Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

HEAS: Hierarchical Evolutionary Agent Simulation Framework for Cross-Scale Modeling and Multi-Objective Search

Ruiyu Zhang, Lin Nie, Xin Zhao

TL;DR

HEAS addresses the lack of tooling for cross‑scale, multi‑level agent‑based modeling by unifying hierarchical modeling with evolutionary search and tournament evaluation in a single framework. It introduces a streams‑and‑layers architecture with a shared execution context, supporting forward simulation, parameter/weight optimization, and rigorous cross‑scenario comparisons without refactoring. The approach is demonstrated on ecological and enterprise decision‑making case studies, highlighting how hierarchical composition enables explicit cross‑scale couplings and robust evaluation. The work promises to enhance reproducibility and comparability in cross‑disciplinary studies by reducing glue code and providing an open, extensible foundation for cross‑scale inquiry.

Abstract

Hierarchical Evolutionary Agent Simulation (HEAS) is a Python framework that unifies layered agent-based modeling with evolutionary optimization and tournament evaluation in a single, reproducible workflow. HEAS represents models as hierarchies of lightweight processes ("streams") scheduled in deterministic layers that read and write a shared context, making cross-scale couplings explicit and auditable. A compact API and CLI-simulate, optimize, evaluate-expose single- and multi-objective evolution, PyTorch policy integration via parameter flattening/unflattening, and general tournament tooling with user-defined scoring and voting rules. The framework standardizes evaluation through uniform per-step and episode metrics, persists seeds, logbooks, and hall-of-fame archives, and provides plotting helpers for traces, Pareto fronts, and comparative outcomes, reducing glue code and improving comparability across studies. HEAS emphasizes separation of mechanism from orchestration, allowing exogenous drivers, endogenous agents, and aggregators to be composed and swapped without refactoring, while the same model can be used for forward simulation, optimization, or systematic comparison. We illustrate usage with two compact examples-an ecological system and an enterprise decision-making setting. HEAS offers a practical foundation for cross-disciplinary, multi-level inquiry, yielding reliable, reproducible results.

HEAS: Hierarchical Evolutionary Agent Simulation Framework for Cross-Scale Modeling and Multi-Objective Search

TL;DR

HEAS addresses the lack of tooling for cross‑scale, multi‑level agent‑based modeling by unifying hierarchical modeling with evolutionary search and tournament evaluation in a single framework. It introduces a streams‑and‑layers architecture with a shared execution context, supporting forward simulation, parameter/weight optimization, and rigorous cross‑scenario comparisons without refactoring. The approach is demonstrated on ecological and enterprise decision‑making case studies, highlighting how hierarchical composition enables explicit cross‑scale couplings and robust evaluation. The work promises to enhance reproducibility and comparability in cross‑disciplinary studies by reducing glue code and providing an open, extensible foundation for cross‑scale inquiry.

Abstract

Hierarchical Evolutionary Agent Simulation (HEAS) is a Python framework that unifies layered agent-based modeling with evolutionary optimization and tournament evaluation in a single, reproducible workflow. HEAS represents models as hierarchies of lightweight processes ("streams") scheduled in deterministic layers that read and write a shared context, making cross-scale couplings explicit and auditable. A compact API and CLI-simulate, optimize, evaluate-expose single- and multi-objective evolution, PyTorch policy integration via parameter flattening/unflattening, and general tournament tooling with user-defined scoring and voting rules. The framework standardizes evaluation through uniform per-step and episode metrics, persists seeds, logbooks, and hall-of-fame archives, and provides plotting helpers for traces, Pareto fronts, and comparative outcomes, reducing glue code and improving comparability across studies. HEAS emphasizes separation of mechanism from orchestration, allowing exogenous drivers, endogenous agents, and aggregators to be composed and swapped without refactoring, while the same model can be used for forward simulation, optimization, or systematic comparison. We illustrate usage with two compact examples-an ecological system and an enterprise decision-making setting. HEAS offers a practical foundation for cross-disciplinary, multi-level inquiry, yielding reliable, reproducible results.

Paper Structure

This paper contains 7 sections, 1 figure, 5 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Architecture of HEAS
  4. Package structure
  5. Demonstration of Application in Science: Ecological Case
  6. Applications in Social Science: Enterprise Decision Making
  7. Innovation and Conclusion

Figures (1)

  • Figure 1: Abstract stream–layer architecture in HEAS. Models are composed of minimal step() streams scheduled in deterministic layers that read and write a shared key–value context $C_t$. Within each tick, upstream layers update $C_t$, intermediate layers read from $C_t$ and write actions/states, and downstream layers aggregate metrics $M_t$; the scheduler then advances to $C_{t+1}$. Dashed connectors indicate external hooks for evolutionary search and tournament evaluation.