Angelo Accetta

Sensorless Control of PMSM Fractional Horsepower Drives by Signal Injection and Neural Adaptive-Band Filtering

This paper presents a sensorless technique for permanent-magnet synchronous motors (PMSMs) based on high-frequency pulsating voltage injection. Starting from a speed estimation scheme well known in the literature, this paper proposes the adoption of a neural network (NN) based adaptive variable-band filter instead of a fixed-bandwidth filter, needed for catching the speed information from the sidebands of the stator current. The proposed NN filter is based on a linear NN adaptive linear neuron (ADALINE), trained with a classic least mean squares (LMS) algorithm, and is twice adaptive. From one side, it is adaptive in the sense that its weights are adapted online recursively. From another side, its bandwidth is made adaptive during the running of the drive, acting directly on the learning rate of the NN filter itself. The immediate consequence of adopting a variable-band structure is the possibility to enlarge significantly the working speed range of the sensorless drive, which can be increased by a factor of five. The proposed observer has been tested experimentally on a fractional horsepower PMSM drive and has been compared also with a fixed-bandwidth structure.

MRAS Speed Observer for High-Performance Linear Induction Motor Drives Based on Linear Neural Networks

This paper proposes a neural network (NN) model reference adaptive system (MRAS) speed observer suited for linear induction motor (LIM) drives. The voltage and current flux models of the LIM in the stationary reference frame, taking into consideration the end effects, have been first deduced. Then, the induced part equations have been discretized and rearranged so as to be represented by a linear NN (ADALINE). On this basis, the transport layer security EXIN neuron has been used to compute online, in recursive form, the machine linear speed. The proposed NN MRAS observer has been tested experimentally on suitably developed test set-up. Its performance has been further compared to the classic MRAS and the sliding-mode MRAS speed observers developed for the rotating machines.

Neural Sensorless Control of Linear Induction Motors by a Full-Order Luenberger Observer Considering the End Effects

This paper proposes a neural based full-order Luenberger Adaptive speed observer for sensorless linear induction motor (LIM) drives, where the linear speed is estimated on the basis of the linear neural network: TLS EXIN neuron. With this reference, a novel state space-vector representation of the LIM has been deduced, taking into consideration the so-called end effects. Starting from this standpoint, the state equation of the LIM has been discretized and rearranged in a matrix form to be solved by a least-square technique. The TLS EXIN neuron has been used to compute on-line, in recursive form, the machine linear speed since it is the only neural network able to solve on-line in a recursive form a total least-squares problem. The proposed TLS full-order Luenberger Adaptive speed observer has been tested experimentally on suitably developed test setup.

MRAS speed observer for high performance linear induction motor drives based on linear neural networks

This paper proposes a Neural Network (NN) MRAS (Model Reference Adaptive System) speed observer suited for linear induction motor (LIM) drives. The voltage and current models of the LIM in the stationary reference frame, taking into consideration the end effects, have been obtained. Then, equations of the induced part have been discretized and rearranged so as to be represented by a linear neural network the TLS EXIN neuron, which has been used to compute the machine linear speed on-line and in recursive form. The proposed NN MRAS observer has been tested experimentally on a suitably developed test setup. Its performance has been also compared to the classic MRAS speed observer.

PMSM drives sensorless position control with signal injection and neural filtering

Vector Field Oriented Control (FOC) is one of the best control methods for high-dynamic electrical drives. To avoid the adoption of the speed/position sensor (resolver/encoder), a sensorless technique should be used. Among the various sensorless methods in literature, those based on machine saliency detection by signal injection seem to be most useful for thier giving the possibility of closing the position control loop. This paper proposes a method for enhancing both rotating and pulsating voltage carrier injection methods by a neural adaptive band filter. Results show the goodness of the proposed solution.

Parameter identification of induction motor model by means of State Space-Vector Model Output Error Minimization

This paper proposes a technique for the off-line estimation of the electrical parameters of the equivalent circuit of an Induction Machines (IM), and focuses on the application of an algorithm based on the minimization of a suitable cost function involving the differences between the measured stator current direct (sD) and quadrature (sQ) components and the corresponding estimated by the IM state model. This method exploits an entire start-up transient of the IM to estimate all of the 4 electrical parameters of the machine (Rs, Ls, σLs, Tr). It proposes also a set of tests to be made in order to estimate the variation of the magnetic parameters of the IM versus the rotor magnetizing current as well as the magnetizing curve of the machine. The proposed methodology has been verified experimentally on a suitably developed test set-up.

