Research On Several Key Technologies For Sensorless Control Of Induction Motor Drives

As the core of mechanical energy and electric energy conversion, electric drive system is widely used in industrial and agricultural production, national defense and social life. In the environmental pollution and energy crisis, the application of adjustable speed drives (ASD) have important significances to improve the production and living standards, promote energy saving and emission reduction, and realize sustainable development. Based on literature review and previous works, the subject of this paper is on the speed sensorless control of ASD with the advantages of low cost and high reliability. The paper is focus on the V/f control and direct rotor flux orientation vector control (DRFOC) under adaptive full-order observer (AFO), does lots of relative theoretical analysis, simulation and experimental research, proposes some innovative opinions and solutions.In order to improve the performance of the V/f control induction motor drives (IMD), the researches mainly focus on improving the loading capacity, steady-state accuracy and oscillation suppression. First, the steady-state model and small-signal model of the V/f control IMD are developed to analyze the stability and convergence rate of the system. Based on the models, the influences of various induction motor (IM) parameters on the IMD’s stability are analyzed. Then the paper proposes a method to suppress oscillation with wide application range and small parameter sensitivity. Second, based on vector theory, stator resistance compensation is added to keep a constant steady-state stator flux in d axis, contributing to boost the loading capacity of the drive system in low speed. And the speed accuracy is improved by using the precise slip frequency compensation. Besides, the steady-state and small-signal models of the proposed V/f control IMD are developed, and the influences of IM parameters on the system’s stability and accuracy are analyzed.In order to design high performance of sensorless IMD under DRFOC with AFO, the researches mainly focus on controller and observer. First, the precise control models of current, flux and speed loop are developed, and the corresponding discrete transfer functions are derived. Then, based on sampling frequency, expected bandwidth and motor parameters, a method is proposed for designing the proportional-integral (PI) parameters used in DRFOC regulators. The proposed method is suitable for various levels of IM, with wide applicability. Second, the small-signal models of the flux and speed estimation under AFO are developed by using the method of perturbation and linearization, and used to analyze the stability and convergence rate of the two estimations.The state equation of the small-signal model can accurately judge the unstable range of the AFO in low-speed regenerative mode. Third, under the assumption that the motor parameters are accurate and the stator frequency isn’t used in feedback gains, the necessary and sufficient conditions for complete stability speed estimation system are derived, and kinds of feedback gains are designed based on the conditions. In designing the feedbacks, a new strategy is proposed that adops different sets of feedback gains in different operating regions to achieve a better performance. Forth, a method is developed to design adaptive gains changes with IM’s operating state, to further improve the dynamic performance of the speed estimation.In order to design the stable and high accuracy sensorless IMD under parameter mismatches even in low-speed regenerative mode. First, the steady-state model of sensorless IMD with AFO is developed under mismatched motor parameters. Based on the steady-state model, a new unified parameter sensitive equation is proposed which is suitable for sensorless IMD with AFO. In IM’s operating area, the equation can directly calculate the estimated rotor-speed error caused by any parameter mismatches with arbitrary feedback gains. Second, a feasible condition is proposed for judging the drive system’s stability when parameter mismatches exist. Based on the condition, the conclusion can be drawn that the sensorless IMD with AFO easily involves in instability in low-speed regenerative mode with parameter mismatches. Third, a stategy is proposed to use the robust constant current variable frequency (CCVF) control method to replace the DRFOC in unstable circumstances, inorder to guarantee the stability of the IMD in full operating area. The paper proposes a precise slip frequency compensation to improve the steady-state accuracy of IMD under CCVF method. Besides, the corresponding steady-state model is developed to analyze the parameter sensitivity of IMD under CCVF method.In order to further improve the performance of sensorless IMD, the influences of the observer’s discretization, current measurement error and inverter nonlinearity are studied. First, the conclusion is drawn that voltage-model observers are more appropriate in low sampling frequency and output frequency ratio, based on the current voltage characteristics under the digital signal processor controlled inverter. And a suitable discretization method for Holtz observer is proposed to achieve high accuracy estimation of stator flux and rotor speed. Second, the impact of the current sampling ratio and offset error is analyzed, providing a theoretical basis for current sampling online adjustmanet. Through the careful analysis of the ADC sampling trigger point set, nonideal delay and the symmetrically channel arrangement, the current sampling setting method is systematically recommended. Third, the inverter nonlinear error caused by dead time and insulated gated bipolar transistor (IGBT) non-ideal characteristics is analyzed. A conclusion is drawn that the rise or fall time of output voltage is inversely proportional to current value in IGBT turn-off process when bus voltage keeps constant. Based on the relationship, a new inverter nonlinearity compensation method is proposed. The compensation voltage changes with the current value, particularly at low current.