Mono Posted July 29, 2015 Share Posted July 29, 2015 I wonder if anyone knows, or can derive, the optimal control law to achieve minimal braking distance? Is the law for the shortest braking time (to zero velocity) different? Do the results depend on whether a balanced end position is required?EDIT: in case anybody wonders, the actuator (the controller acts upon) would be the (longitudinal) inclination angle of the wheel, i.e. the pedals. Link to comment Share on other sites More sharing options...
esaj Posted July 29, 2015 Share Posted July 29, 2015 I cannot offer you any formal proof, but this is what I've figured using plain common sense (and might be just wrong ):So, when moving forwards there's your forward momentum and gravity, causing the actual force to be angled towards the ground. To stop as fast as possible, I lean back, push the wheel in front of me and push the pedals down to tell the motor to start braking very strongly (what I call "power braking"). My common sense says that by pushing towards the actual direction of force, the most effective (and fastest) braking is achieved, but maybe someone who knows physics better could explain it better (or just point out that my thought is plain wrong ). Link to comment Share on other sites More sharing options...
Mono Posted July 29, 2015 Author Share Posted July 29, 2015 I already figured out (common sense ) that the first thing to do is apply maximal forward inclination angle on the pedals. This is what you describe as "lean back, push the wheel in front of me" (the feeling of leaning back is just the consequence of forward inclination angle on the pedals). But after that? "Pushing towards the actual direction of force" seems to beg the question, as the only original force is down (assuming no friction, which seems a good enough model for this purpose). So, you need to create the force in the first place in order to push against it and the question remains what is the optimal course of force to create. However, this question seems indeed to have an answer: it is the force which just can be handled by the tire without starting to slide, i.e., it must be constant. To generate constant deceleration force (at the limit of what the tire can handle), the lean-back angle needs to increase during breaking, because the speed goes down (and therefore the force, given constant angle). From this it seems feasible to compute the required lean-back angle which is more or less already the solution to the problem. Link to comment Share on other sites More sharing options...
Gimlet Posted July 29, 2015 Share Posted July 29, 2015 You all seem to be forgetting that just as you can easily out lean even the most powerful wheels on acceleration so you can out brake the wheel. When this happens they tend to go into freewheel mode and shoot off in front of you. I've had it happen on my IPS's and my Gotway M10's and mcm2s but not with my M18 yet, but that's probably because I've learnt to be careful going down hills and I don't fancy the heavy weight M18 shooting off on it's own. Link to comment Share on other sites More sharing options...
esaj Posted July 29, 2015 Share Posted July 29, 2015 You all seem to be forgetting that just as you can easily out lean even the most powerful wheels on acceleration so you can out brake the wheel. When this happens they tend to go into freewheel mode and shoot off in front of you. I've had it happen on my IPS's and my Gotway M10's and mcm2s but not with my M18 yet, but that's probably because I've learnt to be careful going down hills and I don't fancy the heavy weight M18 shooting off on it's own.Yes, this is an important point; unlike in "normal" deceleration (regenerative braking), in power braking you are using the motor to brake (ie. asking it to start rolling in opposite direction by tilting the pedals backwards), and that requires LOTS of power from the motor. It hasn't happened to me with the Firewheel, but with the 14" generic, doing this on low battery is downright dangerous, as the motor/batteries cannot give enough power. It never shot from under me, but I had a few close calls... Link to comment Share on other sites More sharing options...
Mono Posted July 29, 2015 Author Share Posted July 29, 2015 That's really terrible and scary, though in answering the question it doesn't seem to change much: in this case one would have to be as close and as long as possible in the limit deceleration the engine offers (instead of the limit grip the tire offers). Do you get warnings before this happens? Link to comment Share on other sites More sharing options...
esaj Posted July 29, 2015 Share Posted July 29, 2015 Do you get warnings before this happens? Not on my wheels, the 14" generic only has that beeping when going above 12km/h, and the Firewheel has never played back any messages or beeped when doing this (but I've never had a situation where it would even feel like I was overpowering the Firewheel during power braking). Might be different with other wheels. Probably the warning would need to be current-related, as demanding high power from the batteries drops the voltage, which leads to even more current needed to provide the same amount of watts (power) than with higher voltage (P = U*I). Link to comment Share on other sites More sharing options...
