Modello bayesiano per la riduzione dell'incertezza nella previsione delle piene del fiume Reno

Many efforts have been devoting since last years to reduce uncertainty in hydrological modeling predictions. The principal sources of uncertainty are provided by input errors, for inaccurate rainfall prediction, and model errors, given by the approximation with which the water flow processes in the soil and river discharges are described. The aim of the present work is to develop a bayesian model in order to reduce the uncertainty in the discharge predictions for the Reno river. The ’a priori’ distribution function is given by an autoregressive model, while the likelihood function is provided by a linear equation which relates observed values of discharge in the past and hydrological TOPKAPI model predictions obtained by the rainfall predictions of the limited-area model COSMO-LAMI. The ’a posteriori’ estimations are provided throw a H∞ filter, because the statistical properties of estimation errors are not known. In this work a stationary and a dual adaptive filter are implemented and compared. Statistical analysis of estimation errors and the description of three case studies of flood events occurred during the fall seasons from 2003 to 2005 are reported. Results have also revealed that errors can be described as a markovian process only at a first approximation. For the same period, an ensemble of ’a posteriori’ estimations is obtained throw the COSMO-LEPS rainfall predictions, but the spread of this ’a posteriori’ ensemble is not enable to encompass observation variability. This fact is related to the building of the meteorological ensemble, whose spread reaches its maximum after 5 days. In the future the use of a new ensemble, COSMO–SREPS, focused on the first 3 days, could be helpful to enlarge the meteorogical and, consequently, the hydrological variability.

Model predictive control in thermal management of multiprocessor systems-on-chip

MultiProcessor Systems-on-Chip (MPSoC) are the core of nowadays and next generation computing platforms. Their relevance in the global market continuously increase, occupying an important role both in everydaylife products (e.g. smartphones, tablets, laptops, cars) and in strategical market sectors as aviation, defense, robotics, medicine. Despite of the incredible performance improvements in the recent years processors manufacturers have had to deal with issues, commonly called “Walls”, that have hindered the processors development. After the famous “Power Wall”, that limited the maximum frequency of a single core and marked the birth of the modern multiprocessors system-on-chip, the “Thermal Wall” and the “Utilization Wall” are the actual key limiter for performance improvements. The former concerns the damaging effects of the high temperature on the chip caused by the large power densities dissipation, whereas the second refers to the impossibility of fully exploiting the computing power of the processor due to the limitations on power and temperature budgets. In this thesis we faced these challenges by developing efficient and reliable solutions able to maximize performance while limiting the maximum temperature below a fixed critical threshold and saving energy. This has been possible by exploiting the Model Predictive Controller (MPC) paradigm that solves an optimization problem subject to constraints in order to find the optimal control decisions for the future interval. A fully-distributedMPC-based thermal controller with a far lower complexity respect to a centralized one has been developed. The control feasibility and interesting properties for the simplification of the control design has been proved by studying a partial differential equation thermal model. Finally, the controller has been efficiently included in more complex control schemes able to minimize energy consumption and deal with mixed-criticalities tasks

Development and experimental validation of a knock induced damage model and real-time implementation of model-based strategies to control knock intensity in boosted gasoline engines

This PhD thesis reports the main activities carried out during the 3 years long “Mechanics and advanced engineering sciences” course, at the Department of Industrial Engineering of the University of Bologna. The research project title is “Development and analysis of high efficiency combustion systems for internal combustion engines” and the main topic is knock, one of the main challenges for boosted gasoline engines. Through experimental campaigns, modelling activity and test bench validation, 4 different aspects have been addressed to tackle the issue. The main path goes towards the definition and calibration of a knock-induced damage model, to be implemented in the on-board control strategy, but also usable for the engine calibration and potentially during the engine design. Ionization current signal capabilities have been investigated to fully replace the pressure sensor, to develop a robust on-board close-loop combustion control strategy, both in knock-free and knock-limited conditions. Water injection is a powerful solution to mitigate knock intensity and exhaust temperature, improving fuel consumption; its capabilities have been modelled and validated at the test bench. Finally, an empiric model is proposed to predict the engine knock response, depending on several operating condition and control parameters, including injected water quantity.

Obstructive sleep apneas naturally occur in mice during REM sleep and are more prevalent in a mouse model of Down syndrome

Study Objectives. The use of mouse models in sleep apnea research is limited by the belief that central (CSA) but not obstructive sleep apneas (OSA) occur in rodents. With this study we wanted to develop a protocol to look for the presence of OSAs in wild-type mice and, then, to apply it to a mouse model of Down Syndrome (DS), a human pathology characterized by a high incidence of OSAs. Methods. Nine C57Bl/6J wild-type mice were implanted with electrodes for electroencephalography (EEG), neck electromyography (nEMG), diaphragmatic activity (DIA) and then placed in a whole-body-plethysmographic (WBP) chamber for 8h during the resting (light) phase to simultaneously record sleep and breathing activity. The concomitant analysis of WBP and DIA signals allowed the discrimination between CSA and OSA. The same protocol was then applied to 12 Ts65Dn mice (a validated model of DS) and 14 euploid controls. Results. OSAs represented about half of the apneic events recorded during rapid-eye-movement sleep (REMS) in each experimental group while almost only CSAs were found during non-REMS. Ts65Dn mice had similar rate of apneic events than euploid controls but a significantly higher occurrence of OSAs during REMS. Conclusions. We demonstrated for the first time that mice physiologically exhibit both CSAs and OSAs and that the latter are more prevalent in the Ts65Dn mouse model of DS. These findings indicate that mice can be used as a valid tool to accelerate the comprehension of the pathophysiology of all kind of sleep apnea and for the development of new therapeutical approaches to contrast these respiratory disorders.

