Buang masa

Hari ni, abg buang masa buat banner site ni. Guna 'Seashore', Gimp-based picture editing software. Boring nya...
My simulation is still taking a long time to run. I don't know what's wrong. The ODE solution seems too close to singularity. Eventhough my matrix is full rank. Eventually I force to stop the program when the error tolerance is too small.

Why? why? why?It is because;

1. the ODE is 'stiff' type
2. my matrix isn't actually full rank, but due to processor limit, it says it is full rank. ??
2. the 'kappa' term after I tie it up to dhdq term, results in zero matrix. is it logic? (how can I get the x,y,phi position if it is not driven by the input tau.)