AISP
A comprehensive software suite for crystal structure prediction and analysis, integrating symmetry constraints, genetic algorithms, and machine learning, with four core modules: AispGen,AispOP,AispML and AispAgent. dedicated to advancing the discovery and research of novel materials.
Structure Generation
AispGen
A crystal structure generation module based on space group
symmetry constraint methods, enabling efficient creation of
initial crystal structures with precise symmetry requirements.
Structure Search
AispOp
A crystal structure search module utilizing genetic algorithms
for exploring stable and metastable configurations, optimized
for discovering novel materials with desired properties.
Machine Learning
AispML
A module dedicated to machine learning potential generation
and advanced machine learning algorithms, accelerating
materials discovery through data-driven approaches.
Agent Orchestration
AispAgent
A unified agent layer that connects external scientific tools (Matlab, R, Python libraries)、Computational workflows、and mainstream AI agents,
providing a consistent RESTful/WebSocket API with extensible adapters, robust error handling, and structured logging.
Key Features
Flexible Framework: JSON-based input parameter system
for user-friendly configuration and automation
Modern Architecture: Completely refactored codebase
with enhanced performance and maintainability
Integration of Multiple Tools: Seamless integration with a variety of density functional theory (DFT)
softwares and machine learning potential functions for training and optimization networks.
Extensible Workflow Engine: A high-performance framework that seamlessly interfaces with various high-throughput computational workflows,
such as quantum chemistry and molecular dynamics
AISP: Advanced tools for materials discovery