An efficient battery pack-level thermal management system was crucial to ensuring the safe driving of electric vehicles. To address the challenges posed by insufficient heat dissipation in traditional liquid cooled plate battery packs and the associated high system energy consumption. This study proposes three distinct channel liquid cooling systems for square battery modules, and compares and analyzes their heat dissipation performance to ensure battery safety during high-rate discharge. The results demonstrated that the extruded multi-channel liquid cooled plate exhibits the highest heat dissipation efficiency. Subsequently, response surface experiments were conducted to analyze the width parameters of various flow channels in the liquid cooled plate Finally, the Design of Experiment (DOE) was employed to conduct optimal Latin hypercube sampling on the flow channel depth (H), mass flow (Q), and inlet and outlet diameter (d), combined with a genetic algorithm for multi-objective analysis. The Tmax of the battery module decreased by 6.84% from 40.94°C to 38.14°C and temperature mean square deviation decreased (TSD) by 62.13% from 1.69 to 0.64. Importantly, the battery thermal management model developed in this study successfully met heat dissipation requirements without significantly increasing pump energy consumption.
In advanced manufacturing, accurate and reliable 3D geometry measurement is vital for the quality control of large-sized components with multiple small key local features. To obtain both the geometric form and spatial position of these local features, a hybrid robot-assisted laser scanning strategy is introduced, combining a laser tracker, a fringe-projection 3D scanner, and a mobile robotic unit that integrates an industrial robot with an Automated Guided Vehicle. As for improving the overall measurement accuracy, we propose an accuracy-enhanced calibration method that incorporates both error control and compensation strategies. Firstly, an accurate extrinsic parameter calibration method is proposed, which integrates robust target sphere center estimation with distance-constrained-based optimization of local common point coordinates. Subsequently, to construct a high-accuracy, large-scale spatial measurement field, an improved global calibration method is proposed, incorporating coordinate optimization and a hierarchical strategy for error control. Finally, a robot-assisted laser scanning hybrid measurement system is developed, followed by calibration and validation experiments to verify its performance. Experiments verify its high precision over 14 m (maximum error: 0.117 mm; mean: 0.112 mm) and its strong applicability in large-scale scanning of key geometric features, providing reliable data for quality manufacturing of large-scale components.
Large-sized components with numerous small key local features are essential in advanced manufacturing. Achieving high-precision quality control necessitates accurate and highly efficient three-dimensional (3D) measurement techniques. A flexible measurement system integrating a fringe-projection-based 3D scanner with an industrial robot is developed to enable the rapid measurement of large object surfaces. To enhance overall measurement accuracy, we propose an enhanced calibration method utilizing a multidimensional ball-based calibrator to simultaneously calibrate for hand-eye transformation and robot kinematic parameters. Firstly, a preliminary hand-eye calibration method is introduced to compensate for measurement errors at observation points, leveraging geometric-constraint-based optimization and a virtual single point derived via the barycentric calculation method. Subsequently, a distance-constrained calibration method is proposed to jointly estimate the hand-eye transformation and robot kinematic parameters, wherein a distance error model is constructed to link parameter errors with the measured deviations of a virtual single point. Finally, calibration and validation experiments were carried out, and the results indicate that the maximum and average measurement errors were reduced from 1.053 mm and 0.814 mm to 0.421 mm and 0.373 mm, respectively, thereby confirming the effectiveness of the proposed method.
Large-sized components with numerous small key local features are essential in advanced manufacturing. Achieving high-precision quality control necessitates accurate and high-efficiency three-dimensional (3D) measurement techniques. A flexible measurement system integrating a fringe-projection-based 3D scanner with an industrial robot is developed to enable the rapid measurement of large object surfaces. To enhance overall measurement accuracy, we propose an online calibration method utilizing a multidimensional ball-based calibrator to simultaneously calibrate for hand-eye transformation and robot kinematic parameters. Firstly, a preliminary hand–eye calibration method is introduced to compensate for measurement errors at observation points, leveraging angular-constraint-based optimization and a virtual single point derived via the barycentric calculation method. Subsequently, a distance-constrained calibration method is proposed to jointly estimate the hand–eye transformation and robot kinematic parameters, wherein a distance error model is constructed to link parameter errors with the measured deviations of a virtual single point. Finally, calibration and validation experiments were carried out, and the results indicate that the maximum and average measurement errors were reduced from 1.041 mm and 0.809 mm to 0.428 mm and 0.394 mm, respectively, thereby confirming the effectiveness of the proposed method.