To investigate the two most important pigments of human, melanin and hemoglobin, an image-channel-difference of optical density space based algorithm was proposed for automatically extracting melanin and hemoglobin concentration distribution map from single color image. The algorithm built mathematic model between pigment and digital image based on theoretical foundation of skin structure and its optical property. The input image firstly was divided into several sub-regions. Independent component analysis (ICA) technology was performed in every sub-region to calculate Separation Vector, which is successively verified by specified rules. All the valid Separation Vectors were then re-combined to form new vectors, from which the final separation vector with minimal deviation is selected. The pigment concentration distribution maps were displayed after obtaining the final global separation vector. The experiments show the effectiveness and great robustness of the proposed algorithm.

Cavity pressure and temperature difference are two important quality criteria. Considering that most injection molded parts have a sheet like geometry, a fast strip analysis model based on mechanics equations for viscous fluid, was adopted as a surrogate model to approximate the time-consuming computer simulation software for predicating the above quality criteria. According to the predicted quality criteria, a particle swarm optimization algorithm was employed to find out the optimum injection parameters. The proposed optimization method can optimize the injection parameters in short time and it does not rely on any knowledge of molding process. Finally, two experiments were employed to validate the surrogate model and the proposed optimization method. Experimental results show that the cavity pressure predicted by the surrogate model agree well with the experimental data, with the relative error being less than 8.41%, and the results of the proposed optimization method are nearly identical to that of response surface method, while the required time of the proposed method is only 0.02% of that of response surface method.

A thorough analysis and cross-comparison of recent relevant works was provided, outlining a closed-loop process for EEG data analysis based on deep learning. EEG data were introduced, and the application of deep learning in three key stages: preprocessing, feature extraction, and model generalization was unfolded. The research ideas and solutions provided by deep learning algorithms in the respective stages were delineated, including the challenges and issues encountered at each stage. The main contributions and limitations of different algorithms were comprehensively summarized. The challenges faced and future directions of deep learning technology in handling EEG data at each stage were discussed.

The porous materials with low dielectric constant are suitable for the applications in integrated circuits. From the aspects of composition and structure, preparation method and dielectric properties, this work introduced the porous low-dielectric-constant materials with different matrix such as inorganic materials, organic materials, inorganic and organic composite separately, whose dielectric constants can be reduced to 1.99, 1.50, 1.99, respectively. The using temperature of the porous low-dielectric-constant materials with organic matrix can reach 450 ℃. The flexural strength of the porous low-dielectric-constant materials with inorganic matrix can reach 136 MPa. The introduction of cave into the materials leads to the decrease of mechanical properties and the increase of dielectric loss. The effort to get a low-dielectric-constant and improve the above properties can broaden the application scope of the porous materials．

At present, the manufacturing enterprises are extending to the service industry, and product service system(PSS) is developing rapidly as a new manufacturing system mode. The background of PSS was analyzed, in view of the demand of users for the service, the product innovation through the service from enterprise, the demand for the difference competition, the demand for the protection of surroundings, and the role of information and web technology. The development of relative theory and practice of PSS was summarized. The characteristics of various product services were analyzed. The theories of PSS were researched, which include: the theory and method for the management of lifecycle of product services, the optimization theory for organization and process of PSS et al. The key technologies were studied, such as: the product design for the product services, the technology of mining the user demand for the product services, the technology of product maintaining service, the technology of collection information of product services, the humanness technology of product services, et al. The function and characteristics of some PSS were summarized.

A new multimodal sentiment analysis model (MTSA) was proposed on the basis of cross-modal Transformer, aiming at the difficult retention of the modal feature heterogeneity for single-modal feature extraction and feature redundancy for cross-modal feature fusion. Long short-term memory (LSTM) and multi-task learning framework were used to extract single-modal contextual semantic information, the noise was removed and the modal feature heterogeneity was preserved by adding up auxiliary modal task losses. Multi-tasking gating mechanism was used to adjust cross-modal feature fusion. Text, audio and visual modal features were fused in a stacked cross-modal Transformer structure to improve fusion depth and avoid feature redundancy. MTSA was evaluated in the MOSEI and SIMS data sets, results show that compared with other advanced models, MTSA has better overall performance, the accuracy of binary classification reached 83.51% and 84.18% respectively.

