Simulation Device Handling Non-Ideal Observation Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing simulation technologies face challenges in achieving high-resolution and high-accuracy simulations over wide ranges due to non-ideal and discontinuous observation data, particularly in scenarios with incomplete or peculiar data sets, leading to inaccuracies and errors in state estimation.
Innovation Solution
A simulation device that employs a system model, data selection processing, observation models, posterior distribution creation, and unification units to handle multiple types of observation data, categorizing and unifying posterior distributions to improve accuracy and handle missing data, using ensemble approximation and Bayesian methods for stochastic modeling.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If data assimilation is performed to improve simulation accuracy, then measurement precision improves, but device complexity increases due to multiple observation models and posterior distribution processing
Solution Approach 1:
The system segments the complex data assimilation process into distinct functional modules: observation model selection unit, posterior distribution creation unit, and posterior distribution unification unit. Each module handles a specific aspect of the processing, making the overall complex system manageable and maintainable while preserving accuracy improvements.
Solution Approach 2:
The patent introduces posterior distributions as an intermediary mathematical construct between observation data and simulation results. This intermediary allows for systematic handling of uncertainties and multiple data sources without directly complicating the core simulation engine, thereby improving measurement precision while containing complexity growth.
2Measurement precision
If multiple types of observation data are processed to reduce errors, then measurement precision improves, but loss of time increases due to extensive data processing
Solution Approach 1:
The system performs preliminary actions by pre-establishing observation models and their corresponding relationships with state variables before actual data assimilation. This preparation work is done once, allowing subsequent processing of multiple observation data types to proceed more efficiently, reducing the time penalty while maintaining accuracy improvements.
Solution Approach 2:
The patent implements a selective approach where not all possible observation models are applied to all data types. Instead, the system selects appropriate observation models based on the specific characteristics of each observation data type, performing only the necessary processing steps required to achieve accurate results without unnecessary computational overhead.
3Measurement precision
If simulation resolution is increased to improve accuracy, then measurement precision improves, but productivity decreases due to large amount of calculation required
Solution Approach 1:
The system applies local quality by focusing computational resources on regions and parameters where observation data are available and most informative. Rather than uniformly increasing resolution across the entire simulation domain, the posterior distribution-based approach concentrates computational effort where it most directly improves measurement precision, maintaining productivity while achieving high resolution where needed.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The present invention performs a high-resolution and high-precision simulation over a wide range in consideration of observation data that is non-ideal, discontinuous, or peculiar. A simulation device 100 includes a system model 21, a data selection processing unit 30, a plurality of observation models 31, a post-distribution creating unit 40, a post-distribution unifying unit 50, and a determining unit 51. The system model 21 calculates a time evolution of a state vector. The data selection processing unit 30 selects multiple items of observation data. The observation model 31 converts the state vector from the system model 21 on the basis of the relationship with the observation data. The post-distribution creating unit 40 creates, on the basis of the state vector from the observation model 31 and the selected observation data, a first post-distribution based on all pieces of the observation data or a second post-distribution based on absent observation data. The post-distribution unifying unit 50 unifies the first and second post-distributions. The determining unit 51 determines which of the second post-distribution or the unified post-distribution is to be used.