# US 2019/0171942 A1 · Self-evolving agent-based simulation system and method thereof

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2019/0171942 A1
- Document ID: US20190171942A1
- Status: Published application
- Kind: A1
- Application number: 16/196,602
- Filed: 2018-11-20
- Published: 2019-06-06
- Earliest claimed: 2017-11-20
- Assignees: Electronics and Telecommunications Research Institute
- Applicants: Electronics and Telecommunications Research Institute
- Inventors: Eui Hyun PAIK; Dong-Oh KANG; Jang Won BAE; Chun Hee LEE; Joon Young JUNG; Ki Ho KIM; Ok Gee Min
- Web page: https://patentgrep.com/patent/US20190171942A1

## CPC

- G06N 3/126: Evolutionary algorithms, e.g. genetic algorithms or genetic programming
- G06F 30/20: Design optimisation, verification or simulation (optimisation, verification or simulation of circuit designs G06F30/30)
- G06N 3/006: based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
- G06N 5/043: Distributed expert systems; Blackboards
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)

## Abstract

A self-evolving agent-based simulation system generates model evolution strategy for applying a difference between real-data and a simulation resulting value to a simulation model, and reconstructing components included in the simulation module using the model evolution strategy to evolve the simulation model when the difference between the real-data and the simulation resulting value of the agent-based simulation model does not satisfy a value in a predetermined error range.

## Claims

### Claim 1

A self-evolving simulation method of an agent-based simulation system, the method comprising:
outputting a simulation resulting value by executing an agent-based simulation model;
when a difference between real-data and the simulation resulting value of the simulation model is outside a predetermined error range, generating a model evolution strategy for applying the difference between the real-data and the simulation resulting value to the simulation model; and
evolving the simulation model by reconstructing components included in the simulation model using the model evolution strategy.

### Claim 2 (depends on 1)

The method of claim 1, further comprising:
adjusting the number of agents that participate in the simulation model prior to the executing of the simulation model.

### Claim 3 (depends on 2)

The method of claim 2, wherein the adjusting of the number of the agents includes:
grouping agents with similar characteristics to one agent according to a predetermined reference; and
indicating density to the grouped agent.

### Claim 4 (depends on 1)

The method of claim 1, further comprising:
updating the predetermined error range in consideration of tendency and conditions of data.

### Claim 5 (depends on 1)

The method of claim 1, wherein the generating of the model evolution strategy includes generating the model evolution strategy using at least one of change results between the real-data and the simulation resulting value, a simulation model structure, a change history, and the simulation resulting value.

### Claim 6

A self-evolving agent-based simulation system comprising:
a simulation model execution unit for executing an agent-based simulation model and outputting a simulation resulting value;
a change recognition unit for checking a difference between real-data and the simulation resulting value; and
a model self-evolving unit for evolving the simulation model by applying the difference when the difference is outside a predetermined error range.

### Claim 7 (depends on 6)

The self-evolving agent-based simulation system of claim 6, further comprising:
a model evolution strategy generator for generating model evolution strategy based on the difference.
wherein the model self-evolving unit reconstructs components of the simulation model based on the model evolution strategy.

### Claim 8 (depends on 6)

The self-evolving agent-based simulation system of claim 6, wherein the change recognition unit includes:
a complex domain data analyzer for converting the real-data into a form in which the real-data is comparable with the simulation resulting value;
a change recognition processor for recognizing the difference between the real-data and the simulation resulting value at the same time; and
a data tendency analyzer for generating a change recognition results according to the difference when the difference is outside the error range.

### Claim 9 (depends on 8)

The self-evolving agent-based simulation system of claim 8, wherein the model self-evolving unit generates the model evolution strategy using at least one of the change recognition results, a simulation model structure, a change history, and the simulation resulting value.

### Claim 10 (depends on 6)

The self-evolving agent-based simulation system of claim 6, further comprising:
a simulation preparation processor for adjusting the number of agents that participate in the simulation model prior to execution of the simulation model.

### Claim 11 (depends on 10)

The self-evolving agent-based simulation system of claim 10, wherein the simulation preparation processor includes an agent grouping unit for grouping agents having similar characteristics to one agent according to a predetermined reference and indicating density to the grouped agents.
