Mission
To develop differentiable computational models integrating differential equations, scientific machine learning, and experimental data to solve frontier engineering challenges.
"Computational Engineering evolves through the seamless integration of physical principles and machine learning. At PHI Lab, we build differentiable models where conservation laws, numerical methods, and computational intelligence share a unified optimization graph."
The Concept of PHI (φ)
The symbol φ (phi) stands for Physics-informed Hybrid Intelligence and evokes fundamental continuum mechanics and multiphase flow parameters — such as void fraction, porosity, and phase variables —, symbolizing the convergence of physical knowledge and artificial intelligence.
Mission & Vision
Vision
To establish Physics-informed Hybrid Intelligence as a leading international paradigm for computational science, multiphase simulation, and high-performance academic-industrial software.
Scientific Principles
- 01 Physics Priority: Conservation laws ground and constrain data-driven models.
- 02 Differentiable Modeling: Numerical solvers and neural networks share gradient backpropagation.
- 03 Complementary Learning: Machine intelligence calibrates and accelerates physics without replacing it.
- 04 Experimental Loop: Physical experiments close the optimization loop for virtual sensors.
Permanent Research Programs
1. Differentiable Computational Physics
CFD, Multiphase Flows, and Differentiable Numerical Solvers.
2. Scientific Machine Learning
Physics-Informed Neural Networks (PINNs), Neural Operators, and Surrogate Models.
3. Physical World Models
Latent Fluid Dynamics Modeling and Generative Physics.
4. Cyber-Physical Engineering
Digital Twins, Virtual Sensors, and Real-Time Optimization.
5. High-Performance Scientific Computing
GPU Programming, HPC, and Open-Source Scientific Software.
6. Intelligent Energy Systems
Flow Assurance, Artificial Lift, and Geoenergy.
Global Scientific Collaborations & Research Networks
International Academic Partnerships & Joint Grants
PHI Lab actively fosters international research collaborations with universities, global labs, and centers of excellence in Scientific Machine Learning and Computational Engineering. We welcome joint grant proposals (e.g., FAPESP SPRINT, Horizon Europe, bilateral agreements), visiting scholar exchanges, and co-advising Ph.D. positions.