To provide a favorable temperature for a power battery liquid cooling system, a bionic blood vessel structure of the power battery liquid cooling plate is designed based on the knowledge of bionics and the human blood vessel model. For three different discharge rates of 1C, 2C, and 3C, FLUENT is used to simulate and analyze the heat dissipation performance of the liquid cooling plate with a bionic vascular structure. The influence of the pipe distance (A1 and A2) at the coolant outlet, the thickness of the liquid cooling plate, the inner pipe turning radius R of the pipe in the channel, and the mass flow of coolant on the heat dissipation performance are studied. The results show that the pipe distance (A1 and A2), plate thickness, and inner pipe turning radius R have significant effects on the heat dissipation of the liquid cooling plate, especially under a 3C discharge. In addition, the channel area at the coolant outlet also has great influence on the heat dissipation performance of the liquid cooling plate, and the variable width optimization of the channel area at the outlet greatly improves the heat dissipation performance of the liquid cooling plate. Increasing the inlet mass flow rate can improve the heat dissipation capacity, but at the expense of a pressure drop. A verification experiment is designed for 3C discharge. The results show that the error between the experiment and simulation results is within 9.8%; therefore, the simulation is accurate, and the liquid cooling plate has a significant heat dissipation effect.
29 November 2018 in Neural Computing and Applications
Tire slip angle is a vital parameter in vehicle dynamics control. This article proposes a tire slip angle estimation method using intelligent tire technology and machine learning algorithms. First, a finite element model (FEM) of an intelligent tire with five polyvinylidene fluoride (PVDF) piezoelectric film sensors (PVDF sensor) attached to the inner liner is built using ABAQUS software, and the validity of the FEM is verified. Then, the signal responses of sensors in different slip angle, load, tire pressure, vehicle speed, slip ratio, and tread wear under the tire rolling state are analyzed. Subsequently, promising input feature values are selected based on signal response analysis and dimensionally reduced using the principal component analysis theory. Finally, these feature values are used for training three different machine learning techniques with the purpose of slip angle online estimation. The test results show that the slip angle estimation model using Gaussian process regression has the best performance, with a tenfold cross-validation mean error of 5.47%. At the same time, the estimation model also can accurately estimate the tire slip angle in extreme maneuvers, with the mean absolute percentage error of 7.16%. In general, the proposed slip angle estimation method is feasible and has great potential to improve the vehicle motion control performance especially in extreme maneuvers.
In the study of autonomous obstacle avoidance of intelligent vehicles, the traditional artificial potential field method has the problem that the vehicle may fall into the local minima and lead to obstacle avoidance failure. Therefore, this paper improves the traditional potential field function. Based on the vehicle dynamics model, a strategy of jumping out of local minima based on smaller steering angles is proposed. By finding a smaller steering angle and setting a suitable jump out step length, the intelligent vehicle is enabled to jump out of the local minima. Simulation experiments by MATLAB show that the improved method can jump out of the local minima. By comparing the planned trajectories of the traditional method and the improved method in static and dynamic obstacles situations, the trajectory planned by the improved method is smooth and the curvature is smaller. The planned trajectory is tracked by the Carsim platform, and the test results show that the improved method reduces the front steering wheel angle while the intelligent vehicle satisfies the vehicle dynamics constraints during active obstacle avoidance, which verifies the stability and rationality of the improved method.
The real-time monitoring of the vertical force of the tire is the basis for ensuring the driving safety, handling stability, fuel economy and the riding comfort of the vehicle. An algorithm for the vertical force of the tire estimation based on the combination of intelligent tire technology and neural network theory is herein proposed. Firstly, the finite element model of the 205/55/R16 radial tire was established by ABAQUS and the validity of the finite element model was verified through static experiment and dynamic experiment. Secondly, the effects of inflation pressure, speed, load and tread wear on the tire contact patch length and the radial displacement at the virtual acceleration sensor were analyzed based on the finite element analysis method and control variable method. Finally, three kinds of the vertical force prediction algorithms based on the GA-BP neural network algorithm were established, and the network performance of each prediction model was tested. The results show that the vertical force prediction model with inflation pressure and the peak value of radial displacement as the characteristic input parameters has the highest prediction accuracy and the shortest calculation time. At the same time, the mathematical formula of the Model 3 was built.
The current vehicle stability judgment methods in the stability control of distributed drive electric vehicles focuses on using the phase plane method, but the existing phase plane stability boundary delineation methods ignore the influence of the front wheel rotation angle. Therefore, a new phase plane partitioning method (NPPPM) is proposed in this paper, and a new phase plane stability domain library is established. Based on this, the direct yaw moment control of distributed drive electric vehicle through phase plane analysis is established. Firstly, the influence of the front wheel turning angle on the phase plane is considered in the design of the phase plane stability boundary, and a new stability boundary function is fitted by the least squares method. This approach avoids the lateral instability caused by too large a front-wheel turning angle and improves the accuracy of vehicle stability judgment. Next, a hierarchical direct transverse moment controller is constructed using Matlab/Simulink. Among them, the upper layer controller adopts sliding mode control with the sideslip angle error of the center of mass as the tracking target. The lower controller adopts the optimal configuration algorithm with the optimal road adhesion utilization as the objective function. Finally, the effectiveness and superiority of the proposed phase plane partitioning method (NPPPM) for distributed drive electric vehicles with sinusoidal and angular order inputs for the front wheel rotation angle are verified by simulation.