 |
 |
 |
         |
 |
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Main
Former Projects
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
Power Electronics In modern society, mobility is worldwide at the basis of economic and social growth. However, mobility has its price. And the irony of fate, the growing demand for transport is accompanied by problems that threaten the social system that they should promote. First, in terms of externalities such as air pollution and noise pollution, and certainly in terms of cost of energy sources. Just think that 25% of global energy needs and the same percentage of anthropogenic carbon dioxide is produced annually by transport systems. This scenario requires therefore that the growth of transport systems for both these aspects is sustainable. To date, the technology of electric traction has proven to be the cornerstone of sustainable mobility, not without problems. In particular, electric vehicles as an alternative to vehicles with internal combustion engine sees the main limit in the limited range and duration of accumulators to store energy (accumulator aging). Consider that the ratio between the cost of kWh at the wheel produced by oil and by electricity is approximately equal to 10, that is reduced to little more than 2 whereas the degradation of batteries. It is therefore strategic in the development of electric vehicle technology the reliability of batteries and more generally the management of incoming and outgoing energy flow from them. At state of the art there is no modeling of accumulators, nor empirical nor analytical, which gives a consolidated characterization of their behavior in time. While this investigation concerns the chemical of constituting components of batteries, both the characterization and their optimized management involve the design and development of new architectural solutions, electronic control, i.e. digital power management. The priority of electric traction on the road as a primary factor of sustainable mobility is widely admitted worldwide by China, the United States and more recently by Europe, becoming a priority in their programs of research and development in the short / medium term. In recent years the use of rechargeable batteries is facing an increasingly large spread. Although in the past they were used more as a source of energy for portable devices such as laptops and mobile phones are getting popular recently for use in all-electric or hybrid vehicles, or as a reservoir of energy to combine, for instance, with systems of energy generation from renewable sources. The explosion of the use of rechargeable accumulators is due to the increase in energy density on the one side and to the cost reduction on the other. Suffice it to say that the recent lithium cells have an energy density up to an order of magnitude larger than the lead acid cells. That means less space for the same autonomy in the case of mobile devices and more autonomy for the same space in energy storage applications. A fundamental role of the accumulation of energy today is undoubtedly played within the context of distributed generation, which allows for having devices within the distribution system that can be both users and generators. This significantly improves the performance of distribution networks, which, as is well known, can suffer severe limitations in energy and power. Resources are addressed in the design and realization of modular energy storage architectures able to exchange energy with the distribution network in both directions. This topic is also investigated in the frame of renewable energy sources such as photovoltaic power generation. In partnership with:
|
|
|
Radiation detectors, read-out and processing electronics Electronics for adaptive digital processing of pulses In the last years digital processors for pulse processing have been intensively investigated and developed in many fields of application as an alternative to classic analogue systems. This interest is due to the intrinsic adaptivity, easiness of calibration and capability to obtain signal-to-noise ratioes very close to the optimum one. This research deals with the design, realization and test of a general purpose digital signal processor for random pulses analysis with throughput up to 100kevents/sec, whose features and performances are comparable or superior to state-of-the-art analogue designs. In first implementations, the system has been configurated as a high resolution amplitude spectrometer (X- and gamma-ray spectroscopy), capable also to time the occurrence of pulses of random amplitude arriving randomly in time. The timing resolution is better than a sampling interval. The set-up is based on FPGA and DSP technology and runs at a relatively low sampling frequency (below 100MHz). In order to get advantages of spatial computing in programmable devices, data-path structures of temporal computing process techniques have been revised and new processing architectures have been conceived. Among relevant improvements consequent to these optimizations are the reduction of processing speed, time-continuous processing operation and adaptive dynamic management of numeric filters length. Automatic optimal initialization of digital processors A completely automatic procedure to derive the coefficients of numerical filters of digital pulse processors corresponding to the optimum weight functions has been implemented. The shape of the synthesised filter can be customized for yielding the optimum filter in the actual experimental conditions with arbitrary constraints: e.g., necessity of time-limited filters for input signals of arbitrary shape, lorentzian noise spectral density