Redundant Automation Processing Units with Buffered Task Synchronization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing redundant automation systems with multiple processor units face challenges in synchronization and data exchange, limiting the performance advantage of multi-core processors and requiring complex breakpoint insertion, which increases response times and complicates the distribution of tasks across hardware units.
Innovation Solution
The method involves distributing tasks across multiple processor units within hardware units, using buffers for data exchange between groups of tasks to maintain path synchronicity and optimize communication links, allowing for efficient synchronization and data exchange without requiring all software components to share the same data or interruption points.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If tasks are distributed across multiple processor units in redundant automation systems, then processing performance and system responsiveness are improved, but synchronization complexity and data exchange overhead increase
Solution Approach 1:
The patent segments tasks into different groups (first group, second group, third group, fourth group) and distributes them across multiple processor units (first processor unit, second processor unit) within hardware units. This segmentation allows parallel processing while maintaining clear boundaries for synchronization, resolving the contradiction between improved processing performance and increased synchronization complexity.
Solution Approach 2:
The patent introduces communication links as intermediaries between processor units and hardware units. These communication links facilitate controlled data exchange and synchronization between distributed tasks, enabling parallel processing without overwhelming synchronization overhead. The communication links act as mediators that manage the complexity of coordinating multiple processor units.
2Stability of the object's composition
If all software components share the same data and interruption points for synchronization, then path synchronicity is maintained, but response times increase and task distribution is limited
Solution Approach 1:
The patent allows different task groups to have different interruption point configurations tailored to their specific synchronization needs. The first and second task groups can be synchronized differently from the third and fourth task groups. This local quality approach maintains path synchronicity where needed while avoiding unnecessary synchronization overhead elsewhere, thereby reducing response times.
Solution Approach 2:
By segmenting tasks into different groups with different synchronization requirements, the patent enables selective synchronization. Not all software components need to share the same interruption points - only those that require path synchronicity. This segmentation reduces the time loss associated with universal synchronization while maintaining synchronicity where necessary.
3Productivity
If communication links are optimized for data exchange between processor units, then task distribution efficiency is improved, but system complexity and data exchange overhead increase
Solution Approach 1:
The communication links in the patent are designed to serve multiple functions: they enable data exchange between different processor units, support synchronization of distributed tasks, and facilitate the redundant automation system's failover capabilities. This multi-functionality improves task distribution efficiency without proportionally increasing system complexity, as the same communication infrastructure serves multiple purposes.
Data Source
Figure 1
Figure 2
Figure 3~4
AI summary
Several hardware units (2, 3) of a redundant automation system each comprise several processor units (4 to 7). The processor units (4 to 7) each execute a group (G1 to G4) of tasks. Groups (G1 to G4) of tasks executed on different hardware units (2, 3) receive the same input signals (E1, E2) directly or indirectly from a controlled industrial process (1) and determine the same output signals (A1, A2) for the controlled industrial process (1). However, only the tasks of each of the groups (G1 to G4) executed on different hardware units (2, 3) output their output signals (A1, A2) to the controlled industrial process (1). The groups (G1 to G4) executed on different hardware units (2, 3) synchronize with each other via a respective communication link (14, 15).The groups (G1 to G4) of tasks running on the same hardware unit (2, 3) exchange data with each other in pairs via a respective pair of intermediate storage locations (16 to 19).