基于ISMA-ELM混合模型的選擇性激光燒結(jié)工藝參數(shù)優(yōu)化

打開文本圖片集

關(guān)鍵詞：選擇性激光燒結(jié);黏菌算法;極限學(xué)習(xí)機；Levy飛行;隨機反向?qū)W習(xí);高度破壞性多項式變異DOI：10.15938/j. jhust.2025.02.002中圖分類號：TP18 文獻標(biāo)志碼：A 文章編號：1007-2683（2025）02-0011-11

Abstract：Anew hybrid model is proposed toaddesstheisueofsrinkage inselectivelasersinteringparts，whichcombines the Improved Slime mould Algorithm（ISMA）andExtreme Learning Machine（ELM）to predictthe shrinkagerateofthe partsusing limitedinputdata.Fistly，threeimprovementstrategiessuchasLevyflight，andomopposition-basedleaingandhiglydisuptive polynomialmutationareusedtoimproveteperformanceoftheviscousbacteriaoptimizationalgorithminallaspects.Subsequently， ISMAisusedtooptimizethekeyparametersofELM，andanSMA-ELMmodelisproposedtopredicttheshrinkagerateofSSparts. SimulationresultsdemonstratethattheproposedISMA-ELMobtainsoptimalpredictionresultscompared tothestandardandother algorithm-optiizedELmodels.Finall，theoptimalprocsingparameterspredictedbyteISMA-ELMmodelareusedtoguidethe machining，and the dimensional accuracy of the obtained molded parts is improved by 29.62% compared to the ELM model and 18.02% comparedto the SMA-ELM，which shows thatthe modelcan provideoptimalproess parameters for SLSmolding processing and guide the machining effectively.

Keywords：selective laser sintering；slimemouldalgorithm；extreme learning machine；levy flight；randomopposition-basec learning；highly disruptive polynomial mutation

0 引言

選擇性激光燒結(jié)（selectivelasersintering，SLS）是一種利用高能量激光束逐層燒結(jié)堆積成型的快速成型技術(shù)，近些年得到了廣泛的關(guān)注[1-2]。（剩余12378字）