基于IPSO-K-means的拖拉机丘陵地区运输作业工况构建方法

Construction Method of Tractor Transportation Operation Condition in Hilly Areas Based on IPSO-K-means

摘要:

为构建能够真实全面反映拖拉机丘陵地区运输工况的标准化运行曲线,提升拖拉机作业效率、降低能耗,本文提出一种基于改进粒子群优化算法(IPSO)与 K-means 融合的拖拉机运输作业工况构建方法(IPSO-K-means)。 首先对采集的原始运行数据进行运动学片段分解处理,选取能够表征拖拉机运动特性的特征参数,构建多维特征矩阵进行量化分析。 采用主成分分析(PCA)进行数据降维处理,基于 IPSO-K-means 算法对降维后的特征空间进行聚类分析,以所有特征维度的平均相对误差之和为目标函数,并基于特征相似性选取代表性片段,合成能够全面反映拖拉机实际运输工况的标准化运行曲线。 最终的特征参数匹配度验证结果表明,构建的工况较原始样本特征参数的相对误差均保持在 10% 以下,整体平均误差 6. 33% 。 同时,构建的拖拉机运输作业工况与原始总体样本数据的速度、加速度联合概率在主要分布区域上表现相近,整体的分布趋势一致。 表明本文所构建的拖拉机运输作业工况较好地反映了拖拉机丘陵地区真实的运输作业情形。

Abstract:

In order to construct a standardized operation curve that can truly and comprehensively reflect the transportation conditions of tractors in hilly areas, a tractor's transportation condition construction method was proposed based on the integration of the improved particle swarm optimization algorithm (IPSO)and K-means. The research firstly conducted kinematic fragment decomposition processing on the collected raw operation data, selected the characteristic parameters that can represent the tractor's motion characteristics, and constructed a multi-dimensional feature matrix for quantitative analysis. Principal component analysis (PCA)was adopted for data dimension reduction processing. The IPSO-K-means algorithm was used to conduct cluster analysis on the feature space after dimension reduction. The sum of the average relative errors of all feature dimensions was taken as the objective function, and representative segments were selected based on feature similarity to synthesize a standardized operation curve that can comprehensively reflect the actual transportation condition of tractors. The final verification results of the matching degree of characteristic parameters showed that the relative errors of characteristic parameters of the constructed working conditions compared with the original samples remained below 10% , and the overall average error was 6. 33% . Meanwhile, the joint probabilities of speed and acceleration of the constructed tractor transportation operation condition and the original population sample data showed similarities in the main distribution areas, and the distribution trend was relatively consistent. It indicated that the tractor transportation operation conditions constructed well reflected the real transportation operation situation of tractors in hilly areas.