Regression-Based Correction and I-PSO-Based Optimization of HMCVT’s Speed Regulating Characteristics for Agricultural Machinery

Abstract

To improve the speed regulating characteristics of continuously variable transmission for agricultural machinery, in order to meet the engineering and technical requirements of precision agriculture and intelligent agriculture, the paper researches and proposes a method combining the analysis of speed regulating characteristics, regression-based correction, and the improved particle swarm optimization (I-PSO) algorithm. First, the paper analyzes the degree of deviation between the linearization degree and the theoretical value of the speed regulating characteristics of the variable-pump constant-motor system of agricultural machinery according to the measurement results of the bench test. Next, the paper corrects the speed regulating characteristics and compares the regression results based on four models. Finally, the paper proposes a design method for the expected speed regulating characteristics of agricultural machinery and it completes the optimization of speed regulating characteristics and the matching of transmission parameters with the I-PSO algorithm. Results indicate that the speed regulating characteristics of the variable-pump constant-motor system show high linearization (with a coefficient of determination of 0.9775). The theoretical and measured values of the speed regulating characteristics have a certain deviation (with a coefficient of determination of 0.8934). Therefore, correcting the speed regulating characteristics of the variable-pimp constant-motor system is highly necessary. In addition, the second reciprocal function model proposed has the highest correction precision (with a coefficient of determination of 0.9978). The I-PSO algorithm is applicable to the design and application of hydro-mechanical continuously variable transmission (HMCVT) for agricultural machinery. The new method proposed can improve the HMCVT’s speed regulating characteristics efficiently and quickly. It also ensures that the speed regulating characteristics are highly consistent with the expected design characteristics (with a mean error of 1.73%). Thus, the research offers a theoretical direction and design basis for the research and development of continuously variable transmission units in agricultural machinery.