Research on the efficiency characteristic of the hydromechanical continuous variable transmission (HMCVT) in tractors is key to obtaining optimal transmission, developing control strategies, and assessing efficiency. To ease and improve the accuracy of obtaining the efficiency completely based on test measurements or theoretical calculation, this study proposes a method for building the HMCVT efficiency model. The method is based on an improved simulated annealing (SA) algorithm according to a small amount of test data. The study uses 8 groups of transmission efficiency values under different operating conditions obtained from bench tests. By theoretical analysis of the HMCVT, this study divides the total transmission efficiency into (i) the transmission efficiency from the output power of the power source to the confluence mechanism, (ii) the transmission efficiency of the confluence mechanism, and (iii) the transmission efficiency of the output part after confluence. The formulas for the three parts of transmission efficiencies are then derived. This study improves the SA algorithm and uses it to identify the three key parameters of hydraulic systems of the transmission efficiency calculation model. Research results indicate that the efficiency model built using the proposed method exhibits high accuracy with an error of about 1.90%. The improved SA algorithm can rapidly complete key parameter identification with an error of about 2.16%; when the displacement ratio is 0, the efficiency values at the same stage are approximately equal under different operating conditions. The HMCVT efficiency model can be built rapidly and effectively with only five groups of efficiency measurement tests.
To build a more accurate motor efficiency model with a strong generalization ability in order to evaluate and improve the efficiency characteristics of electric vehicles, this paper researches motor efficiency modeling based on the bench tests of two motor efficiencies with differently rated powers. This paper compares and analyzes three motor efficiency modeling methods and finds that, when the measured values in motor efficiency tests are insufficient, the bilinear interpolation method and radial basis kernel function neural networks have poor generalization abilities in full working conditions, and the precision of polynomial regression is limited. On this basis, this paper proposes a new modeling method combining correlation analysis, polynomial regression, and an improved simulated annealing (I-SA) algorithm. Using the mean and the standard deviation of the mean absolute percentage error of the 5-fold Cross Validation (CV) of 100 random tests as the evaluation indices of the precision of the motor efficiency model, and based on the motor efficiency models with verified precision, this paper makes a comparative analysis on the full vehicle efficiency of electric tractors of three types of drive in five working conditions. Research results show that the proposed novel method has a high modeling precision of motor efficiency; tractors with a dual motor coupling drive system have optimal economic performance.
A wet clutch is the key component to realize power uninterrupted in agricultural machinery operation. To reduce impact of the system and improve engagement quality, this paper studies and establishes the dynamic load characteristics model of a wet clutch and analyzes three kinds of tractor working conditions. This paper proposes and adopts the method of combining ‘PLS analysis-Improved SA—Comparison of various models-Actual test data’. The results show that with the limit of 100 Nm, the relationship between dynamic load characteristics and oil pressure is opposite. Load is highly inversely correlated with dynamic load, and it has enough precision to build a power curve model only by load (MAPE is 4.5929%). Take a certain type of tractor for example, oil pressure should be maintained at a low level, plowing resistance should be greater than 1600 N and the mass of transportation should avoid 600~1800 kg. This study provides a direct basis for the control, design and performance improvement of agricultural machinery.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.