Bidirectional Prediction of Process Parameters and Quality Indicators of Reconstituted Bamboo Based on Multi-Strategy Traffic Jam Optimization and Support Vector Regression
1.College of Control and Information Engineering,Northeast Forestry University,Harbin 150040,China
2.College of Mechanical and Electrical Engineering,Northeast Forestry University,Harbin 150040,China
Show less
文章历史+
Received
Published
2026-01-03
2026-05-20
Issue Date
2026-06-11
PDF (6947K)
摘要
针对重组竹热压成型过程中工艺参数多因素耦合、依赖经验调参且试错成本高的问题,亟需建立工艺参数与质量指标之间的定量映射关系,并进一步实现面向目标性能的参数反推。为此,提出一种基于多策略(multi-strategy,MS)交通拥堵优化算法(traffic jam optimization,TJO)优化的支持向量回归(support vector regression,SVR)方法——MS-TJO-SVR,用于构建重组竹工艺参数与质量指标的双向预测模型。正向预测以密度、含水率、施胶量和保压时间为工艺参数输入,以静曲强度、水平抗剪强度、吸水宽度膨胀率和吸水厚度膨胀率为质量指标输出;反向预测以质量指标为输入,反推相应工艺参数。通过MS-TJO对SVR关键超参数进行联合寻优,提高模型对非线性关系的表征能力与预测稳定性。结果表明,所构建的MS-TJO-SVR在正向与反向预测任务中均具有较高的拟合精度和较小的预测误差,且在与传统SVR及其他优化SVR方法的对比中表现出更优的综合性能。研究结果可为重组竹热压成型参数优化与质量预测提供有效的建模工具与方法参考。
Abstract
In the hot-press molding process of reconstituted bamboo, the coupling of multiple process parameters, reliance on empirical parameter adjustment, and high trial-and-error costs make it necessary to establish a quantitative mapping relationship between process parameters and quality indicators, and further realize reverse prediction of parameters for target performance. To address this issue, a support vector regression(SVR) method optimized by a multi-strategy (MS) traffic jam optimization(TJO) algorithm, namely MS-TJO-SVR, is proposed to develop a bidirectional prediction model for the process parameters and quality indicators of reconstituted bamboo. In the forward prediction, density, moisture content, adhesive content, and pressure holding time are used as input process parameters, while modulus of rupture, horizontal shear strength, water absorption width swelling rate, and water absorption thickness swelling rate are used as output quality indicators. In the reverse prediction, the quality indicators are used as inputs to predict the corresponding process parameters. By jointly optimizing the key hyperparameters of SVR, MS-TJO enhances the model’s ability to characterize nonlinear relationships and improves prediction stability. The results indicate that MS-TJO-SVR achieves high fitting accuracy and low prediction error in both forward and reverse prediction tasks, and outperforms traditional SVR and other optimized SVR methods in overall performance. This study provides an effective modeling tool and methodological reference for process parameter optimization and quality prediction in the hot-press molding of reconstituted bamboo.
HUN, LIX J, WUY Q,et al.Preliminary study on the production technology of the reconsolidated bamboo and reconsolidated bamboo-wood composite panels for outdoor use[J].Forestry Machinery & Woodworking Equipment,2011,39(11):17-18,21.
ZUOY F, WUY Q, XIAOJ H,et al.Effect of preparation technology on mechanical properties of reconstituted bamboo[J].Journal of Southwest Forestry University (Natural Sciences),2016,36(2):132-136.
ZHANGY H, MAH X, MENGF D,et al.Effects of impregnating methods on the properties of bamboo-based fiber composites[J].China Forest Products Industry,2017,44(9):24-27.
XUY, LIUQ, SUNZ J,et al.Study on the physical and mechanical properties of high-density bamboo scrimbers[J].China Forest Products Industry,2024,61(6):20-26.
LIQ, YANGM J, CHENL B,et al.Effect of technological parameter on performance of recombined bamboo materials[J].Journal of Fujian College of Forestry,2011,31(2):189-192.
[23]
BRERETONR G, LLOYDG R.Support vector machines for classification and regression[J].Analyst,2010,135(2):230-267.
DUJ Y, HOUY B.Parameters selection of support vector regression by genetic algorithms[J].Systems Engineering and Electronics,2006,28(9):1430-1433.
[29]
KAVOUSI-FARDA, SAMETH, MARZBANIF.A new hybrid modified firefly algorithm and support vector regression model for accurate short term load forecasting[J].Expert Systems with Applications,2014,41(13):6047-6056.
WANGX, HUANGK, ZHENGY H,et al.Short-term forecasting method of photovoltaic output power based on PNN/PCA/SS-SVR[J].Automation of Electric Power Systems,2016,40(17):156-162.
LIJ Y, ZHANGW B, ZHAOX Z,et al.Global maximum power point tracking for PV array based on support vector regression optimized by improved whale algorithm[J].Transactions of China Electrotechnical Society,2021,36(9):1771-1781.
XUJ N, NIY L, ZHUC B.Remaining useful life prediction for lithium-ion batteries based on improved support vector regression[J].Transactions of China Electrotechnical Society,2021,36(17):3693-3704.
[37]
丁禹程.重组竹热压温度自抗扰控制及工艺参数质量双向预测研究[D].哈尔滨:东北林业大学,2024.
[38]
DINGY C.Research on active disturbance rejection control of hot-pressing temperature and bidirectional prediction of process parameters and quality for reconstituted bamboo[D].Harbin:Northeast Forestry University,2024.
[39]
WANGJ P, SHANGZ Y.Traffic jam optimizer:A novel swarm-based metaheuristic algorithm for solving global optimization problems[J].Applied Mathematical Modelling,2026,150(PB):116410.
[40]
QINK A, HUANGL V, SUGANTHANN P.Differential evolution algorithm with strategy adaptation for global numerical optimization[J].IEEE transactions on evolutionary computation,2009,13(2):398-417.
[41]
ADEWUMIO A, ARASOMWANM A.An improved particle swarm optimiser based on swarm success rate for global optimisation problems[J].Journal of Experimental & Theoretical Artificial Intelligence,2016,28(3):441-483.
[42]
JIAH M, LUC H.Guided learning strategy:A novel update mechanism for metaheuristic algorithms design and improvement[J].Knowledge-Based Systems,2024,286:111402.