Package jline.solvers.auto

package jline.solvers.auto
Automatic solver selection and configuration for queueing networks.

This package provides an implementation of SolverAUTO. It implements intelligent solver selection algorithms that automatically choose the most appropriate solution method based on network characteristics, model size, and accuracy requirements.

Selection Criteria

  • Network topology and structure analysis
  • Problem size and complexity assessment
  • Required accuracy and performance trade-offs
  • Available computational resources
  • Model-specific feature requirements

Supported Solvers

  • MVA - For Product Form closed networks
  • JMT - For complex networks requiring simulation
  • CTMC - For small state-space models
  • SSA - For steady-state analysis
  • Fluid - For large population models

Decision Algorithms

  • Rule-based selection for common patterns
  • Performance prediction models
  • Machine learning-based recommendations
  • Hybrid solver cascading strategies

Configuration Features

  • Automatic parameter tuning
  • Convergence criteria optimization
  • Resource allocation management
  • Fallback strategy implementation
Since:
LINE 2.0
See Also:
  • Classes
    Class
    Description
    AUTO
    AUTO is an alias for SolverAUTO (Automatic solver selection).
    AUTOptions
    Options specific to the AUTO solver
    LINE
    LINE solver - Java/Kotlin implementation equivalent to MATLAB LINE.m
    ModelAnalyzer
    Helper class to analyze model characteristics for solver selection
    SolverAUTO
    Automatic solver selection for queueing network models.