Dual-source heat pump systems combining photovoltaic-thermal (PVT) and air-source technologies have attracted considerable research interest due to their energy complementarity. Based on the climatic characteristics of the Dalian region, this study conducted field measurements and data analysis on a developed dual-source heat pump system incorporating three adaptive operational modes: (1) PVT mode, (2) PVT/air dual-source mode, and (3) photovoltaic (PV)/air-source mode. Compared to Mode (3), Mode (1) achieves a 5.76% higher heating capacity and an 11.56% greater electrical efficiency. Meanwhile, Mode (2) demonstrates a 12.23% increase in heating capacity, and a 9.14% improvement in electrical efficiency relative to Mode (3). A data-driven methodology is provided to quantify the system’s evaporation temperature, the thermal efficiency of PVT mode, and the coefficient of performance (COP) of the PVT heat pump. The economic assessment demonstrates that the proposed dual-source heat pump system achieves a heating cost as low as RMB 0.1125/kWh and a payback period of 6.4 years, indicating favorable economic benefits. This study provides fundamental data and computational methods for the optimized operation of the PVT/air dual-source heat pump.
Leakage identification in district heating networks (DHNS) is challenging because leakage-induced transient pressure waves often overlap with pressure disturbances triggered by routine operations such as valve regulation, pump speed variation, and emergency shut-off. In addition, the scarcity of high-quality labeled leakage samples limits the robustness of data-driven models under small-sample conditions. To address these issues, this study proposes a compact and moderately interpretable framework for multi-condition identification from transient pressure-wave signals, integrating signal preprocessing, handcrafted statistical feature extraction, multiclass ReliefF-based feature selection, and class-wise generative adversarial network (GAN) augmentation in the selected feature space. A dataset containing four representative conditions, namely leakage, valve regulation, pump speed regulation, and emergency valve shut-off, was constructed using an integrated indoor district heating network testbed. After Hampel-based spike suppression and zero-phase Butterworth band-pass filtering within 0.5 to 300 Hz, time- and frequency-domain statistical features were extracted, and a compact subset was selected by multiclass ReliefF. A class-wise GAN was then used to augment the training set in feature space, while all evaluations were performed strictly on real samples. The results show that feature-space augmentation improves robustness and generalization under operational disturbances and noise. Using random forest as the representative classifier, Accuracy and Macro-F1 increased from 0.960 to 0.985, while leakage recall improved from 0.920 to 0.980. Further comparisons confirmed that the ReliefF-selected subset outperformed representative alternatives such as LASSO and mRMR. Overall, the proposed framework provides an effective solution for distinguishing leakage events from operational disturbances and offers practical support for online monitoring and intelligent operation of DHNS.
Sustainable development is paid more and more attention all over the world nowadays. Kitchen environmental improvement and energy saving are closely related to national economy, household consumption and people\'s health, which is one of the important contents in the sustainable development. The conventional ventilation mode is only installing a range hood in a residential kitchen. The distribution law of contaminants in a residential kitchen in conventional ventilation mode is studied via experimental test, and the results are shown in tables and figures in the paper. We found that locations of contaminants accumulation are high near to ceiling above the location of cooking utensil (range) in kitchen section planes with range in it, and closer to the walls in horizontal planes. According to the contaminant distribution patterns, for creating a clean and energy saving kitchen, this project set three different control schemes: (1) installed a range-hood only; (2) installed a range-hood and a natural air make-up system (a fresh air inlet under one of walls); (3) installed a range-hood, a natural air make-up system, and an exhaust system (a louver in the ceiling for mechanical ventilation). When energy consumption is the same, the contaminant removing capabilities of three schemes are compared with exergy theory and exergy intensity and sustainable development analysis. The results are: (1) To keep the kitchen cleanness, the best effectively control method is scheme 3, followed by scheme 2, and then scheme 1; (2) It conforms to the sustainable development strategies that the scheme 3 is proposed as an alternative of the conventional scheme 1, since the scheme 3 is the best effective and does not increase energy consumption compared with others.
Leakage identification in district heating networks is challenging because leakage-induced transient pressure waves often overlap with pressure disturbances triggered by routine operations such as valve regulation, pump speed variation, and emergency shut-off. In addition, the scarcity of high-quality labeled leakage samples limits the robustness of data-driven models under small-sample conditions. To address these issues, this study proposes a compact and moderately interpretable framework for multi-condition identification from transient pressure-wave signals, integrating signal preprocessing, handcrafted statistical feature extraction, multiclass ReliefF-based feature selection, and class-wise generative adversarial network augmentation in the selected feature space. A dataset containing four representative conditions, namely leakage, valve regulation, pump speed regulation, and emergency valve shut-off, was constructed using an integrated indoor district heating network testbed. After Hampel-based spike suppression and zero-phase Butterworth band-pass filtering within 0.5 to 300 Hz, time- and frequency-domain statistical features were extracted, and a compact subset was selected by multiclass ReliefF. A class-wise generative adversarial network was then used to augment the training set in feature space, while all evaluations were performed strictly on real samples. The results show that feature-space augmentation improves robustness and generalization under operational disturbances and noise. Using random forest as the representative classifier, Accuracy and Macro-F1 increased from 0.960 to 0.985, while leakage recall improved from 0.920 to 0.980. Further comparisons confirmed that the ReliefF-selected subset outperformed representative alternatives such as LASSO and mRMR. Overall, the proposed framework provides an effective solution for distinguishing leakage events from operational disturbances and offers practical support for online monitoring and intelligent operation of district heating networks.