What statistics in battery pack BMS would you like to track? My company is developing some new features in the Emus BMS and one of them is usage statistics. It should be something simple, described in a few numbers or a minimal chart. For example, what is an average SOC (state of charge) at the end of trip (before charging), how often battery goes to LVC condition, how long in average it usually rests between charge/drive, ... etc. So I'd like to ask you, the end users, what would be the most usage-and-health describing parameters?
Battery usage statistics calculation
- Thread starter circuit
- Start date
theRealFury
10 kW
Hello,
Only others i can think of right now that are east to log would be.
1) Recharge Cycles completed.
2) Total ah/wh supplied by pack since purchase.
Only others i can think of right now that are east to log would be.
1) Recharge Cycles completed.
2) Total ah/wh supplied by pack since purchase.
jonescg
100 MW
Some kind of datalogging of average currents being drawn from the battery would be good. Then from this information, you could determine how fast the watt-hours were being used. How many watt-hours used per cycle would also be handy information. If not for the benefit of the user, then definitely the supplier of the battery!
Kirk
100 W
The current Voltage and Ah of the pack
dogman dan
1 PW
First of all, voltage of each cell group.
then like the other say, cycles number, total wh used to date, current pack voltage, wh used this cycle, max amps discharged.
In other words, the bms would be a cycleanalyst, but have some cell level data too.
then like the other say, cycles number, total wh used to date, current pack voltage, wh used this cycle, max amps discharged.
In other words, the bms would be a cycleanalyst, but have some cell level data too.
Voltage of each cell and total pack voltage is already done:
[youtube]7sUlY9MaTL4[/youtube]
This includes Ah state and SOC (State Of Charge), which is displayed in % (calculated total Wh of the pack)
"max amp discharge" would be useful, I think, only to the user, in some sort of "amp competition". This parameter has no information about how the battery was used, as that amp peak may have occurred only for one second in whole usage period of, say, 5 years. Some kind of "smarter" statistical value should be used. Or a deviation graph.
Cycle counting is, also, not that easy. What do we count as a cycle? 100% charge-discharge? 80%? What about 20? There might be a count of "trips" and "charges" (not essentially full), but next to it there should be an average DOD per trip and some statistical value to determine how "sharp" the deviation peak (the average) is. Or just provide whole DOD deviation graph...
Any more suggestions?
[youtube]7sUlY9MaTL4[/youtube]
This includes Ah state and SOC (State Of Charge), which is displayed in % (calculated total Wh of the pack)
"max amp discharge" would be useful, I think, only to the user, in some sort of "amp competition". This parameter has no information about how the battery was used, as that amp peak may have occurred only for one second in whole usage period of, say, 5 years. Some kind of "smarter" statistical value should be used. Or a deviation graph.
Cycle counting is, also, not that easy. What do we count as a cycle? 100% charge-discharge? 80%? What about 20? There might be a count of "trips" and "charges" (not essentially full), but next to it there should be an average DOD per trip and some statistical value to determine how "sharp" the deviation peak (the average) is. Or just provide whole DOD deviation graph...
Any more suggestions?
ptd
100 W
how about what amount of time was spent at what rate of discharge (AND charge, be it charger, or regen). small equalizer graph kinda thing. could be mined a lot of different ways. total wattage used. expected deterioration, and the resulting cycle life. an order of magnitude more accurate volt meter might give you good data about efficiency, capacity, AND health. could also be useful in realizing a cell was going bad BEFORE it went bad, and then you figure out a way to limit drain of that cell earlier on. i mean, theoretically, you want all the cells to reach the LVC at the same time, right ? so is there a way to proportionally limit current ? using the BMS ? methinks so. like something proportional to sag, or it's deviation from the mean. or is it more like a weak link kinda thing ? would it be possible to just have an effect similar to limiting the throttle instead ? so that instead of pwm, you just have an upper level, dropping off, as individual cells kicked off ? you'd have reduced power, but at least you'd have power. and it seems like you wouldn't be pushing any cell past it's limit. ok, maybe sleep deprivation is making me loopy. anyways, love the bluetooth, and the real screen thing. justin needs to take the CA to that level, for $125. and the datalogger should be integrated into it. and a fixed memory and bluetooth might not be too much more, versus SD slot and removeable card. bit more weatherproof, too.
Similar threads
