Process Control Function Updates with State-Based Switchover Validation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for updating process control functions in network centric process control systems are inadequate, as they often rely solely on comparing output signals and do not ensure that the new version operates identically to the original, and there is a need for seamless transition and fault-tolerance during updates.
Innovation Solution
A method involving parallel testing of new and old process control function versions, comparing states at multiple points in the cycle, and switching to the new version only after satisfactory alignment is confirmed, utilizing a testing and switchover control module to manage updates in a network centric process control system.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a new version of the process control function is tested by comparing only output signals with the old version, then the testing process is simple, but the reliability of the update is insufficient because it does not ensure identical operation across all states
Solution Approach 1:
The patent extends the testing comparison from a single dimension (output signals) to multiple dimensions by including intermediate states and input states. This dimensional expansion ensures comprehensive verification that the new version operates identically to the old version across all states, not just final outputs, thereby resolving the contradiction between testing thoroughness and complexity.
Solution Approach 2:
The patent segments the process control function execution cycle into distinct states (input states, intermediate states, and output states). By comparing each segmented state separately between the old and new versions, the system achieves comprehensive reliability verification while maintaining a structured, manageable testing approach that doesn't overwhelm system complexity.
2Reliability
If the new version is tested thoroughly by comparing multiple states including intermediate states, then the reliability of the update improves, but the time required for testing increases
Solution Approach 1:
The patent performs preliminary testing of the new version against the old version's recorded execution traces before actual deployment. By pre-comparing intermediate and input states during the testing phase using historical data, the system verifies reliability in advance, allowing for confident updates without extended real-time testing that would cause operational delays.
Solution Approach 2:
The patent implements periodic testing where the new version is evaluated at multiple discrete points (periods) during the execution cycle, comparing states at specific intervals rather than continuously. This periodic sampling approach verifies reliability across the entire cycle while maintaining efficient testing time by focusing on critical state comparison points.
3Reliability
If redundancy is implemented with two identical process control functions running in parallel, then fault-tolerance improves, but the resource requirements on the controller node increase
Solution Approach 1:
The patent implements dynamic redundancy management where the system transitions from static parallel execution of two identical functions to a dynamic approach. The new version is first tested and validated against the old version's execution traces, then selectively activated. This dynamic lifecycle management (test phase, validation phase, operational phase) maintains fault-tolerance while reducing continuous resource consumption compared to permanently running both versions in parallel.
Solution Approach 2:
The patent introduces an intermediary testing and validation mechanism that mediates between the old and new versions. Rather than directly running both versions in parallel for redundancy, the system uses execution traces and state comparisons as intermediaries to verify the new version's correctness before deployment. This intermediary validation layer ensures reliability while avoiding the continuous resource overhead of parallel execution.
Data Source
AI summary
A method for network centric process control system, where a first version of the process control function is being run in a controller node and has a number of original states and a second version of the process control function is provided in the process control system, the second version having a number of new states corresponding to the original states. The method comprises testing the second version in relation to a run of the first version, which testing comprises comparing new states of the second version with corresponding original states of the run of the first version and switching over to using the second version instead of the first version, after the second version has been found to be satisfactory, which finding has been made based on the comparison of the states of the first and second versions of the process control function.