5 th Condition Monitoring & Fault Diagnosis Conference March, 2011, Petroleum University of Technology, Abadan, Iran Chart Pattern Recognition using the Bees Algorithm Afshin Ghanbarzadeh 1 Mechanical Engineering Department, Faculty of Engineering, Shahid Chamran University Ghanbarzadeh.A@scu.ac.ir Abstract Control charts are employed in manufacturing industry for statistical process control (SPC). It is possible to detect incipient problems and prevent a process from going out of control by identifying the type of patterns displayed by the control charts. Various techniques have been applied to this control chart pattern recognition task. This paper presents the use of radial basis function networks for recognising patterns in control charts in order to determine if the process being monitored is operating normally or if it shows gradual changes (trends), sudden changes (shifts) or periodic changes (cycles). The radial basis function networks were trained, not by applying standard training algorithms, but by employing a new optimisation algorithm developed by the authors. The algorithm, called the Bees Algorithm, is inspired by the food foraging behaviour of honey bees. The paper briefly explains the Bees Algorithm and gives the results obtained. Keywords: Bees Algorithm, Neural Network, Pattern Recognition, Control Charts, Swarm Intelligence. 1 Assistant Professor Introduction Statistical Process Control (SPC) employs statistical means such as control charts to show how consistently a process is performing and whether it should be adjusted [1]. SPC control charts enable a manufacturing engineer to compare the actual performance of a process with customer specifications and provide a process capability index to guide and assess quality improvement efforts. By means of simple rules, it is possible to determine if a process is out of control and needs corrective action. However, incipient problems could be detected before the process goes out of control from the type of patterns displayed by the control charts. Various techniques have been applied to this control chart pattern recognition task [2, 3]. This paper presents the use of Radial Basis Function (RBF) networks for recognising patterns in control charts in order to determine if the process being monitored is operating normally or if it shows gradual changes (trends), sudden changes (shifts) or periodic changes (cycles) (Figure 1). The RBF