Heavy metals are the primary pollutants in agricultural soil and have hindered the sustainable development of agriculture. To control heavy metal pollution, it is essential to identify the pollution sources, particularly the prior source, in agricultural soils. In the current study, Qianguo Irrigation District, a typical agricultural region in Northeast China, was selected to be investigated for the source apportionment of soil heavy metals and identify the prior pollution source. The results showed that the study area was at a moderate pollution level with considerable ecological risk, while Hg and Cd were the main pollutants. Human-health risk assessment indicated that the non-carcinogenic risk for all populations was acceptable (HI < 1), and the carcinogenic risk was not negligible (10⁻⁶ < TCR < 10⁻⁴). The main pollution sources were concluded to be of lithogenic origin (35.5%), livestock manure (25.4%), coal combustion (21.5%), and chemical fertilizers (17.7%). Coal combustion was identified as the prior pollution source, accounting for 47.69% of the RI contribution. This study can provide scientific support for environmental management and pollution control of soil heavy metals in agricultural regions.

With the development of artificial intelligence, the large-scale access of intelligence equipment with complex heterogeneity will bring unpredictable spectrum limitation and complex interferences to traditional cellular networks. This situation can be alleviated by device-to-device (D2D) technique which improves spectrum efficiency by reusing cellular resources. In this paper, a resource allocation framework comprising channel allocation and power control based on Stackelberg game is proposed for distributed interference coordination between D2D and cellular communications with quality-of-service guarantee. First, base station matches the channel to D2D pairs on the basis of potential throughput gain. Second, interferences from D2D pairs to cellular users are converted into a penalty for D2D pairs through the interlayer price, and the optimization problem of system throughput is decoupled into multiple subproblems those can be solved in distributed and iterative manner in each D2D pair. Simulation results show that two proposed distributed algorithms of channel allocation and power control on interference coordination perform well in convergence and overall system throughput.

Real-time machine-vision monitoring of group-housed juvenile sika deer is challenging because targets are often small, densely distributed and partially occluded, while edge deployment requires compact models. This study developed a lightweight system for non-contact behavior recognition under commercial farm-pen conditions. A fixed-camera dataset was constructed with five annotated categories: resting, standing, feeding, social and abnormal. The dataset was limited to one month of recordings from one fixed camera monitoring six juvenile sika deer in a single pen at one farm in Jilin Province, China. To reduce background interference during training, a Multi-target Bionic Foveal Augmentation (MBFA) strategy was proposed to generate target-guided, spatially nonuniform degradation based on the distribution of deer instances, without introducing additional inference cost. A Lightweight Selective Alignment Reorganization and Efficient Detection Network (LSARE-Net) was then designed to enhance local behavioral cues, contextual feature representation, and cross-scale fusion. On the test set, LSARE-Net achieved 91.26% recall, 93.47% mAP@0.5 and 80.24% mAP@0.5:0.95 with 0.70 M parameters and a 1.7 MB model size. After TensorRT conversion, it reached 62.64 FPS on an NVIDIA Jetson AGX Orin. These results demonstrate the feasibility of compact frame-level behavior detection under the evaluated pen conditions. The accompanying prototype organizes detections into behavioral summaries and abnormal-cue records for manual review.

The development and application of network teaching system is conducive to promoting the construction of online courses, teaching methods and means. This paper designs a network teaching system framework, functions, the establishment of the network teaching system database, the realization of the network teaching system. This system mainly realizes the function of teaching courseware and teaching documents, students' autonomous learning, after class communication, job submission, question answering and testing, etc., which can provide ideas for the construction of the open teaching platform.

At present, there are problems in the mainstream network education system in domestic. It is important to carry out autonomous learning and collaborative learning support to solve that. In combination with the relevant education theory, the teaching mode under network environment can be transplanted into the network environment, and a network teaching system which can fully play the teaching mode is developed.

The purpose of teaching evaluation is to improve the quality of teaching. The evaluation method of teaching is numerous, and has achieved good results. Due to the particularity of the curriculum, the teaching evaluation of different courses should be different. In this paper, the evaluation model is established based on grey correlation theory, and the course of SPSS statistical analysis is evaluated, and the evaluation index and model are established. According to the experimental results, the course's setting, teaching process and evaluation model were analyzed.