Research And Development Of Remote Operation And Maintenance System For Hollow Elevator Guide Rail Assembly Production Line

With the development and progress of society,more and more high-rise buildings.In order to improve people ’s quality of life and save physical strength,elevators have been more and more widely used in high-rise buildings.Elevator guide rail is an indispensable part in elevator installation,and the safety of elevator operation is closely related to its quality.In the past,the production of elevator guide rails relied on manpower.Each processing step was done manually,with low production efficiency and uneven quality of guide rails.Nowadays,with the development of intelligent technology,modern industrial production tends to be more and more automated,which greatly improves production efficiency and ensures product quality.However,there are many stations in the production line of hollow elevator guide rail assembly,and each station contains many mechanical equipment.Any failure of any equipment will cause the whole production line to stop running.Therefore,it is of great significance to find out the causes of production line faults in time and solve them quickly to improve the operation efficiency of production line.However,the traditional fault diagnosis relies on the accumulated experience of technicians,and the diagnosis time is long and the diagnostic accuracy is difficult to guarantee.In summary,in the face of a whole industrial production line,the use of intelligent diagnosis technology for equipment fault diagnosis,the development of remote operation and maintenance system for real-time monitoring of production line operation status and fault warning,has become an important research topic in the automation industry.Aiming at the hollow elevator guide rail assembly production line,this thesis proposes a fault diagnosis method based on deep learning and develops a remote operation and maintenance system for the hollow elevator guide rail assembly production line.The main research contents of this thesis can be divided into the following aspects:(1)Based on the composition of the production line stations for the hollow elevator guide rail assembly,the processing process of each station was introduced,and the possible faults and symptoms of important equipment at each station were analyzed.(2)Aiming at the problems that the traditional convolutional neural network can not fully extract the fault feature information in a single dimension and the extracted fault feature information contains redundant information,this thesis proposes a method of combining 1DCNN with 2D-CNN to diagnose the important equipment of the production line.Firstly,the input original vibration signal is divided into two channels for feature extraction.The upper channel uses 1D-CNN to extract the feature of the one-dimensional input signal,and the lower channel uses the improved HHT to process the one-dimensional input signal to obtain the time-frequency map,and then uses 2D-CNN to extract the feature,and the attention mechanism module is added to the upper and lower channels.Finally,the features of the two channels are fused for fault diagnosis.(3)Based on the introduction of the production line processing process and the fault diagnosis of important equipment earlier in this article,a B/S architecture based remote operation and maintenance system for the production line assembly has been developed.Firstly,the overall layout of the remote operation and maintenance system was carried out,followed by hardware selection for data collection and transmission based on the layout.Finally,a remote operation and maintenance system for the hollow elevator guide rail assembly production line was developed using My SQL database,PHP language,and Apache server,achieving functions such as user management,monitoring of production line operation status,and fault diagnosis of key components in the production line.