Riassunto analitico
Nowadays the electric vehicles are still a small segment of the automotive market, and one of the main causes is that they are much more expensive than the internal combustion engine vehicles In the present is discussed the possibility of decreasing the costs related to the energy storage system, applying cheaper battery cells, i.e. LiFePO4 cells, and reducing the tests required to obtain a correct battery management. In Chapter 1 a quite large overview of the electrochemical storage systems is given and the LiFePO4 battery cells are introduced with their benefits and limitations. In addition, the theoretical basis to build a cell/battery model are given. In Chapter 2, a SOC estimator through Extended Kalman Filter is discussed and it is then applied to some real current profiles performed on al LiFePO4 cell. The Kalman filter (from R.E. Kalman, 1960) is a set of mathematical equations that provides an efficient computational (recursive) way to estimate the state of a process, in a way that minimizes the mean of the squared error. The filter is very powerful in several aspects: it supports estimates of states even when the precise nature of the modelled system is unknown, and can handle noisy signals. Important features include: Provides a statistically optimal estimate of the underlying system states from noisy observations; If the noise follows a Gaussian distribution, the Kalman filter minimizes the mean square error of the estimated states and parameters; It is a recursive method in which new measurements can be processed upon arrival; The filter can not only “clean up” the measurements from noise, but it also utilizes these measurements for state estimate. One of the key factors that must be available when designing a Kalman filter is the knowledge of the mathematical (linear) model of the dynamic system. In a battery, the entity of the instantaneous drop in voltage due to the application of current, that is the most relevant variation in voltage, is related to a physical characteristic of the cell, the R_dc. In practical application, R_dc is usually mapped in relation to SOC, temperature, state of health and current direction. This comports the necessity of a huge quantity of tests to cover the totality of possible conditions in which the cell will operate. In Chapter 3 an alternative method is showed. The R_dc is mathematically estimated from the measures of current and voltage through two method: Kalman Filter; Least Squares. The LS method has a high susceptibility to measurement noise and errors. In Chapter 4 this problem is analysed and a solution is discussed. Chapter 5 describes how to get the Ccodes of the R_dc estimators from the Matlab/Simulink model and then how to implement them on a real BMS. Finally, tests in laboratory are performed in order to verify the possibility of implementing the Rdc estimator in an operating battery pack.
