Evaluation Of Assembly Time Based On Artificial Neural Network And Virtual Assembly

Thanks to the development of computer aided design and manufacturing technology, the manufacturing has promoted efficiency of product research and design as well as product manufacturing. But the development of automation and assembly efficiency during the assembling process has not been given the deserved attention, so the paper put forward the way to process preassembly and Assembly process analysis on a product in the design phase, which plays a great significance in the improvement of assembly efficiency.Traditional assembly process design needs to establish a physical model for mould test to determine the design, resulting in long cycle, high cost, and poor effect of the product assembly. Virtual assembly technology provides a new research idea to solve the above problems. Based on DELMIA simulation platform, this article has simulated the assembly path and assembly time, which makes the assembly process optimization more intuitive, more accurate and lower cost. Moreover, this article utilizes artificial neural network algorithm to evaluate measured assembly time, making the assembly process optimization more persuasive.Based on the virtual assembly, based on a model engine as an example, use the matrix method to generate a removed constraint matrix according to Interference conditions generated in the parts movement, and then process a plan on the assembly sequence and assembly path under the condition of virtual assembly. At last the paper uses the neural network toolbox in MATLAB to estimate the assembly planning and evaluation method proposed in this paper to a more reasonable assembly sequence and assembly path.