基于多代理的负载均衡算法在新药研发网格中的应用

设计了利用因特网上多个集群的空闲资源来处理生物医学领域数据密集型计算的新药研发网格（DDG）。采用兼备网格和对等网络双重特点的混合资源管理系统架构，解决了传统的主从结构无法避免的单点失效、性能瓶颈等问题。系统使用基于多代理协同计算的负载均衡算法，根据计算结点的实时工作负载状态进行资源的管理和调度，蛋白质分子对接实验验证了设计的有效性，系统可以保持良好的负载均衡状态，同时可以大幅缩短药物筛选周期。

Application of Multi-Agent Based Load Balancing in Drug Discovery Grid

Utilizing the idle resources donated by the clusters that scatter over the Internet, drug discovery grid (DDG) can implement efficient data-intensive biologic applications. With P2P technologies, the high-level resource management framework with a Grid-P2P hybrid architecture can solve some problems caused by traditional master-slave structure, such as single point of failure and performance bottleneck. A novel multi-agent technology based resources scheduling algorithm was designed to schedule the idle computational resources dynamically. The solid molecular docking experimental results show the load balance of the proposed system, which can greatly speed up the process of protein molecules docking.