Momentum resolution study lingzhu

This MR uses a different parameterization for momentum and improves the momentum resolution. The short of it is: use dev_velo_kalman_endvelo_states as the input for velo state and the state at z_mid_t as input for SciFi state. The performance(momentum resolution and throughput) of this new parameterization can be seen in this report on WP2 meeting

• The origin qop calculation in lf_quality_filter and the related information(parameters, states, tracks) are gone
• The new qop calculation is done in scifi_consolidate_tracks

this branch relies on end_velo_state_maxime

Closes #126 (closed)