In this investigation, we study the M/M/\(\infty\) queueing system with server breakdown and two different types of customers called loss and delay customers. The system is subject to random service interruptions due to server breakdown; the broken-down server will be recovered after repair provided by a repairman according to exponential distribution. Due to server breakdown, there may be either fully or partially interruptions in service, i.e., during breakdown each server may stop providing service or provide service with compromised efficiency. The customers’ balking behavior is incorporated during the service interruptions. The stochastic decomposition property for the system size has also established. The sensitivity analysis has performed by taking numerical examples. Adaptive neuro-fuzzy inference system (ANFIS) technique is implemented to develop ANFIS controller so as to evaluate the simulation results by including the realistic feature of fuzzy descriptors which may be noticed at several queueing scenarios.

在本次调查中，我们研究了 M/M/ \(\infty\)服务器故障和两种不同类型的客户（称为丢失客户和延迟客户）的排队系统。系统可能因服务器故障而随机中断服务；故障的服务器将在维修人员按照指数分布进行维修后恢复。由于服务器故障，服务可能全部或部分中断，即在故障期间每个服务器可能停止提供服务或提供服务以降低效率。客户的犹豫行为在服务中断期间被纳入。系统规模的随机分解性质也已建立。通过数值例子进行敏感性分析。