Discrete Event Simulation Replication Distribution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Discrete event simulations (DES) are time-consuming due to the need for synchronizing processor clocks, leading to poor performance and reduced predictive accuracy, especially as model complexity and size increase, making it difficult to analyze results within project timelines.
Innovation Solution
Implementing a simulation scenario that distributes multiple replications across multiple processors, allowing each processor to execute replications with randomness factors to simulate system variations, and using a master processor to compile results, thereby reducing processing time and improving accuracy without sacrificing simulation accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple replications are executed to improve predictive accuracy, then accuracy is improved, but processing time increases significantly
Solution Approach 1:
The patent divides the simulation replications into separate tasks that can be executed in parallel across multiple processors. Each processor handles a subset of replications independently, allowing the total processing time to be reduced while maintaining the number of replications for accurate predictive analysis.
Solution Approach 2:
The patent introduces a temporal dimension by executing replications simultaneously across multiple processors rather than sequentially. This transforms a single-dimensional time-based execution into a multi-dimensional parallel execution space, reducing total processing time while preserving accuracy through sufficient replication count.
2Speed
If HLA is used to split replications into dependent components for parallel execution, then processing speed improves, but system overhead increases due to clock synchronization requirements
Solution Approach 1:
The patent extracts the clock synchronization requirement from the parallel execution architecture. By using independent task execution where processors don't need to maintain synchronized clocks, the system eliminates the complex synchronization overhead while still achieving parallel speedup through independent replication execution.
Solution Approach 2:
The patent introduces a master processor as an intermediary that collects results from multiple independent processors. This mediator approach allows independent execution without direct coordination between workers, reducing system overhead while maintaining parallel processing benefits.
3Measurement precision
If the simulation model becomes larger and more complex to improve realism, then model accuracy is improved, but execution time increases proportionally
Solution Approach 1:
The patent segments the simulation execution across multiple processors, allowing complex models to be analyzed more efficiently. Each processor handles a portion of the computational load, reducing the execution time for large-scale simulations while maintaining model complexity and accuracy through comprehensive replication analysis.
Data Source
AI summary
A simulation system includes a receiver component that receives multiple replications of a simulation scenario. The simulation scenario includes a plurality of interacting components representing devices and elements of the real-world system or abstract process to be simulated. Each of the interacting components has one or more numerical parameters associated therewith, to represent an operational parameter of the system device. During the execution of the replications, each replication operates with a randomness factor that modifies a respective numerical parameter, in accordance with the sorts of variations that can be expected in the actual system. An assignment component is included that assigns a subset of the replications to multiple processors such that a processor within the multiple processors executes at least one replication.