Embedded Model Generation via Relation Replication
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Highly computationally intensive mathematical models used in simulations are limited for use on devices with limited processing capabilities, such as embedded systems, due to their complexity and resource requirements, which can lead to system instability and inefficiency.
Innovation Solution
A system comprising an evaluation manager, relation manager, and model generator manager that identifies relevant input parameters and relations in a complex model, generates a simplified model by replicating these relations, and approximates the behavior of the original model using lookup tables and mathematical approximations, suitable for execution on embedded devices.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a highly accurate non-embedded model is used, then measurement precision is improved, but device complexity increases and cannot be executed on embedded systems
Solution Approach 1:
The patent creates a simplified copy of the complex non-embedded model that replicates its essential behavior and output parameters while using reduced computational complexity. The simplified model maintains the functional relationships and key output characteristics of the original model but with fewer computational resources required for execution on embedded systems
Solution Approach 2:
The patent transforms the model by changing computational parameters and reducing the complexity of mathematical operations while preserving the essential input-output relationships. This involves modifying the computational methods used in the model to achieve acceptable accuracy with reduced processing requirements suitable for embedded systems
2Device complexity
If a simplified model is generated for embedded systems, then device complexity is reduced, but measurement precision deteriorates
Solution Approach 1:
The patent extracts and retains only the essential relations and parameters from the complex non-embedded model that are necessary for achieving acceptable accuracy in the simplified version. By identifying and preserving only the critical computational relationships, the patent creates a simplified model that maintains sufficient precision for embedded system applications
Solution Approach 2:
The patent segments the complex model into essential functional components and relations, separating the critical computational elements from the less important details. This segmentation allows the simplified model to focus on maintaining accuracy for key output parameters while eliminating unnecessary computational complexity
3Measurement precision
If a complex mathematical model is executed, then measurement precision is improved, but use of energy increases and productivity decreases
Solution Approach 1:
The patent creates a simplified copy of the complex model that maintains the essential computational relationships and output accuracy while significantly reducing the execution time and processing speed requirements. This simplified copy can be executed efficiently on embedded systems with limited computational resources
Solution Approach 2:
The patent changes the computational parameters and mathematical operations in the model to achieve acceptable accuracy with reduced processing requirements. This involves transforming the computational methods to balance accuracy and execution speed for embedded system constraints
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Generation of models in real time embedded systems that approximate non-embedded models while reducing a complexity associated with the non-embedded models is provided herein. A system (100) can comprise a memory (108) coupled to a processor (110) that stores and executes executable components comprising an evaluation manager component (102) that identifies an input parameter (112) of a first model (114) based on a defined output parameter (116) of the first model (114) and a relation manager component (104) that determines one or more relations (118) in the first model (114). Further, the executable components can comprise a model generator manager component (106) that generates a second model (120) that approximates the first model (114) and includes a replication of the one or more relations (122) of the first model (114).