<p>Mobile Edge Computing (MEC) is considered one of the key technologies for enabling large-scale network services, addressing the challenges posed by cloud computing and traditional computing methods. Considering the load imbalance in the edge computing network, the task latency and the uncertainties of device-side mobility, this paper proposes a cloud-edge collaborative task scheduling model that accounts for device mobility,in our cloud-edge collaborative model, we propose task offloading and task scheduling algorithms. During task scheduling, we construct an optimization model using mathematical formulations to simultaneously optimize system task completion delay and edge server load balancing in different time slots. On this basis, a enhanced chaotic mapping marine predation algorithm(ECMPA) is proposed in this paper in order to iteratively optimize our scheduling model. In the ECMPA ,chaotic sequences is introduced in the construction of the elite matrix to avoid the algorithm from falling into local optimal solutions. Furthermore, to balance the exploration and exploitation of ECMPA, we introduce a guided learning strategy.In addition,ECMPA introduces an adaptive step-size control factor to speed up the convergence. Finally, we validate the performance of the cloud-side collaborative task scheduling model,and compare the ECMPA with existing algorithms and the classic marine predators algorithm(MPA) through simulation experiments.The experimental results show that the cloud-side collaborative task scheduling model can be effectively applied to dynamic mobile edge network scenarios,it reduces task completion delays by 5.2% to 19.1%.And experiment results demonstrate that ECMPA not only optimizes task completion latency but also achieves load balancing across edge nodes.</p>