异步电机低开关频率的模型预测高性能控制研究

发布时间：2018-08-19 13:36

【摘要】：降低逆变器开关器件的开关频率可以增加其输出功率，但会增大谐波畸变。大容量变频器低开关频率的高性能控制，涉及采用适合的电机控制策略，使得在低开关频率下获得较小谐波畸变的同时，又能使系统具有快速响应能力，是交流电机中压大功率传动高性能控制方面的一个难题。本文以国家自然科学基金项目（51377102）和台达环境与教育基金会《电力电子科教发展计划》项目（DREG2013009）为背景和支撑，对异步电机低开关频率（200~300Hz）下的高性能控制方案进行了较为全面深入的研究，主要工作包括：将模型预测控制引入到逆变器驱动电机控制领域，提出一种单步模型预测直接转矩控制（MPDTC）方法，以磁链和转矩偏差平方和作为价值函数，优先惩罚其中较大的偏差，减小了磁链和转矩脉动。通过在价值函数中添加开关跳变次数约束适当降低了开关频率。仿真结果验证了该MPDTC方法的有效性，同时也发现单步预测控制难以权衡多个控制目标。在单步MPTDC的基础上，以NPC三电平逆变器驱动异步电机系统为控制对象，引入状态输出轨迹外推形成长预测范围，提出一种新颖的低开关频率MPDTC方法。以平均开关频率作为价值函数，将控制问题描述为一个带约束条件的有限状态集滚动时域优化问题。针对优化求解的不可行性，，采用更新优化准则价值函数的策略。该方法可使系统运行在300Hz左右的低开关频率下获得快速的动态响应和较理想的电流谐波畸变。针对4kW和1.6MW异步电机驱动系统的仿真研究对其有效性进行了评估验证。在本文确立的MPTDC方法基础上，提出一种低开关频率的模型预测直接电流控制（MPDCC）方法，使得逆变器开关频率最小化且保持电流轨迹在给定滞环范围内。该方法可将NPC三电平逆变器开关频率降低至300Hz以下，同时获得了较理想的电流谐波畸变和动静态性能。与已有单步预测电流控制的对比仿真结果验证了其有效性。针对所提低开关频率MPDTC和MPDCC方法谐波性能不如采用优化PWM时理想，优化PWM不能直接应用于高性能闭环控制系统，本文深入研究了一种基于自控电机定子磁链轨迹跟踪控制（FTTC）的优化PWM闭环方案。提出结合SHEPWM特点的脉冲实时修正策略，实现了磁链轨迹跟踪控制。仿真结果表明，该方法既能在200Hz~300Hz的低开关频率下获得较小谐波畸变，又具有快速响应能力，相比本文所提的低开关频率MPDTC和MPDCC方法，其电流谐波性能更优。研究了一种基于模型预测磁链轨迹跟踪的优化PWM新型闭环控制方案，在不需要估计基波分量的前提下实现优化PWM的闭环控制。将控制问题构造为一个带边界约束条件的二次目标函数型最优化问题，设计了基于无差拍（DB）和二次规划（QP）的模型预测磁链轨迹跟踪控制器。基于该新型闭环系统的仿真结果表明，QP法能尽可能小地修正优化PWM同时消除磁链偏差，电流谐波性能比DB法略好。相比基于自控电机模型的FTTC闭环系统，该系统结构相对简单，两者动静态性能相当，电流总谐波畸变率都维持在5%以内。
[Abstract]:Reducing the switching frequency of inverter switching devices can increase the output power, but will increase the harmonic distortion. High-performance control of large-capacity inverter with low switching frequency involves the adoption of suitable motor control strategy, which can achieve less harmonic distortion at low switching frequency, and at the same time, it can make the system have fast response capability and is AC. A difficult problem in high performance control of medium voltage and high power drives for induction motors. This paper compares the high performance control schemes of induction motors under low switching frequency (200~300 Hz) with the background and support of the National Natural Science Foundation (51377102) and the Delta Environment and Education Foundation (DREG2013009) projects. Comprehensive and in-depth research, the main work includes:
Model predictive control (MPDTC) is introduced into the field of inverter-driven motor control. A one-step model predictive direct torque control (MPDTC) method is proposed, in which the sum of the squares of flux and torque deviations is taken as the value function, and the larger deviations are given priority punishment, thus reducing the flux linkage and torque ripple. The simulation results show that the MPDTC method is effective and the single-step predictive control is difficult to balance multiple control objectives.
On the basis of single-step MPTDC and NPC three-level inverter-driven asynchronous motor system as the control object, a novel low-switching-frequency MPDTC method is proposed by introducing state output trajectory extrapolation to form a long prediction range. The control problem is described as a finite state set roll with constraints by taking the average switching-frequency as the value function. Dynamic-time domain optimization problem. In view of the infeasibility of the optimization solution, the strategy of updating the optimization criterion value function is adopted. This method can make the system run at low switching frequency of about 300 Hz to obtain fast dynamic response and better current harmonic distortion. It has been evaluated and verified.
A low switching frequency model predictive direct current control (MPDCC) method is proposed to minimize the switching frequency of the inverter and keep the current trajectory within a given hysteresis range. This method can reduce the switching frequency of the NPC three-level inverter to less than 300 Hz, and obtain an ideal current. Harmonic distortion and dynamic and static performance are compared with the existing single-step predictive current control.
Because the harmonic performance of the proposed MPDTC and MPDCC methods with low switching frequency is not as good as that of the optimized PWM, the optimized PWM can not be directly applied to the high performance closed-loop control system. An optimized PWM closed-loop scheme based on the stator flux tracking control (FTTC) of the automatic control motor is studied in detail in this paper. The simulation results show that the proposed method can not only achieve small harmonic distortion at low switching frequencies of 200 Hz to 300 Hz, but also has fast response ability. Compared with the proposed low switching frequency MPDTC and MPDCCC methods, its current harmonic performance is better.
A new closed-loop control scheme of optimal PWM based on model predictive flux trajectory tracking is studied. The closed-loop control of optimal PWM is realized without estimating fundamental component. The control problem is constructed as a quadratic objective function type optimization problem with boundary constraints, and the deadbeat (DB) and quadratic programming (QP) are designed. The simulation results based on the new closed-loop system show that the QP method can modify the optimized PWM and eliminate the flux deviation as little as possible, and the current harmonic performance is slightly better than that of DB method. The total harmonic distortion is maintained within 5%.
【学位授予单位】：上海大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TM343

【参考文献】