Knowledge Graph Control Architecture for Dynamic Industrial Networks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional techniques for modeling and controlling industrial systems, such as water supply networks, are inefficient and lack automation, making it difficult to dynamically adjust upstream and downstream flow and manage complex relationships between resources and control states.
Innovation Solution
A knowledge graph-based system architecture that uses graph searches to generate control query results, allowing for efficient operation of industrial actuation and sensor systems by mapping relationships between resources, control states, and entities, enabling autopiloting and dynamic adjustments within the network.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If conventional techniques for modeling and controlling industrial systems are used, then system complexity is reduced, but automation capability and efficiency deteriorate
Solution Approach 1:
The patent introduces a knowledge graph as an intermediary layer between the control system and the industrial systems. This knowledge graph stores relationships between resources, control states, and entities, enabling automated query processing and control decisions without requiring complex direct modeling. The knowledge graph acts as a mediator that transforms unstructured system relationships into structured, queryable data, thereby improving automation while managing complexity.
Solution Approach 2:
The patent replaces conventional mechanical/control system approaches with information-based processing. Instead of using complex control algorithms and mathematical models to manage system relationships, the system uses knowledge graph storage and query mechanisms to automatically retrieve and process control information. This substitution of information processing for mechanical control logic enhances automation while reducing the complexity of control system design.
2Adaptability or versatility
If conventional control methods are used, then system simplicity is maintained, but dynamic adjustment capability deteriorates
Solution Approach 1:
The patent implements dynamic adjustment capability through the knowledge graph's ability to store and retrieve relationships between control states and resources. The system can dynamically query the knowledge graph to determine current system states and adjust control parameters in real-time based on changing conditions. This dynamic querying and adaptation mechanism enables flexible response to system changes without requiring complex reconfiguration of control logic.
Solution Approach 2:
The patent enables dynamic adjustment by storing multiple control state parameters in the knowledge graph and retrieving appropriate parameters based on current system conditions. The system can change control parameters such as flow rates, pressure settings, and operational modes by querying the knowledge graph for the appropriate control state associated with current resource conditions, thereby achieving adaptability through parameter retrieval and modification.
3Measurement precision
If detailed resource mapping is implemented, then control precision is improved, but information processing complexity deteriorates
Solution Approach 1:
The patent creates a simplified copy of the industrial system's relationships and control logic in the form of a knowledge graph. Instead of processing complex real-time system data directly, the system stores structured representations of resource relationships, control states, and entity associations in the knowledge graph. This copying approach allows precise control queries to be executed against the simplified knowledge structure rather than the complex actual system, thereby improving control precision while reducing information processing complexity.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A system maintains, generates, and manages graphs that map resources to control states of a robotic apparatus, The system may receive system control queries and produce search results and contextual information in response. The system may reference the system control queries against the graphs to determine the search results and contextual information. The contextual information may include operator-interactive tools that may be used to control the robotic apparatus. To control the apparatus, the system may generate control state update messages responsive to the operator interactions. The control state update messages may be sent to a control interface of the robotic device. The robotic device may execute an action responsive the receipt of the control state update message.