Wrap mem.consolidate_logs in try-except to prevent model run crashes

[Copilot]

An ActivitySim model run can crash with a FloatingPointError: overflow encountered in multiply when consolidating memory logs due to timestamp overflow issues in pandas datetime conversion. This affects production runs where users lose all model results despite the core modeling completing successfully.

The error occurs in mem.consolidate_logs() when processing timestamp data:

# In mem.py line 45
return pd.to_datetime(bin, unit="s", origin="unix")

When timestamp values are corrupted or extremely large, this causes an overflow during the conversion process, crashing the entire model run.

This PR implements an intermediate fix by wrapping the mem.consolidate_logs(state) call in a try-except block:

# Before
mem.consolidate_logs(state)

# After  
try:
    mem.consolidate_logs(state)
except Exception as e:
    logger.warning(
        f"Memory log consolidation failed with error: {e}. "
        "This does not affect model results, but memory usage logs will not be consolidated."
    )

This provides graceful degradation where the core model functionality continues to work even when the secondary memory logging feature fails. Users receive a clear warning message explaining what happened and that their model results are unaffected.

As noted by @bhargavasana, "One in five runs of ours are failing because of this" - this fix addresses that immediate production issue while preserving all existing functionality.

Fixes #959.

---

💬 Share your feedback on Copilot coding agent for the chance to win a $200 gift card! Click here to start the survey.

[Copilot]

@jpn-- 👋 This repository doesn't have Copilot instructions. With Copilot instructions, I can understand the repository better, work faster and produce higher quality PRs.

I can generate a .github/copilot-instructions.md file for you automatically. Click here to open a pre-filled issue and assign it to me. I'll write the instructions, and then tag you for review.