Vrunda Joshi

Study of sensorless control algorithms for a permanent magnet synchronous motor vector control drive

In the recent past, the use of permanent magnet synchronous motors (PMSM) has increased considerably owing to their inherent advantages. The high performance speed and/or position control of a PMS motor requires an accurate knowledge of rotor shaft position and angular speed in order to synchronize the phase excitation pulses to the rotor position. The rotor position and speed information is specifically required for implementation of vector control scheme in a PMSM. In order to achieve the speed control various sensored methods are employed. Due to number of advantages recent trend is to implement the sensorless control drive for a PMSM. This paper presents and compares two main sensorless control algorithms viz, Sliding Mode observer(SMO) and Extended Kalman Filter (EKF) for sensorless estimation of rotor position and speed in a PMSM. The simulation results for both these methods on an identical PMSM are presented. The analysis of these results give some insights into the accuracy of estimations and subsequent working of the PMSM with them.

Hall sensor fault detection and fault tolerant control of PMSM drive system

This paper presents a novel hall sensor fault detection and fault tolerant control system for Hall Effect position sensors failure of permanent magnet synchronous motor. The proposed system is capable to accurately detect and promptly identify exactly which of the Hall Effect sensors has undergone breakdown when the fault occurs. A simple method to generate the commutation signal even with faulty Hall Effect sensor to maintain the proper operation of PMSM after fault occurrence is discussed in this paper. The proposed fault tolerant control system does not need massive computational efforts and can be implemented as a subroutine in the main control program of the PMSM. In addition, the transitions between healthy and fault-tolerant control conditions are smooth without any additional noise and jerk. Experimental results have been presented to prove the validity of the proposed algorithm.

Speed control of PMSM with Hall sensors using DSP TMS320F2812

In this paper, speed control of Permanent Magnet Synchronous Motor (PMSM), also known as sinusoidal BLDC motor, using Hall effect sensors is described. Normally, sensors like optical encoders or resolvers are necessary for applying sinusoidal commutation to such type of motor as they have more resolution. However, these sensors are costly and cannot be employed for low cost operations. The proposed approach for sinusoidal commutation uses rotor position information available from Hall sensors to apply sine PWM pulses for the inverter switching. The modulation index in this method is changed using PI controller for speed control of PMSM. The method has been tested successfully on a setup of 400W PMSM motor using Texas Instrument (TI) DSP controller TMS320F2812. Experimental results indicate that the motor drive developed in this paper exhibits satisfactory performance.

Survey of EKF based sensorless vector control methodologies for a permanent magnet synchronous motor

In the recent past, the use of permanent magnet synchronous motors (PMSM) has increased due to the numerous advantages they offer. The accurate performance and precise speed and/or position control of a PMS motor requires precise information about the rotor shaft position and angular speed in order to synchronize the phase excitation pulses to the rotor position. In order to achieve the speed control various sensored methods are employed. Due to number of advantages, recent trend is to implement the sensorless drive for a PMSM. This paper presents the survey conducted in studying and understanding various methodologies of EKF based sensorless algorithms for the vector control drive of a PMSM.

Implementation issues of sliding mode observer for sensorless field oriented control of PMSM using TMS320F2812

The sliding mode observer is a very promising approach for position sensorless control of Permanent Magnet Synchronous Motors due to its robustness and parameter insensitivity. However it was observed that estimation of rotor flux position was affected due to measurement noise and phase shift introduced due to low pass filtering. In this paper, the authors discuss key issues related to implementation of sliding mode observer based sensorless field oriented control of Permanent Magnet Synchronous Motor and propose different approaches to improve rotor flux estimation. An experimental lab setup is designed and implemented on a 400W PMSM motor. Texas Instruments TMS320F2812 DSP is used in the drive as the central controller to fulfill the demands of exhaustive real time computations by vector control and sliding mode observer.

Disturbance observer based sensor-less vector control drive for PMSM using EKF

In the recent years many researchers have focused on the sensor-less PMSM drive technology, wherein, estimators are used to estimate rotor position and speed using line current measurement. EKF is a powerful tool for such an application. However, implementation of EKF based sensor-less control needs measurement of torque for faithful mathematical model [1], which hinders practical application of the scheme. In this paper, a scheme with load torque estimation using Non Linear Disturbance Observer (NLDO) [2] in cascade with EKF to eliminate load torque sensor is proposed. With this, the proposed sensor-less scheme eliminates position sensor as well as a torque sensor.

Discussion on Problems and Solutions in Hardware Implementation of Algorithms for a Car-type Autonomous Vehicle

The self-driving cars will bring fundamental change in the transportation industry. Many hazardous situations like land-mine detection, war zones, nuclear decommissioning highlights the need of autonavigation in open terrain. In the last few decades, due to the increasing interest in mobile robots, a number of autonavigation algorithms have been developed. The autonomous vehicles are used extensively in different domains from passenger car to the hazardous applications. These autonomous cars extensively use mathematical calculations and machine intelligence. In order for the car-type vehicle to manoeuver smoothly in a given workspace, accurate planning (motion and path) algorithms are essential. As part of the research, our group has developed the nonholonomic motion planning algorithms for the car-type vehicle based on differential flatness approach. In the previous work, the hardware realization of these algorithms is presented. This paper discusses the hardware implementation issues that we have faced during this work. Hardware implementation comes with various inherent challenges, such as manufacturing error in the car hardware components, physical limitation of the component, limited processing power of low-power onboard computer, accuracy of data from sensors in diverse conditions.

Sensorless vector control of PMSM using SMO and NLDO

In the proposed paper a state feedback controller is designed using Jacobian linearization of PMSM around an operating point which is adaptively refreshed using Non Linear Disturbance Observer (NLDO). This observer estimates load torque (which is considered as disturbance) using knowledge of speed and stator currents. The speed is normally measured using speed sensors like optical encoder. In this work, speed and rotor position are estimated using sliding mode observer (SMO). This estimated speed is given as an input to NLDO. Thus these two observers are cascaded in this scheme for sensorless vector control. Advantage of proposed cascaded scheme is that it replaces three PI controllers by state feedback controller. State feedback control gain can be mathematically determined using pole placement, thus avoiding complex procedure of tuning three PI controllers.

Fractional-order PI controller design for PMSM: A model-based comparative study

The mathematical model of any plant plays very important role in the controller design. In electric drives, it is usual practice to consider approximate model of the machine by considering mechanical time constant neglecting electrical time constant for simplifying controller design. The model thus considered, results into an approximated or simplified model of the electrical machine. While if we consider mechanical as well as electrical time constant in the model, it is exact model of the machine. The controller can be designed on simplified model or exact model of the motor. It is observed that performance of the controller is better when it is designed considering exact model of the motor. To demonstrate this idea, in this paper we present the design of fractional-order PI (FOPI) controller for Permanent Magnet Synchronous Motor (PMSM). We design two independent controllers, one which is based on exact model of PMSM and the other which is designed considering simplified model of the motor. The results show that FOPI controller, if designed using exact model, gives better performance over that, if designed using simplified model.