| With the rapid development of China’s aerospace industry,the frequency of rocket launches has increased year by year,and the quality requirements for rockets have gradually improved.The electro-hydraulic servo system is the core part of the rocket,which directly affects the success rate of the launch.The nozzle baffle servo valve is the core component of the rocket electro-hydraulic servo system,and its characteristics are related to the quality of the coil winding of the servo valve.At present,the coil winding of the nozzle baffle servo valve depends largely on manpower,and it is very difficult to realize automatic winding.Although there are some winding devices at the moment,it is difficult to achieve tight winding with many layers for the thin wire diameter wire.Therefore,the disturbance of coil in the winding process requires manual monitoring and adjustment.The large workload and the long coil winding period result in higher production cost.On the other hand,during the coiling process of artificial monitoring coil,it is impossible to ensure that the failure in the winding line can be eliminated in time,which may cause the failure of the whole winding process because of the delay of brake.According to the above requirements in the coil winding process of the nozzle flapper servo valve,the automatic winding equipment is developed on the basis of the analysis of the winding process of the coil.The user control software and the visual inspection system is established.Optimize the control parameters with experiments.After all,the winding system can realize the coil winding automation with high quality and efficiency.According to the basic principle of coil winding movement,the design and construction of the hardware and tension control system of coil winding equipment are completed.On the basis of the winding characteristics of the coil,the linear feed motion is carried out via speed planning,and the control system based on STM32 is built.Based on the edge detection algorithm,a visual inspection system is established by using industrial camera and Open CV function library.We develop a computer software which integrates functions containing image acquisition,detection,user control and results display.The return difference of linear feed motion is measured by laser displacement sensor and a compensation scheme is proposed.In view of the resistance value of finished coil,the orthogonal test of four factors at four levels on the resistance value of finished coil is designed and completed.The optimal control parameter combination can meet the requirement of the resistance value.Based on the neural network algorithm,taking the experimental data as the sample,the prediction system of the number of winding layers is designed and trained for the control parameters of the automatic winding equipment,which estimates the number of layers through the input control parameters.The accuracy of the prediction algorithm is evaluated by experimental verification,and the optimal control parameters are defined to achieve the finished coil that meets the requirements of resistance and number of layers.In view of the low success rate and high artificial cost of the current winding scheme of the nozzle baffle servo valve coil,the visual inspection system is designed and the control parameter prediction method based on the neural network algorithm is proposed to realize the automatic winding of the high quality coil.Shorten the adapting period and reduce the cost of test.Reduce the work difficulty of the operator and improve the winding efficiency of the coil.Produce the high quality coil satisfying the requirement with a higher success rate. |