric,therefore,the research on this driving system localization is of great practical significance.Taking the finishing mill main driving system as an example,an in-depth study on sensorless control and optimized on-load control of LCI-fed synchronous motor control system is done in this paper,and the main research contents are as follows:The rotor position estimation method for synchronous motor at standstill and low speed is studied.For the synchronous motor drive system,real-time rotor position information is the basic condition of reliable operation.The installation and use of traditional mechanical sensors are greatly limited in many situations,and it also has low reliability.To solve this problem,a rotor position estimation methods based on field current harmonic response at standstill and low speed is proposed in this paper,and the phase shift caused by discrete sampling is corrected by the envelope prediction algorithm,improving the estimation accuracy of rotor position.To improve the reliability and accuracy of rotor position estimation further,in this paper,a signal processing method which directly utilizes the field current harmonics to assist the rotor position information extraction is proposed.Since the LCI drive system cannot superimpose harmonic signal with appropriate frequency for rotor position estimation,the topological structure and control method of LCI driving system are analyzed in this paper,then,an improved excitation control method is proposed.Furthermore,the filters and phase-locked loop used in the rotor position estimation methods are analyzed and designed,and the rotor position estimation error is analyzed and compensated,so as to further improve the estimation accuracy of rotor position estimation methods at standstill and low speed.The speed estimation method at medium and high speed load commutation control and the vector control strategy for inverter control are studied.In order to improve the control accuracy of the LCI in the load commutation mode,the real-time three-phase stator current information is essential,however,adding current sensors in high-power drive system will significantly increase the system cost.To solve this problem,a three-phase stator current reconstruction method is proposed in this paper,and this method can realize the real-time three-phase stator current estimation only by detecting the DC-link current.In order to achieve reliable commutation and sensorless control under medium and high speed operation,a BEMF phase and speed continuous estimation method based on stator current reconstruction at medium and high speed is proposed,furthermore,the sensorless speed estimation in full operating range is realized.Then,based on the BEMF phase continuous estimation method,a vector control strategy based on the minimum remainder commutation angle is proposed,which realizes the decoupling control between rectifier and inverter,and can improve the load capacity of drive system on the basis of ensuring the LCI reliable commutation.The model predictive control strategy suitable for the system with step load characteristic is studied.The load characteristic of finishing mill main driving system is analyzed in this paper.Then,taking the finishing mill main driving system with step load characteristic as an example,the limitations on improving torque transient characteristics of traditional double closed-loop control and feedforward control under step load disturbance are analyzed and explained.In view of the shortcomings of the existing control methods in LCI-fed synchronous motor system,such as poor dynamic response and high commissioning cost,a sensorless model predictive control method which is matched with the minimum remainder commutation angle control is proposed in this paper.In this method,the mathematical model of controlled object is established which combines the LCI direct current control circuit with the motor motion equation.This system mathematical model can reduce the order of the model and ensure that the coefficients of the established state space model are constant,thus the operation efficiency of model predictive control is greatly improved.According to the engineering application characteristics of finishing mill system,an appropriate constraint control optimization method is presented,and the system stability is analyzed.The sensorless control strategy of LCI-fed synchronous motor proposed in this paper is verified.A 3.75 kW LCI-fed synchronous motor on-load test platform is built in this paper,and the hardware design scheme and software control flow of this experimental platform are also given.Taking the engineering application characteristics and requirements of finishing mill as an example,the corresponding experiment scheme is designed and tested in this paper,which contains the rotor position estimation method at standstill and low speed,the BEMF phase and speed estimation method at medium and high speed,the minimum remainder commutation angle control strategy and the sensorless model predictive control strategy.Compared with the typical control strategy of finishing mill control system,the LCI-fed synchronous motor sensorless control strategy proposed in this paper is verified to have better load capacity and dynamic response characteristic.