Implement comprehensive conjunction assessment capabilities with:

- **ConjunctionEvent**: Data structure for storing conjunction analysis results
  - TCA (Time of Closest Approach)
  - Miss distance with RIC (Radial-Intrack-Crosstrack) components
  - Relative velocity
  - Combined covariance matrices
  - Probability of collision
  - Mahalanobis distance calculation
  - High-risk event detection

- **ProbabilityOfCollision**: Pc calculation using Chan's 2D method
  - Projects covariance onto encounter plane (B-plane)
  - Computes 2D probability density integration
  - Handles edge cases (low relative velocity)
  - Analytical approximations for performance
  - Covariance matrix combination

- **ConjunctionAssessment**: Main workflow orchestrator
  - Supports TLE or state vector inputs
  - Multiple propagator options (SGP4, RungeKutta89)
  - TCA finding using golden section optimization
  - Covariance propagation using sigma-point method
  - Leverages existing TLE quality assessment
  - High-fidelity force model integration

Key Features:
- Historical TLE accuracy-based covariance initialization
- Regime-specific covariance aging (LEO/MEO/GEO/HEO)
- Flexible propagator selection for accuracy/speed tradeoff
- Comprehensive test coverage
- Example implementations for common use cases

The workflow integrates with existing OOTK capabilities:
- CovarianceSample for sigma-point propagation
- RIC coordinates for relative geometry
- GoldenSection optimizer for TCA search
- Multiple propagator types (SGP4, RK4, RK89)
- Force models for high-fidelity prediction

Example usage in conjunction-assessment-example.ts demonstrates:
1. Basic TLE-based conjunction assessment
2. High-fidelity propagation with covariance
3. Custom covariance matrix input
4. Multi-object screening workflows