I have been developing a control software for the three phase voltage source inverter for three phase induction motor control. There is a requirement to implement regenerative braking algorithm where the kinetic energy of the drive is transformed into the electrical energy which is then disipated in a braking circuit (resistor along with transistor in series) in parallel to the dc link capacitor.

enter image description here

My question regards the control algorithm of the braking circuit. I am going to implement the control algorithm of the braking circuit in following manner:

enter image description here

Do you think that the suggested control algorithm could work? Does anybody know better approach how to control the braking circuit? Thanks in advance for any ideas.

  • 2
    \$\begingroup\$ Regenerative braking doesn't dissipate into a resistor - it reuses the energy. Also, what does this mean: kinetic energy of the drive \$\endgroup\$
    – Andy aka
    Commented Jul 2, 2020 at 7:40

2 Answers 2


As the user Andy already told you, regeneration means that you dump the energy back into the mains. This is rather dynamic braking with braking resistor. There is no need of PI controller or something other complex algorithm, rather an on/off controller with hysteresis. It turns the transistor on at certain DC link threshold value and then it turns off, when the DC link voltage drops to a certain value.

  • \$\begingroup\$ thank you for your reaction. I am sorry for wrong terminology. Do you see any problem in case I will use the suggested control scheme? \$\endgroup\$
    – Steve
    Commented Jul 2, 2020 at 11:00
  • \$\begingroup\$ Yes, the control loop has to be stable. The control loop doesn't need to have a precise regulation and has to have a very quick response. The on/off control is always stable, you can limit the switching frequency by setting an appropriate hysteresis It can be hardware implemented by using a comparator, without MCU jitter/delay. \$\endgroup\$ Commented Jul 2, 2020 at 13:07

In addition to the basic on/off control with hysteresis as described in another answer, you might want to consider another alternative. If the input voltage is unregulated, it may be necessary to distinguish regeneration from the inverter from variations of the supply voltage. If the input voltage is rectified AC from the utility, you can measure the differential between the DC link and a small separately rectified supply. Another alternative is to detect the direction of the DC link current.

Other Considerations

Normal motor and inverter losses will absorbed some braking energy. That may reduce the need for quick braking response. Also, the inverter will tolerate some rise in voltage, so the DC link capacitor will reduce the need for quick response.

Braking torque that is required to reduce the speed of an inertia is proportional to the rate of speed reduction. Therefore, limiting the rate of deceleration can reduce or eliminate the need for braking.

Traction Drive

Since the motor is labeled "traction-motor" this may be a vehicle traction drive. In that case, you might want to consider a torque control loop for the inverter. The torque command would come from a manual speed control loop for which the operator would observe the performance of the vehicle and adjust the driving or braking command to suit the observed vs. desired performance.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.