Battery Management System (BMS) SOC

The complex electrochemical and physical processes that take place in a cell lead to a range of externally observable behaviors. These behaviors exist in the relationship between voltage and current (and, to a certain extent, temperature). Since the primary purpose of a lithium-ion battery is to be an energy storage device in a circuit, it is often useful to represent these behaviors as an equivalent circuit. Other types of models can be used based on the basic physical and electrochemical processes that take place inside the cell. A basic assumption in equivalent circuit modeling is the existence of an observable state variable (or set of observable state variables) as a function of the state of charge. In most cases, the relationship between the completely relaxed open circuit voltage and the charge state is the one used. It is usually assumed that this relationship is constant with temperature, degree of aging, and cycle life. In battery management system implementations, the goal of battery modeling is to create a way to convert easily measured quantities (current, voltage, and temperature) into accurate representations of internal states that are not easily measured, such as charging states. Many techniques are used to determine the charging status of other types of batteries, such as simple voltage lookup tables for simple applications and load profiles.

The open circuit voltage is assumed to be a function of the charge state defined by VOC(SOC). This relationship is monotonous and is defined as a unique SOC mapping to a unique VOC. This relationship need not be linear, exponential, or even conveniently approximated by a closed mathematical relationship. This means that if VOC is known, SOC is also known. Furthermore, if the error in VOC can be bounded, so can the error in SOC. The purpose of modeling the equivalent circuit cell is to calculate the dynamic overpotential (the difference between the open circuit voltage and the actual measured terminal voltage). If the overpotential is equal to Vd, VOC=Vt−Vd, allowing the SOC to be calculated from the measured voltage and calculated overpotential. Physics-based models may not take advantage of empirical SOC-OCV relationships.

Battery Battery voltage also depends on SOC; Thus, like capacitors, battery soCs do not measure energy storage. Battery voltage decreases as SOC decreases, initially at low slope, and then faster at DOD(DOD=1−SOC) reaching 1. Some chemicals (e.g., lithium iron phosphate) have a very flat voltage with the MOD curve until the MOD is large and the curve drops sharply. Others (such as the PbC) follow a fairly linear curve to the MOD, similar to a capacitor. All batteries show DOD voltage drop. As with capacitors, the capacity of a battery depends on the operating voltage range. The battery is fully charged when it maintains its chemical Vh specified battery voltage (e.g., 4.2V for lithium-ion batteries) in a steady state and at room temperature. This can be achieved by, for example, trickling charging for a reasonably long time at Vh. Vh voltages are limited by side reactions that damage and reduce battery life at high pressures. Fully discharged batteries provide a maximum ampere hour when fully charged, with a steady-state voltage of Vl at room temperature (for example, 3.0V for lithium-ion batteries). The Vl voltage is also selected to limit battery damage. The voltage curve with DOD is sharply skewed at high DOD, so operating at reduced Vl may not result in a significant increase in stored energy. High discharge rates, low temperatures, and aging can significantly reduce battery capacity. Compared to the C/30, a battery with 80%SOC can only provide rated ampere-hours because the battery voltage drops rapidly to Vl before reaching the rated capacity. However, if the high rate discharge stops, the voltage can be restored and further capacity can be removed from the battery. The nominal capacity of old batteries, or those that operate at low temperatures, may be significantly lower than the nominal capacity of fresh batteries at room temperature. Similarly, an 80% SOC of an old or cold battery may mean that only 20% of room temperature and 20% of the capacity of a fresh battery can be discharged.

From a systems perspective, the SOC mirrors the gas meter of a conventional vehicle powered by ice. The distance traveled on a quarter tank, or distance, depends on the rate of fuel consumption. In an ideal situation, speed is proportional to the rate of fuel consumption. In practice, however, range estimation is complicated by many unknowns. Vehicle weights and road classes are not known in advance. Rink test efficiency and air resistance depends on speed. At stop lights or traffic jams, efficiency drops to zero. However, unlike a battery SOC, a gas meter measures the amount of energy left in the tank. Many decisions made by the HEV control system depend on the SOC. The BMS must decide in real time whether to use mechanical or regenerative braking during deceleration or to use battery packs or ICE to extract power during acceleration. These decisions are usually based on SOC. Unlike a gas meter, however, no sensor can measure SOC directly, so estimates must be used. The accuracy of such estimates is critical to the proper, safe and effective performance of HEV.

Battery voltage and temperature are rarely measured, and some packages may not be able to measure individual battery voltage and temperature. Batteries and batteries can be formed by series and parallel connections of batteries and batteries respectively. However, the battery and battery soCs are averaged over individual cells and battery soCs, respectively. We also assume that the BMS uses only charge and discharge currents to meet application requirements. The SOC estimation scheme does not control the battery current and introduces pulse or sinusoidal scan into the battery current to aid in SOC estimation. This would require complex and expensive equipment and limit its availability for intended uses. Finally, we assume that the sensor is sampling at a limited bandwidth, typically around 10Hz. Therefore, high-frequency dynamics do not need to be modeled and cannot be relied upon to aid SOC estimation.

SOC provides C/30 amps for 30 hours for 100% battery. This does not necessarily mean that the battery can provide 0.5 hour amperes of 2C. The diffusion process in the cell produces uneven concentration and potential distribution. Ions can only move so fast in a battery, and if the current current is too high, the voltage drops sharply, leading to a sudden loss of power. Just like the fuel in the tank of an ICE vehicle flows through a very small tube, the rate limit. However, the rate-limiting process in ICE is usually not fuel supply, but combustion. Increase fuel flow above rate limit (depending on displacement, compression ratio, fuel injection, etc.). It doesn't increase the power output, it just floods the engine. The ring-limiting nature of batteries means that their design and integration into the powertrain is more critical and complex than a simple fuel tank.