Model Predictive Controller Based on Online Obtaining of Softness Factor and Fusion Velocity for Automatic Train...

Area of Science:

• Control Systems Engineering
• Artificial Intelligence
• Transportation Systems

Background:

• Traditional model predictive controllers (MPC) for automatic train operation (ATO) face limitations in tracking precision.
• Real-time adaptation of control parameters is crucial for optimizing ATO performance.

Purpose of the Study:

• To develop an improved model predictive controller (MPC) for automatic train operation (ATO) that enhances tracking control performance.
• To introduce adaptive strategies for softness factor and fusion velocity estimation.

Main Methods:

• An online adaptive adjustment method for the softness factor using fuzzy logic based on system output and velocity trajectory.
• An improved whale optimization algorithm for solving adjustable parameters.
• An online velocity sampling method using a fusion velocity model and intelligent digital torque sensor.

Main Results:

• The proposed improved strategies demonstrate good tracking precision and simplicity.
• The controller ensures real-time accomplishment of computational tasks, considering hardware limitations.
• Matlab/Simulink and hardware-in-the-loop simulation (HILS) verified the controller's effectiveness.

Conclusions:

• The improved MPC offers superior tracking control effectiveness compared to traditional methods for ATO.
• The adaptive and online strategies enhance the robustness and efficiency of automatic train control.