基于模型参考自适应的感应电机温度辨识方法研究

发布时间：2018-10-13 09:37

【摘要】：由于一些感应电机需要运行在复杂工况下(例如潜油电机),由于这一类复杂工况下工作的感应电机温度不易直接测量,其电机内腔温度也不易散发。此时电机温升情况关系到电机长期稳定运行的安全性,这就需对复杂工况下运行的感应电机温度进行状态监测。尤其在电机负载突变的瞬间,其电流会发生突变,但电机的温度却不会发生突变。因此,本文针对以上问题进行了研究,将电机定转子电阻参数辨识方法与电机等效热模型进行结合,以实现对电机定转子侧温度的在线辨识。将模型参考自适应理论应用于电机定转子电阻的参数辨识当中,根据两相静止坐标系下的感应电机数学模型进行推导,可到得到等效转子磁链的电压与电流模型。通过等效转子磁链的电压与电流模型可以确定定子和转子电阻辨识方案的参考模型与可调模型,所做工作可为定转子电阻各自自适应率的设计提供理论依据。为了保证定转子电阻辨识系统的稳定性,将超稳定不等式应用于定转子电阻辨识系统误差方程,得到定转子电阻对应的各自自适应率。由于电机温度变化对定转子电阻的影响,利用电机温度与金属电阻的近似线性关系对自适应率中的电阻阻值进行了修正,得到了跟随电机温度变化的定转子电阻辨识方案。通过建模仿真可验证所建立模型的稳定性。由于电机的温升是因为其所产生的损耗引起的,跟据感应电机结构得到感应电机两热源的等效热模型。在此基础上将感应电机散热系统对电机温度的影响因素进行考虑,得到修正后的感应电机等效热模型,利用该模型可以实现对感应电机定转子侧温度的实时辨识。改进了感应电机温度辨识系统的平台,完善了算法平台。然后进行了恒负载转矩,变负载转矩,变压变负载转矩以及变频变负载转矩的实验以模拟实际电机运行的复杂工况。通过对不同实验条件下的实验辨识温度与实测温度的对比分析,二者间的最大误差小于8℃,且负载突变瞬间温度辨识结果无突变,体现了算法模型的具有良好的动态跟随性,同时也验证了本文所提出的电机温度辨识方法的可行性与正确性。
[Abstract]:Since some induction motors need to operate under complex operating conditions (such as submersible motors), the temperature of induction motors operating under this kind of complex working conditions is not easy to be measured directly, and the temperature of the inner cavity of the motors is not easy to emit. In this case, the temperature rise of the motor is related to the safety of the motor running steadily for a long time, so it is necessary to monitor the temperature of the induction motor operating under complex conditions. Especially in the moment of the sudden change of the motor load, the electric current will change, but the temperature of the motor will not change. Therefore, this paper studies the above problems and combines the identification method of resistance parameters of stator and rotor with the equivalent heat model of motor to realize the on-line identification of stator and rotor side temperature of motor. The model reference adaptive theory is applied to the parameter identification of the stator and rotor resistance of the motor. According to the mathematical model of the induction motor in the two-phase stationary coordinate system, the voltage and current model of the equivalent rotor flux can be obtained. The reference model and adjustable model of stator and rotor resistance identification scheme can be determined by the voltage and current model of equivalent rotor flux. The work can provide a theoretical basis for the design of self-adaptive rate of stator and rotor resistance. In order to ensure the stability of the stator and rotor resistance identification system, the superstability inequality is applied to the error equation of the stator and rotor resistance identification system, and the corresponding adaptive rates of the stator and rotor resistance are obtained. Because of the influence of the temperature change of the motor on the stator and rotor resistance, the resistance in the adaptive rate is modified by using the approximate linear relationship between the temperature of the motor and the metal resistance, and the identification scheme of the stator and rotor resistance following the variation of the temperature of the motor is obtained. The stability of the established model can be verified by modeling and simulation. Due to the loss caused by the temperature rise of the motor, the equivalent heat model of the two heat sources of the induction motor is obtained according to the structure of the induction motor. On the basis of this, the influence factors of induction motor heat dissipation system on motor temperature are considered, and the modified equivalent heat model of induction motor is obtained, which can be used to realize the real-time identification of stator and rotor side temperature of induction motor. The platform of induction motor temperature identification system and the algorithm platform are improved. Then the experiments of constant load torque, variable voltage load torque and variable frequency load torque are carried out to simulate the complex operating conditions of the actual motor. Through the comparison and analysis of the experimental identification temperature and the measured temperature under different experimental conditions, the maximum error between them is less than 8 鈩，

本文编号：2268134