components, presence of 1/f current noise smoothed-to-white at low frequency, timing filters, etc. The method can be easily translated into computer programs and has been used as a tool for optimising a digital signal processing spectroscopy set-up in its digital filter section. For example, a test structure to perform non destructive repetitive readout of the signals from high resolution detectors has been considered. By implementing the proposed technique, a characterization of the device has been performed, in terms of effects of the repetitive readout method on noise contributions, both for sinusoidal signal induced on the readout electrode by the periodic oscillation of the signal charge and for signal approximated by a series of delta pulses of opposite areas. The theory can be extended to different fields of application of optimal filtering. Multichannel applications A specific research is focused on the problem of the spatial localization of energy releasing events (hits) in segmented large-volume HPGe detectors. The shape of the signal generated at the detector electrodes depends on the drift path of the charge cloud generated in the interaction between the g photon and the detector. The problem is extremely complex. The pulse shapes induced on the electrodes strongly depend on the three spatial coordinates of the hit. This is true for the pulses at the electrodes collecting charge as well as for the zero-area pulses induced on the neighbor electrodes. An innovative algorithm for radial, angular, longitudinal coordinates estimation of events occurring at the same time in the same segment has been conceived. In order to design the algorithm, pulse shapes at the electrodes of coaxial HPGe detectors have been calculated in closed form. The effects on pulse shapes of the detector encapsulating metallic cluster and of the gaps between sensing electrodes in terms of crosstalks due to capacitive coupling have been also investigated and taken into account. The algorithm was designed bearing in mind that, with up to thousands of parallel channels, it is mandatory to perform the measurements in each channel with the least possible number of input data, while losing no important information and, possibly, operating on line in real time. For the sake of simulation, multiple coincident hits with energy ranging from 60 keV up to 2 MeV are taken into account and electronic noise and signal finite bandwidth effects are considered. Resources are invested in the acquisition and processing of signals from detector arrays for medical and space imaging. For instance, the DAQ and processing digital electronics for a compact and high resolution Anger camera to be used in clinical and research environments has been recently developed. Main features of the system are re-configurability, linearity, low noise and processing data rate through hardware and firmware architecture solutions that also allow the system to operate in presence of disturbances and electromagnetic noise. The use of Silicon Drift Detectors coupled to CsI(Tl) scintillators offers a high intrinsic spatial resolution (less than 1 mm), an overall spatial resolution of about 2.5 mm at 5 cm, and an appropriate sensitivity. With such features, this gamma camera is well suited for clinical and research applications where high overall spatial resolution and system compactness are required. While the state-of-art in the field of medical imaging instrumentation is represented by commercial systems with large field detectors (order of 40x50 cm2) and with an overall effective spatial resolution typically of 10-16 mm at an imaging distance greater than 20 cm, this Anger camera has a smaller field of view but thanks to its compactness the distance of the imaged target could be greatly reduced, also allowing easier positioning in case of space constraints during acquisition and improving patient comfort during the diagnostic procedure. A small field of view camera with high resolution has potential also in several molecular imaging on small animals, in applications used for evaluation of pharmaceutical distribution or functional effects resulting from medical treatments. In partnership with: 
  
|
 |
Image processing HARDWARE TOPICS - R&D of digital configurable systems
on uC, DSP and FPGA for processing and
transmission of audio-video signals. FIRMWARE AND SOFTWARE TOPICS - Driver development at kernel level for
LINUX compliant with V4L, V4L, ALSA. In partnership with: |
 |
Audio processing The research deals with the development
of new techniques for acoustic feedback
("Larsen effect") suppression in case of
on-stage applications. A fully digital
solution has been focused and the
prototypation of the processing channel
has been under construction for the
on-field test of the proposed technique.
The designed hardware platform is based on
a Xilinx Virtex-5 FPGA device. In
partnership
with:
|
|
|
High performance scientific computing In the analysis of the dynamics of many
real complex systems, descriptive
mathematical models permit to simulate
their behaviour in different specific
operating conditions. However, quite all
applications require mathematical models
that cannot be handled by means of closed
form calculations but just through
numerical resolution techniques. For
instance this is the case of models of
very fine grained systems, such as human
body organs, which are represented by very
sparse matrix of extremely high
dimensionality (several millions of
elements). In partnership with:
|
Partner for sponsored
joint research agreements with private and
public Companies
| © Dipartimento di Elettronica, Informazione e Bioingegneria DEIB - Politecnico di Milano University | Updated: 11/03/2021 |