A trajectory linearisation-based offset-free continuous control-set model predictive control algorithm is proposed in this paper. The problem formulation is obtained by means of a full-incremental velocity form, which brings along an integral-like action that ensure zero offset tracking without the aid of disturbance observers. The adopted motor model is based on trajectory linearization rather than constant motor parameter or parameter adaptation from tabled data. The proposed strategy simplifies the tuning of the model predictive control algorithm, meanwhile boosting the current control performances to achieve optimal control. Several tests were carried out on a synchronous reluctance motor characterised by highly nonlinear magnetic flux linkages characteristics. Comparisons with conventional full-incremental velocity form MPC implementation are reported to highlight the benefit that the proposed algorithm brings along.

In the paradigm of more-intelligent ac drives, the adoption of a system parameter-free algorithm for the current control is a substantial innovation, as it makes modelling and identification of the physical system obsolete. The prediction of current space vector trajectory is based on an adaptive model instead of a traditional motor model. A newly tailored optimisation problem exploits the current prediction to generate the optimal voltage vector reference within an hexagonally limited continuous set. The paper provides several hints for design and implementation, while also providing a remedy for stagnation phenomena that can appear due to the application of a constant voltage during transients. The proposed current control algorithm has been tested extensively on an experimental test rig, featuring a full bus voltage, and two different synchronous motors, either with or without permanent magnets. It is not necessary to fine-tune any controller or to know in advance the motor connected.

Stray flux analysis is an interesting source of information for the diagnosis of Induction Motors (IMs). The widespread use of these motors in industry leads to a necessity of additional tools and methods for their predictive maintenance. On the other hand, soft-starters are increasingly used to reduce the high consumption of IMs at start-up. In this work, AI techniques based on convolutional neural networks are applied to detect rotor faults in soft-started motors. The objective is the automatic early detection of broken bars, avoiding the necessity of user intervention to interpret the obtained results. This work proves the potential of the methodology, including a successful set of experimental results.

Under the thrust of the increasing demand of energy-efficiency and reliability, sensorless permanent magnet synchronous motor drives are showing compelling advances. The presence of a salient magnetic structure (anisotropic rotor) reduces the presence of magnets and it is fundamental for operations at zero and very low speed, obtained by injecting and processing high frequency signals. But the excitation signals cause unwanted torque ripple, noise and vibrations. This work deals with the lowering of the lower speed limit, above which the estimation techniques can work without any signal injection. Specifically, the proposed method mitigates the instability problems of back electromotive-based observers when the drive operates in generating mode. The speed and position estimates are based on a modified phase locked loop observer. It trusts on direct phase voltage measurements, even if it works also with standard compensated reference voltage signals. An accurate laboratory measurements campaign, focused on showing the injection-free speed range improvements, is included in the paper.

Sensorless drives for permanent magnet and reluctance synchronous motors are an interesting option in modern drives. Modified observers, such as the ones based on extended electromotive force or active flux, are used to get a unified model of the motor. The fundamental frequency based observer is of easy implementation, although the reliability of the model is undermined by parameter variations and corrupted measurements, especially when anisotropic motors are adopted. This study aims at analysing the influence of all motor model uncertainties, by means of an analytical sensitivity analysis. The resulting functions and graphic plots are useful to draw interesting considerations about the sensorless drive, in terms of motor selection and control design.

The paper proposes the novel application of Radial Basis Function (RBF) networks to the magnetic modelling of any PM synchronous motor, with special emphasis on internal PM rotors. The advantages are the use of already available measurements, the precision and completeness of the continuous 2D model, which inherently includes all the nonlinearity and cross-coupling effects and that does not use any interpolation algorithm. The paper includes the theory of operation and experimental results, obtained on a laboratory prototype.

Energy saving will be the key-factor of the next-generation electrical drives. Extremum-seeking techniques, and the derived injection-based Maximum Torque per Ampere (MTPA) strategies, are a step towards the optimal drive efficiency, combining implementation easiness and robustness against parameter variations. Nevertheless, due to both the heavy motor non-linearity and the presence of mixed frequency signal dynamic, not all the implementation issues have been successfully addressed yet. Along this path, the paper offers a comprehensive theoretical and experimental validation of the MTPA tracking algorithm based on high frequency signal injection, including the stability numerical proof and an extended batch of commented experimental results.

Environmental-friendly technologies are growing up fast and they are getting more and more attention. In such a scenario, Interior Permanent Magnet (IPM) synchronous motors have all the peculiarities to satisfy the drive efficiency requests. So far, techniques like the Maximum Torque-per-Ampere (MTPA) were a stimulating field of research, because of the great results in term of energy savings. The MTPA locus brings along additional information that is still worth deepening. A proper rearrangement of the torque expression in MTPA condition, obtained by describing the q-axis flux linkage by a polynomial approximation, merged with a simple least-square algorithm, returns a fast and accurate on-line estimation of the electromagnetic torque. The effectiveness of the proposed method is proved by a worked out example, based on a IPM drive model fitted with finite-element analysis and experimental measurements, and including motor saturation effects.

This paper proposes two novel approaches to the simplification of multi-input fuzzy logic controllers (FLC). By these approaches either a hierarchical or a parallel structure is derived for a multi-input FLC, leading to the implementation of faster controllers with reduced memory demand. Both these features can be used profitably in the realization of hardware and software effective FLC for the industrial field applications. Experimental tests and comparative considerations are also included.