![]() This creates very stable orientation data and outputs a single yaw, pitch & roll that doesn't drift. I have worked with IMU's (Inertial measurement unit) in the past and needed to use directional cosine matrix's to mesh the accelerometer, gyroscope and compass readings together. My thought was that I would need to mesh the data together using a Kalman filter. I love the SSC-32 it is super great! Sensors Triple Axis Accelerometer (ADX元35) When I started off I thought that I would need a 3-axis gyroscope in addition to the accelerometer. Once all the calculations are complete it sends the appropriate commands to the SSC-32 servo sequencer. Controller/CPU: Arduino MegaMini 2560, Lynxmotion SSC-32 The Arduino Mega handles all of the computation along with taking input from the accelerometer & blue-tooth. Here is the link to the ideal joysticks from servocity. I found some 3-funtion joysticks at servo city that will replace the parallax joysticks I have now. As I continue to progress I intend to build more functionality into my remote and robot. Thanks, & thanks for your email responses! I have built a blue-tooth remote control which has worked great so far. A big thanks to the people that put that tool together, it was true to its name! Also big thanks to "tom_chang79" his discussion, "Let's Discuss Kinematics, Shall We?" was very helpful. Instead they had an already completed sketch for a quad which is what I modeled my program after. I am not using the micro-controller is was designed for so I could not use their python tools to generate the code I borrowed from. I also don't believe in re-inviting the wheel I used the rotation matrix & inverse kinematics from NUKE (Nearly Universal Kinematic Engine). It has been some time since I have been enrolled in a math class. This allows the robot to stay level on uneven surfaces. The first two modes are pretty self explanatory accelerometer mode takes the x-axis & y-axis from the accelerometer, maps the values to (-20, 20), then use those values as the body's pitch & roll respectively. A few modes exists, 1: Body Translation/Body Position 2:Walking Mode 3:Accelerometer Mode. I am using a ripple gait for walking although I am still fine tuning that portion of the code. That in turn feeds into the Inverse Kinematics equations that position the legs. Locomotion: The body's roll, pitch & yaw are calculated using a rotation matrix. I love the UBEC and will definitely be the regulator of choice for future robot projects. Dimensions 8 inches, 8 inches, 7 inches (L:W:H) Run Time 15-30 minutes Programing Language Arduino IDE Control Method Blue-tooth Remote Cost ~$540.00 Powersource: LiPo 1x 7.4V 1650Mah 2S LipPo I am powering everything using a 2S LIPO & Turnigy 15A UBEC (Ultimate battery elimination circuit). Build Time 8 weeks Weight 2 Pounds 14 Oz. ![]() I chose a quadruped to help keep the cost down. This is my first robot that relies on calculations to position the servos vs. ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |