Heterogeneous Redundancy Control for AEB Error Mitigation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In safety-relevant control systems, such as Autonomy Emergency Braking (AEB) systems, the extensive and complex image processing software makes it difficult to fully test and detect design errors, leading to potential incorrect results due to algorithm or hardware errors, which can cause unsafe conditions in autonomous technical systems.
Innovation Solution
Implementing diverse redundancy by using two independent fault containment units with self-checking hardware and a weighted average algorithm to calculate new control values, prioritizing the control value that quickly leads to a safe state, and storing error messages for diagnostic purposes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If extensive image processing software is used to analyze sensor data and calculate control values, then the system can process complex scenarios, but the software cannot be fully tested due to its scope, leading to undetected design errors
Solution Approach 1:
The system divides the control system into two independent fault containment units (FCU 110, 120), each with its own sensor evaluation and control value calculation. This segmentation allows independent testing and error isolation, addressing the inability to fully test extensive software by creating smaller, independently verifiable units.
Solution Approach 2:
The patent creates a duplicate diverse system (second FCU) that mirrors the first FCU's functionality but uses different hardware and algorithms. This copying approach enables cross-validation of results and identifies undetected errors in either system by comparing their outputs.
2Reliability
If two independent diverse systems are used to calculate control values, then the risk of simultaneous errors is reduced, but the device complexity increases
Solution Approach 1:
The actuator control unit performs multiple functions: it receives control values from both FCUs, compares them for equality, handles error cases through weighted average calculation, and manages actuator activation. This multi-functionality reduces overall system complexity by consolidating diverse operations into a single universal controller.
Solution Approach 2:
The actuator control unit serves as an intermediary between the two diverse FCUs and the actuator. It mediates the control values, applying the weighted average algorithm when discrepancies occur, and manages the actuator activation based on the calculated control value, thereby simplifying the interaction between diverse components.
3Reliability
If a weighted average algorithm is used to calculate a new control value when control values differ, then the system can quickly transition to a safe state, but the calculation complexity increases
Solution Approach 1:
The system changes the parameter of control value calculation by applying a weighted average algorithm instead of simple averaging. This parameter change (using weights) allows the system to prioritize safety-oriented control values while maintaining a relatively simple computational approach that can be efficiently implemented in real-time control.
4Reliability
If self-checking hardware is used for actuator control, then hardware errors can be detected immediately, but the manufacturing cost increases
Solution Approach 1:
The actuator control unit performs self-checking by comparing control values from the two diverse FCUs and detecting discrepancies. This self-service approach to error detection enables the system to identify potential hardware or software errors without requiring external monitoring systems, thereby reducing overall manufacturing costs while maintaining high reliability.
Data Source
Figure 1
AI summary
The invention relates to a method and a device for limiting the risk of errors in a control system with heterogeneous redundancy, in particular for a braking system (autonomous emergency braking AEB) for a motor vehicle, in which method two fault-containment units (FCU) with heterogeneous redundancy determine an adjustable variable and if said variables differ, an actuator control (AST) applies different weightings to the adjustable variables using a weighted mean value algorithm and calculates a new adjustable value from said process. The new adjustable variable guides the object to be controlled into a safe state.