Artificial Reference Signal for Model Predictive Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Engine control systems face challenges in immediate responsiveness to inputs due to the varying response times of actuators, leading to delays in achieving desired operating parameters, and existing multiple single input single output (SISO) controllers often prioritize system stability at the expense of fuel consumption and are costly to calibrate.
Innovation Solution
Implementing a model predictive control (MPC) system that generates target values for actuators using an artificial reference signal, allowing for torque reserve management and faster response to predicted inputs, while selecting the most cost-effective set of target values that satisfy constraints.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Stability of the object's composition
If multiple SISO controllers are used to control actuators, then system stability is improved, but fuel consumption increases and calibration costs increase
Solution Approach 1:
The patent combines multiple SISO controllers into a single model predictive control (MPC) system that uses a unified cost function to simultaneously control multiple actuators. This integration eliminates the need for separate controllers while achieving the same stability objectives, thereby reducing fuel consumption and calibration complexity.
Solution Approach 2:
The MPC system serves as a universal controller that can control multiple different actuators (throttle valve, spark timing, fuel injection) through a single unified framework. This multi-functional approach replaces multiple specialized SISO controllers, improving fuel efficiency while maintaining system stability.
2Stability of the object's composition
If multiple SISO controllers are used to control actuators, then system stability is improved, but calibration costs increase
Solution Approach 1:
The patent merges multiple independent calibration processes into a single unified calibration procedure for the MPC cost function. Instead of calibrating each SISO controller separately, the unified cost function requires only one calibration process, significantly reducing calibration costs while maintaining system stability.
Solution Approach 2:
The universal MPC framework provides a single calibration approach that works for all actuators simultaneously. This eliminates the need for multiple separate calibration procedures required by SISO controllers, making the calibration process more economical and easier to implement.
3Stability of the object's composition
If traditional control systems are used, then system stability is maintained, but response time to inputs increases
Solution Approach 1:
The MPC system performs preliminary action by predicting future system states and proactively adjusting actuators before deviations occur. The cost function anticipates future torque requests and pre-positions actuators accordingly, enabling faster response times while maintaining stability through predictive rather than reactive control.
Solution Approach 2:
The patent implements dynamic control by continuously updating the cost function based on real-time system state and predicted future conditions. This dynamic adaptation allows the system to respond quickly to changing conditions while maintaining stability through ongoing optimization rather than fixed control parameters.
4Adaptability or versatility
If actuators with varying response times are used, then system flexibility is improved, but achievement of desired operating parameters is delayed
Solution Approach 1:
The MPC cost function performs preliminary action by predicting which actuators will need adjustment and preparing control signals in advance. This allows the system to account for varying actuator response times proactively, ensuring all actuators arrive at their target positions simultaneously despite different inherent speeds, thereby eliminating delays while maintaining flexibility.
Solution Approach 2:
The patent adds the time dimension to the control problem by using predictive control over a future time horizon. The cost function optimizes actuator commands across multiple future time steps, allowing it to coordinate actuators with different response times effectively. This temporal dimension enables the system to achieve desired parameters faster while maintaining the flexibility to handle diverse actuator characteristics.
Data Source
AI summary
A control system includes a control module that receives a first request corresponding to a control value for at least one of a plurality of actuators, selectively receives a second request associated with a predicted future control value for at least one of the plurality of actuators, determines a target value for the actuator based on the first request if the second request was not received, and generates a reference signal representing the second request if the second request was received. The reference signal indicates at least one of a predicted increase in the control value and a predicted decrease in the control value. A model predictive control module receives the reference signal and adjusts one of the plurality of actuators associated with the predicted future control value based on the reference signal.