摘要：

针对充电设施监控系统在大并发数据处理上的难点,提出基于M / M / c 排队论模型的数据并发控制算法。对数据帧在监控系统中的传输过程进行分析,确定解决通信服务器中数据传输的实时性和并发性问题是提高监控系统性能的关键。根据随机过程理论得到充电机数据帧进入通信服务器的过程服从泊松分布的结论,利用排队论建立并发控制算法的数学模型,采用多线程和多队列技术实现数据并发处理算法。根据性能指标,对损失代价和最优网关数量的关系进行定量分析,得到求解最优网关数量的优化方法。将并发处理算法应用到大型充电站中进行分析实验,结果表明,该算法不仅能够满足数据帧传输的实时性和并发性需求,且证明了网关优化方法可准确推导出最优网关的数量。

关键词: 电动汽车, 充电设施, 监控系统, 排队论, 多线程, 并发控制

Abstract:

Due to the difficulty of handling data concurrency in supervisory and control system of charging facility,this paper proposes a data concurrency control algorithm based on M / M / c queuing theory. The real-time and concurrency of data transmission is identified as the key to improve performance of supervisory and control system. The process of data frame into the communication server is substantially Poisson distribution according to stochastic process theory so that the concurrency control theory model is based on queuing theory. The data concurrency control algorithm is implemented by the multithread and multi-queue. Based on performance indicators,it presents quantitative analysis for the relationship between costs and the number of optimal gateway,obtains optimization method for the optimal number of gateways. The concurrent processing algorithm is applied to large-scale charging station for analysis,and experimental results show that the algorithm not only can meet the needs of real-time and concurrency data frame transmission,but also proves that the gateway optimization method can accurate derive optimal number of gateways.

Key words: electric vehicle, charging facility, supervisory and control system, queuing theory, multithread, concurrency control

中图分类号: 

• TP391