前回，マウスが高速ターンではタイヤの向いている向きには進まないと書きました．図のように左ターンしているとすると，ロボットはθrの向きに向いていますが，実際の移動方向はθmの向きです．この角度差がスリップ角です．通常，マウスの向きはジャイロセンサの出力を積分して得るので，直接推定できるのはθrです．スリップ角β=θr-θmを予測できれば，θmがわかるので，前回の一般的なオドメトリで自己位置推定ができます．
Last time, I wrote that a micromouse does not proceed in the direction of the tire in the high-speed turn. When it is assumed that a mouse turns to the left as shown in the figure, the direction of the robot is θr, but the direction of actual movement is θm. This angular difference is the slip angle. Usually, the direction of the mouse is obtained by integrating the output of the gyro sensor, so you can estimate θr directly. If you can predict the slip angle β=θr-θm, the general odometry will give more precise self-position written last time.
ところで，マウスが走りながら旋回するためには向心力が必要ですが，これはタイヤが変形したときの反発力です．そしてこの変形分がスリップ角になるわけです．簡単のため向心力（遠心力）とスリップ角が比例関係にあるとすれば，遠心加速度acenとスリップ角βは比例係数をkとして次式の関係になります．
acen = k × β
これを用いてスリップ角を推定します．つづきは次回．
By the way, centripetal force is required for the mouse to turn, and this is a repulsive force when the tire is deformed. This deformation causes the slip angle. Assuming a proportional relationship between slip angle and centripetal force (centrifugal force) for simplicity,
acen = k × β,
where k is the proportional coefficient and acen is centrifugal acceleration.
Slip angle can be estimated using this. Continued next time.
4 件のコメント:
Love it, especially you put English version as well!
Thank you.
First, I wrote it in Japanese. And I tried google translate for this article, but the translated English was terrible! So, I modified the English sentences and put them.
If use a gyro, doesn't the gyro take care of the slip angle?
A gyro sensor gives the direction of the mouse, which equals to the direction of the tires. But, the direction of tires is different from the actual moving direction of the mouse. The difference is slip angle. So, the output of gyro sensor is not sufficient. Estimating the slip angle with a gyro and a rotary encoder is required. This is the next topic.
コメントを投稿