Mono Posted July 29, 2015 Author Share Posted July 29, 2015 @esaj, I am not quite sure I understand why braking becomes non-regenerative. Is the maximal current one can drag from the battery (much) larger than the maximal current it can digest? And kudos to the engineers/programmers, if they can produce this switch in a smooth manner. Link to comment Share on other sites More sharing options...
esaj Posted July 29, 2015 Share Posted July 29, 2015 @esaj, I am not quite sure I understand why breaking becomes non-regenerative. Is the maximal current one can drag from the battery (much) larger than the maximal current it can digest? I don't know enough about DC motors and electronics to actually answer, but I do know that the braking power is MUCH less on the 14" when the battery is low. So I've just concluded that it must use the power from the batteries (on top of regenerative energy?), as if the regenerative energy alone was enough, the braking power shouldn't drop with the battery charge. And usually the Li-Ion battery cells have higher discharge C-ratings than charge C-ratings, for example the LG 18650MH1's that are being used for my (yet to arrive) custom battery packs have 1C (3.2A) max charge current, 3+C (around 10A) continuous discharge current and pulse discharge current up to 20A. Link to comment Share on other sites More sharing options...
Gimlet Posted July 29, 2015 Share Posted July 29, 2015 Just as the magnetic attraction between the permanent magnets and electro magnetic coils that drive the wheel has it's limits in driving the wheel it can be overcome by too much force driving it. Nothing has limitless power to either drive or resist. Link to comment Share on other sites More sharing options...
Rotator Posted July 29, 2015 Share Posted July 29, 2015 @esaj, I am not quite sure I understand why breaking becomes non-regenerative. Is the maximal current one can drag from the battery (much) larger than the maximal current it can digest? And kudos to the engineers/programmers, if they can produce this switch in a smooth manner. A shorcircuit in a motor, brakes the motor, but only to a given point.Even if the power mosfet transistors are very low resistance, and battery can eat all the current generated by the motor, it could be insufficient to stop the euc. So it might be necessary to try to run the motor in the opposite direction to brake. And this will use a lot of energy from the battery. Link to comment Share on other sites More sharing options...
hobby16 Posted July 29, 2015 Share Posted July 29, 2015 Even if the power mosfet transistors are very low resistance, and battery can eat all the current generated by the motor, it could be insufficient to stop the euc. So it might be necessary to try to run the motor in the opposite direction to brake. And this will use a lot of energy from the battery.Are we sure of that ? What I know for sure, for having read some flowcharts and source code of BLDC controllers that the decision between regenerative braking and dissipative braking (that is a short of the motor coil) is decided for each commutation cycle (the selection is simply made by deciding which mosfet pair among the six to commute). I thought running in the opposite direction for braking is inefficient (because of emf) or even impossible (because of voltage surge) unless braking down to zero speed. Anyway, it must be a hell lot of brain twisting things to write.If only I had been more studious at the advanced control courses in my student years (we had for example to study the control loop of the Leclerc tank, which has the notable peculiarity of being able to run AND shoot with high accuracy, and of giving headache to students). Link to comment Share on other sites More sharing options...
esaj Posted July 29, 2015 Share Posted July 29, 2015 Are we sure of that ? What I know for sure, for having read some flowcharts and source code of BLDC controllers that the decision between regenerative braking and dissipative braking (that is a short of the motor coil) is decided for each commutation cycle (the selection is simply made by deciding which mosfet pair among the six to commute). I thought running in the opposite direction for braking is inefficient (because of emf) or even impossible (because of voltage surge) unless braking down to zero speed. Anyway, it must be a hell lot of brain twisting things to write.If only I had been more studious at the advanced control courses in my student years (we had for example to study the control loop of the Leclerc tank, which has the notable peculiarity of being able to run AND shoot with high accuracy, and of giving headache to students).Ok, so apparently the control software chooses BETWEEN regenerative and "power braking". That would explain why the more effective power braking loses power when the batteries are more empty, although I cannot claim to really understand what happens in the motor... Link to comment Share on other sites More sharing options...
