Its been tested a ton, in the industrial arena. I'm an EE and used to work for GE Energy building their Wind Turbines and testing their driver boards and IGBT's. My father also works there and we spoke briefly about controller design and test. I'm currently set to discuss with him a design on the weekend of the 4th that takes into account: Regen, PWM of the drivers, PWM of the forward and reverse relays, software control, dynamic braking (brake resistor) and current, torque and speedlimit.
My idea is to drive the drivers PWM as normal, but when coasting, regen. When the brake is applied, apply a small resistor to act as a brake. it'l dissipate energy, and waste it, but it'l also allow you to stop without use of the brakes, or at least in adddition to (bigger resistors will lock the motor and throw you off). Also, things to consider are current limit, torque limit and speed limiting. I work as a controls engineer now, and worked a ton with VFD's. They're AC but they use braking resistors to dissipate the regen energy when trying to stop a motor. Usually, under normal operating the motors are driven. If they slow down, the regen charges the DC bus and is actually reused. When the machine is told to STOP, or the DC bus hits a peak, the DC brake is applied at a certain Rate (ramp down) until it stops. Regen is seen from the leads of the motor supply, What you do with it (brake or pour the energy back into the batteries) is up to you. its wasted energy, and you could run your accessories off it. Technically, the batteries WILL recharge some, but for a small bike, there just isn't enough force to regen anything more than an amp or so. It DOES extend the range, but not a ton. BUT if you can do it now, once batteries and controllers are more efficient, it'l be easier to integrate into a design that already allows for it. Anyone interested in building a controller? I work with engineers, and have a ton of resources for a side project like this.