Feedback Linearizing Control of Induction Motor Considering Magnetic Saturation Effects

This paper presents an input–output feedback linearization (FL) control technique for rotating induction motors, which takes into consideration the magnetic saturation of the iron core. Starting from a new formulation of the dynamic model taking into consideration the magnetic saturation expressed in a space-state form in the rotor-flux-oriented reference frame, the corresponding FL technique has been developed. To this aim, a particular care has been given to the choice of nonlinear functions interpolating the magnetic parameters versus the rotor magnetizing current and the corresponding magnetic characteristic. The proposed FL technique has been tested experimentally on a suitably developed test setup and compared, under the same bandwidths of the speed and flux closed loops, with both the FL technique not taking into consideration the magnetic saturation and with the industrial standard in terms of high performance control of the induction motor: field-oriented control. A further comparison with the unique example present in the scientific literature of the FL control technique including the magnetic saturation has been shown, highlighting consistent improvements achievable by the proposed FL in terms of reduced computational demand and better dynamic performance.

PEM Fuel Cell System Model Predictive Control and real-time operation on a power emulator

Fuel Cell Systems (FCS) seem to be among the most reliable devices to produce clean energy, although they still suffer for many problems, mostly related to the fragility of the Polymer Electrolyte Membrane (PEM). Particularly, this paper focuses on the oxygen starvation, that leads both a decrease of the FCS performance and a shortening in the FCS lifetime. The purpose is to use the Model Predictive Control (MPC) and its capacity of accounting for linear constraints for managing the air system without risking to damage the fuel cell. Two control inputs and no static feed-forward actions have been used. Results show that the MPC is able to avoid the oxygen starvation, even with a sudden increase of load current. Results obtained in previous works are confirmed with a FCS emulator and show the usefulness of the Model Predictive Control applied to the FCS.

Closed-Loop MRAS Speed Observer for Linear Induction Motor Drives

This paper presents a closed-loop model reference adaptive system (CL-MRAS) speed observer developed for linear induction motor (LIM) drives. Starting from the structure of the CL-MRAS speed observer developed in the literature for rotating induction motors, a corresponding speed observer for LIMs has been developed here. It is based on the LIM dynamic model taking into consideration its dynamic end effects. In particular, the following aspects are original: 1) It employs the voltage and current models of the LIM considering its dynamic end effects, 2) it proposes a net thrust model including the braking force caused by the dynamic end effects; 3) it models the friction load of the LIM drive experimentally obtained with off-line tests. The proposed sensorless technique has been tested on a purposely developed experimental set-up. Results show that the CL-MRAS observer permits the drive to operate down to the speed of 0.01 m/s, i.e., about 0.15% of the rated speed, which is about 20 times lower than the minimum working speed achieved with other observers implemented on the same LIM drive, such as the total least squares (TLS) MRAS observer, the TLS EXIN full-order Luenberger observer, the extended Kalman filter, and, finally, the TLS EXIN Kalman filter.

Neural sensorless control of linear induction motors by a full-order Luenberger observer considering the end-effects

This paper proposes a neural based full-order Luenberger adaptive speed observer for sensorless linear induction motor (LIM) drives, where the linear speed is estimated with the total least squares (TLS) EXIN neuron. A novel state space-vector representation of the LIM has been deduced, taking into consideration its dynamic end effects. The state equations of the LIM have been rearranged into a matrix form to be solved, in terms of the LIM linear speed, by any least squares technique. The TLS EXIN neuron has been used to compute online, in recursive form, the machine linear speed. A new gain matrix choice of the Luenberger observer, specifically taking into consideration the LIM dynamic end effects, has been proposed, overcoming the limits of the gain matrix choice based on the rotating-induction-machine model. The proposed TLS full-order Luenberger adaptive speed observer has been tested experimentally on an experimental rig. Results have been compared with those achievable with the TLS EXIN MRAS, the classic MRAS, and the sliding-mode MRAS observers.