Discrete Time Rate-Based Model Predictive Controller for Diesel Engine Air Path
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing model predictive control methods for diesel engines require multiple operating ranges, leading to increased computation time and memory storage requirements, while also facing issues with overshoot restraints of controlled engine variables.
Innovation Solution
A discrete time rate-based predictive model controller is developed to optimize VGT and EGR valve positions using intake manifold pressure and EGR rate measurements, reducing the number of operating ranges and employing partial non-linear inversion to minimize computational complexity and overshoot constraints.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If multiple operating ranges are used in model predictive control, then zero state-to-state tracking error is achieved, but computation time and memory storage requirements increase
Solution Approach 1:
The patent combines multiple separate predictive controllers (each designed for a specific operating range) into a single unified predictive controller. This is achieved by formulating a single optimization problem that incorporates constraints and objectives valid across the entire operating range, eliminating the need to switch between multiple controllers and reducing computational overhead.
Solution Approach 2:
The patent creates a universal predictive controller that functions across all operating ranges simultaneously. The controller uses a single set of matrices and optimization algorithms that adapt to different operating conditions through measured variables, rather than requiring separate specialized controllers for each range.
2Manufacturing precision
If multiple operating ranges are used in model predictive control, then zero state-to-state tracking error is achieved, but memory storage requirements increase
Solution Approach 1:
The patent merges the storage requirements for multiple controllers into a single controller structure. Instead of storing separate sets of matrices and parameters for each operating range, the unified controller stores one comprehensive set of matrices that work across all ranges, significantly reducing memory requirements.
Solution Approach 2:
The universal controller design eliminates the need to maintain multiple copies of controller data structures in memory. A single optimized set of matrices and algorithms serves all operating conditions, reducing the overall memory footprint of the control system.
3Manufacturing precision
If integral type action is used in predictive controller, then zero state-to-state error is guaranteed, but overshoot restraints of controlled engine variables are compromised
Solution Approach 1:
The patent implements a feedback mechanism within the predictive control framework that actively monitors controlled variables and adjusts predictions to prevent overshoot. The controller uses measured variables and update rates to adapt predictions in real-time, ensuring both accuracy and constraint satisfaction without relying solely on integral action.
Solution Approach 2:
The predictive controller performs preliminary action by forecasting future states and adjusting control inputs in advance to prevent overshoot before it occurs. The optimization problem incorporates constraints that proactively limit predicted trajectories, avoiding the need for reactive integral correction that may cause overshoot.
Data Source
AI summary
A discrete time rate-based model predictive controller for air path control for a diesel engine regulates VGT position and EGR valve position to specified set points by coordinated control of intake manifold air pressure and EGR rate. The controller may be configured to measure or estimate at least one of the intake manifold pressure and EGR rate. A non-linear discrete time rate-based predictive model may be used, as developed by the controller.