Automated Optimization of Train Control Systems

Project Overview

As lead inventor, I developed an automated, server-based system designed to solve a non-trivial problem in rail operations: determining the ideal control parameters for a train's driving system to achieve highly specific, often conflicting, performance goals. Unlike static, one-size-fits-all control files, this system determines the optimal driving strategy unique to every trip.

This project sits at the intersection of complex systems modeling, computational optimization, and real-time control implementation.

The Challenge: The Myriad Variables of Train Control

The optimal way to drive a train (throttle changes, braking application, coasting) depends on numerous variables that are highly dynamic and unique to each journey:

• Train Makeup: Radical variations in weight, length, and distributed power configuration (the plant).
• Route Profile: Grade, curvature, and speed limits of the track.
• Performance Metrics: The operating railroad's goal, which requires balancing competing metrics (e.g., maximizing fuel economy or prioritizing on-time arrival).
• Environmental Factors: Real-time weather data affecting traction and resistance.

Attempting to tune these control system parameters manually is practically impossible due to the sheer volume of scenarios.

The Bespoke Automata Solution: Scenario-Based Optimization

The patented system utilizes an Optimization Server and a high-speed, parallel Simulator to solve this parametric optimization challenge through a simulation-and-score process.

Methodology

• Data Acquisition: The system receives all Operational Parameters (Train Consist, Track Profile, Operational Rules) and User-Specified Performance Metrics (e.g., minimum transit time, maximum fuel usage).
• Scenario Generation: The server generates a plurality of scenarios, each using the same operational data but a different set of Control Parameters (which govern when the engineer should apply dynamic braking, change throttle notches, etc.).
• Parallel Simulation: The simulator, comprising multiple modules operating in parallel, runs each scenario to predict the train's performance over the route.
• Scoring and Iteration: The server reviews the simulation results, scores how closely each scenario achieved the desired performance metrics, and identifies the best-performing control parameter set.
• Robustness Check: Additional scenarios are generated using slight variations (perturbations) of the operational parameters to ensure the optimized control set remains robust against real-world measurement errors (e.g., slight errors in reported train weight).
• Deployment: The final, optimal control parameters are transmitted to the train control system (TCS) on the lead locomotive for use by the engineer, dramatically improving the train's efficiency and reliability according to the railroad's current business goal.

System Impact

This invention transforms a highly variable, complex control problem into an automated, data-driven optimization process. It allows rail operators to flexibly adjust their automation strategies instantly based on business needs (e.g., prioritizing speed during peak demand and fuel savings during off-peak times) while ensuring safety and compliance.