With the continuous advancement of urbanization, the socioeconomic development and environmental damage in Zhengdong New District of Zhengzhou City seriously threaten the development of green space. The paper focused on Zhengdong New District and selected four important nodes in 2007, 2011, 2015, and 2019 to interpret the remote sensing image, explore the local spatial evolution through the method of land use transfer model construction, extract the green space area conversion analysis information of the study area, and analyze the influence factors of its transformation. The research shows that from 2007 to 2019, the internal structure of green space in Zhengdong New District that were cultivated land, woodland, grassland, and water bodies, its showed different trends, and the dynamic changes were different. Among them, the area of cultivated land continued to decrease, the area of construction land continued to increase, the total area of woodland increased, the area of grassland and water bodies fluctuated, and the total green space was greatly reduced; the conversion of land types in Zhengdong New District mainly focused on the conversion of cultivated land to construction land, woodland, and grassland. The results shows that the green space of Zhengdong New District is affected by social, economic, natural, and other factors and various factors interact with each other. However, social and economic development factors are obviously stronger than natural and administrative factors on the evolution of green space area.

Carbon sinks have been widely recognized as critical components of climate change mitigation and achieving carbon neutrality. With rapid urbanization, protecting and optimizing urban carbon sinks remain major challenges. This study uses Zhengzhou as a case study and analyzes 2000–2023 land-use data with the InVEST model to quantify carbon stocks and identify high-value carbon-sink areas. Circuit theory was further integrated to delineate ecological security patterns and inform optimization strategies. The results show a net decrease of 19.12 × 10⁶ t in carbon storage from 2000 to 2023, with the most rapid decline occurring during 2015–2020. Spatially, high-value carbon storage clustered in forested, high-elevation areas in the southwest, whereas low values predominated in the urban core. Carbon-sink source areas continued to shrink: fragmentation increased in the east, the west remained relatively stable, and the central area was highly fragmented. Corridor analysis indicated that the mean corridor length first increased and then decreased, accompanied by an expansion of pinch points and barrier areas. The study developed a systematic optimization framework that establishes a “Two Cores, Five Carbon-Sink Areas, Multiple Corridors” security pattern and proposes targeted conservation measures. The proposed methodology and findings offer a transferable basis for managing urban carbon sinks in rapidly developing regions and support both ecological security and climate-change mitigation, promoting sustainable urban development.

Feature optimization, which is the theme of this paper, is actually the selective selection of the variables on the input side at the time of making a predictive kind of model. However, an improved feature optimization algorithm for breath signal based on the Pearson-BPSO was proposed and applied to distinguish hepatocellular carcinoma by electronic nose (eNose) in the paper. First, the multidimensional features of the breath curves of hepatocellular carcinoma patients and healthy controls in the training samples were extracted; then, the features with less relevance to the classification were removed according to the Pearson correlation coefficient; next, the fitness function was constructed based on K-Nearest Neighbor (KNN) classification error and feature dimension, and the feature optimization transformation matrix was obtained based on BPSO. Furthermore, the transformation matrix was applied to optimize the test sample’s features. Finally, the performance of the optimization algorithm was evaluated by the classifier. The experiment results have shown that the Pearson-BPSO algorithm could effectively improve the classification performance compared with BPSO and PCA optimization methods. The accuracy of SVM and RF classifier was 86.03% and 90%, respectively, and the sensitivity and specificity were about 90% and 80%. Consequently, the application of Pearson-BPSO feature optimization algorithm will help improve the accuracy of hepatocellular carcinoma detection by eNose and promote the clinical application of intelligent detection.