Research And Design Of Graphical Programming System For Metal Bending

With the continuous improvement of the automation and intelligence level of the sheet metal bending industry,the requirements for the accuracy and production efficiency of the bending process are also gradually increasing.Since the bending programming system exceeds the traditional manual bending in terms of scalability,robustness,and safety,the development and design of an efficient programming system has become an important development direction in the field of bending processing.Based on the analysis of the current domestic and foreign sheet metal bending process algorithms and system software design ideas,this paper designs a graphical programming as the interactive form,neural network and biological evolution for the movement trajectory of the workpiece and the machine tool in sheet metal bending processing.The algorithm is a programming system supported by operations.The main research contents are as follows:Firstly,based on the limitations of traditional bending processing methods and process parameter solving algorithms,the method of using a graphical programming system to assist bending processing is determined.By studying the characteristics of sheet metal bending processing technology,the overall scheme of the bending processing programming system is designed,and its functional requirements and system structure are clarified.On this basis,combined with the lack of accuracy of the process parameter solution formula,a neural network model of the multilayer perceptron was built to predict and calculate the radius of the inner arc of the bending angle.Secondly,adopting the modular design idea,with the process data as the communication bridge,the human-computer interaction module,the database module and the three-dimensional simulation module are organically integrated to form an overall graphical programming system.In order to simplify the bending process,when designing the human-computer interaction module,the projected side section is used as the support of the system data structure,and a graphical parameter input interface is established in combination with supporting platforms such as Visual Studio.In addition,in order to ensure the real-time synchronization of data,the corresponding relationship between different process data was constructed when designing the database module,which optimized the process calculation process.When constructing the 3D simulation module,a two-step research method is adopted,and the processing scene reproduction is realized according to the workpiece state before and after bending.Then,aiming at the low efficiency of the traditional sheet metal bending process planning algorithm,a backtracking algorithm that can effectively reduce the calculation space is proposed.In addition,the positional relationship between the part and the machine tool mold during sheet metal bending processing is analyzed,and variables such as interference collision,production accuracy,and production time are extracted,and the corresponding fitness value function is constructed based on this,which transforms the process planning problem.is a function optimization problem.Aiming at this optimization problem,a process planning algorithm based on particle swarm is designed.On the basis of adaptive inertia weight,crossover and mutation operators are introduced to optimize the poor process in each generation of population.Finally,the radius of the inner arc of the bending angle corresponding to different processing environments is collected from the production data,and the prediction model of the neural network is trained.Finally,the maximum error and average error of the model are obtained to be 0.16 mm and 0.05 mm,respectively.In addition,particle swarm and genetic algorithm are used to plan the sheet metal bending process respectively.The experimental results show that when there are more than 10 bending steps,the improved particle swarm algorithm can quickly calculate the more accurate bending process,the fitness result is about 11% lower than that of the traditional particle swarm.