A front multi-vehicle target tracking algorithm optimized by DeepSort was proposed in order to improve the awareness of autonomous vehicles to the surrounding environment. Gaussian YOLO v3 model was adopted as the front-end target detector, and training was based on DarkNet-53 backbone network. Gaussian YOLO v3-Vehicle, a detector specially designed for vehicles was obtained, which improved the vehicle detection accuracy by 3%. The augmented VeRi data set was proposed to conduct the re-recognition pre-training in order to overcome the shortcomings that the traditional pre-training model doesn't target vehicles. A new loss function combining the central loss function and the cross entropy loss function was proposed, which can make the target features extracted by the network become better in-class aggregation and inter-class resolution. Actual road videos in different environments were collected in the test part, and CLEAR MOT evaluation index was used for performance evaluation. Results showed a 1% increase in tracking accuracy and a 4% reduction in identity switching times compared with the benchmark DeepSort YOLO v3.

The research progress of direct air capture CO₂ adsorbents was reviewed. The advantages and disadvantages of alkali/alkaline metal based adsorbents, metal organic framework adsorbents, amine loaded adsorbents and moisture swing adsorbents were compared. Meanwhile, the properties of adsorbents from the aspects of adsorption capacity and amine efficiency, kinetics and supporters, regeneration mode and energy consumption, thermal stability and resistance to degradation were evaluated. Additionally, the related engineering demonstration projects and economic evaluation were briefly discussed. Finally, the problems existing in the current research were summarized, and the future research direction was prospected.

In order to analytizing neurons release pattern of the primary motor cortex of rats accurately and predicting corresponding body movement, the activities of the neurons ensemble spike activities in rats primary motor cortex and the forelimb pressure were recorded simultaneously in the experiment. K-means and principal component analysis were used to classification of neurons, then the partial least squares was used to analyze the relations between the neurons activities of the primary motor cortex of the rat and the forelimb motion parameters, and the results were compared with Wiener filter and Kalman filter. The experimental results indicate that the activities of neurons ensembles began a trend of increase 0.6 second before lever pressing, Which hints the neurons distributed activities of the primary motor cortex in rats can be used to analysis and prediction its forelimb movement and the correlation coefficient between the predicted value and real pressure value is more than 085 using the partial least squares, with a better decoding results than those using the Wiener filtering and Kalman filtering．

For the poor real-time detection capability of the current object detection model in the production environment of electronic components, GhostNet was used to replace the backbone network of YOLOv5. And for the existence of small objects and objects with large scale changes on the surface defects of electronic components, a coordinate attention module was added to the YOLOv5 backbone network, which enhanced the sensory field while avoiding the consumption of large computational resources. The coordinate information was embedded into the channel attention to improve the object localization of the model. The feature pyramid networks (FPN) structure in the YOLOv5 feature fusion module was replaced with a weighted bi-directional feature pyramid network structure, to enhance the fusion capability of multi-scale weighted features. Experimental results on the self-made defective electronic component dataset showed that the improved GCB-YOLOv5 model achieved an average accuracy of 93% and an average detection time of 33.2 ms, which improved the average accuracy by 15.0% and the average time by 7 ms compared with the original YOLOv5 model. And the improved model can meet the requirements of both accuracy and speed of electronic component surface defect detection.

A refrigerant-based battery thermal management system with compact structure and high heat efficiency was proposed in order to solve the heat dissipation problem of high specific energy and superior energy density power battery. The coupling model of air-conditioning and battery thermal management was constructed by AMESim based on the whole vehicle system. The temperature drop and temperature uniformity of the single cell and battery module, the system’s COP and exergy efficiency were analyzed from the point of view of system temperature response characteristics and system energy consumption. Results show that the refrigerant-based system has a fast temperature response characteristic. The battery can be quickly cooled, and a better temperature uniformity under high temperature and high speed steady state and dynamic conditions can be achieved. The energy analysis was conducted for a stable working condition, and a higher system energy efficiency ratio with a COP of 4.19 was obtained. The exergy efficiency of system was 46.17%, and there’s the promotion space of system exergy efficiency.

An all digital architecture was presented to realize the logarithm modeled automatic gain control (AGC) loop based on field programmable gates array (FPGA) platform. The architecture is mainly made up of four modules, including envelope detector, coordinator rotation digital computer (CORDIC) based iterative realization of effective hyperbolic function, infinite impulse response (IIR) based loop filter, and reconfigurable control logic generator. Envelope detector can meet the filter performance requirement only with finite impulse response (FIR) filter of low order. Symmetric coefficient based transposed architecture reduces the number of multipliers in filter to the half. By approximation to analog integrator and suitable process, IIR based filter can simplify the architecture of digital integrator. The logarithm convertor is easy to be realized by CORDIC based iterative hyperbolic function processor. The architecture can also output with random precision, and realize the compromise of circuit resources and speed, which overcomes the high demand for random access memory (RAM) in traditional algorithms. With fully parallel pipeline architecture, the maximum working clock can achieve 206 MHz. Finally, the realization results based on the FPGA processor and the hardware simulation were given. Experimental results were consistent well with the theoretical analysis．