Daan Posted July 29, 2015 Share Posted July 29, 2015 Are we sure of that ?I spoke recently with Shane Chen and he confirmed that on the Solowheel at least, when braking, the wheel first uses regeneration and when even more braking power is needed it will actually switch to driving the wheel backward. This is why when going downhill with a full battery you will get a battery overcharge warning, but if you go really slow (ie. brake a lot) then the warning goes away since you are actually using power. Just my 2c. Link to comment Share on other sites More sharing options...
hobby16 Posted July 29, 2015 Share Posted July 29, 2015 Ok, so apparently the control software chooses BETWEEN regenerative and "power braking". That would explain why the more effective power braking loses power when the batteries are more empty, although I cannot claim to really understand what happens in the motor... Indeed, depending on the braking power required, the controller can choose between regenerative braking (weak) and dissipative braking (strong). But dissipative braking is not "power braking", it consists just of shorting the coils and requires no power, so it should not depend on the battery's state of charge.I don't brake enough to confirm what you see about less braking power with empty batteries. If it's true, then there is really a "power braking" which should be something more than just regenerative or dissipative braking. Link to comment Share on other sites More sharing options...
esaj Posted July 29, 2015 Share Posted July 29, 2015 Indeed, depending on the braking power required, the controller can choose between regenerative braking (weak) and dissipative braking (strong). But dissipative braking is not "power braking", it consists just of shorting the coils and requires no power, so it should not depend on the battery's state of charge.I don't brake enough to confirm what you see about less braking power with empty batteries. If it's true, then there is really a "power braking" which should be something more than just regenerative or dissipative braking.I've only ever noticed it on the 14" generic, when braking strongly with low battery (usually 1 led, or 2 but close to one), the wheel doesn't "resist" moving forwards as much as with fuller battery, and there's a clear danger of kicking it out from under you (when doing the leaning-back/pushing wheel forwards/pushing pedals to backward tilt -kind of braking). With Firewheel, this has never happened.If the motor is normally driven by switching on the coils to create a force "pulling" towards the magnets (attracting each other), could the "third kind" of braking then be the coils activating "in reverse", so that they try to push away from magnets (repulsing each other) to gradually start turning the engine in reverse (of course first bringing it to full stop)? Or then I've misunderstood the entire idea of how brushless DC motors work... EDIT: Consider this animation: http://educypedia.karadimov.info/library/2pole-bldc-motor031102.swfIf the motor was already turning and the coils were switched in "reverse" (attraction/repulsion), wouldn't it use power and slow down the turning? Link to comment Share on other sites More sharing options...
Daan Posted July 29, 2015 Share Posted July 29, 2015 Indeed, depending on the braking power required, the controller can choose between regenerative braking (weak) and dissipative braking (strong). But dissipative braking is not "power braking", it consists just of shorting the coils and requires no power, so it should not depend on the battery's state of charge.I don't brake enough to confirm what you see about less braking power with empty batteries. If it's true, then there is really a "power braking" which should be something more than just regenerative or dissipative braking.Huh, not sure I fully understand all this. I got the impression that the wheel is actually put 'in reverse' and therefore 'brakes'. This means that it will take regular battery power right? Perhaps I misunderstood?? Link to comment Share on other sites More sharing options...
hobby16 Posted July 30, 2015 Share Posted July 30, 2015 If the motor is normally driven by switching on the coils to create a force "pulling" towards the magnets (attracting each other), could the "third kind" of braking then be the coils activating "in reverse", so that they try to push away from magnets (repulsing each other) to gradually start turning the engine in reverse (of course first bringing it to full stop)? Or then I've misunderstood the entire idea of how brushless DC motors work... EDIT: Consider this animation: http://educypedia.karadimov.info/library/2pole-bldc-motor031102.swfIf the motor was already turning and the coils were switched in "reverse" (attraction/repulsion), wouldn't it use power and slow down the turning? I just asked a collegue at work, quite experienced in sophisticated control algorithms. He said in fact it's quite possible to do both, dissipative braking and rotation in reverse to increase braking power. Dissipative braking is triggered when the rotor lags or is in phase with the stator, it's where braking force is at its peak. But when the rotor leads the stator, power can be applied to try to "pull back" the rotor, thus the braking. And then again for the next cycle, ect, it's decided on a cycle to cycle basis.Normally, dissipative braking is made using an external big shunt resistor to dissipate energy by Joule effect, it's an ubiquitous mechanism found on nearly all big motors, machining stations, elevators... But here, he told me if the mainboard does not have such shunt resistor, it's not even sure dissipative braking would be possible without grilling the mosfets or the motor's coil. Link to comment Share on other sites More sharing options...
Gimlet Posted July 30, 2015 Share Posted July 30, 2015 I was under the impression that IPS used that method for braking?Not sure, it's just from something @Jason McNeil said elsewhere. Link to comment Share on other sites More sharing options...
