Engine Controller Dynamic Learning Speed for Air-Fuel Hunting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing internal combustion engine systems face challenges in preventing hunting of the sub-feedback learning value, which affects the convergence speed of air-fuel ratio control, leading to potential air-fuel ratio imbalances and engine inefficiencies.
Innovation Solution
An internal combustion engine system controller is configured with a sub-feedback learning portion, a state determining portion, and a learning update-speed setting portion to dynamically adjust the update speed of the sub-feedback learning value based on the fluctuating state, preventing hunting and ensuring quick convergence by changing the update speed when transitions between stable, intermediate, and unstable states occur.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If the learning speed of the sub-feedback learning value is increased to achieve faster convergence, then the convergence speed is improved, but hunting occurs causing instability
Solution Approach 1:
The patent applies dynamics by making the learning speed variable rather than fixed. The learning speed is dynamically adjusted based on the fluctuation state of the sub-feedback learning value. When fluctuations are large, the learning speed is reduced to prevent hunting; when fluctuations are small, the learning speed is increased to achieve faster convergence. This dynamic adjustment resolves the contradiction between speed and stability.
Solution Approach 2:
The patent changes the parameter of learning speed based on the fluctuation state. By monitoring the fluctuation magnitude and adjusting the learning speed parameter accordingly, the system achieves both fast convergence and stability. The learning speed parameter is modified in response to system state changes, allowing the system to adapt between speed and stability requirements.
2Loss of time
If the learning speed is set too high, then convergence is achieved faster, but air-fuel ratio imbalances among cylinders occur
Solution Approach 1:
The learning speed parameter is changed based on the fluctuation state to prevent air-fuel ratio imbalances. When the system detects large fluctuations that could lead to cylinder imbalances, it reduces the learning speed. This parameter adjustment prevents precision degradation while still achieving convergence, resolving the contradiction between convergence time and air-fuel ratio consistency.
Solution Approach 2:
The system uses feedback from the fluctuation state to adjust the learning speed. By continuously monitoring whether the fluctuation exceeds predetermined thresholds and adjusting the learning speed accordingly, the system maintains air-fuel ratio consistency across cylinders while achieving convergence. This feedback mechanism prevents the harmful effects of too-high learning speeds.
3Stability of the object's composition
If the learning speed is set too low, then hunting is prevented and stability is maintained, but convergence becomes slow
Solution Approach 1:
The learning speed is made dynamic rather than statically low. The system starts with a lower learning speed to ensure stability and prevent hunting, then progressively increases it as the sub-feedback learning value converges and fluctuations decrease. This dynamic approach maintains stability while improving convergence efficiency over time.
Solution Approach 2:
The learning speed is periodically adjusted based on the fluctuation state. The system transitions from a conservative low learning speed to a higher learning speed as convergence progresses. This periodic adjustment of the learning speed parameter allows the system to maintain stability initially while achieving efficient convergence ultimately.
Data Source
AI summary
Provided is an internal combustion engine system controller, including a sub-feedback learning section, a state determining section, and a learning update-speed setting section. The state determining section determines, to which of at least three states including: (a) a stable state in which a fluctuating state of a sub-feedback learning value is stable; (b) an unstable state in which the fluctuating state greatly fluctuates; and (c) an intermediate state between the stable state and the instable state (may be referred to as sub-stable state), the fluctuating state corresponds. The learning update-speed setting section sets an update speed of the sub-feedback learning value in accordance with the result of determination by the state determining section. Further, the learning update-speed setting section suppresses the occurrence of hunting of the sub-feedback learning value.