随着光网络规模的快速扩张，光网络的运维与管理在可靠性、灵活性和安全性等方面面临着越来越严峻的挑战。为降低光网络决策控制的复杂度，提升光网络的智能化控管能力，文章结合人工智能技术提出了自优化光网络的概念，并基于边缘计算与云计算协同的思想，提出了边缘计算与云计算协同的自优化光网络实现机制，通过光传送节点和控制节点的协同机制，为光网络动态提供人工智能所需的计算资源。

With the rapid expansion of optical networks, the network operation and management are facing increasingly severe challenges in terms of reliability, flexibility, and security. In order to reduce the complexity of decision making and improve the intelligent control ability of optical networks, this paper proposes the concept of Self Optimizing Optical Networks (SOON). In addition, this paper proposes a SOON implementation mechanism based on edge computing and cloud computing. The cooperative mechanism among optical transport nodes and control nodes is able to satisfy the computing requirement for artificial intelligence application.