Electric Skateboards and Wind Turbines?

jens_overby · November 8, 2024, 7:40pm

So this is the wind-“MPPT” algorithm I have used (some years ago). It kind of worked…

It makes sense, because the power in the wind increases with the cubed windspeed (windspeedwindspeedwindspeed).

It’s an approximation of the physics… My tip-speed-ration TSR was high, because it was a two-blade turbine.

jens_overby · November 8, 2024, 7:51pm

Here’s a link to the arduino-code, and sorry it’s a mess. The algorithm is in there. (the code was not running with any Vesc-hardware at that time)

github.com

JensOverby/moelle/blob/main/turbinecontrol.ino


unsigned long timeStamp_verbose = 0;
unsigned long rpm_time = 0;

bool positive = true;
void sample(float k)
{
  float k1 = 1. - k;
  Vin_raw = (analogRead(ADC_VOLTAGE_IN)/1024.) * vcc / voltageDividerVoltsPerVolt;
  Vin_filter = k1*Vin_filter + k*Vin_raw;
  Vout_raw = (analogRead(ADC_VOLTAGE_OUT)/1024.) * vcc / voltageDividerVoltsPerVolt;
#ifdef DEBUG
  Vout_filter_dbg = k1*Vout_filter_dbg + k*Vout_raw;
  Vout_raw = 12.3;
#endif
  Vout_filter = k1*Vout_filter + k*Vout_raw;
  //Iout_raw = ((analogRead(ADC_CURRENT_OUT) - 509)/1024.) * 5. / acs712VoltsPerAmp;// - 0.02;
  Iout_raw = (analogRead(ADC_CURRENT_OUT) - 53) / 9.5;
  Iout_filter = k1*Iout_filter + k*Iout_raw;

This file has been truncated. show original

jens_overby · November 8, 2024, 8:15pm

jens_overby · November 8, 2024, 8:28pm

Last comment:
From the above it appears that you DON’T need to subtract the 35W from the calculated power, because losses are built into the Cp constant. So:

Power = POWER_FACTOR * windspeed³

Tune the algorithm to fit your turbine by modifying the TSR and Cp values.

If you want to get an explanation of the algorithm, just feed the whole thing to ChatGPT

usus · November 11, 2024, 4:03pm

Thank you for the feedback.Are a lot of questions by let’s take them one by one.

So wind speed is just a teoretical one? I dont understend why is devided by 35 (valFloat)
windspeed = rpm * RPM_TO_TIP_SPEED_FACTOR / valFloat[TipSpeedRatio_ID]; // Unit is meter/sec

The cage with the motor is probably reaching 40kg so the mass will act like a dealy.

Meanwhile I work on the mechical side. Most of the parst are fabricated and need to assemble them.This is a sneak peek (is top botom)

Online calculator Wind Power Calculators for various wind turbines－HAWT/VAWT – Smart Servo Home

From this I can see that cage speed is quiet height.
Then the voltage is 1 v/10 rpm for this generator reachign 200V.
In the same time i was looking inside generator that is in Star and i can change wires in Delta.

For now my target is to install and see if the turbine is self limiting RPM of the cage.

usus · November 11, 2024, 9:12pm

I was wrong before “I dont understend why is devided by 35 (valFloat)” is a nonsense.

usus · November 12, 2024, 10:26am

Is RPM_TO_TIP_SPEED_FACTOR 0.120427718 caluclated or determine by testing?

jens_overby · November 24, 2024, 9:14pm

No, it’s a straight forward figure given by: Circumference / 60 sec.

The Circumference of the swept area is 2*PI *Radius

(Circumference is the length the tip travels in 1 rotation, hence the unit: rotations per 60sec, or simply rpm)

jens_overby · November 24, 2024, 9:18pm

And in theory a two-bladed turbine can generate as much energy as a 3-bladed. It’s the swept area that matters.

BUT obviously you have a VAWT… Some of the constants might be different here. You can try give the equations for the HAWT to e.g. chatgpt, and then ask it to give you the formulae for a VAWT

usus · November 26, 2024, 3:40pm

Hi, the equation are the same.
Here a testing rig don’t give too hard