Assessing the fit of unidimensional IRT models for binary data under model misspecification

Model misspecification affects the classical test statistics used to assess the fit of the Item Response Theory (IRT) models. Robust tests have been derived under model misspecification, as the Generalized Lagrange Multiplier and Hausman tests, but their use has not been largely explored in the IRT framework. In the first part of the thesis, we introduce the Generalized Lagrange Multiplier test to detect differential item response functioning in IRT models for binary data under model misspecification. By means of a simulation study and a real data analysis, we compare its performance with the classical Lagrange Multiplier test, computed using the Hessian and the cross-product matrix, and the Generalized Jackknife Score test. The power of these tests is computed empirically and asymptotically. The misspecifications considered are local dependence among items and non-normal distribution of the latent variable. The results highlight that, under mild model misspecification, all tests have good performance while, under strong model misspecification, the performance of the tests deteriorates. None of the tests considered show an overall superior performance than the others. In the second part of the thesis, we extend the Generalized Hausman test to detect non-normality of the latent variable distribution. To build the test, we consider a seminonparametric-IRT model, that assumes a more flexible latent variable distribution. By means of a simulation study and two real applications, we compare the performance of the Generalized Hausman test with the M2 limited information goodness-of-fit test and the Likelihood-Ratio test. Additionally, the information criteria are computed. The Generalized Hausman test has a better performance than the Likelihood-Ratio test in terms of Type I error rates and the M2 test in terms of power. The performance of the Generalized Hausman test and the information criteria deteriorates when the sample size is small and with a few items.

Design and testing of innovative, artificial intelligence-based techniques to model and control the combustion process and to reduce the Co2 emissions in modern, high-performance, gasoline direct injection engines

This work deals with the development of calibration procedures and control systems to improve the performance and efficiency of modern spark ignition turbocharged engines. The algorithms developed are used to optimize and manage the spark advance and the air-to-fuel ratio to control the knock and the exhaust gas temperature at the turbine inlet. The described work falls within the activity that the research group started in the previous years with the industrial partner Ferrari S.p.a. . The first chapter deals with the development of a control-oriented engine simulator based on a neural network approach, with which the main combustion indexes can be simulated. The second chapter deals with the development of a procedure to calibrate offline the spark advance and the air-to-fuel ratio to run the engine under knock-limited conditions and with the maximum admissible exhaust gas temperature at the turbine inlet. This procedure is then converted into a model-based control system and validated with a Software in the Loop approach using the engine simulator developed in the first chapter. Finally, it is implemented in a rapid control prototyping hardware to manage the combustion in steady-state and transient operating conditions at the test bench. The third chapter deals with the study of an innovative and cheap sensor for the in-cylinder pressure measurement, which is a piezoelectric washer that can be installed between the spark plug and the engine head. The signal generated by this kind of sensor is studied, developing a specific algorithm to adjust the value of the knock index in real-time. Finally, with the engine simulator developed in the first chapter, it is demonstrated that the innovative sensor can be coupled with the control system described in the second chapter and that the performance obtained could be the same reachable with the standard in-cylinder pressure sensors.

Pharmacological targeting of P-selectin to halt the progression of myelofibrosis in the Gata1low mouse model

Primary myelofibrosis is a clonal hematopoietic disorder characterized by marked degrees of systemic inflammation. The release of pro-inflammatory factors by clonal hematopoietic cell populations cause the remodeling of a specialized microenvironment, defined niche, in which the hematopoietic stem cells reside. The main source of pro-inflammatory cytokines is represented by malignant megakaryocytes. The bone marrow and spleen from myelofibrosis patients, as well as those from the Gata1low mouse model of the disease, contain increased number of abnormal megakaryocytes. These cells express on their surface high levels of the adhesion receptor P-selectin that, by triggering a pathological megakaryocyte-neutrophil emperipolesis, lead to increased bioavailability of TGF-β1 in the microenvironment and disease progression. Gata1low mice develop with age a phenotype similar to that of patients with myelofibrosis. We previously demonstrated that deletion of the P-selectin gene in Gata1low mice prevented the development of the myelofibrotic phenotype in these mice. In the current study, we tested the hypothesis that pharmacological inhibition of P-selectin may rescue the fibrotic phenotype of Gata1low mice. To test this hypothesis, we have investigated the phenotype expressed by old Gata1low mice treated with the anti-mouse monoclonal antibody against P-selectin RB40.34, alone or in combination with the JAK2 inhibitor Ruxolitinib. The results showed that the combined therapy normalized the phenotype of Gata1low mice with limited toxicity by reducing fibrosis, TGF-β1 and CXCL1 content in the BM and spleen and by restoring hematopoiesis in the bone marrow and the normal architecture of the spleen. In conclusion, pharmacological inhibition of P-selectin was effective in targeting malignant megakaryocytes and the microenvironmental abnormalities that affect the hematopoietic stem cell compartment in this model. These results suggest that P-selectin and JAK1/2 inhibitors in combination may represent a valid therapeutic option for patients with myelofibrosis.