Sensorless controlling techniques of AC motor drivesAnam Mughees | July 01, 2020
AC motors are widely used in in both industrial and domestic environments for motion control. The induction motor (IM) and the permanent magnet synchronous motor (PMSM) are two types of AC motors that serve a wide variety of applications. Many applications, particularly in the industrial sector, require a high degree of accuracy, speedy dynamic response and high efficiency in the design and implementation of processes. Accurate position and speed control is essential in these cases. In general, a high-resolution mechanical position speed sensor is a prerequisite for precise control of drives but it increases the cost and complexity while compromising the reliability of the system.
The obvious solution to this problem is the sensorless control of AC drives. Lower costs, greater reliability, lower hardware complexity, better immunity to noise and less maintenance make sensorless motor drives a very attractive option. Sensorless control strategies vary depending on the type of motor. This article will take a look at controlling techniques of both IM drives and PMSM drives.
Induction motor drives
In general, the rotor flux vector control is used as the basic control strategy for IM drives. The basic drive model consists of a rectifier that converts the three-phase alternating current into direct current of DC-link and an inverter that converts this direct current to three-phase alternating current again for driving the motor. An important part of this model is the estimation of rotor flux angle and motor speed. Most of the effective and popular control methods of IM drives can be divided into three categories:
1. Flux and speed estimation-based strategy
2. Field-weakening control strategy
3. Current loop control strategy
Flux and speed estimation, without sensors, is obviously an important part of sensorless control strategies. One strategy to estimate these parameters is based on signal injection. Data regarding the position of the rotor is obtained by injecting a signal that determines the desired information using rotor slot harmonic and rotor saturated and leakage inductance. In this way, the stability of the induction motor is guaranteed at zero frequency stator current. However, signal injection-based methodologies require obvious magnetic field anisotropy and depend greatly on the nature of the motor. Furthermore, certain problems such as torque and noise make it difficult for them to be widely utilized in industry. Another group of strategies for observing flux and speed of an induction motor is based on an IM model. First, the IM mathematical model is established and then, according to this model, the rotor flux and speed are calculated. Kalman filter observer is a robust method that falls into this category. Overall, the main focus of all researchers in this field is to maintain stability at zero speed and zero frequency.
There are certain applications like locomotive traction and common numerical control (CNC) systems that require high-speed control capability of IM drives. In addition to being highly accurate, these drive systems must be capable of high-speed operation, have fast start-stop ability, and demonstrate a high dynamic response. Field weakening control strategies are employed in these systems because reduction in the motor’s magnetic field is required in order to reduce back EMF of motors. Reducing back EMF increases the accelerating abilities of the motor, which is essential for high-speed operation. Motor model-based open-loop calculations can be used for field-weakening control strategies. Another field-weakening control technique is based on voltage limiting closed-loop regulation.
Current loop control strategy is also ideal for high-speed applications like CNC and drilling machines. A commonly used method for controlling IM current loops is hysteresis control. The basic concept is that the three-phase stator currents are controlled by three distinct hysteresis comparators. Two methods of proportional integral (PI) control are also used: stationary frame PI control and synchronous rotating frame PI control. The most widely used current loop control method is the synchronous frame PI control. However, in recent years another controlling technique has gained much attention: predictive control (PC). The fundamental idea of predictive control is to obtain the optimal control output through the prediction of the change in the control object on the basis of the mathematical model of the control object. Four common types of predictive control are:
1. Model PC
2. Trajectory-based PC
3. Hysteresis-based PC
4. Dead-beat PC
Permanent motor synchronous drives
Sensorless control methods of PMSM drives are categorized as model-based methods (adaptive or non-adaptive) and high-frequency signal injection (HFSI) methods. HFSI is used for low-speed applications while model-based methods are used for medium- to high-speed applications. Unlike the sensorless IM drive control, saturation and structural saliencies contribute to low-speed and zero-speed position tracking. Acoustic noises, however, restrict HFSI methods to areas such as household equipment.
Harmonics can adversely affect the estimated rotor position for model-based sensorless control methods. A relatively new technique of adaptive linear neural-network-based filter (ADALINE) can be used to overcome this problem. Some of the common control techniques that fall under the model-based methods are listed as under:
1. Extended Kalman filter
2. Model reference adaptive system
3. Sliding mode observer
Model-based methods work by detecting the fundamental component of back EMF. In low-speed range, the signal-to-noise ratio of this EMF is too low to be picked up, so, model-based strategies only work for high-speed applications. HFSI methods, on the other hand, can easily calculate the position at low speeds (even at zero speed). However, due to limited observer bandwidth, HSFI strategy does not work when dealing with high-speed applications. Researchers are now moving towards developing hybrid techniques to provide a wide speed range for sensorless control of drives.