Digital microfluidic biochip whole-process comprehensive method, device, equipment and medium
By combining full-process coding and machine learning, the coupling problem of scheduling, layout and routing in the integrated design of digital microfluidic biochips was solved, achieving more efficient chip integration scheme optimization and reducing experimental time.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- SOUTHERN UNIVERSITY OF SCIENCE AND TECHNOLOGY
- Filing Date
- 2026-03-09
- Publication Date
- 2026-07-17
AI Technical Summary
In existing integrated design methods for digital microfluidic biochips, the step-by-step integration method cannot fully consider the coupling effects of operation scheduling, module layout and droplet routing, resulting in the transmission of suboptimal constraints and affecting the instruction and routing process of the integrated scheme.
A comprehensive approach is adopted, which generates an initial chromosome population through full-process coding, performs conflict repair based on position and operation dependencies, uses machine learning to predict droplet routing time, and performs genetic operations based on fitness values to iteratively optimize the chip integration scheme.
This effectively avoids the problem of constrained suboptimal transitivity, improves the quality and computational efficiency of the integrated scheme, and shortens the completion time of biological experimental tasks.
Smart Images
Figure CN122417128A_ABST