With the rapid development of industrial environment

edge computing plays a crucial role in the intelligence of industrial park. In this paper

the self-organizing perception-control function migration method of edge node clusters under abnormal events is studied with purpose of improving the resilience and reliability of industrial park. Firstly

the abnormal scenarios are modeled in this paper to simulate the impact of both natural disasters and general fault abnormal events on the intelligent units. Then

the traditional maximum coverage problem is transformed into a perception system maximum range coverage problem

and corresponding algorithms and strategies are developed. On this basis

the perception-control function migration algorithm is designed and a multi-stage optimization method for resilience enhancement is proposed. It is verified through case analysis that the proposed method can quickly respond and reorganize the perception and control functions of the edge node cluster under abnormal events so to effectively enhance the observable and controllable range of the perception system of the industrial park. This study has important practical significance and theoretical value for ensuring the production safety

improving production efficiency and reducing operational costs of the industrial park.