Adaptive Throttle Position Prediction for Engine Control

The patent applies dynamics by making the prediction method adaptive rather than static. The system dynamically switches between different prediction strategies (first strategy using setpoint gradient, second strategy using measured gradient) based on real-time conditions such as throttle position, engine load, and transient state detection. This dynamic adaptation allows the system to maintain high prediction accuracy across varying operating conditions while managing complexity through conditional logic rather than overly complex algorithms.

The patent changes parameters by introducing multiple prediction strategies with different characteristics. The first strategy uses setpoint gradient which is more accurate for steady-state conditions, while the second strategy uses measured gradient which responds better to rapid transients. The system adjusts which strategy to use based on parameter thresholds (e.g., throttle position rate of change, time horizon) thereby optimizing prediction accuracy without requiring a single complex model.

The system implements dynamic adaptability by continuously monitoring operating conditions and switching between prediction strategies. The control unit detects transient states, throttle position ranges, and load conditions to dynamically select the most appropriate prediction method. This maintains implementation simplicity through modular strategy design while achieving high adaptability through conditional selection based on real-time sensor data and pre-defined thresholds.

The patent segments the prediction problem into multiple distinct strategies, each optimized for specific operating conditions. The first strategy (setpoint gradient) is segmented for use during steady-state and slow-transient conditions, while the second strategy (measured gradient) is segmented for rapid transient conditions. This segmentation allows each strategy to be relatively simple while the combination provides comprehensive adaptability across all operating conditions.

The patent applies dynamics by detecting transient conditions and adapting the prediction approach accordingly. During transient phases, the system uses measured gradient-based prediction which captures the actual dynamic behavior rather than assuming frozen phenomena. The control unit monitors parameters such as throttle position rate of change and engine load variations to identify transients, then switches to the more responsive prediction strategy, maintaining reliability without excessive complexity.

The system uses feedback from actual measurements to improve prediction quality during transients. The measured gradient strategy incorporates real-time sensor data (actual throttle position, pressure, flow rates) to adjust predictions dynamically. This feedback mechanism allows the system to respond to actual transient behavior rather than relying on static assumptions, improving reliability while keeping calculations manageable through efficient use of available sensor data.