Actuation ECU Fault Tolerance via Pre-computed Command Sequences

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Safety-critical applications in vehicles face challenges in maintaining operation for a limited time when faults occur in sensors, serial data links, or control ECUs, leading to disruptions in actuation command communication.

Innovation Solution

A method is developed to compute and store a sequence of actuation commands based on pre-fault sensor data, allowing the actuation ECU to continue controlling the actuator for a limited time by using estimated commands if updated sequences are not received due to communication or ECU failures.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If real-time sensor data and actuation commands are continuously transmitted through serial data links between control ECUs and actuation ECU, then the actuator can be precisely controlled under normal conditions, but the system fails to maintain operation when communication disruptions or ECU failures occur

Engineering Contradiction:
Improvecontinued operation during faultVSAvoidfault tolerance mechanism
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The control ECU pre-computes a sequence of actuation commands S(tx) = C(tx,tx), C(tx,tx+1), ..., C(tx,tx+n) based on current sensor data before time tx, and transmits this entire sequence to the actuation ECU. This preliminary action ensures that even if communication is disrupted after transmission, the actuation ECU has pre-calculated commands ready for future time intervals (tx, tx+1), (tx+1, tx+2), ..., (tx+n, tx+n+1), maintaining system operation during faults without requiring complex real-time fault detection and recovery mechanisms.

Inventive Principle:
Principle #10Preliminary action

2Duration of action of moving object

If the actuation ECU continuously receives updated actuation commands from the control ECU, then the actuator control remains current and accurate, but the system cannot operate if communication is disrupted

Engineering Contradiction:
Improvecontrol duration during faultVSAvoidcommand reception dependency
Core Design Contradiction:
Duration of action of moving objectVSEase of operation

Solution Approach 1:

The control ECU calculates not only the current actuation command but also a sequence of future commands C(tx,tx), C(tx,tx+1), ..., C(tx,tx+n) based on sensor data available at time tx. This preliminary calculation extends the control horizon, allowing the actuation ECU to execute pre-computed commands for time intervals up to (tx+n, tx+n+1) even if communication is completely disrupted, thereby extending the duration of autonomous operation without requiring continuous command reception.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The actuation ECU is designed to automatically execute the pre-received sequence of actuation commands from S(tx) when communication with the control ECU is disrupted. This ensures continuous useful action by switching from real-time command execution to pre-computed command execution seamlessly, maintaining actuator control without interruption or manual intervention during communication faults.

Inventive Principle:
Principle #20Continuity of useful action

3Reliability

If a sequence of estimated actuation commands is pre-computed and stored in the actuation ECU for future time intervals, then the actuator can continue operating during communication disruptions, but the control precision may degrade over time intervals

Engineering Contradiction:
Improveoperation continuityVSAvoidcontrol accuracy
Core Design Contradiction:
ReliabilityVSMeasurement precision

Solution Approach 1:

The control ECU computes a sequence of actuation commands S(tx) based on sensor data obtained before time tx, creating estimated commands for future time intervals. These pre-computed commands are transmitted to the actuation ECU, which executes them when communication is disrupted. This approach accepts controlled degradation of precision over time intervals (tx, tx+1), (tx+1, tx+2), etc., in exchange for guaranteed operation continuity, with the precision degradation being a known and acceptable trade-off for achieving fault tolerance.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentUS8843218B2Method and system for limited time fault tolerant control of actuators based on pre-computed values
Publication Date: 2014.09.23 GM GLOBAL TECHNOLOGY OPERATIONS LLC
  • US8843218B2 patent drawing
  • US8843218B2 patent drawing

AI summary

A method of controlling an actuator includes developing a sequence of actuation commands S(tx)=C(tx, tx), C(tx, tx+1), . . . , C(tx, tx+n) obtained from data sensed before time tx for controlling the actuator at different time intervals (tx, tx+1), (tx+1, tx+2), . . . , (tx+n−1, tx+n). The sequence of actuation commands is transmitted to and stored in memory of an actuation ECU, which then applies the actuation command for time interval (tx, tx+1). If a fault affects a sensor, a control ECU or data communication therebetween, the actuation ECU will not receive an updated sequence of actuation commands S(tx+1) at time tx+1. If the updated actuation command sequence, the actuation ECU applies the actuation command for the time intervals (tx+1, tx+2), . . . , (tx+n−1, tx+n) from the sequence of actuation commands S(tx) that is stored in the memory of the actuation ECU.