Why don’t the batteries of electric cars accurately show how much they have left?

So, back to the original question, why do electric vehicles (and lead batteries) seem inaccurate? This is because the power of electric vehicles (usually called SOC, or state of charge) is more difficult to measure than the power of mobile phones.

There are several reasons why electric cars are more difficult to estimate than mobile phones. Here are some deeper points:

Commonly used SOC estimation methods:

Let us start with one we know better: GPS. Now, mobile phone-based GPS positioning is accurate to the order of meters. As far as missiles are concerned, such positioning is not enough. Two things are missing: accuracy and real-time (ie, the number of seconds to successfully locate). So the missile has another compensation system: gyroscope.

Gyroscopes are a complete complement to satellite GPS positioning-they are accurate (at least on the millimeter scale) and real-time, but the problem is that the errors are cumulative. For example, if you blindfold a person and ask him to walk in a straight line, you may not see tens of meters, but you may be able to walk several kilometers and turn 180 degrees.

Is there a way to fuse the information between GPS and gyroscope to complement each other to obtain the most accurate position? The answer is yes, Kalman filtering is good, that’s it.

What does this have to do with the battery’s SOC estimation? There are two commonly used methods of measuring SOC:

The first method is the open circuit voltage method, which measures the state of the battery based on the open circuit voltage of the battery. This is easy to understand, but full high battery voltage, low battery power, power and voltage are corresponding. This method is similar to satellite GPS positioning, there is no cumulative error (because it is based on conditions), but the accuracy is low (due to various factors, the previous response has now been explained).

The second type is called ampere hour integrator, which measures the state of the battery by integrating the voltage (flow) of the battery. For example, if you charge a 100-kilowatt battery, and each time you measure the current and increase it to 50 degrees, the remaining power will be 50 degrees. This method can be compared with high-precision gyroscopes (instantaneous measurement accuracy is less than 1%, 0.1% measurement cost is low and common), but the cumulative error is larger. In addition, even if the ammeter is a god brand and completely accurate, the integral method and the open circuit voltage method are inseparable when the ammeter is used. Why? Because the nature of the battery itself will change.

Academically, maybe some advanced car companies combine open circuit voltage and ampere-hour integration with Kalman filter algorithm to get the most accurate SOC estimation, but they often make mistakes because they don’t understand how to deepen the evolution of batteries. nature.

Electric vehicles developed by domestic automobile companies generally use the open circuit voltage method and the LAMV integration method: the following car leaves enough time (for example, the car is only turned on in the morning), and the open circuit voltage method is used to estimate the force start, SOC_start said. After the car is started, the battery status becomes chaotic, and the open circuit voltage method is no longer valid. Then, use the SOC_start-based amV integration method to estimate the current battery power.

An important factor that makes the SOC of electric vehicles difficult is the difficulty in modeling lithium battery packs. Or in other words, the research field that can make electric vehicle SOC estimation become more and more accurate. The nature of batteries and battery packs is the key direction. Improving instrument accuracy is not the current research direction that is accurate enough, and no matter how accurate it is, it is useless.