Reliability based Preventive Maintenance Modeling for Workforce Assignment in Energy Systems

Maintaining network nodes availability in a cellular mobile network (MN) with a remarkable and widespread number of nodes is one of the main challenges of any operational team that demands a highly proactive maintenance method. In this research, an operational research model is presented taking into account resources constraints that will select the tasks for an optimum set of base transceiver stations (BTS). The objective is to achieve the lowest probable financial and non-financial losses. The non-financial factor of BTS is a qualitative factor that is categorized and converted to a quantitative factor. Data samples are extracted from the network operation control monitoring system, which contains information on several hardware and software alarms, the time interval between last corrective and preventive maintenance activities, and outage status shown in binary True/False. Results illustrate significant domination of the neural networks in comparison with conventional parametric method.

Keywords:

Mobile Network , Density Estimation , Neural Network , Bayesian Regularization , Preventive Maintenance .