Log of OpenHRP3 1

Gain file format
P D
8000 100
8000 100
8000 100
8000 100
8000 100
8000 100

Angle format
Time Joint 1 Joint 2 ........................
0 0.00E+00 -3.60E-02 0 7.85E-02 -4.25E-02
0.002 0.00E+00 -3.60E-02 0 7.85E-02 -4.25E-02

Velocity format is similar to angle.

Digest of paper and Simulation

I read the paper regarding ZMP control and did the Matlab M file to testify some equations combined with my concerns. The program is shown bellow. After the simulation, I have a better, deeper understanding=)

function trajectory_avs
offset=0; g=9.8; Zc=500
t=0:0.1:(2*pi);
a=0.5*sin(t);
for i=1:length(a)
v(i)=sum(a(1:i));
end
for j=1:length(v)
Xcom(j)=sum(v(1:j));
end
Xcom=Xcom+offset;
figure(1);
subplot(3,1,1); plot(a);
subplot(3,1,2); plot(v);
subplot(3,1,3); plot(Xcom);

coe=a/g;
for k=1:length(Xcom)
Xzmp(k)=Xcom(k)-coe(k)*Zc;
end
figure(2); plot(Xzmp,'r'); hold on; plot(Xcom,'b+');

end

% 仿真结果表明，用sine曲线规划加速度肯定无法避免ZMP的超调。同时，同样的加速度轨迹下，超调量与重心的高度直接相关。这个也符合常理的认知，一个人在不改变支撑面的情况下，
% 单纯增加双脚高度自然增加了行走的困难，一点过多的加速度就可以导致前倾或者后倾。所以，目前研究人员用控制理论的方法得到传递函数进行控制，能达到更好的控制效果和性能。
% 因为本文的这种在a v 底层进行简单的sine曲线规划是远远不够的。用控制理论的方法，得到的加速度曲线，如果画出来的话将是一个非线性高阶的曲线，这样的加速度控制得到的最后结
% 果是良好的速度，位置控制轨迹，是以性能为衡量标准的。这样的加速度曲线也更为复杂！本文也只是探索性质，看来这个课题没有这么简单说是用什么曲线规划加速
% 度或者速度就能达到良好的指标，这种单纯的想法、简单的控制根本没有动态适应性。

So this is the reason why so many researchers apply complicated control methods in order to solve the balancing control algorithms=)

Pelvic motion

the trochanter is highest during supporting phase and lowest in swinging phase

Website of openHRP3

Link gives an over view and introduction
Last Update
(2005.10.12) a bit old

http://www.is.aist.go.jp/humanoid/openhrp/English/indexE.html

Official web Latest one

http://www.openrtp.jp/openhrp3/en/index.html

Download OpenHRP3

Download OpenHRP3
Thank you for agreeing with Software license.
Please download OpenHRP3, using following links;
OpenHRP3 Source Archive (Common for Linux and Windows)
version 3.0.3 (28/11/2008)
What`s new in version 3.0.3


version 3.0.2 (19/09/2008)
What`s new in version 3.0.2

Not easy, speed quite slow ya......... Jam

Ultimate Research Assistant

http://ultimate-research-assistant.com

A good website for mining the internet resources=)

Curve Fitting Sample Code For Christian

% first the data of knee is stored as txt file
kneef=load('knee_abstract.txt');
%%%% curve fitting
x=1:length(kneef); x=x';
% knee
a = POLYFIT(x,kneef,18);
t=1:(( length(kneef))/250): length(kneef);
z=polyval(a,t);

%%%%%%%%%%%%%%
plot(kneef,'r*');hold on;
plot(z,'b+');hold on;
legend('Original','Processed');