Research On Several Key Technologies Of High Performance Induction Motor Drives

Nowadays, induction motor drives are quite essential for both the industrial applications and the social consumptions. The research and promotion of the variable frequency motor drives with high performance is an effective way to deal with the energy crisis and the environment problem, improve the quality of the industry and the living, and reduce the grid power pollution. So the improvement of the motor drives is important for not only the technology progress, but also the sustainable development of human society. Based on the investigation of the induction motor control and the motor drives' hardware, this dissertation performs a detail research on the speed sensorless vector control technologies, which help to reduce the hardware cost and improve the system reliability, and the active front end related technologies, which help to improve the grid power quality and the efficiency. Some novel solutions are also proposed for performance improvement.First, several types of flux observer for speed sensorless control are analyzed and compared, and the speed adaptive full order flux observer is selected for further investigation. With the linearized small signal method, the instability problem of the conventional full order observer operating at the low speed and regenerative braking state is indicated. In order to deal with this problem, an improved strategy to design the feedback gain matrix of the full order observer. The indicated issue and the proposed solution are both verified by the theoretical analysis and the experimental results that are carried out on a prototype of induction motor drives with speed sensorless vector control. The results prove that the proposed solution is able to improve the stability of the sensorless control scheme at the low speed operating range.Secondly, this dissertation investigates the practical implementation issues for the full order observer in speed sensorless motor drives, and some novel solutions are proposed and verified experimentally. Taking the coupling effect between the flux and the rotor speed into consideration, the relationship between the error of motor parameters and the error of the flux estimation and the rotor speed estimation is analyzed in detail. Based on the analysis and the simulation results, the influence on estimation accuracy with respect to each motor parameter is concluded. Then, solutions for self-commissioning at standstill and on-line parameter identification are investigated and experimented respectively. Practical implementation issues of the inverter system in the motor drives are also investigated, including digital discretization process, sampling process, and dead time effect. The merits and demerits of the basic discretization methods are analyzed in the condition that they are implemented for the full order observer. Then a compound strategy using partial forward-difference method and partial bilinear transformation method is proposed. The proposed discretization strategy is able to realize better accuracy and lower computing burden simultaneously, which is quite essential for applications that require high switching frequency. In order to emphasize the key points for sampling process design, the influence of the stator current sampling noise and error on the accuracy of flux and speed estimation is also investigated. Furthermore, the dead time effect of voltage source inverter is analyzed in detail, especially its impact on the reconstruction of the stator voltage. Then, a novel method to determine the status of the inverter output current, together with a correspondent dead time compensation strategy are proposed. The proposed solution is verified by both the theoretical analysis and the experimental results, and it realizes better accuracy by selecting the different compensating terms based on the different current status, which effectively reduces the influence of stator voltage reconstruction error on the flux observer.Finally, this dissertation researches on the design and the control strategy of the active front end with LCL filter. The design criteria of the filter and the control scheme is investigated and concluded. According to the design criteria, a motor drives prototype is implemented with the LCL filter and the active front end as the grid interfacing stage, and both the motoring and the regenerating mode are tested on the prototype. Then, the harmonic suppression performance of the system in the condition of distorted grid voltage is analyzed in detail, and the limitation of conventional control scheme caused by the resonance problem of the LCL filter is also indicated, which leads to the research on the improved control scheme. Different structures of compound repetitive control scheme are discussed in detail, and are compared with each other in stability criteria, required compensation process, and harmonic suppression performance. Together based on the analysis and the experimental results, the merits and demerits of different compound repetitive control scheme are concluded. Furthermore, active damping is also analyzed in this dissertation. A novel active damping method that fed back the capacitor voltage estimated by a reduced-order observer to suppress the resonant peak of the LCL filter is proposed. The proposed method is able to implement active damping without increasing the hardware cost and complexity, and is verified by the experimental results. The performance of the proposed active damping method is also compared with the compound repetitive control scheme by the experimental results.In a word, this dissertation performs researches on several critical aspects of the induction motor drives, in order to achieve some improvement on the performance and the practical implementation. Some novel solutions are proposed in the dissertation, and a great quantity of experimental work is also carried out for verification.