UKJ Posted November 18, 2015 Share Posted November 18, 2015 Can anyone explain how to, if possible, to slow down the video. I would really like to see and fully understand this, both for E U Cs and my Tiny oh so strong and fast R C heli motors. Thanks. ukj Link to comment Share on other sites More sharing options...
Villac Posted November 18, 2015 Share Posted November 18, 2015 8 minutes ago, UKJ said: Can anyone explain how to, if possible, to slow down the video. I would really like to see and fully understand this, both for E U Cs and my Tiny oh so strong and fast R C heli motors. Thanks. ukj Click on "step through animation" triangle/arrow, you can run it at the speed you choose by the speed you click. As a non-engineer, I question: aren't all these kinds of braking (and accelleration, and static speed for that matter) all forms of the same thing? The difference is just A) is the battery giving energy or receiving energy and B ) if it is giving (using) energy/work, is it doing work in the direction of travel or against the direction of travel? So I don't think the board "decides" to do anything, the driver decides whether to make the motor idle, work with the vector of travel, or against the vector of travel. The only "decision" the board makes is to shut down the whole show to darkness if a low voltage threshold is crossed. And from a personal observation, I think my EU has given me a warning sign that it was about to give an under-voltage shutoff during braking: There is a little bit of lurching, shuddering forward on hard braking that I (maybe erroneously) interpreted as the stator magnets "skipping" polarity from one coil to the next, an then locking back on the next magnetic polarity. Link to comment Share on other sites More sharing options...
UKJ Posted November 18, 2015 Share Posted November 18, 2015 @Villac Thanks! Now to study. ukj Link to comment Share on other sites More sharing options...
MarkoMarjamaa Posted November 18, 2015 Share Posted November 18, 2015 Could someone with accurate (voltage?) logging check if when power braking, do we start the braking with accelerating the wheel more in front? If you are already riding at max speed, you maybe cannot move the wheel in front of you fast enough. If you know the max power of wheel and rider's weight, it should be possible to calculate the most effective braking. So could someone make the formula ? :) Link to comment Share on other sites More sharing options...
Jurgen Posted November 19, 2015 Share Posted November 19, 2015 On Wednesday, November 18, 2015, Villac said: On Wednesday, November 18, 2015, Villac said: Click on "step through animation" triangle/arrow, you can run it at the speed you choose by the speed you click. As a non-engineer, I question: aren't all these kinds of braking (and accelleration, and static speed for that matter) all forms of the same thing? The difference is just A) is the battery giving energy or receiving energy and B ) if it is giving (using) energy/work, is it doing work in the direction of travel or against the direction of travel? So I don't think the board "decides" to do anything, the driver decides whether to make the motor idle, work with the vector of travel, or against the vector of travel. The only "decision" the board makes is to shut down the whole show to darkness if a low voltage threshold is crossed. And from a personal observation, I think my EU has given me a warning sign that it was about to give an under-voltage shutoff during braking: There is a little bit of lurching, shuddering forward on hard braking that I (maybe erroneously) interpreted as the stator magnets "skipping" polarity from one coil to the next, an then locking back on the next magnetic polarity. Click on "step through animation" triangle/arrow, you can run it at the speed you choose by the speed you click. As a non-engineer, I question: aren't all these kinds of braking (and accelleration, and static speed for that matter) all forms of the same thing? The difference is just A) is the battery giving energy or receiving energy and B ) if it is giving (using) energy/work, is it doing work in the direction of travel or against the direction of travel? So I don't think the board "decides" to do anything, the driver decides whether to make the motor idle, work with the vector of travel, or against the vector of travel. The only "decision" the board makes is to shut down the whole show to darkness if a low voltage threshold is crossed. And from a personal observation, I think my EU has given me a warning sign that it was about to give an under-voltage shutoff during braking: There is a little bit of lurching, shuddering forward on hard braking that I (maybe erroneously) interpreted as the stator magnets "skipping" polarity from one coil to the next, an then locking back on the next magnetic polarity. Jason Mc Neil made a datalog of hard accelerating at http://forum.electricunicycle.org/topic/748-ips-zero/?page=21 doesn't the same happen during hard braking (apart from different behaviour of the batteries; hysteresis effect)? just my 2c I Link to comment Share on other sites More sharing options...
Jurgen Posted November 19, 2015 Share Posted November 19, 2015 Jason Mc Neil made a datalog of hard accelerating at http://forum.electricunicycle.org/topic/748-ips-zero/?page=21 doesn't the same happen during hard braking (apart from different behaviour of the batteries; hysteresis effect)? just my 2c Link to comment Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.