Design Of Welding Monitoring System Based On Wireless Transmission And Research On Real-time Data Compression Algorithm

The variation of welding arc during welding affects the quality of welding directly.In order to alleviate the adverse effects of arc light,smoke and dust on the health of equipment operators,it is an effective solution to realize long-distance monitoring of welding signals through high-speed wireless data transmission technology.Based on the above engineering requirements,a welding monitoring system based on wireless transmission is proposed in this paper.And a real-time data compression algorithm is designed and developed to solve the problem that the bandwidth of hardware system is limited and can not meet the requirement of high-speed wireless data transmission.Firstly,the development status of on-line monitoring and data compression at home and abroad is investigated and analyzed.According to the requirement of high-speed and real-time transmission,the overall scheme design of welding wireless monitoring system is carried out and design idea of data compression algorithm.Modular design of monitoring system.The system includes data acquisition module,wireless transmission module and networking module.In order to meet the high precision and real-time synchronous acquisition of welding signals,the interface circuit and software control program based on AD7606 and DSP are designed.In order to realize data wireless transceiver function,a wireless transmission module based on nRF24L01 is developed and its interface circuit and software program with DSP are designed.Aiming at the scattered distribution of multi-welders in welding workshop,a distributed communication network is established by CAN bus and the information is transmitted to the host computer through the Ethernet controller W5500.The above modules are debugged to verify that the working accuracy and scope of each functional module can meet the monitoring requirements.In order to realize hardware system to meet the need of wireless high-speed data transmission,considering the periodic data characteristics of welding signals,a compression algorithm based on lifting wavelet piecewise constant-adjacent copy is proposed.The signal data is compressed by three steps:wavelet threshold compression,piecewise constant compression and adjacent segment copy.The algorithm is simulated in MATLAB environment,and the distribution rules of wavelet threshold u,subsection threshold v and similarity threshold s,affecting the compression performance are analyzed.Three parameters are optimized,and the maximum compression ratio reaches 4.5222.Compared with the existing algorithm,the compression ratio and precision have been improved greatly.The research provides a good solution for the wireless data transmission in welding monitoring.Several groups of actual welding experiments were carried out under different welding conditions.By comparing the original signal with the compression and reduction signal,it is proved that the algorithm is suitable for the compression of welding signals.Finally,the system is applied to the prediction of weld deviation in narrow gap seam tracking.The reason of deviation is analyzed and a neural network model for estimating deviation is established based on a large number of welding signals collected.The training of neural network can accurately estimate the deviation trend and quantity of weld seam.The actual welding test proves that the developed monitoring system can make accurate deviation judgment and deviation prediction for narrow gap welding seam deviation.This provides technical support for the follow-up research of seam tracking based on arc signal.