fix(control): limit native math threads

[xiaolefang-dm]

Summary

• limit OpenBLAS/OMP/MKL/NumExpr native worker threads to 1 before importing numpy/scipy in cmd_vel_control
• prevents small matrix operations in the low-rate Python control node from spawning busy native worker threads on Jetson

Testing

• python3 -m py_compile tinynav/platforms/cmd_vel_control.py
• tested manually on current integration test machine by applying the same change in-place and restarting nav nodes
• observed cmd_vel_control CPU drop from ~170-180% to ~3-4% after warm-up while /cmd_vel remained ~12 Hz

before

after

[github-actions]

cloc	github.com/AlDanial/cloc v 2.04 T=0.17 s (254.3 files/s, 37966.0 lines/s)

Language	files	blank	comment	code
Python	13	621	265	3469
Markdown	4	155	6	426
C++	4	60	38	409
Bourne Shell	15	41	9	272
Dockerfile	1	31	30	202
JSON	2	8	0	58
TOML	1	4	0	58
make	1	18	15	55
CMake	1	4	0	17
--------	--------	--------	--------	--------
SUM:	42	942	363	4966

[dvorak0]

@xiaolefang-dm let's make it global, by setting single thread by default in Dockerfile.

[github-actions]

cloc	github.com/AlDanial/cloc v 2.04 T=0.18 s (235.7 files/s, 35152.6 lines/s)

Language	files	blank	comment	code
Python	13	620	262	3464
Markdown	4	155	6	426
C++	4	60	38	409
Bourne Shell	15	41	9	272
Dockerfile	1	31	30	203
JSON	2	8	0	58
TOML	1	4	0	58
make	1	18	15	55
CMake	1	4	0	17
--------	--------	--------	--------	--------
SUM:	42	941	360	4962

[github-actions]

cloc	github.com/AlDanial/cloc v 2.04 T=0.18 s (237.4 files/s, 35405.2 lines/s)

Language	files	blank	comment	code
Python	13	620	262	3464
Markdown	4	155	6	426
C++	4	60	38	409
Bourne Shell	15	41	9	272
Dockerfile	1	31	32	203
JSON	2	8	0	58
TOML	1	4	0	58
make	1	18	15	55
CMake	1	4	0	17
--------	--------	--------	--------	--------
SUM:	42	941	362	4962