Function ```normalize``` in ```KnotVector``` should consider float precision

Take the knotvector

```python
from compmec.nurbs import KnotVector
knotvector = (0.0, 0.0, 1.8678173101311735, 1.8678173101311735)
knotvector = KnotVector(knotvector)
knotvector.normalize()
print(knotvector)  # (0.0, 0.0, 0.9999999999999999, 0.9999999999999999)
```

It's expected to get the knotvector ```(0.0, 0.0, 1.0, 1.0)```