Abstract:

[Objective] As a large-scale complex system, megacities possess significant characteristics such as high-dimensional globality, dynamic complexity, and rapid growth. Scientific governance of megacities requires efficient cross-domain data circulation across different fields, levels, and entities. [Methods] In response to the major challenges in megacities such as massive data sources, data decentralization, and complex scenarios, this paper proposes a data-efficient cross-domain circulation technology framework. The framework includes key technologies such as distributed data identification based on directory generation, cross-domain scheduling based on data routing, and secure collaborative computing based on algorithm slicing. [Results] Based on this framework, a multi-level and multi-granularity directory chain system is constructed for Beijing, which supports efficient cross-domain data circulation for thousands of scattered data sources, tens of thousands of heterogeneous data structures, and real-time data in the scale of hundreds of billions. [Conclusions] The framework is validated in scenarios such as intelligent dispatching and license verification, providing a feasible technical path for data-efficient cross-domain circulation in megacity governance.

Key words: complex system, data circulation, data identification, cross domain scheduling, collaborative computing